2026 年 51 巻 2 号 p. 237-248
Pancreatic acinar cells (PACs) are specialized for the production and secretion of digestive enzymes. PACs are highly polarized, with rough endoplasmic reticulum (ER) located in the basal and lateral regions, while zymogen granules (ZG) are localized apically. Furthermore, a distinct ER compartment containing inositol 1,4,5-trisphosphate receptors (IP3Rs), which are Ca2+ channels, is positioned underneath the filamentous-actin and tight junctions in the apical region. This structurally polarized architecture underlies the localized elevation of Ca2+ levels in the apical region, which supports the efficient secretion of granules. In this report, we show that Jaw1, a regulatory protein of IP3Rs, is a new component of the distinct ER compartment in PACs, colocalizing with IP3R2 and IP3R3 rather than IP3R1. Importantly, PACs from Jaw1-deficient mice exhibited an accumulation of ZG and cell hypertrophy without disrupting both the arrangement of the distinct ER compartment and the localization of IP3Rs. These findings shed more light on the underlying molecular components of the distinct ER compartment in the apical region of PACs, offering insights into their physiological functions in the homeostatic control of ZG abundance in PACs.
Key words: cell polarity, ER, IP3Rs, Jaw1, pancreatic acinar cells
Graphical Abstract
The pancreas is a unique organ that has both endocrine and exocrine functions. The endocrine portion consists of the islets of Langerhans, which produce and secrete hormones such as insulin, glucagon, and somatostatin to regulate blood glucose levels. In the exocrine portion, acini secrete digestive enzymes, such as amylase, lipase, and trypsinogen, through the pancreatic duct to the duodenum of the small intestine.
Acini are composed of highly polarized acinar cells arranged around a central lumen. As shown in Supplementary Fig. S1, in pancreatic acinar cells (PACs), the rough endoplasmic reticulum (ER) responsible for protein synthesis is predominantly distributed in the basal and lateral regions, while zymogen granules (ZG) containing the digestive enzymes are stored in the apical region (Motta et al., 1997). In contrast to this basolateral ER, a distinct ER compartment enriched with inositol 1,4,5-trisphosphate receptors (IP3Rs), which are Ca2+ channels, is arranged underneath filamentous-actin and tight junctions in the apical region, thus in close proximity to the plasma membrane (PM) (Fallon et al., 1995; Fujimoto et al., 2011a; Lee et al., 1997; Lur et al., 2011; Nathanson et al., 1994; Yule et al., 1997). Tight junctions restrict the passage of digestive enzymes in the interstitial space, and Zonula Occludens-1 (ZO-1), a component of tight junctions, is connected to filamentous-actin (Fallon et al., 1995; Fanning et al., 1998, 2002). Furthermore, Ki-ras-induced actin-interacting protein (KRAP) functions as a bridge between IP3Rs and filamentous-actin (Fujimoto et al., 2007, 2011a; Thillaiappan et al., 2021). Upon G protein-coupled receptor (GPCR) stimulation by acetylcholine and cholecystokinin, inositol 1,4,5-trisphosphate (IP3) is generated, and its binding to IP3Rs triggers Ca2+ release (Cancela, 2001; Williams, 2019). Importantly, the specific apical localization of IP3Rs facilitates localized Ca2+ elevation in this region, which in turn drives efficient and directional secretion of ZG into the lumen (Kasai and Augustine, 1990; Lee et al., 1997; Orabi et al., 2012). Disruption of this system has been reported to cause functional defects in PACs, accompanied by changes in the Ca2+ signal. For example, the loss of KRAP leads to cell hypertrophy and the accumulation of granules in PACs (Fujimoto et al., 2011b; Miyasaka et al., 2011; Thillaiappan et al., 2021). Furthermore, the loss of IP3R2 and IP3R3 causes a severe disorder in the secretion of ZG (Futatsugi et al., 2005; Orabi et al., 2012). Thus, the coordinated organization between the distinct ER compartment - which contains IP3Rs - together with KRAP, filamentous-actin, and tight junctions are crucial for the proper function and homeostasis of PACs.
