key: cord-0742856-2rzt0400
authors: Chang, Wei-Ting; Ragazzi, Eugenio; Liu, Ping-Yen; Wu, Sheng-Nan
title: Effective block by pirfenidone, an antifibrotic pyridone compound (5-methyl-1-phenylpyridin-2[H-1]-one), on hyperpolarization-activated cation current: An additional but distinctive target
date: 2020-06-07
journal: Eur J Pharmacol
DOI: 10.1016/j.ejphar.2020.173237
sha: 2de55052d14aa603b291e99d82ac897112eea8c9
doc_id: 742856
cord_uid: 2rzt0400

Pirfenidone (PFD), a pyridone compound, is well recognized as an antifibrotic agent tailored for the treatment of idiopathic pulmonary fibrosis. Recently, through its anti-inflammatory and anti-oxidant effects, PFD based clinical trial has also been launched for the treatment of coronavirus disease (COVID-19). To what extent this drug can perturb membrane ion currents remains largely unknown. Herein, the exposure to PFD was observed to depress the amplitude of hyperpolarization-activated cation current (I(h)) in combination with a considerable slowing in the activation time of the current in pituitary GH(3) cells. In the continued presence of ivabradine or zatebradine, subsequent application of PFD decreased I(h) amplitude further. The presence of PFD resulted in a leftward shift in I(h) activation curve without changes in the gating charge. The addition of this compound also led to a reduction in area of voltage-dependent hysteresis evoked by long-lasting inverted triangular (downsloping and upsloping) ramp pulse. Neither the amplitude of M-type nor erg-mediated K(+) current was altered by its presence. In whole-cell potential recordings, addition of PFD reduced the firing frequency, and this effect was accompanied by the depression in the amplitude of sag voltage elicited by hyperpolarizing current stimulus. Overall, this study highlights evidence that PFD is capable of perturbing specific ionic currents, revealing a potential additional impact on functional activities of different excitable cells.

to a reduction in area of voltage-dependent hysteresis evoked by long-lasting 1 inverted triangular (downsloping and upsloping) ramp pulse. Neither the amplitude 2 of M-type nor erg-mediated K + current was altered by its presence. In whole-cell 3 potential recordings, addition of PFD reduced the firing frequency, and this effect 4 was accompanied by the depression in the amplitude of sag voltage elicited by 5 hyperpolarizing current stimulus. Overall, this study highlights evidence that PFD is 6 value of activation time constant (τ act ) of the current, in response to sustained 1 hyperpolarization, was evidently raised during exposure to the compound (Fig. 1A  2 and 1B). For example, the application of 3 µM PFD to the bath was effective at 3 increasing the τ act value to 834±54 ms (n=9, P<0.05) from a control value of 576±38 4 ms (n=9). 5

The depressant effect of PFD with different concentration on I h elicited by 6 2-s-long sustained hyperpolarization was further examined. The 7 concentration-dependent relationship of the inhibitory effect of this compound on I h 8 amplitude in GH 3 cells is illustrated in Fig. 1C . According to the Hill equation 9 indicated under Materials and Method, the IC 50 value of PFD which was needed for 10 the inhibition of I h was thereafter calculated to be 3.65 µM with the Hill coefficient of 11 1.3. 12

zatebradine, zatebradine plus PFD on I h amplitude 14

Next, whether the subsequent addition of PFD, but still in the continued 15 presence of ivabradine or zatebradine, was able to perturb the block by ivabradine or 16 zatebradine of I h inherently in GH 3 cells was investigated. Ivabradine or zatebradine 17 has been reported to decrease HCN-encoded currents effectively (Chang et al., 2019; (1 µM) or zatebradine (1 µM), was capable of producing an additional decline in I h 2 amplitude elicited by 2-s-long hyperpolarizing step. The experimental results 3 obtained demonstrated that either ivabradine or zatebradine could act in concert 4 with PFD to decrease I h amplitude, and could thus be interpreted to indicate that 5 both inhibitory effects of ivabradine or zatebradine and that of PFD on I h amplitude 6 appear to be additive in these cells. 7

