Characterization of the expression of hiPSC-derived sensory neurons

HiPSC-derived sensory neurons (cat #RCDN004N, Reprocell.Inc) were used in this study. To identify characteristics of human iPSC-derived sensory neurons, we confirmed sensory neuron-related genes and proteins expression by real-time PCR and immunocytochemistry (ICC). Real-time PCR shows an increase in Peripherin, Brn3a, TRPV1, TRPM8, Nav1.7, Nav1.8, Piezo2, TRKA, TRKC, TRKB, P2X3, H1R, MrgprX1, CGRP and TAC1 compared with hiPSC cultured 14days in vitro (DIV) (Fig.1). The nociceptor phenotype consists of A-fibers, and C-fibers. C-fibers respond to both peptidergic and non-peptidergic neurotransmitters. HiPSC-derived sensory neurons expressed TRKA (nociceptor marker), IB4 (A-fibers marker), CGRP and TAC1 (peptidergic neurotransmitters) and P2X3 (ATP (non-peptidergic neurotransmitter) receptor) (Figs. 1i,l,o,p and 2k). Furthermore, TRPV1, TRPM8, Nav1.7 and Nav1.8 which are nociceptors receptors were also expressed (Fig.1c,eg). TRPV1 is known to be activated by capsaicin and noxious heat (43C)14. TRPM8 is known to be activated by menthol, noxious (<15C) and non-noxious (2815C) heat15,16. Nav1.7 and Nav1.8 are known to be subtype of voltage-gated sodium channels which is preferentially expressed in nociceptors17,18,19. The mechanoreceptor phenotype consists of relatively large diameter cells that are A-fibers. HiPSC-derived sensory neurons expressed TRKB (mechanoreceptor marker), NF200 (A-fibers marker), TRPM8 and Piezo2 (mechanoreceptor receptors) (Figs. 1e,h,j and 2j)4. TRKC, a proprioceptor marker was expressed in hiPSC-derived sensory neurons (Fig.1k). Thus, these data suggests that the hiPSC-derived sensory neurons generated constitute a heterogeneous population of sensory neuronal subclasses. The expression of TRPA1 in hiPSC-derived sensory neurons was lower than the one in hiPSC (Fig.1d). The expression levels of Brn3a, TRPM8, TRKB and MrgprX1 in hiPSC-derived sensory neurons were comparable to those in human DRG, whereas the others were lower than in hDRG. The reason some genes of hiPSC-derived sensory neurons showed lower expression than hDRG might be due to the immature nature of the hiPSC-derived sensory neurons20. Although we cultured them for a long time, the expression levels of Peripherin, TRPV1, TRPA1, Nav1.7, Nav1.8, H1R, and CGRP were not comparable to the ones in hDRG (Supplementary Fig. S1). Therefore, we confirmed proteins expression by ICC.

Expression of sensory neuron related genes in hiPSC, hiPSC-derived sensory neurons and human DRG. Real-time PCR showed expression of (a) Peripherin, (b) Brn3a, (c) TRPV1, (d) TRPA1, (e) TRPM8, (f) Nav1.7, (g) Nav1.8, (h) Piezo2, (i) TRKA, (j) TRKB, (k) TRKC, (l) P2X3, (m) H1R, (n) MrgprX1, (o) CGRP, (p)TAC1. The square marker, the circle marker and triangle marker indicate expression of genes in hiPSC, hiPSC-derived sensory neurons and human DRG respectively. Three different lot of hiPSC-derived sensory neurons were examined. The line marker represents the mean expression of genes in hiPSC-derived sensory neurons.

Expression of sensory neuron related proteins in hiPSC-derived sensory neurons and their morphology. The cells are stained for (a) TUBB3, (b) Peripherin, (c) Brn3a, (d) TRPV1, (e) TRPM8, (f) Nav1.7, (g) TRKA, (h) TRKB, (i) TRKC, (j) NF200, (k) IB4. DAPI stain of nuclei is shown in blue. (l) Image of iPSC-derived sensory neurons which exhibit a bipolar (red arrowhead), pseudounipolar (yellow arrowhead), or multipolar morphology (green arrowhead). Scale bar represents 50m.

