Naunyn-Schrnjedeberg‘s Arch Pharmacol (2004) 370: 203-210
`DOI 10.1007/s00210-004-0954-1
`
`
`
`Janos Magyar - Norbert Szentandrassy -
`Tamas Banyasz - Valéria Kecskeméti - Péter P. Nanasi
`
`Effects of norfluoxetine on the action potential
`and transmembrane ion currents in canine ventricular
`
`cardiomyocytes
`
`Received: 6 May 2004 I Accepted: 14 June 2004 1 Published online: 26 August 2004
`(Q) Springer-Verlag 2004
`
`important active
`Abstract Norfluoxetine is the most
`metabolite of the widely used antidepressant compound
`fluoxetine. Although the cellular electrophysiological
`actions of fluoxetine are well characterized in cardiac
`cells, little is known about the effects of its metabolite. In
`this study, therefore, the effects of norfluoxetine on action
`potential
`(AP) configuration and transmembrane ion
`currents were studied in isolated canine cardiomyocytes
`using the whole
`cell configuration of patch clamp
`techniques. Micromolar concentrations of norfluoxetine
`(l—l0 u.M) modified AP configuration: amplitude and
`duration of the AP and maximum velocity of depolariza-
`tion were decreased in addition to depression of the
`plateau and elimination of the incisura of AP. Voltage
`clamp experiments revealed a concentration-dependent
`suppression of both L—type Ca2+ current, IQ, (EC50=l.l3
`:|:0.08 uM)
`and transient outward K""
`current,
`I“,
`(EC5o=1.19:t0.l7 uM) having Hill coefficients close to
`unity. The midpoint potential of the steady-state inactiva-
`tion of IC, was shifted from ---20.9:t0.7S mV to ---127.7
`:|:1.35 mV by 3 uM norfluoxetine (P<0.05, n=7). No such
`shift in the steady—state inactivation curve was observed in
`the case of Ito. Similarly, norfluoxetine caused no change
`in the steady—state current—voltage relationship of the
`membrane or in the density of the inward rectifier K+
`current, [K]. All these effects of norfluoxetine developed
`rapidly and were fi.1lly reversible. Comparing present
`results with those obtained previously with fluoxetine, it
`can be concluded that norfluoxetine displays stronger
`suppression of cardiac ion channels than fluoxetine.
`
`J. Magyar- N. Szentandrassy - T. Banyasz - P. P. Nanasi (El)
`Department of Physiology, University Medical School of
`Debrecen,
`PO Box 22, 4012 Debrecen, Hungary
`e-mail: nanasi@phys.dote.l1u
`Tel.: +36-52-416634
`Fax: +36-52-432289
`
`V. Kecskeméti
`Department of Pharmacology and Pharmacotherapy,
`Semmelweis University,
`PO Box 370, 1445 Budapest, Hungary
`
`the majority of the cardiac side effects
`Consequently,
`observed during fluoxetine treatment are likely to be
`attributed to its metabolite norfluoxetine.
`
`Keywords Fluoxetine - Norfluoxetine - Cardiac cells -
`Electrophysiology — Action potentials - Calcium current -
`Potassium current - Antidepressant drugs
`
`Introduction
`
`is a widely used antidepressant
`Fluoxetine (Prozac)
`compound,
`its primary action is based on inhibition of
`serotonin reuptake in the central nervous system. Recent
`studies indicate, however,
`that fluoxetine has several
`additional effects on neuronal (Stauderman et al. 1992;
`Tytgat et al. 1997; Pancrazio et al. 1998), cardiac (Pacher
`et al. 2000; Magyar et al. 2003), smooth muscle (Farrugia
`1996), and epithelial (Rae et al. 1995) cells. These effects
`appear to involve direct inhibition of the ion channels in
`the cell membrane.
`It
`is generally believed that a
`significant part of the therapeutic activity of fluoxetine is
`attributable
`to its most
`important
`active metabolite
`norfluoxetine
`(Fuller
`and Snoddy
`1991), which is
`produced via demethylation of fluoxetine by cytochrome
`P450-2C9 enzyme (von Moltke et al. 1997; Ingelman-
`Sundberg 2004). Norfluoxetine was shown to inhibit
`neuronal K+ channels (Choi et al. 1999, 2001), serotonin-
`mediated currents
`(Choi
`et al. 2003) and nicotinic
`acetylcholine receptors (Lopez-Valdés and Garcia-Colun-
`ga 2001), however, in contrast to the well-characterized
`cellular electrophysiological actions of fluoxetine, little is
`known about such effects of norfluoxetine in cardiac
`
`membranes. The goal of the present study was, therefore,
`to characterize the cellular electrophysiological actions of
`norfluoxetine in isolated canine ventricular cardiomyo-
`cytes, comparing these effects of norfluoxetine to our
`previous results obtained with fluoxetine. Since norfluox-
`etine caused much stronger suppression on cardiac ion
`channels than fluoxetine,
`it was concluded that
`the
`majority of the cardiac side effects observed during
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2052 - 0001
`
`

`
`204
`
`fluoxetine treatment may likely be attributed to its
`metabolite norfluoxetine.
