Electrophysiological analyses of human dorsal root ganglia and human induced pluripotent stem cell-derived sensory neurons from male and female donors

Abstract

Human induced pluripotent stem cell-derived sensory neurons (hiPSC-SNs) and human dorsal root ganglia (hDRG) neurons are popular tools in the field of pain research; however, few groups make use of both approaches. For screening and analgesic validation purposes, important characterizations can be determined of the similarities and differences between hDRG and hiPSC-SNs. This study focuses specifically on electrophysiology properties of hDRG in comparison to hiPSC-SNs. We also compared hDRG and hiPSC-SNs from both male and female donors to evaluate potential sex differences. We recorded neuronal size, rheobase, resting membrane potential, input resistance, and action potential waveform properties from 83 hiPSCs-SNs (2 donors) and 108 hDRG neurons (9 donors). We observed several statistically significant electrophysiological differences between hDRG and hiPSC-SNs, such as size, rheobase, input resistance, and several actional potential (AP) waveform properties. Correlation analysis also revealed many properties that were positively or negatively correlated, some of which were differentially correlated between hDRG and hiPSC-SNs. This study shows several differences between hDRG and hiPSC-SNs and allows better understanding of the advantages and disadvantages of both for use in pain research. We hope this study will be a valuable resource for pain researchers considering the use of these humanin vitrosystems for mechanistic studies and/or drug development projects.

Machine learning elucidates electrophysiological properties predictive of multi- and single-firing human and mouse dorsal root ganglia neurons

“…hDRG-N culture was performed as previously described (Valtcheva et al, 2016; Zurek et al, 2024). hDRG were obtained from consenting recently deceased organ donors at University of New Mexico Hospital in coordination with New Mexico Donor Services.…”

“…Study activities were approved by the Human Research Review Committee at the University of New Mexico Health Sciences Center; approval numbers #21-412 or #23-205. Cultures of hDRG-N were prepared as described previously and cultured for up to 11 days in vitro (DIV) (Zurek et al, 2024) and electrophysiological recordings took place between DIV 3 to 11. We used high-quality recordings where all 22 electrophysiological properties could be extracted from 94 untreated hDRG-N.…”

“…We used high-quality recordings where all 22 electrophysiological properties could be extracted from 94 untreated hDRG-N. The data from 8 of 10 donors was previously published in (Zurek et al, 2024). Donor demographics are described in Table 1.…”

“…Neuronal hyperexcitability is a hallmark of chronic pain and understanding the electrophysiological mechanisms that lead to neuronal excitability is crucial for the development of pain treatments (Berta et al, 2017; Alles and Smith, 2018). Sensory neurons isolated from human and mouse dorsal ganglia (hDRG-N and mDRG-N) are important tools in the study of neuronal excitability, chronic pain, and nociception (Davidson et al, 2014; Zhang et al, 2017; Emery and Ernfors, 2018; North et al, 2019; Zheng et al, 2019; Zurek et al, 2024). While the transcriptomes of hDRG and mDRG have been compared (Ray et al, 2018), few studies have compared the electrophysiological features of these tools.…”

“…In this study, we compared the electrophysiological features of multi- and single-firing hDRG-N and mDRG-N and applied several machine learning algorithms to elucidate which combinations of features were most predictive of multi-firing neurons. Because we have collected a relatively high number of hDRG-N recordings from a diverse demographic of donors (Zurek et al, 2024), we aimed to see if machine learning algorithms can predict whether an hDRG-N will be single- or multi-firing based on other intrinsic, phenotypic, and AP waveform electrophysiological features. While many machine learning studies use a single model to make predictions, we used several different models.…”