**3. Functional Properties and the Role of Piezo1 vs. Piezo2 in Nociception**

After discovery, isolation and cloning of Piezo1 and Piezo2 channels [8], the next challenging step would be to describe the functional properties of these channels in different cell types [9]. The natural activation of Piezo1 and Piezo2 channels can be achieved by different types of mechanical stimulation: stretching, pulling, pushing, exposure to hypo- or hyper-osmotic solutions, and flow-induced shear stress [8,13,14]. The kinetic properties of Piezo channels, studied by recording of the whole cell-inward currents in HEK cells, showed the fast activation and inactivation of these channels with recovery in the range of hundred milliseconds [6].

The comparison of two subtypes of Piezo channels showed that Piezo2 has a faster kinetics, and typically mediates a rapid membrane response, whereas Piezo1 channels are characterized by slower kinetics [8]. Thus, Piezo2 are more specified for detection of transient mechanical forces, whereas Piezo1 can react to more persistent activation [15]. Notably, these properties are sensitive to the channel's environment in the membrane. Thus, the reduction of cholesterol in the cell membrane largely slowed down the inactivation kinetics of Piezo1 channels [16], which can support the persistence of signaling via this receptor type.

Consistent with fast kinetics, Piezo2 channels mediate the short-lasting mechanosensitive processes such as touch [6]. Piezo2 channels, implicated in this sensory modality, have been found in dorsal root ganglion (DRG) neurons [6]. Mechanical pain represents a different sensory modality, which is also expected to be mechanistically linked either to Piezo1 or Piezo2 channels if they are expressed in neurons mediating somatic and visceral pain. However, there are some controversies regarding expression of Piezo channel subtypes in nociceptive neurons. Indeed, one study showed the presence of Piezo2 but not Piezo1 channel in DRG neurons [17]. Another group [18] presented the evidence for the expression not only of Piezo2 but also of Piezo1 in DRG neurons. They found Piezo2 transcripts in sensory neurons of different sizes, including the largest diameter neurons mediating touch and proprioception. In contrast, Piezo1 was preferentially expressed in small size nociceptive neuron suggesting their role in pain [18]. Co-expression of Piezo1 and Piezo2 in the same neurons raises an interesting issue of functional interactions between these channels. The functional interplay between Piezo1 and Piezo2 channels was analyzed in the study where they found that the deletion of Piezo2 from the low-threshold mechanoreceptors in mice impaired touch but surprisingly sensitized mechanical pain, suggesting a negative interaction between Piezo1 and Piezo2 [19].

Our studies indicated that both Piezo1 and Piezo2 channels are expressed in trigeminal sensory neurons [5], which innervate head and face tissues and implicated in generation of migraine pain. Most of electrophysiological studies of Piezo channels in sensory neurons were performed by recording signaling from somas of these cells, representing a surrogate model of nerve terminals. However, more physiologically relevant approach to study the role of Piezo channels in migraine pain would be the recording of electrical spiking activity directly from the trigeminal nerve terminals in brain meninges [20]. Cranial meninges, comprising abundant blood vessels that are densely innervated by somatic and autonomous nerves, represent the so-called 'trigeminovascular system', which is considered as the origin site of primary headaches including migraine [21]. Application of the pro-inflammatory compounds to meninges induces mechanical sensitization of meningeal nociceptors [22], suggesting involvement of the professional mechanotransducers such as Piezo1 channels in this phenomenon. The repetitive nociceptive traffic in trigeminal neurons is a likely reason for mechanical hypersensitivity and allodynia, typical for migraine pathology [23]. Mechanical hypersensitivity can be directly linked to activation of Piezo channels in peripheral neurons whereas allodynia, mostly a central phenomenon, nevertheless, can also start from excessive and repetitive activation of peripheral Piezo channels in primary afferents. Taken together, these studies can serve as a background for the hypothesis on the role of Piezo channels in migraine.

## **4. Unusual Chemical Activation of Piezo Channels**

Since the discovery of Piezo mechanotransducers, the main tools to activate Piezo channels were the different types of mechanical stimulation. Unexpectedly, a very efficient alternative approach to activate Piezo channels has been recently found: that is the compound called Yoda1 [24]. The small lipid soluble molecule Yoda1 is able to activate specifically Piezo1 but not Piezo2 channels [24]. Yoda1 interacts with the C-terminal of the Piezo1 protein in the region of 1961–2063 amino acids, also known as the Agonist Transduction Motif (ATM) [25] (Figure 1B). The recent molecular dynamic simulations identified the Yoda1 binding pocket located in the domain approximately 40 Å away from the central pore [26]. Although the Piezo1 channel has three interacting monomers, interestingly, the binding of Yoda1 to only one subunit is already enough to open the ion channel [25], which provides a rationale for the high sensitivity of Piezo1 to this chemical agonist.

The other Piezo1 agonist, Jedi1/2, acts on the blade-beam structure inducing activation of the channel from the peripheral extracellular side [10]. However, it remains to be discovered if there are any endogenous molecules, which, similarly to the synthetic Yoda1 or Jedi1/2 compounds, can activate and/or sensitize Piezo1 channels in the healthy or disease states.

The discovery of chemical agonists of Piezo channels opened a new toolbox to investigate the function of mechanotransduction in different tissues, especially, when the traditional mechanical stimulation is not applicable or when the aim is to provide a widespread activation of Piezo channels in multiple targets. Thus, we found that the application of Yoda1 to the extended receptive field of meningeal afferents induced a massive and prolonged activation of trigeminal nerve fibers [5]. The nociceptive effect of Yoda1 in this study was reproduced by the similar prolonged activation of trigeminal mechanosensitive receptors by the hypo-osmotic solution [5]. Moreover, Yoda1 stimulation triggered the release of CGRP (calcitonin gene-related peptide) from these trigeminal nerve fibers. CGRP, the main migraine mediator, is known as a powerful promoter of meningeal inflammation and sensitization of trigeminal neurons [27–29].

These findings largely supported the proposed role of Piezo channels in peripheral mechanisms of migraine pain. Consistent with the pro-nociceptive role of Piezo channels activated by Yoda1 found in our study, Wang et al. [18] showed that Yoda1 induced a mechanical hyperalgesia with the prolonged time-course which is also typical for migraine pain.