The effect of PFD on I h at various levels of membrane potentials was next 9 investigated. The I h was robustly evoked as the cell was hyperpolarized from -40 10 mV to a series of voltage steps ranging between -120 and -80 mV ( Fig. 3A and 3B) . 11

The mean I-V relationships of I h with or without PFD were established and hence 12 depicted in Fig. 3B . With whole-cell conductance of I h measured at the potentials 13 ranging between -120 and -100 mV, the addition of 3 µM PFD was able to diminish 14 the I h conductance from 4.62±1.17 to 2.68±0.09 nS (n=8, P<0.05). 15

To further investigate the inhibitory effect of PFD on I h in GH 3 cells, the possible 17 changes induced by this compound on the steady-state activation curve of I h were 18 studied. In these voltage-clamp experiments, a two-step voltage pulse was applied 19 in situations where a 2-s conditioning pulse to various membrane potentials were 1 delivered to precede the test pulse (2 s in duration) to -120 mV from -40 mV holding 2 potential. Fig. 3C illustrates the relationship between the conditioning potentials 3 versus the normalized amplitudes (I/I max ) of I h obtained with or without PFD (3 µM). 4

The data satisfactorily fitted to a Boltzmann function, as defined in Materials and 5

Methods. In control, V 1/2 = -85±3 mV, q = 4.7±0.4 e (i.e., elementary charges) (n=8), 6 whereas in the presence of 3 µM PFD, V 1/2 = -96±4 mV, q = 4.8±0.4 e (n=8). The 7 data suggest that the exposure to PFD not only decreases the maximal conductance 8 of I h , but overly shifts the activation curve along with the voltage axis to the negative 9 potential by roughly 11 mV; however, there was devoid of changes in the gating 10 charge of the curve in its presence. 11

3.5 Effect of PFD on the voltage hysteresis elicited responding to triangular ramp 12 pulse 13

The voltage hysteresis of I h has been shown with an impact on electrical 14 behaviors of AP firing (Chang et al., 2020; Männikkö et al., 2005) . Therefore, it was 15 explored whether a possible voltage hysteresis exists in I h measured from GH 3 cells 16 and how the presence of PFD modifies such hysteresis. In this experiment, a 17 long-lasting inverted triangular ramp pulse with a duration of 2 s (i.e., ±0.1 V/s) was 18 delivered in the whole-cell configuration. Of interest, as can be seen in Fig. 4A , the trajectories of I h elicited by the downsloping (i.e., hyperpolarized from -50 to -150 1 mV) and upsloping (i.e., depolarizing from -150 to -50 mV) ramp pulse as a function 2 of time were overly distinguishable between these two limbs. The downsloping 3 (forward) limb elicited current amplitude of the inverted triangular voltage ramp was 4 lower than that by the upsloping (backward) limb, strongly indicating that there was 5 a voltage hysteresis for this current in these cells. As the ramp speed was reduced, 6 the hysteretic magnitude for I h became progressively increased. The degree of 7 voltage hysteresis was quantified according to the difference in area under the curve 8 in the forward (downsloping) and reverse (upsloping) direction, as denoted by the 9 arrows in Fig. 4A . The data indicate that for I h in GH 3 cells, the degree of voltage 10 hysteresis increases with slower ramp speed, and that cell exposure to PFD leads to a 11 substantial reduction in the amount of such hysteresis. 12

In the experiments on PFD or PFD plus oxaliplatin on voltage hysteresis of I h , a 13 long-lasting inverted triangular ramp pulse with a duration of 2 s (i.e., ±0.1 V/s) was 14 undertaken in the whole-cell configuration. Oxaliplatin, a chemotherapeutic antineoplastic agent, has been previously reported 2 to activate I h (Liu et al., 2018; Chang et al., 2020) . 3 3.6 Failure of PFD to affect M-type K + current (I K(M) ) in GH 3 cells 4