`
`Materials and methods
`
`Cell isolation Single canine ventricular cells were obtained
`from hearts of adult mongrel dogs using the segment
`perfusion technique as described previously (Magyar et al.
`2000). Briefly, the animals (10-20 kg) were anesthetized
`with iv.
`injection of 10 mg/kg ketamine hydrochloride
`(Calypsolvet) plus
`1 mg/kg xylazine hydrochloride
`(Rometar). After opening the chest the heart was rapidly
`removed and the left anterior descending coronary artery
`was perfiJsed using a Langendortf apparatus. Ca2+'—fnee
`JMM solution (Minimum Essential Medium Eagle, Joklik
`modification; Sigma Chemicals, product no. —0518, St
`Louis, MO, USA), supplemented with taurine (2.5 g/l),
`pyruvic acid (175 mg/l), ribose (750 mg/1), allopurinol
`(13.5 mg/l) and NaH9_PO4 (200 mg/1), was used during the
`initial 5 min of perfusion to remove Ca2+ and blood from
`the tissue. After addition of NaHCO3 (1.3 g/l), the pH of
`this perfusate was 7.0 when gassed with carbogen (95%
`02 + 5% CO3). Cell dispersion was performed for 30 min
`in the same solution containing also collagenase (660 mg]
`l, Worthington Cls-1), bovine albumin (2 g/l) and CaCl;
`(50 14M). During the isolation procedure the solutions
`were gassed with carbogen and the temperature was
`maintained at 37°C. The cells were rod shaped and
`showed clear
`striation when the external Ca * was
`
`restored. Before use, the cells were stored overnight at
`14°C in modified IMM solution (pH 7.4).
`
`Action potential recording To record action potentials
`from the myocytes, the viable cells were sedimented in a
`plexiglass chamber allowing for continuous superfusion
`with modified Krebs solution (containing in mM: NaCl
`120, KC] 5.4, CaCl2 2.7, MgCl2 1.1, NaH2P04 1.1,
`NaHCO3, glucose 6) having pH adjusted to 7.4i0.05 when
`gassed with carbogen. Transmembrane potentials were
`recorded at 37°C using glass microelectrodes filled with
`3 M KC] and having tip resistance between 20 M52 and
`40 M0. These electrodes were connected to the input of an
`Axoclamp-2B amplifier (Axon Instruments, Union City,
`CA, USA). The cells were continuously paced through the
`recording electrode at steady cycle length of 1,000 ms
`using 1 ms wide rectangular current pulses with 120%
`threshold amplitude. Outputs from the clamp amplifier
`were digitized at 100 kHz using a Digidata 1200 MD card
`(Axon Instruments) and stored for later analysis, which
`was performed under the control of pClamp 6.0 software
`(Axon Instruments).
`
`Voltage clamp Transmembrane ion currents were recorded
`in oxygenated Tyrode solution (containing in mM: NaCl
`140, KC] 5.4, CaCl2 2.5, MgCl2 1.2, Na2HPO,, 0.35,
`HEPES 5, glucose 10, pH 7.4) at 37°C. Suction pipettes,
`fabricated from borosilicate glass, had tip resistance of
`2 Mn after filling with pipette solution (composed of in
`
`mM: KCl 110, KOH 40, HEPES 5, EGTA 10, TEACI 20,
`K-ATP 3 and GTP 0.25 mM, or alternatively, K-aspartate
`[00, KC] 45, MgCl-3 1, EGTA 10, HEPES 5, K-ATP 3,
`when measuring Ca "i or Ki currents, respectively). The
`pH of these pipette solutions was adjusted to 7.2 with
`KOH. Ic, was blocked by 5 pM nifedipine, and 3 mM 4-
`aminopyridine was used to suppress 1,, (both drugs were
`applied externally). Currents were recorded with an
`Axopatch-200B amplifier (Axon Instruments) using the
`whole cell configuration of the patch clamp technique
`(Hamill et al. 1981). After establishing high (1-10 GQ)
`resistance sea] by gentle suction,
`the cell membrane
`beneath the tip of the electrode was disrupted by fi.1rther
`suction or by applying 1.5 V electrical pulses for l—5 rns.
`After this step, the intracellular solution was allowed to
`equilibrate with the pipette solution for a period of 5—
`10 min before starting the measurement. Ionic currents
`were normalized to cell capacitance, determined in each
`cell using short (25 ms) hyperpolarizing pulses from 0 mV
`to ----10 mV. The series resistance was typically 4-8 MS!