ICC showed expression of TUBB3 (mature neuron marker), Peripherin (peripheral neuron marker) and Brn3a (sensory neuron marker) at 14 DIV (Fig.2ac). TRPV1, TRPM8, Nav1.7, TRKA, TRKB and TRKC were expressed at the membrane (Fig.2di). Since TRPV1, and TRPM8 are receptors of noxious and non-noxious stimulation and are expressed at the membrane, we expected them to be available for characterizing their function using MEA (Fig.2d,e). NF200, a A-fibers marker, was expressed at a higher-intensity in relatively large diameter cells than small diameter cells (Fig.2j). Although adult human DRG do not bind IB4 which is a non-peptidergic C-fibers marker, it was expressed in hiPSC-derived sensory neurons (Fig.2k)21. The research showed expression of IB4 in prenatal human DRG at 8-month of gestation22. This data suggests that hiPSC-derived sensory neurons might be immature.

It is known that when observing rat DRG cells in the early stages of development, their morphology changes from bipolar cells to pseudounipolar cells23. Our hiPSC-derived sensory neurons exhibit a bipolar, pseudounipolar and multipolar morphology (Fig.2l). A majority of the hiPSC-derived sensory neurons were bipolar neurons. This image suggests that our hiPSC-derived sensory neurons contained neurons with different degrees of maturity.

Taken together, hiPSC-derived sensory neurons express sensory neuron-related genes and proteins. They constitute a heterogeneous population of nociceptors, mechanoreceptors, and proprioceptors, and they differ in maturity. Thus, we proceeded to characterize their function next.

We confirmed whether hiPSC-derived sensory neurons responded to capsaicin, menthol, noxious heat (4346C), which are noxious stimuli, and bradykinin, and non-noxious heat (3742C), which is a non-noxious stimulus14,15,16,24. Sensory neurons of DRG are used as in vitro model of nociceptive response. DRG responded to 10nM, 100nM, 1M capsaicin25,26. We measured and compared data before and after drug treatment (Fig.3a). We measured the response to treatment with 100nM capsaicin which resulted in increase in Mean Firing Rate (MFR) and Number of Bursts (NOB), whereas vehicle treatment had no effect on them (Fig.3d,e). Capsaicin-evoked activity is known to be rapid in DRG25. This result showed that neural activity was evoked within 10s after treatment with capsaicin in hiPSC-derived sensory neurons as well (Fig.3b). To determine whether capsaicin-evoked neuronal activity is characteristic of hiPSC-derived sensory neuron, we treated with 100nM capsaicin in hiPSC-derived cortical neurons. As a result, hiPSC-derived cortical neurons did not respond to capsaicin (Fig.3ce). Moreover, we added 100nM AMG9810 which is a TRPV1 antagonist for 60min before treating with capsaicin. A response to capsaicin was not observed in the presence of AMG9810 (Fig. S2a,b). These data suggest that capsaicin-evoked activity occurred via TRPV1 in iPSC-derived sensory neurons. Thus, we can conclude that hiPSC-derived sensory neurons specifically respond to noxious stimulus and could be used in functional assays using MEA.

Capsaicin and menthol responsiveness using MEA. (a) Timeline of drug treatment. Baseline and dose response were recorded for 60s when treating with capsaicin or menthol. Capsaicin experiment raster plots of (b) hiPSC-derived sensory neurons and (c) hiPSC-derived cortical neurons. The triangle marker indicates the time of capsaicin addition. (d, h) Mean Firing Rate normalized to the control. Control firing rate is calculated as firing rate before adding vehicle or drug. (e, i) Number of Bursts normalized to the control. Menthol experiment raster plot of (f) hiPSC-derived sensory neurons and (g) hiPSC-derived cortical neurons. n=3 wells.

Menthol activates TRPM8 which is a nociceptive receptor. Since mouse and rat DRG respond to 10M and 100M menthol, we decided to treat with the same concentrations26,27. The high concentration of menthol resulted in suppressing spontaneous neural activity (Supplementary Fig. S3). This may be due in part to the higher expression of TRPM8 in hiPSC-derived sensory neurons than in human DRG (Fig.1e, Supplementary Fig. S1e). Treatment with 100nM menthol resulted in an increase in MFR and NOB in hiPSC-derived sensory neurons whereas hiPSC-derived cortical neurons did not respond to menthol (Fig.3fi). The data show that menthol got a response from nociceptive-like and non-nociceptive-like DRG neurons28. Since our hiPSC-derived sensory neurons responded to menthol, they may also include functionally non-nociceptive like neurons.

