Guardian crypt-base-goblet-cells protect the human colonic stem cell niche by triggering cholinergic calcium signal-dependent MUC2 secretion and luminal flushing

Mucus secreting goblet cells play a vital role in the maintenance of tissue homeostasis. Here we report the discovery of an enigmatic mechanism for the generation of calcium signals that couple cholinergic input to secretion of hydrated mucus in the human colonic stem cell niche. Mechanistic insights for this study were derived from native human colonic crypts and crypt-like organoids expressing MUC2-mNEON. Importantly, we demonstrate that the human colonic stem cell niche is also a cholinergic niche, and that activation of muscarinic receptors initiates calcium signals at the apical pole of intestinal stem cells and neighbouring crypt-base-goblet-cells. The calcium signal ‘trigger zone’ is defined by a microdomain of juxtaposed calcium stores expressing TPC1 and InsP3R3 calcium channels. Co-activation of TPC1 and InsP3R3 is required for generation of cholinergic calcium signals and downstream secretion of hydrated mucus, which culminates in the flushing of the colonic stem cell niche.


INTRODUCTION
The mucosal surfaces in our body fulfil a diverse array of physiological functions under challenging conditions. A notable example is the lining of the human gut. In addition to facilitating digestion and absorption of nutrients and fluid, the intestinal epithelium forms a vital selective barrier between the mucosal immune system and a barrage of microorganisms, ligands and antigens derived from the hostile gut lumen. Preservation of barrier function is underpinned by rapid stem cell-driven tissue renewal, and by secretion of a protective mucus layer 1 . Compromised barrier function undermines tissue homeostasis and is associated with an increased risk of inflammatory bowel disease and colon cancer [2][3][4] . Recent advances have provided important insights into the maintenance of epithelial tissue homeostasis and regulation of mucus production by surface goblet cells 2,5,6 . However, much less is known about how stem cells and their immediate progenitors function to protect the microenvironment of the stem cell niche from gut luminal contents.
The large intestinal stem cell niche is located at the base of epithelial invaginations called colonic crypts. A population of LGR5 + stem cells [7][8][9] self-renew on a daily basis 10 and give rise to progenitor cells that migrate along the crypt-axis and differentiate into absorptive colonocytes or secretory tuft cells, enteroendocrine cells, or goblet cells 11,12 . The intestinal goblet cell population has recently been shown to be heterogenous with diverse functional features 6,11 . Sentinel and intercryptal goblet cell subtypes are strategically placed at the crypt opening and within the surface epithelium, respectively 6,13 . Together with crypt-resident goblet cells, they contribute to the formation of two mucus layers comprising an adherent inner layer that is devoid of bacteria and an outer loose layer 5 , which protect the underlying surface epithelial cells 6 .
The architecture of the colonic crypt adds another dimension to stem cell security by providing a safe harbour for stem cells at the crypt-base, where they are distanced from harmful microbial metabolites and luminal mutagens 14,15 . Approximately 30-40% of cells that form the axis between the crypt-base and surface epithelium are goblet cells 8 . In the upper region of the crypt, goblet cells are more differentiated, whilst those in the lower region of the crypt still hold proliferative potential, intermingle with stem cells at the crypt-base, and secrete antimicrobial peptides 6,8,11 . Mucus secretion by cryptal goblet cells is accompanied by fluid secretion to generate hydrated mucus that flushes the crypt lumen of its contents to maintain a sterile microenvironment 16,17 . These observations beg the question as to how the interrelated physiological processes of mucus and fluid secretion are functionally coupled in the stem cell niche.