Jaw1 is a membrane protein, also known as lymphoid restricted membrane protein (LRMP) or inositol 1,4,5-trisphosphate receptor associated 2 (IRAG2), localizing at the ER and outer nuclear membrane (Behrens et al., 1994; Horn et al., 2013; Kozono et al., 2018). Its expression has been reported to be found in specific cell types across tissues: taste cells on the tongue, tuft cells in the small intestine, immune cells, and PACs (Behrens et al., 1994; Chang et al., 2021; Prüschenk et al., 2021; Shindo et al., 2010). We previously reported that Jaw1 maintains the shape of the nucleus and Golgi apparatus (Kozono et al., 2018; Okumura et al., 2023). Additionally, we demonstrated that Jaw1 interacts with all IP3R subtypes and regulates the Ca2+ signaling augmentatively in response to GPCR stimulation (Kozono et al., 2023, 2025; Okumura et al., 2022). Importantly, cell-based experiments performed in the presence of thapsigargin, an inhibitor for sarco/endoplasmic reticulum Ca2+-ATPase, indicated that Jaw1 enhances Ca2+ release from the ER even under steady-state conditions independent of GPCR stimulation (Okumura et al., 2022). In other words, the expression of Jaw1 confers a higher basal level of cytoplasmic Ca2+. Consistent with our previous work, other groups reported that loss of Jaw1 in PACs results in reduced basal levels of cytoplasmic Ca2+ and basal amylase release, while GPCR stimulation restores the rise in the levels of cytoplasmic Ca2+ and amylase release (Prüschenk et al., 2021). Taken together, these observations suggest that the interaction between Jaw1 and IP3Rs in PACs is crucial for cytoplasmic Ca2+ homeostasis and digestive enzyme secretion at steady-state conditions, rather than in the active state. However, it is unknown how Jaw1 spatially cooperates with IP3Rs in PACs. Furthermore, it remains unclear how the loss of Jaw1 affects the cellular morphology and ultrastructure of PACs.
In this report, we first investigated whether or not Jaw1 is colocalized with IP3Rs at the apical region of PACs, focusing on their spatial arrangement with KRAP, filamentous-actin, and tight junctions. Furthermore, the effects of Jaw1-deficiency on this spatial arrangement in PACs were investigated. Finally, we carried out the ultrastructural analyses to assess morphological impacts of Jaw1 loss on PACs. On the basis of the abovesaid analyses, we suggest that Jaw1 is a new component localizing at the distinct ER compartment together with IP3Rs and KRAP, and contributes to the homeostatic control of the ZG abundance in PACs under steady-state conditions.
Mice aged older than 8 weeks were used for the experiments. The mice were provided ad libitum access to water and chow (CE-2) (CLEA Japan, Inc., Tokyo, Japan) and were kept under a 12:12 h light-dark cycle. All animal experiments were performed in accordance with the Institutional Guidelines for the Care and Use of Laboratory Animals in Research, with the approval of the local ethics committees at Tokyo University of Agriculture and Technology and National Institute of Global Health and Medicine.
Generation of gene-modified miceJaw1–/– mice were generated by CRISPR/Cas9-mediated genome editing technology as previously described (Nitta et al., 2015). Briefly, single-guide RNA (sgRNA) was designed to target the sequence (5'-TTA TGG AGA CTA CAC GAG TG-3') of the exon 7 of the Jaw1 gene. An sgRNA expression vector containing the target sequence under the control of a T7 promoter was synthesized (Eurofins Genomics, Ebersberg, Germany) and subjected to in vitro transcription using the MEGAshortscript T7 Transcription kit (#AM1354, Thermo Fisher Scientific, Waltham, MA, USA). The hCas9 coding sequence was excised from pX330 (#42230, Addgene, Watertown, MA, USA) and subcloned downstream of a T7 promoter. hCas9 mRNA was generated using the mMESSAGE mMACHINE T7 Transcription kit (#AM1344, Thermo Fisher Scientific) and subsequently polyadenylated using a Poly(A) Tailing Kit (#AM1350, Thermo Fisher Scientific). Both sgRNAs and hCas9 mRNAs were purified by phenol–chloroform–isoamyl alcohol extraction followed by ethanol precipitation. For microinjection, fertilized one-cell embryos derived from C57BL/6N (Japan SLC, Hamamatsu, Japan) were obtained by in vitro fertilization. hCas9 mRNA (200 ng/μL) and sgRNA (100 ng/μL) were co-injected into the cytoplasm of pronuclear-stage embryos. Injected embryos were then transferred into the oviducts of pseudopregnant ICR female mice (Japan SLC). The mice were backcrossed to C57BL/6J (CLEA Japan, Inc.) for at least five generations. For genotyping of mouse pups, ear specimens were collected, and genomic DNA was extracted by treatment with Proteinase K at 55°C overnight. The PCR product was then amplified from the extracted genomic DNA using the following primers: forward 5'-GTT TCA TGC TCT GAA GCA TCC-3', reverse 5'-ATA TAC ACA TGC GCA CAC ACG-3'. After confirmation of the amplification on an agarose gel electrophoresis, the sequence was verified by Sanger sequencing using the following primer: 5'-CCA CAG GGC TTC TGG CAT TA-3'.