`before compensation (usually 5(F80%). Experiments were
`discarded when the
`series
`resistance was high or
`substantially increasing during the measurement. The
`applied experimental protocols are described in the Results
`section where appropriate.
`
`Drug application Norfluoxetine (Sigma) was dissolved in
`distilled water and was added to the bath in a cumulative
`
`manner applying each concentration for 2 min. This period
`of time was sufficient to achieve steady-state effects in
`both action potential and ion current measurements.
`
`Statistics All values presented are arithmetic means :1:
`SEM. Statistical
`significance was determined using
`Student’s t-test. Differences were considered significant
`when the P value was less than 0.05.
`
`The entire investigation conforms the Guide for the
`Care and Use ofLaboratory Animals published by the US
`National Institutes of Health (NIH publication no. 85-23,
`revised 1996) and the principles outlined in the Declara-
`tion of Helsinki.
`
`Results
`
`Effect of norfluoxetine on action potential
`configuration
`
`Cumulative concentration-dependent effects of norfluox-
`etine on the configuration and parameters of the action
`potential are shown in Fig.
`1 and Table l. Norfluoxetine
`(1-10 nM) markedly accelerated repolarization,
`i.e.,
`shortened the duration of action potentials measured at
`both 50% and 90% levels of repolarization. The drug also
`attenuated early (phase—l) repolarization, and depressed
`the level of the late plateau (Fig. 1A, B). In addition,
`norfluoxetine decreased action potential amplitude and the
`maximum rate of depolarization (Vnlflx) in a concentration-
`dependent manner, having a 2.42:I:0.36 pM EC;-.0 value for
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2052 - 0002
`
`

`
`205
`
`Current—voltage relations for 13,, were obtained by
`applying a series of test pulses increasing up to +40 mV in
`5-mV steps in the absence and presence of norfluoxetine,
`and peak values of 1.3,. were plotted against their respective
`test potentials. No shift in the current—voltage relationship
`was observed in the seven myocytes after application of
`3 p.M norfluoxetine (Fig. 2D). Ca2+ conductance (Gen)
`was calculated at each membrane potential by dividing the
`peak current by its driving force (the difference between
`the applied test potential and the reversal potential for IQ“,
`estimated to be +55 mV). Ca“ conductance was
`significantly reduced by 3 uM norfluoxetine at each
`membrane potential studied, however, when Gr-3, values
`were nonnalized to the respective Gc, obtained at +30 mV,
`the G(;,,—V,,., relationships were firlly identical (Fig. 2B).
`These results indicate that voltage dependence of activa-
`tion of L3, is not affected by 3 i.LM norfluoxetine.
`In contrast to the unchanged voltage dependence of
`activation,
`the voltage dependence of inactivation was
`altered by norfluoxetine in a reversible manner. In order to
`study the voltage dependence of steady-state inactivation
`of lg-,1, test depolarizations to +5 mV were preceded by a
`set of prepulses clamped to various voltages between
`“"55 mV and +5 mV for 500 ms. Peak currents measured
`
`after these prepulses were normalized to the peak current
`measured after the -55 mV prepulse and plotted against
`the respective prepulse potential. The data were fitted to
`the two—state Boltzmann function (Fig. 2F). Superfusion of
`the cells with 3 [ti.M norfluoxetine shifted the midpoint
`potential by almost 7 mV to the left (fiom the control
`value of -20.9i0.8 mV to -27.7:t1.4 mV, P<0.05, n=7),
`however, no significant difference was observed between
`the respective slope factors (3.9:|:0.3 mV and 4.4:|:0.3 mV,
`NS, n=7).
`
`1C). The effects of the same
`the Vmw,-block (Fig.
`fluoxetine concentrations on Vmx are also depicted in
`this figure for comparison (data taken from Pacher et al.
`2000). As shown in Fig. 1C, the Vmafdepressant effect of
`norfluoxetine was much stronger than that of fluoxetine.
`These
`effects of norfluoxetine on action potential
`characteristics developed rapidly and were firlly reversible.
`Norfluoxetine had no effect on the resting potential at the
`concentrations studied.
`
`Effect of norfluoxetine on the calcium current
`
`a rate of 0.2 Hz using
`Peak IL--,, was measured at
`depolarizing voltage pulses of 400 ms duration clamped
`from the holding potential of -40 mV to the test potential
`of +5 mV. K"" currents were blocked by the externally
`applied 4-aminopyridine and internally applied TEACI.
