Function of the urethral tuft cell deciphered by genetic models allowing cell type-specific activation

Abstract

Brush cells (synonym: tuft cells) are found in the epithelium of the mammalian respiratory, gastrointestinal and urinary tract. They are sensors for hazardous substances, and they release acetylcholine and other mediators (e.g. cysteinyl-leukotrienes and interleukin-25), which activate various cell types including immune cells and nearby sensory nerve fibers. In nose and trachea, these nerve fibers then release the neuropeptide substance P (SP), leading to neurogenic inflammation, a local defense reaction. In the urethra, their stimulation evokes micturition, and previous work of our laboratory using a broadly acting compound also gave first indication of a link of urethral tuft cells (UTC) to neurogenic inflammation. Based on this initial evidence, the present study aims to further investigate their role in inducing neurogenic inflammation by a) establishing a model for the selective activation of brush cells, and b) utilizing a model of brush cell-deficient mice. We used 1) a chemogenetic model, in which brush cells carry a DREADD (designer receptor exclusively activated by designer drugs) 2) an optogenetic approach, using a mouse line (Chat-ChR2-EYFP) expressing a fusion protein of the blue light-sensitive ion channel channelrhodopsin-2 (ChR2) and enhanced yellow fluorescent protein (EYFP) expressed under the control of the choline acetyltransferase (Chat) promoter, and 3) mice lacking UTC due to genetic deletion of the transcription factor Pou2f3. Explanted urethrae from these mouse models were exposed to appropriate stimuli (CNO [clozapine N-oxide] for DREADD, blue LED light for ChR2, denatonium as bitter compound, and capsaicin as UTC-independent activator of nociceptors), and the supernatant was analyzed for content of SP by ELISA. The chemogenetic model proved not to be suitable, since CNO (60 μM) caused SP release also from DREADD control urethrae, likely due to leaky expression of the transgene. In the optogenetic model, UTC-specific expression of the transgene was validated by immunofluorescence with antibodies against the UTC marker DCLK1 (double-cortin like kinase 1) (90% colocalization, 47/52 cells, from 5 animals) and against neuropeptides of sensory nerve fibers (no colocalization). Blue LED light evoked SP release only from urethrae of transgenic but not of control mice. This was fully sensitive to mecamylamine, a general nicotinic acetylcholine receptor inhibitor. The bitter tasting compound denatonium (5 mM) evoked SP release from urethrae of wildtype, but not from urethrae lacking UTC (Pou2f3-/-), whereas the direct nerve fiber activator capsaicin (1 mM) was equally potent in both strains. The data validate Chat-ChR2-EYFP mice, but not Trpm5-DREADD mice, as a suitable model for studying UTC function. This model and the use of the bitter tastant denatonium in conjunction with UTC-deficient mice (Pou2f3-/-) showed that direct stimulation of UTC leads to the release of neuropeptides (SP) from sensory nerve fibers, most probably through the release of acetylcholine from the activated tuft cells acting on nicotinic receptors on nerve fibers. Thus, UTCs not only trigger long-distance reflexes involving the bladder but also evoke neurogenic inflammation, representing a local defense reaction.