Immunofluorescence stainingThe mice were euthanized, and the pancreas was collected. The dissected tissues were immediately wrapped with aluminum foil and frozen in liquid nitrogen. The tissues were then embedded in Tissue-Tek O.C.T. Compound (#4583, Sakura Finetek Japan Co., Ltd., Tokyo, Japan), and the frozen tissue blocks were sectioned at 20 μm thickness on a microtome (Tissue-Tek Polar D; Sakura Finetek Japan Co., Ltd.). The sections on slides were fixed with dry ice-cooled acetone for 10 min. After the fixation, the sections were completely dried. For the immunofluorescence, the sections were washed with PBS for 5 min twice and blocked with Block Ace (#UKB80, KAC Co., Ltd., Tokyo, Japan) for 1 h at room temperature. The sections were then incubated with antibodies diluted in 10-fold diluted Block Ace/PBS overnight at 4°C; anti-Jaw1 rat polyclonal antibody (produced in our laboratory previously described in Kozono et al., 2018) (1:200), anti-ZO-1 rat monoclonal antibody (1:400) (#sc-33725, Santa Cruz Biotechnology, Dallas, TX, USA), anti-IP3R1 rabbit polyclonal antibody (raised against the C-terminal region of human IP3R1 (LLGHPPHMNVNPQQPA) by a commercial vendor (SCRUM Inc., Tokyo, Japan)) (1:200), anti-IP3R2 mouse monoclonal antibody (#sc-398434, Santa Cruz Biotechnology) (1:400), anti-IP3R3 mouse monoclonal antibody (#610312, BD Biosciences, San Jose, CA, USA) (1:400), and anti-KRAP rabbit polyclonal antibody (#14157-1-AP, Proteintech, Rosemont, IL, USA) (1:400). Prior to incubation, the antibodies were labelled with the following dyes; FlexAble 2.0 CoraLite Plus 488 Antibody Labeling Kit for Rat Kappa Light Chain (#KFA621, Proteintech), FlexAble 2.0 CoraLite Plus 555 Antibody Labeling Kit for Rat Kappa Light Chain (#KFA622, Proteintech), FlexAble 2.0 CoraLite Plus 647 Antibody Labeling Kit for Rat Kappa Light Chain (#KFA623, Proteintech), FlexAble 2.0 CoraLite Plus 405 Antibody Labeling Kit for Rabbit IgG (#KFA506, Proteintech), FlexAble 2.0 CoraLite Plus 488 Antibody Labeling Kit for Rabbit IgG (#KFA501, Proteintech), FlexAble 2.0 CoraLite Plus 647 Antibody Labeling Kit for Rabbit IgG (#KFA503, Proteintech), FlexAble 2.0 CoraLite Plus 488 Antibody Labeling Kit for Mouse IgG2a (#KFA541, Proteintech), FlexAble 2.0 CoraLite Plus 488 Antibody Labeling Kit for Mouse IgG2b (#KFA561, Proteintech), and FlexAble 2.0 CoraLite Plus 555 Antibody Labeling Kit for Mouse IgG2b (#KFA562, Proteintech). The F-actin was stained with Rhodamine Phalloidin (#PHDR1, Cytoskeleton, Inc., Denver, CO, USA) (100 nM) diluted in 10-fold diluted Block Ace/PBS by incubating overnight at 4°C with antibodies. After incubation with the antibodies and Rhodamine Phalloidin, the sections were washed with PBS for 5 min three times, and the nuclei were stained with Hoechst33342 in PBS for 20 min at room temperature. After washing with PBS for 5 min three times, the sections were mounted with ProLong Diamond Antifade Mountant (#P36961, Thermo Fisher Scientific) and observed on the confocal laser scanning microscope (AXR; Nikon, Tokyo, Japan) (objective lens; Apo TIRF 100 × /1.49 Oil ∞/0.13-0.20 DIC N2). The excitation lasers at 405, 488, 561, and 640 nm were used, and the emission wavelengths at 429–474, 502–540, 571–605, and 662–737 nm were detected, respectively. For Pearson’s correlation coefficient analysis, the fluorescence images were analyzed using Fiji. The apical region (150 × 150 pixels) with puncta was randomly selected from each image. Pearson’s correlation coefficient between the two channels was then calculated using the Coloc 2 plugin.