Bradykinin activates nociceptors and causes pain,24. Treatment with 100nM bradykinin resulted in significant increase in MFR and NOB compared to vehicle treatment for 60s (n=3, p<0.05) (Fig.4ac). In contrast to capsaicin and menthol, the onset of bradykinin-evoked neural activity was relatively long (Figs. 3b,f and 4a). Bradykinin-evoked activity increased gradually and reached its mean peak at 60s in DRG25. However, hiPSC-derived sensory neurons were able to respond faster than DRG, because they also responded to an additional stimulation which immediately activated them when bradykinin and vehicle were added against the well of the MEA plate. There is expression of TRKB and Piezo2 relevant to touch sensation in hiPSC-derived sensory neurons, explaining why they may have responded to an additional stimulation.

Bradykinin responsiveness. (a) Bradykinin experiment raster plot of hiPSC-derived sensory neurons. The triangle marker indicates the time of bradykinin addition. (b) Mean Firing Rate after addition of Bradykinin normalized to firing rate before addition of Bradykinin. (c) Number of Bursts after addition of Bradykinin normalized to number of bursts before addition of Bradykinin. n=3 wells, *p<0.05.

MFR were observed to increase gradually in DRG, when temperature increases from 37 to 42C via the stage plate heater, part of the recording system25. We increased the temperature from 37 to 46C via MAESTROs system to confirm responsiveness to noxious heat and non-noxious heat. We observed that MFR and NOB increased gradually and reached their mean peak at 45C and 46C respectively, in hiPSC-derived sensory neurons (Fig.5). In the presence of TRPV1 antagonist, AMG9810, MFR were lower than that of vehicle at 4346C (Fig. S2c,d). Because TRPV1 is known to be activated by noxious heat (43C), these results suggest that TRPV1 may contribute to the response to 4346C in iPSC-derived sensory neurons. The relative levels of MFR (1.540.046) and NOB (1.620.01) at 41C, which is non-noxious heat, were significantly higher than those at 37C. However, MFR decreased gradually in hiPSC-derived cortical neurons with an increase in temperature (Fig.5b,c).

Temperature responsiveness. Raster plots of (a) hiPSC-derived sensory neurons and (b) hiPSC-derived cortical neurons when the temperature is gradually increased from 37 to 46C. (c) Mean Firing Rate normalized to the firing rate at 37C. (d) Number of Bursts normalized to the number of bursts at 37C. The data for the number of bursts in hiPSC-derived cortical neurons isnt shown because one of the three wells didnt produce any burst. n=3 wells, *p<0.05, **p<0.01 compared with the corresponding value at 37C. Functional assessment of hiPSC-derived sensory neurons against itching stimuli

These data suggest that the observed and recorded response is specific to sensory neurons and the hiPSC-derived sensory neuron populations generated in this study are likely to include nociceptors that respond to noxious stimuli like capsaicin, menthol, bradykinin, and noxious heat (43C) and to include mechanoreceptors that respond to non-noxious stimuli (41C).

Atopic dermatitis (AD) is the most common chronic skin disease which causes a global disease burden29. AD causes itch (pruritus) and poor non-health-based quality-of-life. It is known that itch occurs via C-fiber in nociceptors30. Recently, investigating itch has been established by using human sensory neurons from stem and other progenitor cells as in vitro model31. Although substances causing itch treat to their cells, their inhibitors effect arent confirmed using hiPSC-derived sensory neurons. Because we demonstrated that our hiPSC-derived sensory neurons expressed nociceptor genes and proteins, and responded to noxious stimuli, we expected that they also responded to an itch stimulus and its inhibitor.

Histamine is one of the substances that cause itching via C-fiber. The histamine receptor is a four G protein-coupled receptor. Histamine H1 receptor (H1R) is involved in the induction of histamine-induced pruritus32. Since we confirmed that the H1R gene is expressed, we examined whether hiPSC-derived sensory neurons respond to histamine, using MEA. Mouse DRG responded to 100M Histamine, as described in the literature33. HiPSC-derived sensory neurons didnt respond to 100M Histamine but responded to 1mM Histamine (Supplementary Fig. S3df and Fig.6a,b). The MFR gradually increased and reached its mean peak at 25min. Pyrilamine is a histamine H1 receptor inverse agonist. We treated the sensory neuron population with 10M Pyrilamine for 60min before adding Histamine. As a result, Pyrilamine inhibited Histamine-evoked activity (Fig.6a,b). These results suggest that Histamine-evoked activity occurred via H1R in iPSC-derived sensory neurons.