An integral component of the gastrointestinal stem cell niche is the enteric nervous system [18][19][20] . Cholinergic input along the crypt-axis has been implicated in the regulation of both mucus and fluid secretion 16,21 , but the mechanism of excitation-secretion coupling is not understood. Previous work in our laboratory has demonstrated initiation of cholinergic calcium signals at the base of human colonic crypts 9,22 but the cellular origins, molecular basis and role in excitation-secretion coupling in the intestinal stem cell niche are not known. Here we utilise native human colonic crypts and MUC2::mNEON crypt-like organoids 23 to address this issue. TPC1 and InsP3R3 cooperation is required for the generation of cholinergic calcium signals in neighbouring colonic stem cells and crypt-base-goblet cells, and also for calcium-dependent luminal secretion of hydrated mucus that flushes the contents of the stem cell niche. Unstimulated GCs exhibit characteristic apical 'bulbs' of MUC2-labelling. The kinetics of GC MUC2 depletion peak at 10-15 min (Figs 3c,d and Supplementary Figs 3a,b). The dependence of GC MUC2 secretion on elevated intracellular calcium levels was examined next. Cholinergic calcium signals were not affected by acute removal of extracellular calcium (Fig. 3e), but could be buffered by pre-incubation with BAPTA-AM (Fig. 3f), which also suppressed CCh-induced GC depletion of MUC2 (Fig. 3g,h). Calcium has been implicated in luminal stimulation of MUC2 secretion by differentiated GCs in the mouse gut 13 and organoids 17 but the origins and mechanism for the calcium signal is not yet fully understood.
Co-operation between endosomal and ER calcium stores is a requirement for cholinergic calcium signal generation in the colonic ISC niche. It is clear that the cholinergic calcium signal originates from an intracellular calcium store (Fig. 3e). The cellular location of endolysosomal and endoplasmic reticular (ER) calcium stores in the stem cell niche was visualised by lysotracker red fluorescence and KDEL IF labelling, respectively. Puncta exhibiting intense Lysotracker fluorescence were prominent at the apical pole and diminished in intensity towards the basal pole of cells in the ISC niche (Fig. 4a). Conversely, IF labelling of the ER marker KDEL was concentrated around the basal nuclei and extended toward the apical pole (Fig. 4a). We next demonstrated the expression of candidate intracellular calcium channels InsP3Rs1-3, RyRs1-3 and TPCs1&2 by RT-PCR analysis of mRNA derived from isolated epithelial crypts (Fig. 4b), the native mucosa and crypt-like colonic organoids (Supplementary Figs 4e,f). IF labelling for InsP3R2, RyR1-3 and TPC2 predominated at the basal pole of crypt-base cells (Fig. 4c,d), whereas InsP3R1 IF labelling extended from the basal pole toward the apical pole (Fig. 4c,d). Strikingly, TPC1 IF co-labelled discrete Rab11 + endosomes that were juxtaposed to InsP3R3 + puncta situated beneath the cellular apical membrane. These so called 'microdomains' were present within cells of slender ISC-like morphology and plump cells of GC-like morphology. TPC2 IF labelling exhibited partial colocalisation at the basal pole with Rab5 + puncta (Fig. 4d). The apical and basal location of TPC1 and TPC2, respectively, was confirmed in isolated single GCs (Fig. 4f). Similar expression patterns for intracellular calcium channels were exhibited by crypt-like organoids (Supplementary Fig. 4a-d). Of note, CD38, an enzyme that can modulate NAADP production (a renowned activator of TPCs) 26 , also co-localises with Rab11 + puncta at the apical pole of MUC2 + GCs and neighbouring slender ISCs (Fig. 4e). We next assessed the roles of ER and endolysosomal calcium stores for generation of cholinergic calcium signals in the stem cell niche. Depletion of ER stores by the SERCA pump inhibitor cyclopiazonic acid (CPA), in the absence of extracellular calcium, markedly suppressed the cholinergic calcium response (Fig 5a). A predominant role for mobilisation from ER stores is supported by sensitivity of the cholinergic calcium response to phospholipase C inhibition (U73122 27 , Fig. 5b) and InsP3R inhibition by caffeine 28,29 (Fig. 5c), but not by Xestospongin C 30 nor 2-APB 28 ( Supplementary Fig. 5a,f). The cholinergic calcium response was also diminished by the RyR blockers dantrolene and ryanodine (Fig. 5d) 31 . Endolysosomes are acidic calcium stores and their calcium load can be dissipated by blocking the proton V-ATPase 32 . The cholinergic calcium response was abrogated by pre-treatment with bafilomycin (Fig. 5e). Calcium release from endolysosomes is commonly mediated by two pore channels, TPC1 and TPC2 32 . TPC channel blockers Ned-19 33 , tetrandrine 34 and diltiazem 34,35 each reduced the amplitude of the cholinergic calcium response (Fig. 5e). TPC channel blockers also abolished the residual response to CCh following CPA-mediated depletion of the ER in the absence of extracellular calcium (see shaded aspects in Figs 5a,f). Calcium responses evoked by TPC agonists were also blocked by TPC inhibitors (Supplementary Fig. 5b).