SEM observationSample preparation using the reduced osmium-thiocarbohydrazide-osmium (rOTO) method for SEM observation was performed according to the previous report with some modifications (Ma et al., 2022). The mice were deeply anesthetized by inhalation of sevoflurane (#193-17791, FUJIFILM Wako Pure Chemical, Osaka, Japan) and intracardially perfused with 0.9% NaCl followed by 2% paraformaldehyde and 2.5% glutaraldehyde (#079-00533, FUJIFILM Wako Pure Chemical) in 100 mM cacodylate buffer (pH 7.4) (#194-04852, FUJIFILM Wako Pure Chemical) containing 3 mM CaCl2. The dissected pancreas was cut into segments (~1 mm3) and fixed with the above fixative solution overnight at 4°C. The tissues were washed with ice-cold 100 mM cacodylate buffer containing 3 mM CaCl2 for 10 min three times at 4°C. The tissues were then post-fixed with 1.5% OsO4 (#300, Nisshin EM, Tokyo, Japan) and 1.5% K3[Fe(CN)6] (#167-03722, FUJIFILM Wako Pure Chemical) in 100 mM cacodylate buffer containing 3 mM CaCl2 for 1 h at room temperature. After washing with ice-cold water for 10 min three times, the tissues were treated with 1% thiocarbohydrazide (#223220, Merck, Darmstadt, Germany) for 30 min at room temperature. The tissues were washed with ice-cold water for 10 min three times, followed by treatment with 1.5% OsO4 solution for 45 min at room temperature. After washing with ice-cold water for 10 min three times, the tissues were stained with EM stainer (#336, Nisshin EM) overnight at 4°C. The tissues were then washed with water for 10 min three times, followed by staining with Walton’s lead aspartate (Walton, 1979) for 1 h at 50°C. After washing with water warmed at 50°C for 10 min three times, the tissues were dehydrated in a graded ethanol series (#057-00451, FUJIFILM Wako Pure Chemical), substituted with propylene oxide (#165-05026, FUJIFILM Wako Pure Chemical) and Epon812 (#T024, TAAB, Aldermaston, UK), and polymerized at 37°C, 45°C, 60°C, 60°C for 24 h each. Ultrathin sections of 50 nm-thickness were prepared using an ultramicrotome (EM UC7; Leica Microsystems, Wetzlar, Germany) and placed on a silicon wafer. The sections were contrasted with an EM stainer (#336, Nisshin EM) and lead (#18-0875, Merck), and thereafter observed using SEM (SUPERSCAN SS-4000; Shimadzu Corporation, Kyoto, Japan) under the following conditions: accelerating voltage (2.5 kV), probe current (800 pA), WD = 6.00, vEM-BSE detector.
Statistical analysisThe collected data were analyzed and graphically presented using GraphPad Prism7 (GraphPad). For the statistics, the significance was determined by two-tailed unpaired Student’s t-test for Fig. 3 and Supplementary Fig. S4–S6, two-tailed unpaired Welch’s t-test for Fig. 4, and two-way ANOVA followed by Sidak’s multiple comparison test for Supplementary Fig. S2. The statistical significances were represented as follows: **P<0.01; ***P<0.001. n.s.: not significant.
A previous report has shown the expression of Jaw1 in PACs and its interaction with IP3Rs (Prüschenk et al., 2021), but it remains unknown how Jaw1 spatially regulates the IP3Rs. Therefore, we first investigated the localization of Jaw1 in PACs. Prior to analysis, we initially generated Jaw1 knockout (Jaw1–/–) mice using a CRISPR/Cas9-mediated genome editing strategy. Jaw1 interacts with IP3Rs via its coiled-coil domain and regulates their activity (Okumura et al., 2022). Therefore, sgRNA was designated to target the sequence upstream of the region coding for the functional domain, within exon 7 (Fig. 1A). Genotyping results showed that one base was inserted within the targeting site in Jaw1–/– mice allele, resulting to the introduction of a premature stop codon upstream of the functional domain due to frameshift (Fig. 1A–C). The body weight (Supplementary Fig. S2A), food intake (Supplementary Fig. S2B), and fertility in Jaw1–/– mice were comparable to those in Jaw1 wild-type (Jaw1+/+) mice. The expression and localization of Jaw1 in acinar cells were then investigated by immunohistochemistry using an anti-Jaw1 antibody against its coiled-coil domain. Confocal imaging showed that Jaw1 is localized at the apical region of PACs, where it appears as discrete puncta positioned in close proximity to the tight junction marker ZO-1 (Fig. 1D). Importantly, this signal was not detected in Jaw1–/– mice. Furthermore, the loss of band corresponding to Jaw1 in the whole tissue lysate from Jaw1–/– pancreas was confirmed by western blotting, validating the reactivity of anti-Jaw1 antibody and the successful generation of Jaw1–/– mice (Supplementary Fig. S3). Thus, these data indicate that Jaw1 is enriched in PACs and specifically localizes apically as puncta in close proximity to tight junctions underneath the PM.