It was additionally evaluated whether the presence of PFD produces any effects on 5 different types of K + currents (e.g., I K(M) ) present in these cells. To measure I K(M) , 6 GH 3 cells were bathed in high-K + , Ca 2+ -free solution with the recording pipette filled 7 with K + -containing solution. The examined cell was voltage-clamped at -50 mV and 8 sustained depolarization to -10 mV was applied (Selyanko et al., 1999; So et al., 2019) . 9

As illustrated in Fig. 5A and 5B, PFD at a concentration of 10 µM did not alone 10 produce any effect on I K(M) responding to maintained depolarization applied from -50 11 to -10 mV. However, as 10 µM PFD was continually present, subsequent application 12 of linopirdine (3 µM) noticeably decreased I h amplitude effectively, notwithstanding 13 the ability of flupirtine (10 µM) alone to enhance current amplitude (Fig. 5B) . 14 Linopirdine can inhibit I K(M) , while flupirtine has been reported to enhance I K(M) (Hsu 15 et al., 2014; Pattnaik and Hughes, 2012) . 16 3.7 Inability of PFD to modify erg-mediated K + current (I K(erg) ) in GH 3 cells 17

In another set of experiments, whether another type of K + current, namely I K(erg) , 18 would be subject to perturbations by PFD was also further examined. With cells 19 bathed in high-K + , Ca 2+ -free solution and the pipette were filled with K + -containing 1 solution (Huang et al., 2011; So et al., 2019) . As shown in Fig. 6A and 6B, cell 2 exposure to 10 µM PFD did not perturb the amplitude of deactivating I K(erg) elicited 3 by membrane hyperpolarization as reported previously (So et al., 2019; Hsu et al., 4 2020) . For example, as the cells were 1-s hyperpolarized to -90 mV from a holding 5 potential of -10 mV, the peak amplitude of I K(erg) did not differ significantly between 6 the absence and presence of 10 µM PFD (742±43 pA [control] versus 739±45 pA [in 7 the presence of 10 µM PFD]; n=9, P>0.05). Moreover, further application of 8

BeKM-1 (1 µM), but still in the continued presence of PFD, was effective at 9 decreasing I K(erg) , as demonstrated by the significant reduction of I K(erg) amplitude to 10 328±23 pA (n=9, P<0.05). BeKm-1 is a scorpion toxin reported to suppress I K(erg) 11 effectively in heart cells and to prolong QT interval in isolated rabbit heart (Korolkova 12 et al., 2001; Qu et al., 2011) . It is conceivable, therefore, that I K(M) or I K(erg) al., 2001) . To achieve these objectives, in a separate set of whole-cell voltage-clamp 2 experiments, cells were bathed in normal Tyrode's solution, which consists of 1.8 3 mM CaCl 2 , 10 mM tetraethylammonium chloride, and 1 µM tetrodotoxin, and the 4 recording pipette was filled with a Cs + -containing solution. As can be seen in Fig. 7A , 5 PFD (3 or 10 µM) decreased the peak amplitude of I Ca,L elicited by the 300-ms 6 depolarizing pulse from -40 to 0 mV. However, no obvious changes were observed 7 in the activation or inactivation time course of I Ca,L elicited responding to the abrupt 8 step depolarization. Furthermore, as the clamp pulses of 300-ms duration from -40 9 mV to various membrane potentials were applied, the overall I-V relationship of peak 10 I Ca,L seen with these cells virtually was not perturbed by the presence of PFD (10 µM) 11 ( Fig. 7B ). As such, distinct from previous reports in heart cells (Ramos-Mondragon 12 et al., 2012), the present results demonstrate that PFD exerts a depressant action of 13 the peak I Ca,L in GH 3 cells.