Cholinergic calcium signals couple mucus secretion with fluid secretion in the human colonic stem cell niche. Pre-treatment of cultured human colonic crypts with intracellular calcium channel inhibitors caffeine, dantrolene, Ned-19 (Fig. 6a,b,c) and tetrandrine (Supplementary 6b) suppressed CCh-induced MUC2 depletion of GCs, as did inhibition of M3AChRs with 4DAMP ( Supplementary Fig. 6c). The InsP3R inhibitor 2-APB, which did not abrogate CCh-induced MUC2 depletion, did not suppress CCh-induced MUC2 depletion ( Supplementary Figs 5a & 6a). Having utilised the MUC2 IF depletion assay to forge functional links between calcium mobilisation and mucus secretion, we next generated MUC2::mNEON human colonic crypt-like organoids to investigate the impact of the cholinergic calcium signalling pathway on coupling mucus and fluid secretion to flush the stem cell niche ( Supplementary Fig. 7a-d) 23,36 . MUC2::mNEON heterozygous crypt-like organoids displayed MUC2-mNEON fluorescence that exhibited a granular appearance at the apical pole and which co-localised with MUC2 immunofluorescence (Fig. Supplementary Fig. 7d 9 ; the cross-sectional area of the crypt lumen increased in size and the apical-to-basal dimension exhibited a secretory volume decrease (Fig. 7e,f) 9 . An increase in luminal MUC2-mNEON fluorescence (Fig. 7c) was accompanied by a pronounced reduction in MUC2-mNEON 'theca' volume ( Fig. 7f) consistent with expulsion of MUC2-mNEON into the crypt lumen. Preincubation with the TPC inhibitor Ned-19 (Fig. 7g) or the InsP3R inhibitor caffeine (Fig. 7h) suppressed all these cholinergic-induced effects. Finally, we sought to determine the fate of mucus expelled into the stem cell niche. Luminal flow was visualised by pre-incubating MUC2-mNEON organoids with a plasma membrane fluorescent dye (Fig. 7i). On stimulation with CCh the number of fluorescent particles in the lumen increased and their position within the lumen was tracked with respect to time ( Fig. 7i; Supplementary Movie 4). Luminal particulate migrated in a collective stream away from the crypt-base towards the opening of the lumen ( Fig. 7i; Supplementary Movie 4). The maximum particle velocity in the central lumen was used to compute the laminar flow rate (Fig. 7j) 16 . By contrast, under control conditions, the luminal particulate was minimal and relatively static. To assess a contribution of fluid secretion to cholinergic-stimulated luminal flushing, closed organoids were selected for organoid swelling assays 37 . CCh stimulation increased organoid luminal cross-sectional area over a period of 2 hours; this was inhibited by Clotrimazole, a blocker of calcium-dependent K + channels and an inhibitor of Clsecretion, the main driver for fluid secretion in the gut ( Supplementary Fig. 8a) 38 . Similarly, suppression of cholinergic calcium signal generation by inhibition of TPCs (with tetrandrine or Ned-19), InsP3Rs (with caffeine), or RyRs (with dantrolene) also abrogated CCh-induced organoid swelling, an established proxy for intestinal fluid secretion ( Fig. 8 and supplementary Fig. 8) 37 .