`Stability of ICE was monitored at least for 5 min before
`cumulative application of norfluoxetine (from 0.1 }.l.M to
`10 uM, each concentration for 2 min). Norfluoxetine
`caused marked suppression of peak IQ, without changing
`the time course of inactivation (Fig. 2A). The decay of IC,
`was fitted as a sum of two exponential components at
`+5 mV, having time constants of 14.211 ms and 831:8 ms
`in control, and 13.7:|:0.9 ms and 7719 ms, respectively, in
`the presence of 3 |.LM norfluoxetine (not significant [NS],
`n=7). The effect of norfluoxetine on Ica developed rapidly
`(within 2 min) and was fully reversible upon washout
`(Fig. 2B). The norfluoxetine-induced block of [(1, was
`concentration—dependent
`(Fig. 2C).
`Inhibition of the
`current was statistically significant from the concentration
`of 0.3 i.LM (inhibition of 20.3d:2.5%, P<0.05, n=7) and
`above. Fitting results to the Hill equation yielded an Ecfifi
`of l.l3:|:0.08 },LM and a Hill coefficient of l.l9:t0.l2. For
`comparison, concentration—dependent effects of 1-100 1.1M
`fluoxetine on 1.3, is also included in this panel (data taken
`from Pacher et al. 2000). The 5.4:|:0.94 LLM EC5{) value
`obtained for fluoxetine indicates that cardiac [Ca is more
`sensitive to norfluoxetine than to fluoxetine.
`
`
`
`
`
`Membranepotential(mV)).
`
`01G
`
`O
`
`_5D
`
`100
`
`o
`
`100
`
`200
`
`300
`
`,_oaD
`
`-23
`-40
`
`0
`
`
`
`Membranepotentiat(mv)
`o8
`
`>~
`
`/ 3 §.I.M
`
`Control W35i‘°U‘
`
`100
`:3‘
`5;,
`g
`— 50
`0 \
`*8‘
`E
`
`_
`
`A Fluuxetlne
`
`ecu = 2.421035 p,M
`n =1.4-4:t.'G.31
`
`. vi. '
`
`
`
`
`10 "M
`
`25
`
`50
`
`0
`
`1-5,""" T
`C 0.1
`
`1
`
`to
`
`Time (ms)
`
`Time (ms)
`
`Concentration (|.tM)
`
`Fig. 1A—C Concentration-dependent effect of norfluoxetine on
`action potential configuration in single canine ventricular myocytes.
`A Superimposed action potentials recorded at 1 Hz before and afier
`Superfusion with 1, 3, and 10 uM norfluoxetine, each for 2 min.
`Finally, the cell was washed in drug-free solution for a further 3 min.
`B The same action potentials are shown on an extended scale to
`demonstrate the drug-induced changes in the notch. C Suppressive
`
`effect of norfluoxetine on the maximum velocity of depolarization,
`V,,,,,, (n=l2). Data were fitted to the Hill equation (solid line).
`Symbols and bars represent mean i SEM. Our earlier results
`obtained with similar concentrations of fluoxetine in five myocytes
`(dotted line) are also included for comparison (data taken from
`Pacher et al. 2000)
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2052 - 0003
`
`

`
`206
`
`Table 1 Cumulative concentration-dependent effects of norf1uox-
`etine on action potential parameters in canine ventricular myocytes
`(n=l2). RP resting membrane potential, APA action potential
`
`amplitude, Vmax maximal velocity of depolarization, APD5,.; and
`APD9n action potential duration measured at 50% or 90% level of
`repolarization respectively
`
`RP (mV)
`
`APA (mV)
`
`V.,,,,, (V/s)
`
`APD5o (ms)
`
`APDgo (ms)
`
`Control
`
`tavr
`1
`3 am
`10 pM
`Washout
`
`—8l.8:|:1.0
`
`—s2.3¢o_7
`—s2.2¢o_9
`—s2.2i1_3
`—s2.4i1.9
`
`ll2.5:|:l.5
`
`111.s¢1_9
`l02.9:I:2.3***
`69.4i4.2""'“"
`110.6=_~1.s
`
`20li:26
`140:|:l8*
`115¢17**
`12r4m
`189¢26
`
`149x12
`l19:|:ll*'”'
`oo¢7***
`76i4*"‘*
`l46:I:l2
`
`221:|:l2
`196:tl3**
`1s7¢s*==*
`124+5-M
`225:l:l4
`
`Mean :t SEM values are given
`Asterisks denote significant changes from control (*P<0.D5, **P<0.D1, ***.P<D.00l) determined using Student’s r-test for paired data
`
`Effect of norfluoxetine on the transient outward
`current
`
`The transient outward current, I,,, was studied at +50 mV
`using voltage pulses of 200 ms duration arising from the
`holding potential of ---80 mV. Each of these test pulses was
`preceded by a 5 ms long prepulse to -40 mV in order to
`inactivate the Na current.