The specific localization of Jaw1 as puncta at the apical region of PACs
(A) Schematic representation of the Jaw1 gene locus and CRISPR/Cas9-mediated genome editing strategy for the generation of Jaw1–/– mice. The sgRNA targeting the sequence within exon 7 encoding the Jaw1 gene was injected with hCas9 mRNA. The exons coding for the Jaw1 gene were shown in boxes with corresponding exon numbers. The areas colored blue and gray indicate the region coding for the coiled-coil domain and the transmembrane domain, respectively, functional domains of Jaw1. (B) The genotyping results show the sequences within exon 7 for Jaw1+/+ and Jaw1–/– mice. The sequences colored yellow and green indicate the sgRNA-targeting sequence and PAM sequence, respectively. One adenine, surrounded by a box, was inserted in the allele of Jaw1–/– mice, resulting to the premature stop codon in exon 9 (A). (C) Schematic representation of mouse Jaw1 and the amino acid sequences of Jaw1 in Jaw1+/+ and Jaw1–/– mice. The areas colored blue and gray indicate the coiled-coil domain and the transmembrane domain, respectively. The letters colored magenta in Jaw1 knockout indicate the newly appeared amino acids due to the frameshift. An anti-Jaw1 antibody against its coiled-coil domain was used in the following assay. (D) Representative images of immunofluorescence staining. Pancreatic tissue sections from Jaw1+/+ and Jaw1–/– mice (N = 4) were co-stained with FlexAble 2.0 CoraLite Plus 488–labelled anti-Jaw1 antibody (cyan), FlexAble 2.0 CoraLite Plus 647–labelled anti-ZO-1 antibody (magenta), and Hoechst33342 (blue). The enlarged images corresponding to the areas surrounded by dotted lines are shown at the bottom. Scale bars: 10 μm (original); 1 μm (enlarged). Arrowheads indicate the representative Jaw1+ puncta (white).
It has been reported that IP3Rs are localized at the apical region of PACs (Lur et al., 2011) and that all three IP3R subtypes (IP3R1–3) interact with Jaw1 (Okumura et al., 2022; Prüschenk et al., 2021). Therefore, we investigated whether Jaw1 colocalizes with each IP3R subtype at the apical region of PACs by immunohistochemistry. Confocal imaging showed that all three IP3R subtypes are localized as puncta in close proximity to tight junctions underneath the PM (Fig. 2A, C, E), consistent with a previous report (Lur et al., 2011). Importantly, most IP3R1-positive puncta did not colocalize with those positive for Jaw1 (Fig. 2A). In contrast, majority of the puncta positive for IP3R2 (Fig. 2C) or IP3R3 (Fig. 2E) completely colocalized with those positive for Jaw1. Consistent with these, the Pearson’s correlation coefficient between IP3R2 and Jaw1 (mean value: 0.86) (Fig. 2D) and IP3R3 and Jaw1 (mean value: 0.86) (Fig. 2F) were much higher than that between IP3R1 and Jaw1 (mean value: 0.26) (Fig. 2B). Furthermore, co-staining for Jaw1 and all three IP3R subtypes showed that almost all IP3R2- and IP3R3-positive puncta exclusively colocalize with Jaw1, whereas IP3R1-positive puncta do not (Fig. 2G). Importantly, the Pearson’s correlation coefficient between IP3R2 and IP3R3 (mean value: 0.80) was much higher than that between IP3R1 and IP3R2 (mean value: 0.15) and IP3R1 and IP3R3 (mean value: 0.15) (Fig. 2H). Thus, these data indicate that the distinct apical ER compartment in PACs is organized into at least two groups: one containing IP3R1 and another containing IP3R2 and IP3R3. Notably, Jaw1 preferentially colocalizes with the group positive for IP3R2 and IP3R3, rather than IP3R1.
The colocalization of Jaw1 with IP3R2 and IP3R3 as puncta at the apical region of PACs
(A, C, E, G) Representative images of immunofluorescence staining. Pancreatic tissue sections from Jaw1+/+ mice (N = 4) were co-stained with antibodies. The enlarged images corresponding to the areas surrounded by dotted lines are shown at the bottom. Scale bars: 10 μm (original); 1 μm (enlarged). (A, C, E) FlexAble 2.0 CoraLite Plus 488–labelled anti-IP3R1 antibody (cyan) (A), anti-IP3R2 antibody (cyan) (C), anti-IP3R3 antibody (cyan) (E), FlexAble 2.0 CoraLite Plus 555–labelled anti-Jaw1 antibody (magenta) (A, C, E), FlexAble 2.0 CoraLite Plus 647 labelled–anti-ZO-1 antibody (yellow) (A, C, E), and Hoechst33342 (blue) (A, C, E). Arrowheads indicate the representative IP3Rs+ Jaw1+ (white), IP3R1+ Jaw1– (cyan), and IP3R1– Jaw1+ puncta (magenta). (G) FlexAble 2.0 CoraLite Plus 405–labelled anti-IP3R1 antibody (green), FlexAble 2.0 CoraLite Plus 555–labelled anti-IP3R2 antibody (cyan), FlexAble 2.0 CoraLite Plus 488–labelled anti-IP3R3 antibody (magenta), FlexAble 2.0 CoraLite Plus 647–labelled anti-Jaw1 antibody (yellow). Arrowheads indicate the representative IP3R1– IP3R2+ IP3R3+ Jaw1+ (white) and IP3R1+ IP3R2– IP3R3– Jaw1– puncta (green). (B, D, F, H) Graphs showing the Pearson’s correlation coefficient between the two channels (IP3R1 and Jaw1) in (A), (IP3R2 and Jaw1) in (C), (IP3R3 and Jaw1) in (E), and (IP3R1 and IP3R2, IP3R1 and IP3R3, and IP3R2 and IP3R3) in (G). Three apical regions were selected per image, resulting in a total of 12 regions analyzed. Error bar shows ±SD.