the pipette was filled with K + -containing internal solution. As depicted in Fig. 8A  1 and 8B, cell exposure to PFD resulted in a progressive decline in the firing frequency 2 of regenerative APs. For instance, the exposure to PFD at a concentration of 1 or 3 3 µM greatly decreased discharge rate of current-clamped cells to 1.07±0.07 Hz (n=9, 4 P<0.05) or 0.74±0.05 Hz (n=9, P<0.05), respectively, from a control value of 1.45±0.09 5 Hz (n=9). Meanwhile, the presence of zatebradine (3 µM) alone was noted to 6 decrease firing frequency effectively. Zatebradine has previously been reported to 7 be an inhibitor of I h (Novella Romanelli et al., 2016; Romanelli et al., 2005; Van 8 Bogaert and Pittoors, 2003) . Therefore, it is conceivable from the current data that 9 PFD-mediated modification in membrane potential tends to be intimately connected 10 to its perturbations on both I h and I Ca,L detected above in these cells. 11

In another stage of current-clamp recordings, the ability of PFD to induce 13 perturbations on sag voltage was further investigated. The magnitude of such 14 voltage responding to hyperpolarizing current injection has been reported linking to 15 the emergence of I h (Chang et al., 2019; Datunashvili et al., 2018) . As depicted in 16 Fig. 9A and 9B , under the current experimental conditions, when the whole-cell 17 potential recordings were set up, the hyperpolarizing current injection at the 18 amplitude of 25 pA was able to induce sag voltage (i.e., drop down to a lower level in the membrane potential applied by such hyperpolarizing current injection). The 1 addition of chlorotoxin (1 µM), a blocker of Clchannels, was not found to exert any 2 effect on the amplitude of sag voltage in response to hyperpolarizing current 3 injection. On the other hand, as cells were exposed to PFD, the amplitude of sag 4 voltage was decreased. For example, the addition of 1 µM PFD caused a reduction 5 in the sag-voltage amplitude from 55±13 to 36±11 mV (n=8, P<0.05). Therefore, the 6 depression of sag voltage caused by PFD can be ascribed in large part to its inhibitory 7 effect on the amplitude and gating of I h detected above in these cells. 8

The I h of pituitary GH 3 cells demonstrated in the present investigation can be 2 observed following the 2-s-long hyperpolarizing pulse applied from -40 mV to the 3 voltages more negative to -80 mV, and the currents were increased in both current 4 amplitude and activation in response to more hyperpolarizing potentials (Chang et al., 5 2019; DiFrancesco and DiFrancesco, 2015; Irisawa et al., 1993; Simasko and 6 Sankaranarayanan, 1997) . The forward and backward amplitudes of this current 7 elicited by long-lasting inverted triangular ramp pulse were also noted to be distinct, 8 strongly reflecting the presence of voltage-dependent hysteresis fo I h (Barthel et al., 9 2016; Männikkö et al., 2005) . This major ion conductance inherently present in GH 3 10 lactotrophs was identified as an I h or I f current (Chang et al., 2019; Simasko and 11 Sankaranarayanan, 1997) . Specifically, this study shows for the first time the 12 effectiveness of PFD in depressing the amplitude of I h in a concentration-and 13 voltage-dependent fashion. 14 In this study, the block by PFD on I h in GH 3 cells are not strictly limited to its 15 inhibition of the current amplitude. Concomitantly, as cells were exposed to PFD, 16 the time course of I h activation and deactivation, in response to sustained membrane 17 hyperpolarization, apparently became slower. The presence of this compound was 18 found to produce a decrease in I h activation in a concentration-and time-dependent 22 manner. The results thus prompted us to reflect that blocking of I h by PFD 1 inherently is not instantaneous, but develops with time after the HCN channel 2 opened and subsequently produces a slowing in current activation. In other words, 3 the PFD molecule tends to represent a higher affinity for the open state in the HCN 4 channel; consequently, the transition from closed to open state turns out to be 5 slower during cell exposure to PFD. It is thus possible that PFD or its structurally 6 related compounds bind to the open state of the channel and/or block a prolonged 7 channel opening. 8

