*2.5. Kinetics of Action*

Observation of the drug effects throughout the series of activations in the drug presence revealed an interesting tendency (Figure 5). Unlike the typical monotonic effect development, in experiments with IEM-2163, the sustained current was strongly inhibited during the first activation in the presence of the drug, but in subsequent activations the inhibition became less pronounced. The washout process was also non-monotonic—in the first activation it demonstrated a significant "overshoot"—the response decay was much slower than in control, resulting in the current at the end of activation being higher than the control one (Figure 5A). The control parameters were eventually reached after 5–7 activations. For IEM-2195, its potentiation developed monotonically but, similarly to IEM-2163, there was also a washout "overshoot"—in the first washout activation the current at the end of the response was even higher than in the last activation with the drug (Figure 5B).

**Figure 4.** Effects of drug application protocol on peak and sustained currents of ASIC3 in different activating pH. Number of \* denotes statistical significance at \* *p* < 0.05, \*\* *p* < 0.01, or \*\*\* *p* < 0.001, respectively, *n* ≥ 5. (**A**) Effects of IEM-2163 on peak current and (**B)** sustained current. (**C**,**D**) Representative examples of ASIC3 responses in different activating pH (6.85 in (**C**) and 6.0 in (**D**)) and application protocols (left panels: application before activation, right panels: continuous application). (**E**–**H**) Same for IEM-2195. IEM-2163 (**A**–**D**) had mostly similar effects regardless of the protocol used, inhibiting peak current at pH 6.85 and sustained current for both pH values, with two notable exceptions. When applied before activation with activating pH 6.85 it had no effect at all, and in the same protocol but with activating pH 6.0 it strongly potentiated sustained current. IEM-2195 (**E**–**H**) mostly had a similar profile, but it typically potentiated the sustained current, although this effect was significantly weaker at pH 6.85, essentially disappearing when applied before activation.

**Figure 5.** Non-monotonic effect development and washout of IEM-2163 and IEM-2195. (**A**,**B**) Representative recordings of the time course of experiments with IEM-2163 (**A**) and IEM-2195 (**B**). To the right are overlaid responses from the main panel. (**C**) The effect was the most pronounced in the protocol of continuous application. Number of \* denotes statistical significance at \* *p* < 0.05, \*\* *p* < 0.01, or \*\*\* *p* < 0.001, respectively, *n* ≥ 5. The average values were calculated as the ratio of amplitudes for the first drug application, last drug application, and first washout to the last control response, respectively.

This phenomenon was the most pronounced with continuous drug application (Figure 5C). However, if the drugs were applied only before the channel activation, the overshoot effect disappeared. In this case both IEM-2163 and IEM-2195 caused potentiation, and recovery from it developed monotonically.

Our explanation of these effects is that inhibition of activation is fast while the effect on desensitization is much slower. Inhibition develops during the first activation in the presence of the drugs and is just as rapidly washed out during the first activation without them, while the reduction of desensitization requires several minutes to develop and wash out. This explanation agrees with the protocol dependence—fast peak inhibition required the drug's presence during activation, whereas a slow effect on desensitization could be obtained during long pre-application and remained even if the drug was absent from the solution during the activation. Thus, the effects on activation and desensitization differed not only in structural determinants (Figure 2) as well as concentration and pH dependence (Figure 3), but also in their kinetics.

#### *2.6. Biphasic Drug Effects, when Applied Exclusively to the Sustained Current*

The fact that ASIC3s do not desensitize completely and can mediate significant sustained current allows for one more type of experiment (Figure 6), which is helpful for the analysis of the mechanism of action. We activated the channels in the absence of a drug and only applied it when the current reached the sustained level. Washout was also performed during this prolonged activation, without returning to the neutral pH. Typical currents are presented in Figure 6A,B. Application of IEM-2195 (Figure 6A) caused fast transient "on current", which then slowly returned to the equilibrium level similar to the value of sustained current potentiation observed in the previous experiments. Removal of the drug resulted in a large transient "tail current" before returning to the control value. Application and removal of IEM-2163 caused similar "on" and "tail" transient currents, although they had smaller amplitude (Figure 6B). The main difference between the drugs was the direction of the change in the sustained current's amplitude at the equilibrium level, which was potentiated by IEM-2195 and inhibited by IEM-2163, respectively. We were especially careful to ensure that this unusual behavior was not an artifact of the solution exchange. Additionally, "tail" currents for both compounds demonstrated clear concentration dependence (Figure 6C), with fitting resulting in EC50 = 269.06 ± 15.35 μM, nH = 1.76 ± 0.11 for IEM-2195 and EC50 = 319.69 ± 44.08 μM, nH = 2.00 ± 0.33 for IEM-2163.

**Figure 6.** IEM-2163 and IEM-2195 cause transient currents when applied to sustained response. (**A**,**B**) Representative recordings. Fast drug application caused transient current decrease ("on" current), while washout induced transient increase ("tail" current). These transients reflect the presence of two opposite effects with different kinetics. (**C**) Concentration dependencies of "tail" currents.

We suggest that the observed "on" and "tail" currents reflect kinetics and complex mechanisms of drug action, which include the inhibition of activation and reduction of desensitization. We suggest that the "on" current appears because inhibitory action develops quickly, while slow modulation of desensitization is responsible for the subsequent equilibrium level of the sustained current. The change of this equilibrium effect (potentiation by IEM-2195 and inhibition by IEM-2163) may depend on the balance between these two opposite actions. Fast inhibition of activation would also be responsible for the "tail" currents, resulting from an acidic shift of activation (see Figure 3B). Purportedly, in this case the drug-bound channels would remain in the resting state even under conditions of acidic pH. Thus, fast removal of a drug would allow protons to bind and activate the channels.

To further check this suggestion, we performed analogous experiments with some other drugs (Figure 7). 9-Aminoacridine and IEM-2044, which inhibit peak and potentiate the sustained component of the response, also demonstrated pronounced "on" and "tail" currents. In contrast, for IEM-2059 and IEM-1755 these transient currents were absent. In analogous experiments with agmatine performed by Li et al. [26], no "on" or "tail" currents were shown, probably because agmatine leads to an alkaline shift of activation and has an overall potentiating effect on ASIC3 currents.

**Figure 7.** "On" and "tail" currents for different drugs. Transient currents were pronounced for the drugs which caused peak inhibition at pH 6.85 due to the activation shift (9AA, IEM-2044, see Figure 2A).