DISCUSSION
A major challenge in the field of human gut epithelial biology is to understand the signalling pathways that coordinate diverse physiological processes to maintain intestinal tissue homeostasis. Here we report the discovery of cholinergic calcium signals that couple the secretion of mucus and fluid to flush luminal contents from the microenvironment of the stem cell niche. Importantly, these outcomes have been derived from a combination of human tissue model systems that preserve tissue topology, polarity and cellular diversity, and from complementary use of reporters that enable the mechanism of calcium signals and cellular secretion to be investigated in space and time.
The concept of a cholinergic niche at the base of human intestinal crypts (see Fig. 1i for schema) was envisaged more than 20 years ago with the discovery that cholinergic input stimulated calcium signals at the base of mouse small intestinal crypts 39 and rat colonic crypts 22 . This was subsequently confirmed in the human colonic epithelium 9,25 . Our current finding that neuronal and non-neuronal sources of acetylcholine (i.e. cholinergic neurons and cholinergic tuft cells) are concentrated at the human colonic crypt base is consistent with recent observations made in the murine gastric mucosa 19 and small intestine 20,40 . We also show that human colonic epithelium M3AChR expression predominates in stem cells and goblet cells in the crypt base. Similarly, M3AChRs have been demonstrated to be expressed in mouse small intestinal Paneth cells and stem cells at the crypt base 20,41 . In accordance with expression by intestinal stem cells, a role for M3AChRs in mucosal regeneration 20,40 and gastrointestinal tumourigenesis 19,42,43 has been proposed. Taken together with our current study, these observations suggest that the cholinergic niche simultaneously regulates both stem cell-driven tissue renewal and antimicrobial defence by secretion of mucus and fluid, the two main contributors to intestinal barrier function. The current study highlights a long suspected fundamental role for calcium in coupling these secretory events and moreover provides a long sort after mechanism for signal generation.
M3AChRs are coupled to cholinergic calcium signals that emanate from a sub-apical microdomain in ISCs and crypt-base-GCs. Juxtaposition of endosomes positive for TPC1 and InsP3R3 positive puncta and the pharmacological profile of intracellular calcium channel inhibitors favour a cooperative cross coupling model for the cholinergic calcium trigger zone 44 . Caffeine was the most potent InsP3R inhibitor of cholinergic calcium signals and has been shown block InsP3R3 29 . Xestospongin C and 2-APB were without effect on the cholinergic response but do block purinergic calcium responses in human colonic crypts (Lee and Williams, unpublished observations). Subapical expression of InsP3R3 has also been demonstrated in cholangiocytes and implicated in bile secretion in health and disease 45 . TPC1 was present in the calcium trigger zone in the intestinal stem cell niche and was accompanied by expression of CD38, which has been implicated in modulation of NAADP levels 26 , an activator of TPCs 46 . The expression of TPC1 in acidic Rab11 recycling endosomes has been documented previously 47 . It is not yet clear whether TPC1 endowed acidic calcium stores prime the neighbouring InsP3R3-positive (ER) calcium stores or vice versa. There is much interest in calcium signalling microdomains and membrane contact sites between different organelles as a means of calcium signal propagation and coordinating inter-organelle cell biology and metabolism 48 . Given the limited diffusion coefficient of cytosolic calcium, and according to our immunolocalisation studies, it would appear that globalisation of the calcium signal from the apical pole to the basal pole of cells within the intestinal stem cell niche is mediated by regenerative calcium-induced-calcium release by InsP3R1,2 and/or RYR1,2,3; TPC2 may also play a role. Accordingly, in the current study, RYR channel blockers suppressed the amplitude of the cholinergic calcium response. The relative contribution of these calcium channel subtypes is under investigation, as is the wider physiological significance.