`
`Similarly to 10,, 1,, was also depressed by norfluoxetine
`without visible changes in current kinetics (Fig. 3A). The
`
`decaying branch of It“ was fitted as a sum of two
`exponentials, yielding time constants of 1.3d:0.l ms and
`6.4i0.4 ms in control, and l;t0.8 ms and 6.6:I:l.2 ms,
`respectively, in the presence of 3 pM norfluoxetine (NS,
`n=5),
`indicating that
`time course of inactivation was
`unaltered in the presence of norfluoxetine. The suppres-
`sive effect of norfluoxetine on I“, completely developed
`within 2 min and was fully reversible (Fig. 3B). [,0 was
`decreased by norfluoxetine in a concentration-dependent
`manner (Fig. 3C). This effect developed at relatively low
`
`
`
`d
`T
`n
`I NS;“Swine
`A Fluoxetine
`
`Time (min)
`
`FC_w=' l.i3 £0.08 uM
`st‘ i.l‘)z0.i2
`
`
`
`-.1.
`
`.
`
`
`
`5" ' '
`C
`0.1
`
`I
`
`as \- r.c,,—s.4n~o.94 an
`
`n'—"l.l10.l4
`1
`'10
`
`100
`
`Concentration (aM)
`
`F
`
`E 1.0
`g
`8
`E 9-5
`‘T
`E
`z
`
`D
`
`Membrane potential (mV)
`-20
`o
`20
`40
`
`-40
`
`E
`
`‘~‘’
`
`-I
`
`I
`
`.1’.
`
`II
`
`5
`G
`‘J
`E O5
`E
`g
`2
`
`M
`
`0
`-2
`
`,..
`”5_
`2
`3 -4
`.3
`-6
`
`-3
`
`-r:r:i:i:r
`-30
`-15
`0
`15
`30
`Membrane potential (mV)
`
`0.0
`
`:—'—r—I—r—-"'r"*"'1
`-60
`-40
`-20
`0
`20
`Prepuise potential (mV)
`
`Fig. 2A—F Effect of norfluoxetine on the calcium current studied in
`seven myocytes. A Superimposed IQ, records, obtained before,
`during, and after 2 min superfusion with 3 uM norfiuoxetine. The
`dashed line denotes zero current. B Representative experiment
`showing the time scale of development and reversion of the
`norfluoxetine-induced changes in 15,. C Cumulative concentration-
`dependent effect of norfluoxetine on peak I(;,,, measured at +5 mV.
`Solid line was obtained by fitting data to the Hill equation. Our
`earlier results obtained with 1-100 HM fiuoxetine in six myocytes
`(dotted line) are also included for comparison (data taken from
`Pacher et a]. 2000). D Current—voltage relationship of peak 15,, in the
`
`absence and presence of 3 }.tM norfluoxetine. E Voltage-dependent
`activation of Ca2+ conductance (Gca) in control and after application
`of 3 MM norfiuoxetine as calculated from the previously shown
`current—voltage curves. At each membrane potential G5,, was
`normalized to that obtained at +30 mV, and the results were fitted to
`a two-state Boltzmann model (.s'olz'd curves). F Voltage dependence
`of steady-state inactivation of [Cu determined using paired-pulse
`protocol in the presence and absence of 3 uM norfluoxetine. Solid
`curves were obtained by fitting data to a two-state Boltzmann
`model. Estimated midpoint potentials and slope factors are given in
`the text. Symbols and bars are mean i SEM values
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2052 - 0004
`
`

`
`207
`
`concentrations (suppression of 28.5i3.9% was observed in
`the presence of 0.3 |.LM nortluoxetine, P<0.05, n=5). The
`Hill equation, used to describe the concentration-depen-
`dency of the norfluoxetine effect, yielded an EC5-0 of 1.19
`;l:0. 17 uM, a value very close to that obtained for Iga (1.13
`$0.08 pM). The Hill coefficient, again, was close to unity
`(0.82:|:0.06). For comparison, 3 uM fluoxetine had no
`significant effect on 1,, (Pacher et al. 2000).
`Current—voltage relations for I“, were obtained in the
`absence and presence of norfluoxetine by applying a series
`of test pulses increasing from -40 mV to +65 mV in 5 mV-
`steps, and peak values of Ito were plotted as a fimction of
`the test potential (Fig. 3D). When the conductance of the
`In, channels (Gm) was calculated in a way similar to that
`applied for I.;, the voltage dependence of activation of I“,
`was generated (Fig. 3E). Although the norfluoxetine—
`induced block of It, showed little voltage dependence, the
`drug caused a small, but statistically significant leftward
`shift on the steady—state activation curve. The midpoint
`potentials, obtained by fitting the data to a two-state
`Boltzmann function, were 10.9d:0.7 mV and 2.9d:3.5 mV in
`the absence
`and presence of
`3
`;.LM norfluoxetine,
`respectively (P<0.05, n=6), while the estimated slope
`factors were practically identical (l4.8:t1.8 mV and 12.7
`$0.7 mV, NS, n=6). Indicating the reversible nature of the
`norfluoxetine effect on activation, the midpoint potential
`obtained after washing out norfluoxetine was 8.7:l:3.5 mV,
`statistically not different
`from the control value (not
`shown).