The steady-state activation curve of I h detected in this study was shifted along 9 the voltage axis toward a negative voltage in the presence of PFD. However, failure 10 of PFD to change the gating charge of the current in situations where the 11 transmembrane electrical field can be crossed during current activation, suggests 12 that PFD action on the HCN channel might consist in opening the gate, not interfering 13 with the region that senses the transmembrane potential. Nonetheless, the 14 sensitivity of electrically excitable cells to PFD could rely not simply on the PFD 15 concentrations given, but also the pre-existing level of the resting potential, the firing 16 pattern of APs, or their combinations, assuming the magnitude of I h is present in the 17 cells examined. 18

Voltage-dependent hysteresis of I h is considered to serve a role in influencing 19 the overall behaviors of electrically excitable cells including GH 3 cells. In accordance 1 with previous observations ((Barthel et al., 2016; Chang et al., 2020; Männikkö et al., 2 2005) , the I h natively in GH 3 cells has been described either to undergo a hysteretic 3 perturbation in its voltage dependence, or to produce a shift of ion-channel mode in 4 which the voltage sensitivity in gating charge movement of the current is dependent 5 on the previous state of the HCN channel involved (Fürst and D'Avanzo, 2015; 6 Männikkö et al., 2005) . In the present study, we also investigated the perturbations 7 of PFD on non-equilibrium property of I h consistently observed in GH 3 cells. In this 8 regard, because of a reduction in ∆area of hysteretic loop between forward and 9 backward limp of the inverted triangular ramp pulse, the experimental data suggest 10 that the presence of this compound is capable of diminishing such hysteresis entailed 11 in the voltage-dependent elicitation of I h . 12

The PFD concentration required for the half-maximal inhibition of I h detected in 13 this study was optimally determined to be 3.65 µM, a value which was within the 14 clinically applied doses reported previously. For example, following intravenous 15 administration of PFD, plasma concentration of this compound was previously 16 reported to range between 0.1 and 10 µg/ml (i.e., 0.54 and 54 µM) (Togami et al., (Ji et al., 2017; Maslanka Figueroa et al., 2020) . In this scenario, the observed 1 effects by PFD presented herein may achieve the concentration of clinical 2 requirements, so that maneuvers are implemented to mitigate the systemic toxicity. 3

Ivabradine or zatebradine, known to inhibit HCN-encoded currents, have been 4 previously reported to reduce diastolic dysfunction and later to ameliorate cardiac 5 fibrosis in animal models (Busseuil et al., 2010; Fang et al., 2017) . In our study, 6 subsequent addition of PFD, still in the continued presence of this compound, can 7 further depress I h amplitude. As such, the inhibitory effect of ivabradine (or 8 zabebradine) and PFD on I h detected in GH 3 cells could be additive. The elucidation 9 whether the specific pharmacological interaction between the compounds is additive 10 or even synergistic would require a more specific approach (i.e. isobolographic 11 assessment), which is beyond the scope of the present study. Therefore, to what extent PFD-mediated depression of I h participates in its 18 antifibrotic activity remained to be further resolved. Whether PFD-mediated block of I h potentially contributes to its cardioprotective action described previously (Aimo et 1 al., 2020; Fang et al., 2017; Nguyen et al., 2010) also needs further detailed 2 investigations. 3