The cholinergic calcium signal trigger zone in the apical pole of crypt-base goblet cells is located at the expected site of mucus granule exocytosis 21 . Complementary MUC2 immunolabelling studies and real-time imaging of MUC2-mNEON secretion demonstrated that cholinergic calcium signals promoted calcium-dependent exocytosis of MUC2 granules at the apical membrane of crypt-base-goblet-cells in the human colonic stem cell niche (see Fig.8d for schema). Exocytosis of secretory granules has previously been described as being restricted by the actin cytoskeleton but can be liberated by oscillations in cytosolic calcium 49 . Elevated calcium oscillations have been shown to promote dissociation of calcium sensors such as KChIP3 from MUC5AC granules thereby triggering calcium-dependent synaptotagmin-driven formation of a SNARE complex and granule fusion to the plasma membrane [50][51][52] . Acidic endolysosomal calcium stores have been implicated in redox stimulated, calcium-mediated mucus granule exocytosis 17 , although the nature of the intracellular calcium release channels was not investigated. It is likely that aspects of the calcium signalosome described by this study of the stem cell niche will be applicable to regulation of calcium-dependent mucus exocytosis from differentiated cryptal and intercryptal goblet cells located at the crypt opening and surface of the gut 6,13,53 . Interestingly, zymogen granules in pancreatic acinar cells 54 and cytolytic T cells 55 not only display calcium signalling microdomains and membrane contact sites between calcium storage organelles, but the zymogen and cytolytic granules serve as a calcium store that regulates their own exocytosis. This remains to be proven to be the case in intestinal mucus granules.
We also demonstrate that cholinergic calcium signals couple mucus secretion to fluid secretion. Cholinergic-induced fluid secretion is highly regulated by calcium and transient in nature, probably to flush the crypt lumen with minimal fluid loss 9,56 . Not surprisingly, organoid swelling assays exhibited a moderate size increase. The molecular machinery for calcium dependent chloride-driven fluid secretion is expressed in colonic crypt base cells, including NKCC1 9 on basolateral membranes and TMEM16A/ANO1 57 on apical membranes. However, closer inspection of cell type specific labelling of participating ion channels and transporters, and modes of intercellular calcium signalling will be required to confirm whether the cholinergic input stimulates GC mucus granule exocytosis and fluid secretion in the same cell or neighbouring cells, and whether it be in a coordinated or independent manner. In conclusion, this study demonstrates for the first time that input from a cholinergic niche elicits calcium signals in ISCs and GCs that are coupled to the secretion of mucus and fluid which flushes the luminal microenvironment of the colonic stem cell niche. The platform utilised in this current study provides a tractable system to investigate the mechanisms by which calcium signals couple to mucus granule exocytosis and fluid secretion, and to study the status and function of this pathway in human health and disease.

MATERIAL AND METHODS
Human colorectal tissue samples. This research study was performed in accordance with a favourable ethical opinion by the Faculty of Medicine and Health Sciences Research Ethics Committee (University of East Anglia; ref. no. 013/2014 -62 HT). Governance of human tissue procurement for this research project was overseen by the Norwich Research Park Biorepository (IRAS No. 130478). Tissue samples were obtained by consultant histopathologists from the healthy mucosa of patients undergoing surgical resection for colorectal cancer.
