Wishing for made-to-order human neurons to model pain? The wait could soon be over. In a paper published online July 1 in Nature Biotechnology, a team led by Lorenz Studer at Sloan-Kettering Institute, New York City, US, reports the generation of nociceptors from human embryonic stem (ES) cells using a combination of five small molecules. The process quickly and efficiently produces cells that express ion channels and other nociceptive markers, and that respond to pain-producing substances. The results suggest the cells are working reproductions of genuine pain-sensing neurons, which could provide researchers a powerful platform for modeling human nociceptor physiology.
Previously, first author Stuart Chambers and the Studer team devised a way to generate neural precursors from human ES and induced pluripotent stem (iPS) cells in high yield using inhibitors of signaling pathways (Chambers et al., 2009)—a different and somewhat simpler strategy than methods that rely on forced gene expression (e.g., see PRF related news story).
One nagging problem remained: It took a month in culture for stem cells to differentiate into neurons. In the new study, Chambers and his colleagues set out to speed that process. They found that adding three more compounds—an inhibitor of key growth factor receptors, a kinase inhibitor that turns on Wnt signaling, and an inhibitor of Notch signaling—steered stem cells to a neuronal fate within 10 days, as judged by expression of postmitotic neuron marker proteins.
The cells were neurons, but what kind? Most expressed markers of sensory neurons, the researchers found, and more than 60 percent expressed the nerve growth factor receptor TrkA, suggesting a nociceptive lineage. Analysis of gene expression over time showed that the cells developed through a neural crest intermediate, the pathway followed by dorsal root ganglion (DRG) nociceptors. By day 15, cells maintained in medium containing neurotrophic factors expressed genes for ion channels involved in pain sensation, including the voltage-gated sodium channels Nav1.7 and Nav1.8, the purinergic receptor P2X3, and transient receptor potential channels TRPV1 and TRPM8. At one month, cell bodies organized into ganglia-like clusters that expressed the neuropeptides Substance P and calcitonin gene-related peptide (CGRP).
Gene expression is not the same thing as neural activity, so the Sloan-Kettering team, with collaborators at the Pfizer unit Neusentis in Great Abington, Cambridge, UK, went on to characterize the cells’ electrical properties. Their studies focused on the function of Nav1.8, a voltage-gated sodium channel that is expressed in most nociceptors and contributes to generation and propagation of action potentials. The researchers observed tetrodotoxin-resistant voltage-gated Na+ currents in one-quarter of cells that were blocked by an inhibitor of Nav1.8, consistent with functional Nav1.8 expression. The cells were also able to fire repetitive action potentials that were blocked by the Nav1.8 inhibitor.
In addition, the neurons responded to nociceptive stimuli from ATP, which mediates inflammatory pain via the P2X3 receptor (see PRF related Forum and PRF news story), and, to a lesser extent, to capsaicin, which activates TRPV1. A P2X3-specific ATP analog induced calcium currents in most cells, while capsaicin evoked a response in a small fraction (1-2 percent) of the cells. Chambers said he has since found that the capsaicin response increases as the cells are cultured longer. All in all, he said, “We get many of the functional characteristics of nociceptors in the dish.”
Much remains to be learned about the lab-made cells, but to the extent that they recapitulate the activity of bona fide nociceptors, they could provide a much-needed source of human cells for studying pain physiology and testing new drugs. Importantly, while most of the current work involved neurons derived from embryonic stem cells, the researchers showed that iPS cells from fibroblasts could also be induced to express sensory neuron markers. That means “we can now model patient-specific diseases, to understand more about the mechanism or look for drugs,” said Chambers.
In vivo, nociceptors are notoriously diverse, which means that the goal of reprogramming is not so much to arrive at a perfect, prototypic nociceptor, but rather to draw a detailed map for manipulating cell fate. TRPV1 expression, for example, is highly variable in nociceptor populations, and is thought to mark functional subtypes (see PRF related news story). Chambers says his focus now is on learning how to control channel expression and other developmental changes to generate cells that precisely model specific nociceptor subsets. “Can we, for instance, add another small molecule to get higher expression of TRPV1?” he asked. “Once you can control what subtype you get, then things become very interesting.”
Top image: Neurons from human pluripotent stem cells arrange into ganglia-like clusters. Credit: S. Chambers.
Comments
Brian Wainger, Massachusetts General Hospital
Stuart Chambers’ article makes two very important contributions. First, the authors have markedly expedited the process of differentiating neurons from human stem cells. Second, for the first time, the scientists in Lorenz Studer’s group have derived neurons with many of the quintessential properties of pain-sensing neurons. Their technique may be very important in the future for disease modeling of human painful conditions.