`
`In contrast to results obtained for L3,, norfluoxetine
`failed to alter the voltage dependence of inactivation of [,0
`(Fig. 3F). The estimated midpoint potentials
`(---35.7
`;t0.7 mV vs.
`---36.1il.9 mV) and slope factors (3.3
`i0.4 mV vs. 4:l:0.6 mV) determined in the absence and
`presence of 3 aM norfluoxetine, respectively, were not
`different statistically (NS, n=6). In these experiments the
`400 ms prepulses were varied from ---60 mV to +10 mV in
`[0-mV steps, while the test potential was set to +50 mV.
`
`Effect of norfluoxetine on the inward rectifier Ki
`current
`
`The steady—state current—vo1tage relationship of the mem-
`brane of canine ventricular cells was determined using
`400 ms long voltage commands clamped to potentials
`ranging from ---135 mV to +45 mV increasing in l0-mV
`steps. Currents measured at the end of these pulses were
`plotted against the respective test potentials. As shown in
`Fig. 4, norfluoxetine (3 uM) failed to modify the steady-
`state current—voltage relationship in canine cardiomyo-
`cytes. The negative branch of the I—V curve depends on
`the amplitude of IKE. The current densities measured at
`---125 mV were ---35.9:l:2.3 pA/pF and ---32.'7:|:2.7 pA/pF,
`respectively, before and after application of 3 HM
`norfluoxetine (NS, n=5) indicating that norfluoxetine—at
`least at this concentration—had no suppressive effect on
`IK1.
`
`A
`
`Control
`
`"—W35h°“1
`
`
`
`B
`
`2
`
`_E__
`-..r
`_3
`
`I
`I
`-
`
`in
`
`1
`
`Nerflucxetine
`3 uM
`0 mm o
`
`C
`100
`
`A
`g
`___§
`-5 50
`_§
`E
`
`ECW-1.!9:D.17uM
`n = D.82 an 06
`
`
`
`r-
`
`T,-mi (min)
`cl Tyrode
`I Nurfluoxatina
`
`10
`
`C
`
`10
`1
`0.1
`Concentration (uM)
`
`100
`
`F
`_ 1.0
`
`E
`3
`E 05
`
`DD
`
`l*'*|?'*l*'?I
`-40
`0
`40
`so
`Membrane potential (mV)
`
`I—-—I—-—I—-—I—'—I
`-so
`-40
`-20
`o
`20
`Prepulse potential (mV)
`
`E
`
`1.0
`
`:9’
`'3
`.N
`E 0 5
`Z
`0.0
`
`
`
`3 ulvl Narllunxallne
`
`——————————————————————————————————-
`
`5m!4
`
`so
`40
`0
`-40
`Membrane potential (mV)
`
`Fig. 3A—F Effect of norfluoxetine on the transient outward current.
`A Superimposed In, records, obtained before, during, and after 2 min
`superfusion with 3 uM norfluoxetine. B Representative experiment
`showing the time scale of development and reversion of the
`norfluoxetine-induced changes
`in the current. C Cumulative
`concentration-dependent effects of norfluoxetine on peak I,” mea-
`sured at +50 mV. The solid line was obtained by fitting data to the
`Hill equation (n=5). D Current—voltage relationship obtained in six
`cells for I“, in the absence and presence of 3 uM norfluoxetine. E
`Voltage-dependent activation of the 1.0 channels, defined as
`
`conductance (Gm), calculated from the current—voltage curves in
`control and in the presence of 3 uM norfluoxetine. At each
`membrane potential Gm was normalized to its value obtained at
`+65 mV, and the results were fitted to a two-state Boltzmann model
`(solid curves). Estimated midpoint potentials and slope factors are
`given in the text. F Voltage dependence of steady—state inactivation
`of I“, determined using paired-pulse protocol in the presence and
`absence of 3 LLM norfluoxetine (n=6). Solid curves were obtained by
`fitting data to a two-state Boltzmann model. Symbols‘ and bars
`represent mean 3: SEM values
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2052 - 0005
`
`

`
`208
`
`A
`Tyrode
`
`MPNFF 100m:
`NorlIun:I91ina(3pM)
`
`B
`
`Membrane potential (mV)
`so
`-100
`-so
`0
`-150
`20
`
`nlnlil‘
`
`a
`
`—l:l—Tymda
`—I— Norfluuxstlne (3 pm
`
`0
`
`E
`:1.