KRAP, a linker between IP3Rs and filamentous-actin, is localized at the apical region of PACs (Fujimoto et al., 2007, 2011a; Thillaiappan et al., 2021). Importantly, it has been reported that KRAP preferentially has affinity and colocalizes with IP3R3, but not IP3R1, in PACs, whereas it has affinity and colocalizes with both subtypes in other tissues or cell lines (Fujimoto et al., 2011a). However, the details regarding how KRAP is spatially localized at the apical region of PACs remain unclear, due to a lack of analyses using high-resolution images. Therefore, we first investigated the spatial arrangement of KRAP, filamentous-actin, tight junctions, and IP3Rs at the apical region of PACs by immunohistochemistry. Confocal imaging revealed a previously unrecognized punctate distribution of KRAP in close proximity to filamentous-actin and tight junctions aligned underneath the PM (Fig. 3A). Furthermore, most puncta positive for IP3R1 did not colocalize with those positive for KRAP (Supplementary Fig. S4A), consistent with a previous report (Fujimoto et al., 2011a). In contrast, almost all puncta containing IP3R2 (Supplementary Fig. S5A) or IP3R3 (Supplementary Fig. S6A) completely colocalized with those containing KRAP. Consistent with these, the Pearson’s correlation coefficient between IP3R2 and KRAP (mean value: Jaw1+/+, 0.78) (Supplementary Fig. S5C) and IP3R3 and KRAP (mean value: Jaw1+/+, 0.82) (Supplementary Fig. S6C) were much higher than that between IP3R1 and KRAP (mean value: Jaw1+/+, 0.42) (Supplementary Fig. S4C). Thus, these data indicate that KRAP preferentially colocalizes with the IP3R2- and IP3R3-positive group of the distinct ER compartment, rather than the IP3R1 group, forming puncta aligned underneath filamentous-actin and tight junctions at the apical region of PACs.
The localization of KRAP as puncta at the apical region of PACs
(A, B) Representative images of immunofluorescence staining. Pancreatic tissue sections from Jaw1+/+ (A) and Jaw1–/– (B) mice (N = 4) were co-stained with FlexAble 2.0 CoraLite Plus 647–labelled anti-ZO-1 antibody (cyan), Rhodamine Phalloidin (magenta), FlexAble 2.0 CoraLite Plus 488–labelled anti-KRAP antibody (yellow), and Hoechst33342 (blue). The enlarged images corresponding to the areas surrounded by dotted lines are shown at the bottom. Scale bars: 10 μm (original); 1 μm (enlarged). Arrowheads indicate the representative KRAP+ puncta (white). (C) Graph showing the Pearson’s correlation coefficient between the two channels (KRAP and F-actin) in (A) and (B). Three apical regions were selected per image, resulting in a total of 12 regions analyzed per genotype. Error bar shows ±SD. Statistics: two-tailed unpaired Student’s t-test. n.s., not significant.
Next, we investigated whether or not Jaw1 controls the spatial arrangement of IP3Rs, KRAP, filamentous-actin, and tight junctions, at the apical region of PACs by immunohistochemistry. Confocal imaging showed that KRAP-positive puncta in the PACs of Jaw1–/– mice remain in close proximity to filamentous-actin and tight junctions aligned underneath the PM, comparable to those observed in Jaw1+/+ mice (Fig. 3A, B). Furthermore, there were no differences observed in the localization patterns of puncta positive for IP3R1 (Supplementary Fig. S4A, B), IP3R2 (Supplementary Fig. S5A, B), and IP3R3 (Supplementary Fig. S6A, B), including their colocalization patterns with KRAP, between Jaw1+/+ and Jaw1–/– mice. Consistent with these, between Jaw1+/+ and Jaw1–/– mice, there were no differences in the Pearson’s correlation coefficient between KRAP and filamentous-actin (mean value: Jaw1+/+, 0.80; Jaw1–/–, 0.80) (Fig. 3C), IP3R1 and KRAP (mean value: Jaw1+/+, 0.42; Jaw1–/–, 0.52) (Supplementary Fig. S4C), IP3R2 and KRAP (mean value: Jaw1+/+, 0.78; Jaw1–/–, 0.81) (Supplementary Fig. S5C), and IP3R3 and KRAP (mean value: Jaw1+/+, 0.82; Jaw1–/–, 0.81) (Supplementary Fig. S6C). Finally, we carried out ultrastructural analyses to investigate how the loss of Jaw1 morphologically impacts PACs. Tight junctions, appearing as electron-dense structures along the PM, as well as ER-PM contacts were similarly observed in both Jaw1+/+ and Jaw1–/– mice (Fig. 4A). Importantly, the number of granules per cell (mean value: Jaw1+/+, 75; Jaw1–/–, 111) (Fig. 4B) and cellular area (mean value: Jaw1+/+, 304 μm2; Jaw1–/–, 404 μm2) (Fig. 4C) were significantly increased in PACs of Jaw1–/– mice compared to those of Jaw1+/+ mice. Thus, these data indicate that the loss of Jaw1 causes the accumulation of ZG accompanied by cell hypertrophy in PACs, without disrupting the spatial organization of the distinct IP3R-containing ER compartment, filamentous-actin, and tight junctions at the apical region, and their close proximity to the PM.