Human colonic crypt isolation and culture. Colonic crypts were isolated as described previously 8,9 . In short, fresh mucosal tissue samples were collected in ice cold PBS, transported to the laboratory, and incubated in Hepes-buffered saline (HBS): (mM) NaCl 140, KCl 5, Hepes (N-2-hydroxyethylpiperazine-N2-ethanesulphonic acid) 10, d-glucose 5.5, Na 2 HPO 4 1, MgCl 2 0.5, CaCl 2 1, and placed in HBS, which was devoid of both Ca 2+ and Mg 2+ , and supplemented with EDTA (diaminoethanetetraacetic acid disodium salt) (1mM), for 1 h at room temperature. Crypts were liberated by serial rounds of vigorous shaking, crypt sedimentation and collection. Sedimented crypts were collected and mixed in Matrigel and a 20 µl droplet containing 50-100 crypts was placed onto no. 0 glass coverslips (VWR) contained within a 12 well plate. After polymerisation at 37˚C for 5-10 mins, crypts were flooded with 0.5 mls of human colonic crypt culture medium (hCCCM), a variant of that described previously for human intestinal stem cell/organoid culture 58 : advanced F12/DMEM containing B27, N2, n-acetylcysteine (1 mM), Hepes (10 mM), penicillin/ streptomycin (100 U/ml), L-Glutamine (2mM), Wnt-3A (100 ng/ml), IGF-1 (50 ng/ml), Noggin (100 ng/ml) or Gremlin-1 (200 ng/ml), RSPO-1 (500 ng/ml), and the ALK 4/5/7 inhibitor A83-01 (0.5 µM). hCCCM was changed every two days and was modified further according to the stated experimental conditions. Human organoid culture. Human colonic organoids were obtained after the passage of colonic crypts grown for 7 days. The samples were detached from the bottom of the plate by manually breaking the Matrigel-containing crypts and these were mechanically dissociated into smaller fragments using a pipette. The suspension containing the crypt fragments was transferred into centrifuge tubes and pelleted at 4˚C. Following the removal of the supernatant containing cell debris, crypt fragments were resuspended in hCCCM and the final pellet was embedded in Matrigel, left to polymerise, and the organoids were flooded with hCCCM. Organoids were cultured at 37˚C and 5% humidified CO2, fed every 3 days and passaged every 5 to 7 days.

Generation of single cells.
Single intestinal epithelial cells were generated from crypt or organoid culture. Briefly, organoid fragments collected from 3 wells of a 24 well plate were incubated with TrypLE express cell dissociation enzyme (ThermoFisher) supplemented with Y-27632 (10 µM, Sigma) for 15 minutes at 37˚C with frequent pipetting. Single cells were washed 3 times and filtered through a 20 µm filter before seeding in Matrigel, left to polymerise and flooded with hCCCM containing Y-27632. Single cells were cultured at 37˚C in 5% humidified CO 2 .
Whole mount immunohistochemistry. On day 1 of culture following embedding in Matrigel, cultured-crypts, organoids or single cells were fixed with 4% PFA for 1 hour or fixed in a methanol:acetic acid solution (3:1) for 5 minutes, and permeabilised with either SDS (1%) or Triton X-100 (0.5% w/v PBS, 30 min). Non-specific binding sites were blocked with 10% donkey serum and 1% bovine serum albumin for 2 h and washed with PBS. Crypts or organoids were incubated with primary antibodies (1:100-200 dilution) overnight at 4˚C. Immunolabelling was visualised by using an appropriate combination of species-specific Alexafluor-conjugated secondary antibodies (488, 568, and 647 nm) raised in donkey (Invitrogen). Crypts or organoids were mounted on glass slides with Vectashield containing DAPI (Vector labs) or SYTOX Blue (Invitrogen).
For the processing of native human fixed sections, a small fraction of the colorectal tissue sample was immediately fixed with 4% PFA for 2 hours at ice cold temperature after the surgical resection. The samples were embedded in OCT compound and frozen on liquid nitrogen in isopentane. Sections of 8 µm thickness were cut using a cryostat and processed for immunohistochemistry as described previously.
Goblet cell depletion assay. Human colonic crypts or organoids on day 1 of culture were pre-incubated at 37 o C with relevant pharmacological agents and/or stimulated with CCh (10 μM) prior to fixation in a methanol:acetic acid solution (3:1) for 5 minutes. Samples were processed for immunohistochemistry as described previously and incubated with primary antibodies against MUC2 and E-cadherin. Images were acquired using laser scanning confocal microscopy (Zeiss 510 META or Zeiss 980 Airyscan) with a ×63 (1.4 numerical aperture) lens. Goblet cell mucus depletion was quantified using ImageJ (NIH) by measuring the fluorescence intensity of MUC2-positive goblet cells' theca of the goblet cells located at the base of the crypt. A minimum of 2 replicates was measured with a minimum of 100 goblet cells quantified per condition.