`-...
`E‘
`E’
`2 20
`3 '
`dc:
`E
`5 -40
`Eas
`E
`
`-60
`
`Fig. 4A, B Effect of norfluoxetine on the inward rectifier Kl’
`current. A Current families, obtained in Tyrode solution and in the
`present of 3 pM norfluoxetine, were elicited with test pulses of
`400 ms duration clamped to voltages ranging from -135 mV to
`*5 mV and increasing in 10-mV steps. B Steady-state current-
`voltage relations obtained in five cells in the absence and presence
`of 3 |.1I\/I norfluoxetine. In these experiments the range of test pulses
`was extended to +45 mV. The current measured at the end of each
`test pulse was plotted as a function of the respective test potential.
`Symbols and bars denote mean :l: SEM
`
`Discussion
`
`Effects of norfluoxetine on cardiac ion currents
`
`Micromolar concentrations of norfluoxetine evoked mul-
`
`tiple effects on action potential configuration in canine
`ventricular cells: the drug decreased the maximum rate of
`rise and amplitude of the action potential, attenuated early
`repolarization, depressed the plateau, and accelerated
`terminal
`repolarization, while the resting membrane
`potential was unaltered. These actions can fully be
`explained by the norfluoxetine-induced changes in ion
`currents, since suppression of IC, (leading to plateau
`depression and action potential
`shortening) and 1',“
`(resulting in reduction of early repolarization) was dem-
`onstrated, while IE1 (responsible for the highly negative
`resting potential) was not affected by norfluoxetine.
`Although IN, was not measured directly in this study,
`reduction of V,,,,,,, is generally accepted as a measure of
`Na"' channel blockade (Hondeghem 1978).
`Interestingly, the EC50 values estimated for [Ca and 1,0
`were practically equal (1.1 p.M and 1.2 uM, respectively),
`furthermore, EC50 of the Vmax-block (2.4 |.LM) was also
`close to these values. Taking into account that the Hill
`coefficients obtained for I,_-,,, 1,0, and the Vm,-block (1.2,
`0.8, and 1.4, respectively) were all close to unity, it seems
`possible that norfluoxetine binds to a single binding site
`which may be a common structure of many 6-TM
`channels, but is absent from the members of the Kir
`superfamily, mediating the 11,-, current. The minor differ-
`ences observed between effects of norfluoxetine on the
`
`kinetic properties of Ic, and I“, (negative shifls in voltage
`dependence of inactivation and activation, respectively)
`may probably be related to different amino acid environ-
`ments of the binding site. Similar differences can be seen
`when comparing the charmel-blocking effect of norfluox-
`etine in various preparations. For instance, the suppressive
`
`effect of the drug was clearly voltage dependent on the
`cloned neuronal potassium channel Kv 3.1 (Choi et al.
`2001), whereas little voltage dependence was observed in
`our cardiac Ito current.
`
`Comparison with effects of fluoxetine
`
`In our previous Work, performed with fluoxetine in various
`mammalian cardiac tissues including canine myocytes,
`fluoxetine was shown to evoke changes in action potential
`morphology similar to those described in the present study
`with norfluoxetine (Pacher et al. 2000). The fluoxetine-
`induced depression of plateau and shortening of action
`potentials in canine ventricular myocytes were attributed
`to inhibition of Ica. Present results indicate that [C3 is more
`sensitive to norfluoxetine than fluoxetine,
`since the
`1.1 pM EC5o value obtained for norfluoxetine was five
`times
`lower
`than the EC,-,0 of 5.4 uM found with
`fluoxetine by Pacher et al.
`(2000). Similar conclusion
`can be drawn regarding the suppression of IN,,, although
`the EC50 value for the fluoxetine-induced V,,,,,,,-block was
`not determined. However, according to Fig. 1C, the EC50
`must be well above 10 uM in the case of fluoxetine, a
`value being again at
`least five times higher than the
`respective EC50 obtained with nortluoxetine. Surprisingly,
`I“, was not inhibited by 10 14M fluoxetine (Pacher et al.
`2000), in contrast with the present results obtained with
`norfluoxetine (EC5.;=1.2 uM). The exact reason for this
`difference remains to be elucidated, however, it can be
`speculated that 11,, channel protein can distinguish between
`the two structures. The most important conclusion of this
`study—based on the comparison aboveeis that not only
`fluoxetine, but also its active metabolite norfluoxetine,
`may exhibit cardiovascular depressant effects in clinically
`relevant concentrations. Since the inhibitory effects of
`norfluoxetine on 15,, IN, and I,,, are much stronger than
`those of fluoxetine, the majority of cardiac side effects,
`being attributed previously to fluoxetine, may likely be
`ascribed to the presence of norfluoxetine. More detailed
`evaluation in a functional assay using isolated hearts might
`be helpful to give further support in this respect.