The accumulation of ZG and cell hypertrophy in PACs due to the loss of Jaw1
(A) Representative electron micrographs of pancreas from Jaw1+/+ and Jaw1–/– mice (N = 3). The magnified images corresponding to the areas surrounded by dotted lines are shown on the right. Arrowheads indicate the representative ER-PM contact. Scale bars: 10 μm (left); 1 μm (right). (B, C) Graphs showing the number of granules per cell (B) and cellular area (C). The parameters were measured in PACs from Jaw1+/+ (71 cells) and Jaw1–/– (66 cells), collected from three mice per genotype (>20 cells each mouse). Error bar shows ±SE. Statistics: two-tailed unpaired Welch’s t-test. **, P<0.01; ***, P<0.001.
In this study, we uncovered that Jaw1 preferentially colocalizes with IP3R2 and IP3R3, but not with IP3R1, as punctate structures at the apical region of PACs. Furthermore, KRAP exhibits a similar punctate colocalization pattern with each IP3R subtype, identical to that observed for Jaw1. These data imply that the distinct ER compartment positive for IP3R2, IP3R3, Jaw1, and KRAP is segregated from the one positive for IP3R1. Importantly, the loss of Jaw1 in PACs led to the accumulation of ZG and cell hypertrophy, whereas the localization patterns of IP3Rs, KRAP, filamentous-actin, and tight junction were comparable between Jaw1+/+ and Jaw1–/– mice. On the basis of these data, we suggest that Jaw1 is a new component that colocalizes with IP3R2, IP3R3, and KRAP at the apical region of PACs and maintains the homeostasis of ZG abundance in PACs under steady-state conditions.
Ca2+-dependent exocytosis of ZG upon GPCR stimulation is mediated by IP3R2 and IP3R3, and their loss leads to extreme accumulation of ZG in PACs (Futatsugi et al., 2005; Orabi et al., 2012). Although our ultrastructural analysis clarified the accumulation of ZG in PACs from Jaw1–/– mice, the extent appears to be less than that of IP3R2–/– and/or IP3R3–/– mice. To date, we have reported that Jaw1 enhances the levels of cytoplasmic Ca2+ at both steady and active states, via the augmentative regulation of IP3R activity (Kozono et al., 2023, 2025; Okumura et al., 2022). Importantly, the elevation of cytoplasmic Ca2+ levels upon GPCR stimulation and the subsequent amylase secretion are comparable between isolated PACs from Jaw1+/+ and Jaw1–/– mice (Prüschenk et al., 2021). In contrast, according to their report, loss of Jaw1 in PACs reduces basal cytoplasmic Ca2+ levels, accompanied by decreased basal amylase secretion. Taken together, our findings suggest a model in which Jaw1-mediated augmentation of IP3R activity in PACs primarily functions maintaining basal cytoplasmic Ca2+ and digestive enzyme levels under steady-state conditions, rather than driving Ca2+-dependent exocytosis in the active state.
However, the molecular mechanism by which the reduced basal cytoplasmic Ca2+ levels due to the loss of Jaw1 leads to the slight accumulation of ZG. In contrast to the secretagogue-induced exocytosis of ZG, which requires elevated Ca2+ levels, basal secretion of digestive enzymes has been reported to occur via two alternative pathways: the constitutive-like pathway, and the minor-regulated pathway (Arvan and Castle, 1987; Castle and Castle, 1996; Hendricks et al., 1992; Huang et al., 2001; Messenger et al., 2013, 2014). These pathways diverge from immature secretory granules by budding and finally transports the digestive enzymes to the PM via endosomes (Messenger et al., 2013). Although it remains debated whether or not the basal Ca2+ levels directly control the activity of traffic-related molecules in these pathways in PACs, the slight accumulation of ZG in the PACs from Jaw1–/– mice might reflect a disturbance in these pathways. In other words, further investigation of digestive enzyme trafficking in Jaw1–/– mice would clarify the significant issue of whether or not basal Ca2+ levels regulate these pathways and contribute to the homeostatic control of the amounts of digestive enzymes in PACs.