Confocal microscopy. Following immunohistochemistry, whole-mounted microdissected (ie, native) and cultured crypts were visualised by laser scanning confocal microscopy (Zeiss 510 META or Zeiss 980 Airyscan). A ×63 (1.4 numerical aperture) objective was used to obtain confocal images of the longitudinal crypt-axis. Image stacks were taken at 1-3 μm intervals which allowed selection of precise focal planes. The same acquisition parameters were used prior to post-hoc comparison of immunolabelling fluorescence intensity. Image analysis was performed using ImageJ. Three dimensional images were rendered in Volocity (Improvision).

RT-PCR.
RNA was extracted from freshly isolated colonic primary mucosa, freshly isolated crypts and from human colonic organoids after their 4 th passage, and isolated using the ReliaPrep™ RNA Miniprep System (Promega) according to the manufacturer's instructions. Complementary DNA was generated using 500 ng of total RNA, oligo-dT primers (Promega) and M-MLV Reverse Transcriptase kit (Thermo Fisher Scientific). RT-PCR was performed using the G-Storm thermal cycler on 25 µl reaction samples containing forward and reverse primers (200 nM) listed in Table 1, dNTPs (200 µM, Promega), 0.04 U/μl of GOTaq® G2 DNA Polymerase (Promega), PCR buffer (Promega) and MgCl2 (2.5 mM). RT-PCR products were run on a 2.5% agarose gel and visualised by ethidium bromide staining.
For the analysis of the knock-in efficiency and genotyping of MUC2-mNEON organoids, genomic DNA was isolated using ReliaPrep™ gDNA Tissue Miniprep System (Promega) according to the manufacturer's instructions and was then amplified using 0.02 U/ μl of Q5® High-Fidelity DNA Polymerase (NEB) according to manufacturer's instructions with the primer sets listed in Supplementary Table 1.
Calcium imaging. Isolated colonic crypts or organoids were loaded at RT with Fura-2/AM (5 μm, 2h; Thermo Fisher Scientific) in HBS, to monitor Ca 2+ as previously described 9 . The loaded specimen was placed in an experimental chamber located on the stage of an inverted epifluorescence microscope (Nikon TE200) using a ×40 1.1 NA objective. Experimental solutions were administered via a two-way tap. The activity of the correspondent calcium channels was blocked using the following inhibitors: Caffeine ( For the analysis of calcium signal initiation at the cellular level, colonic organoids were loaded at RT with Fluo-8/AM (5 μM, 2h; Abcam) or Calbryte™ 630 AM (10 μM, 2h; AAT Bioquest) in HBS. To aid cellular loading, the working solution also contained 0.04% Pluronic® F127 (Sigma) and Probenecid (2.5 mM; AAT Bioquest). Samples were imaged using superresolution confocal microscopy using a Zeiss 980 Airyscan microscope equipped with a x40 1.3 NA objective. Images were noise subtracted using the average of the first 10 frames prior to addition of CCh using ImageJ's Times series analyser plug in and the changes in fluorescence intensity are presented as pseudocolour images.
Organoid swelling assay. Human colonic organoids embedded in Matrigel were preincubated on day 1 with hCCCM containing the relevant pharmacological agents at 37˚C and then stimulated with CCh (10 µM). The whole well containing organoids was live imaged for 2 hours after stimulation, using an inverted microscope (Nikon-Ti, 4x objective). The data was analysed using ImageJ through measurement of the diameter of the intraluminal membrane of the organoid (n≥15) at t=0h and t=2h and obtaining the swelling ratio.