`
`Clinical implications
`
`The norfluoxetine-induced shortening of action potential
`duration is potentially proarrhythmic due to reduction of
`the effective refractory period, and the concomitant
`facilitation of development of re-entry type arrhythrnias.
`In addition, inhibition of [Ca may lengthen atrioventricular
`conduction,
`resulting in atrioventricular block, while
`depression of V,,,,,,, may result in compromised intraven-
`tricular conduction due to inhibition of IN,. At the same
`time, these actions of norfluoxetine (i.e., suppression of
`IN, and It-,,) may also be considered antiarrhythmic (class I
`+ [V type, respectively). Furthermore, the reduction of the
`Ca“ window current, due to the leftward shift of the
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2052 - 0006
`
`

`
`steady-state inactivation curve (shown in Fig. 2F) may
`decrease the incidence of early after depolarizations.
`Clinically, serotonin-reuptake inhibitors, including flu-
`oxetine and its metabolite norfluoxetine, are believed to
`cause less cardiovascular
`side
`effects
`than tricyclic
`antidepressants. However, there is an increasing number
`of case reports on dysrhythmias, like atrial fibrillation or
`bradycardia (Buffet al. 1991; Friedman 1991; Masquelier
`et al. 1993; Drake and Gordon 1994; Hussein and
`Kaufman 1994; Roberge and Martin 1994; Graudins et
`al. 1997; Anderson and Compton 1997) and syncope
`(Ellison et al. 1990; McAnally et al. 1992; Cherin et al.
`1997; Livshits and Danenberg 1997; Rich et al. 1998)
`associated with fluoxetine treatment and overdose. The
`
`upper range of therapeutic plasma concentrations of
`fluoxetine was reported to vary between 0.15 g.LM and
`1.5 11M in humans. In addition, similar concentrations of
`its active metabolite, norfluoxetine is also present in the
`plasma of fluoxetine—treated patients (Orsulak et al. 1988;
`Kelly et al. 1989; Keck and McElroy 1992; Januzzi et a1.
`2002). Under extreme conditions (e.g., decreased metab-
`olism in the elderly, acute overdose or drug interactions),
`these plasma concentrations of fluoxetine and norfluox-
`etine can reach higher levels (Pato et al. 1991; Borys et al.
`1992; Hale 1993; Eap et al. 2001). Furthermore, recent
`data indicate that fluoxetine (and probably norfluoxetine
`as well) can be accumulated in the tissues: 20 times
`accumulation of fluoxetine has been detected in human
`
`brain during chronic fluoxetine treatment (Karson et al.
`1993; Komorski et al. 1994). Considering the rnicromolar
`EC50 values obtained with norfluoxetine for IC,, and the
`I/',,,,,,,, depressed atrioventricular and intraventricular con-
`duction can well be anticipated in patients treated with
`fluoxetine.
`It must be emphasized, however,
`that
`the
`norfluoxetine-induced electrophysiological alterations are
`not necessarily always harmful. They are, of course, in
`patients having deficient
`impulse conduction, but they
`may be beneficial
`in cases with long QT syndrome.
`Therefore,
`in depressed patients having also cardiac
`disorders, ECG control is strongly recommended during
`the fluoxetine therapy.
`
`Acknowledgements Financial support for the studies was obtained
`from grants from the Hungarian Research Found (OTKA-T037332,
`OTKA-T037334, OTKA-T043182) and the Hungarian Ministry of
`Health (ETT-06031/2003, ETT-572/2003).
`
`References
`
`Anderson J, Compton SA (1997) Fluoxetine induced bradycardia in
`presenile dementia. Ulster Med J 661144-145
`Borys DJ, Setzer SC, Ling LJ, Reisdorf JJ, Day LC, Krenzelok EP
`( 1992) Acute fluoxetine overdose: a report of 234 cases. Am J
`Emerg Med 101115-120
`Buff DD, Brenner R, Kirtane SS, Gilboa R (1991) Dysrhythmia
`associated with fluoxetine treatment in elderly patients with
`cardiac disease. J Clin Psychiatry 52:l7tL176
`Cherin P, Colvez A, Deville DE, Periere G, Sereni D (1997) Risk of
`syncope in the elderly and consumption of drugs: a case-control
`study. J Clin Epidemiol 501313-320
`
`209
`
`Choi JS, Hahn SJ, Rhie DJ, Yoon SH, Jo YH, Kim MS (1999)
`Mecha