It has been reported that all IP3Rs subtypes are localized at the apical region of PACs (Lee et al., 1997; Lur et al., 2011; Yule et al., 1997). Importantly, our data demonstrated that the distinct ER compartment is grouped into two distinct subcompartment: one positive for IP3R2, IP3R3, Jaw1, and KRAP, another positive for IP3R1. Consistent with this, previous co-immunoprecipitation and colocalization assays showed that KRAP coexists with IP3R3 but not IP3R1 in PACs, whereas it coexists with both IP3R1 and IP3R3 in other tissues or cell lines (Fujimoto et al., 2011a). Interestingly, PACs from IP3R2–/– and IP3R3–/– double knockout mice completely lacked GPCR-induced Ca2+ signals and Ca2+-dependent amylase exocytosis (Futatsugi et al., 2005), whereas IP3R2–/– or IP3R3–/– single knockout mice, retain partial or normal function (Futatsugi et al., 2005; Orabi et al., 2012). Thus, these findings indicate that the function of IP3R2 and IP3R3 in the secretion of digestive enzymes is redundant, whereas their functions are not compensated for by IP3R1. Taking these into consideration, it raises the possibility that the two distinct ER subcompartment might be functionally different: the IP3R2, IP3R3, Jaw1, and KRAP-positive region supporting both Ca2+-dependent exocytosis of digestive enzyme and homeostatic regulation of ZG abundance under steady-state conditions, and the IP3R1-positive region serving other cellular functions in PACs. Exploration of the differences between subtype-specific IP3R partners in PACs would provide clues to understand the mechanism by which the respective regions are formed and how their functions diverge. As a limitation in this study, the ultrastructural analysis showed the ER compartment being in close proximity to the PM; however, it remains unclear whether or not these ER compartments indeed contain the molecules described above. Therefore, correlative light and electron microscopy would be a strong tool to resolve the morphology and positions of these ER compartments in PACs, directly relating the localization of each IP3R subtype with its associated partners. As an additional limitation, the present study does not address whether Jaw1 directly regulates IP3R2- and IP3R3-mediated Ca2+ signaling in PACs, as we previously demonstrated in a HEK293 cell line (Okumura et al., 2022). Future high-resolution Ca2+ imaging focused on this distinct ER compartment would be required to clarify this point.
This work was supported by the Program on Open Innovation Platform with Enterprises, Research Institute and Academia (OPERA) from Japan Science and Technology Agency (JST) (JPMJOP1833 to AN), JST, the establishment of university fellowships towards the creation of science technology innovation (JPMJFS2111), and Grants-in-aid for Scientific Research from the Japan Society for the Promotion of Science (23K14511 to TK).
Conflict of Interest StatementKriengkamol Tantrakarn is an employee of Shimadzu Corporation. The author declares no potential non-financial conflicts of interest. All other authors declare no competing interests.
Data Availability StatementThe datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Author Contribution StatementConceptualization: AN, TK; Methodology: HS, KT, KN, TO, TK; Validation: TK; Formal analysis: TK; Investigation: HS, TK; Data curation: TK; Writing - original draft: KN, TO, TK; Visualization: TK; Writing - review & editing: MFB, KT, TT, AN; Resources: MFB, KT, KN, TO, YIK, TD; Supervision: YIK, TD, TT, AN, TK; Project administration: TK; Funding acquisition: AN, TK. All authors have read and agreed to the published version of the manuscript.
Ethics Approval and Consent to ParticipateThe animal experiments were conducted with approval from the local ethics committees at Tokyo University of Agriculture and Technology and National Institute of Global Health and Medicine.
Patient Consent for PublicationNot applicable.
We would like to express sincere gratitude to Tamami Honda from RIKEN and Yoshiyuki Kubota from National Institute for Physiological Sciences for the technical advice regarding the rOTO method. The analysis using the confocal laser scanning microscope was performed at Tokyo University of Agriculture and Technology for the Smart-Core-Facility Promotion Organization. Finally, we thank the lab members of the biological chemistry laboratory at Tokyo University of Agriculture and Technology for their technical assistance and valuable discussion.
endoplasmic reticulum
GPCRG protein-coupled receptor
IP3inositol 1,4,5-trisphosphate
IP3Rsinositol 1,4,5-trisphosphate receptors
IRAG2inositol 1,4,5-trisphosphate receptor associated 2
KRAPKi-ras-induced actin-interacting protein
LRMPlymphoid restricted membrane protein
PACspancreatic acinar cells
PMplasma membrane
rOTOreduced osmium-thiocarbohydrazide-osmium
sgRNAsingle-guide RNA
ZGzymogen granules
ZO-1Zonula Occludens-1