Histamine, H1R inhibitor, pyrilamine and chloroquine responsiveness. (a) The left raster plots have been recorded before histamine addition. The right raster plots have been recorded 25min after histamine addition. Upper raster plots are recorded in pyrilamine absence. Lower raster plots are recorded with presence of pyrilamine. (b, e) Mean Firing Rate normalized to firing rate before drug addition. Raster plots of (c) before chloroquine addition and (d) 5min after chloroquine addition. The experiment with histamine and pyrilamine was performed with n=2 wells each. Experiment with chloroquine was performed with n=3 wells. *p<0.05, **p<0.01 compared with the value recorded before addition or with vehicle.

Chloroquine is a drug that has been used in the treatment to prevent malaria. Histamine-independent pruritus is known to be one of the side effects of chloroquine34. Mrgprs are receptors of chloroquine and are activated by it35. Since we confirmed expression of human MrgprX1 by real-time PCR, we investigated the potential response of hiPSC-derived sensory neurons by chloroquine. DRG are reported to respond to 1mM chloroquine, however the MFR gradually decreased at the same concentration in hiPSC-derived sensory neurons (Supplementary Fig. S3g,h)36. 1M chloroquine increased the MFR and reached the mean peak after 5min of incubation (Fig.6ce). The mean peak after stimulation with chloroquine was reached faster than after stimulation with histamine.

These data showed an example of the effect of an itch inhibitor and different responses between itch inducing drugs. HiPSC-derived sensory neurons may be available for drug discovery against AD.

Nav1.7 subtype of voltage-gated sodium channels is expressed in DRG. Mutations in the gene encoding Nav1.7 are associated with either absence of pain or with exacerbation of pain. Recently, Nav1.7 has been an attractive target to pursue treating pain. ProTx-II is a tarantula venom peptide that preferentially inhibits Nav1.7 over other Nav subtypes37. It suppressed spontaneous neural activity in a time dependent manner in hiPSC-derived sensory neurons (Fig.7ae). The MFR and NOB are significantly diminished after 35min of incubation with 1M ProTx-II (Fig.7d,e). After washing out ProTx-II, the suppressed neural activity gradually recovered. Although, responses was completely blocked by 300nM ProTx-II in rodent DRG, the responses in hiPSC-derived sensory neurons were not blocked at the same concentration38. This may be due in part to the lower expression of Nav1.7 in hiPSC-derived sensory neurons than in human DRG.

Nav1.7 channels inhibitor, ProTx-II and Nav1.7 inhibitor responsiveness. Raster plots of (a) before ProTx-II addition (baseline), (b) 35min after adding ProTx-II and (c) 150min after washing ProTx-II, respectively. (d, i) Mean Firing Rate and (e, j) Number of Bursts normalized to mean firing rate and number of bursts before drug addition. Raster plots of (f) before Nav1.7 inhibitor addition (baseline), (b) 50min after adding Nav1.7 inhibitor and (c) 30min after washing Nav1.7 inhibitor, respectively. n=3 wells, *p<0.05, **p<0.01, ***p<0.001 compared to the value recorded before drug addition.

ProTx-II is known to act not only as a Nav1.7 inhibitor but also to act on Nav1.5 channels and on some T-Type Ca2+ channels39,40. Thus we administered a small molecule inhibitor that is more selective for Nav1.741. Although 300nM of a more selective Nav1.7 inhibitor suppressed MFR and NOB in a time dependent manner in hiPSC-derived sensory neurons, the degree of decrease was lower than that of ProTx-II (Fig.7fj). After washing it out, the suppression of the neural activity was lifted.

These results suggested that hiPSC-derived sensory neurons may serve as a drug screening tool for pain.

Go here to see the original:
Characterization of human iPSC-derived sensory neurons and their functional assessment using multi electrode array ... - Nature.com