It is important to mention, from the present observations, that unlike its 4 depressant effectiveness on I Ca,L , block of I h caused by PFD tends to be not 5 instantaneous, but develops with time after the HCN channels become overly 6 opened upon sustained membrane hyperpolarization, thereby producing a 7 conceivable increase in τ act value detected from whole-cell current activation. In 8 the presence of this compound, deactivating current of I h was also notably observed 9 to exhibit a blunted peak followed by a slowing in the decay, reflecting that the 10 closing (i.e., deactivating process) of channels virtually became slower by unbinding 11 of the PFD molecule. PFD or its structurally similar compounds may hence be an 12 intriguing tool to probe HCNx channels in the structural and functional perspectives, 13

given that the pore region of the channel protein to which they bind appears to be of 14 particular relevance for an open-channel blockade. 15

The present observations revealed the effectiveness of PFD on the decline in the 16 firing frequency of spontaneous APs observed under current-clamp potential 17 measurements. The amplitude of sag voltage elicited by 2-s hyperpolarizing current 18 stimuli was also decreased in its presence. The inhibition by PFD of I h amplitude 19 and gating may confer its effectiveness on different cellular functions (e.g., 1 stimulus-secretion coupling) in various types of electrically excitable cells (Lee et al., 2 2017; Stojilkovic et al., 2017; Zhang et al., 2008) . Whether similar findings occur in 3 different types of native cells in vivo still remains to be further delineated. 4

It is also important to emphasize that the treatment with PFD has been 5 described to enhance the peak amplitude of I Ca,L in heart cells (Ramos-Mondragon et 6 al., 2012) . However, in the present study, vastly different from the depressant 7 action of PFD on I h , the addition of this compound was able to decrease the peak I Ca,L 8 detected in GH 3 cells. Moreover, neither activation nor inactivation time course of 9 the current by abrupt membrane depolarization presumably was not modified. A 10 previous study has indeed reported that quercetin, a bioflavonoid, could either 11 increase or decrease the amplitude of I Ca,L in different types of electrically excitable 12 cells (Wu et al., 2003) . Nonetheless, in combination with the depression of I h , 13 PFD-mediated depression of I Ca,L might also contribute to its effectiveness in altering 14 functional activities of endocrine or neuroendocrine cells (Lee et al., 2017) , despite 15 inability of PFD to alter the amplitude of I K(M) or I K(erg) . However, how PFD-perturbed 16 changes on I h or other ionic currents demonstrated here could be linked to its 17 pathological or morphological changes (e.g., fibrosis) remained to be further studied.

Taken together, these results suggest that PDF is able to produce a decrease in I h 1 activation in a concentration-dependent manner. Also, the substance presents the 2 characteristic of opening HCN channel and counteracting a prolonged channel 3 opening. Our findings on GH 3 cells shed light on the evidence that PFD perturbs 4 specific ionic currents which may be linked to additional effects, potentially useful for 5 therapeutic application, when elicited on different excitable cells. No conflicts of interest, financial or otherwise, are declared by the author(s). 7

The funders had no role in the design of the study; in the collection, analyses, or 8 interpretation of data; in the writing of the manuscript, or in the decision to publish 9 the results. Kanayama, M., Mori, M., Matsumiya, H., Taira, A., Shinohara, S., Kuwata, T., Imanishi, 2 N., Yoneda, K., Kuroda, K., and Tanaka, F., 2020 . Perioperative pirfenidone treatment 3 for lung cancer patients with idiopathic pulmonary fibrosis. Surg Today 50, [469] [470] [471] [472] [473] [474] Y.V., Kozlov, S.A., Lipkin, A.V., Pluzhnikov, K.A., Hadley, J.K., Filippov, A.K., 5 Brown, D.A., Angelo, K., Strobaek, D., Jespersen, T., Olesen SP, Jensen BS, Grishin EV., 6 2001. An ERG channel inhibitor from the scorpion Buthus eupeus. J Biol Chem 276, 7 9868-9876. 8 Krämer, M., Markart, P., Drakopanagiotakis, F., Mamazhakypov, A., Schaefer, L., 9 Didiasova, M., and Wygrecka, M., 2020 

Pirfenidone is a cardioprotective drug: Mechanisms Significantly different from control

The authors declare no competing financial interests.