IRAG2 Interacts with IP3-Receptor Types 1, 2, and 3 and Regulates Intracellular Ca2+ in Murine Pancreatic Acinar Cells

The inositol 1,4,5-triphosphate receptor-associated 2 (IRAG2) is also known as Jaw1 or lymphoid-restricted membrane protein (LRMP) and shares homology with the inositol 1,4,5-triphosphate receptor-associated cGMP kinase substrate 1 (IRAG1). IRAG1 interacts with inositol trisphosphate receptors (IP3 receptors /IP3R) via its coiled-coil domain and modulates Ca2+ release from intracellular stores. Due to the homology of IRAG1 and IRAG2, especially in its coiled-coil domain, it is possible that IRAG2 has similar interaction partners like IRAG1 and that IRAG2 also modulates intracellular Ca2+ signaling. In our study, we localized IRAG2 in pancreatic acinar cells of the exocrine pancreas, and we investigated the interaction of IRAG2 with IP3 receptors and its impact on intracellular Ca2+ signaling and exocrine pancreatic function, like amylase secretion. We detected the interaction of IRAG2 with different subtypes of IP3R and altered Ca2+ release in pancreatic acinar cells from mice lacking IRAG2. IRAG2 deficiency decreased basal levels of intracellular Ca2+, suggesting that IRAG2 leads to activation of IP3R under unstimulated basal conditions. Moreover, we observed that loss of IRAG2 impacts the secretion of amylase. Our data, therefore, suggest that IRAG2 modulates intracellular Ca2+ signaling, which regulates exocrine pancreatic function.


Introduction
Pancreatic acinar cells have been a model for the investigation of Ca 2+ signaling and enzymatic secretion over a long period. Changes in Ca 2+ signaling in these cells can lead to the altered secretion of digestive enzymes and are associated with severe pancreatic diseases like acute pancreatitis due to excessive Ca 2+ release from intracellular stores [1].
Inositol trisphosphate receptors (IP 3 receptors /IP 3 R) are intracellular Ca 2+ release channels that play an important role in Ca 2+ signaling. The effect of different subtypes of IP 3 receptors and the extent of their contribution to Ca 2+ signaling and Ca 2+ fluctuation is still widely unknown [2]. In pancreatic acinar cells, it is described that IP 3 R2 and IP 3 R3 are the major Ca 2+ release channels that are responsible for exocrine secretion [3]. However, all of the three subtypes of IP 3 R are expressed at the secretory pole of pancreatic acinar cells [4]. Upon muscarinic stimulation with pancreatic secretagogues, phospholipase C is activated, which cleaves phosphatidylinositol-4,5-bisphosphate (PIP 2 ) into the second messenger inositol 1,4,5-trisphosphate (IP 3 ). IP 3 binds to IP 3 receptors and induces Ca 2+ release from intracellular stores [5].
Expression of IRAG2 was found in B-cell lines, T-cell lines, spleen, thymus, and intestinal tuft cells [6,9,13,14]. IRAG2 was also detected in sinoatrial nodes and in sweet, bitter, and umami taste-responsive cells [8,9]. It has been previously described that IRAG2 expression is regulated by p53 in intestinal tuft cells and that IRAG2 ensures Ca 2+ flux in these cells, which is a critical trigger for IL-25 release in response to parasitic infections [13]. A previous study also reported that IRAG1 and IRAG2 are isoform-specific interaction partners and modulators of HCN4, with opposing effects on these channels [8]. In sweet, bitter, and umami taste-responsive cells, IRAG2 is co-expressed with IP 3 R3 [9].
Based on these data, it is possible that IRAG2 also interacts with IP 3 R in other tissues and modulates intracellular Ca 2+ signaling. However, physiological and molecular functions of IRAG2, especially its possible role as a modulator of the IP 3 -Ca 2+ cascade, are still widely unknown. In the pancreas, no information exists about the expression pattern or function of IRAG2 as a modulator of intracellular Ca 2+ signaling. In our study, we analyzed the expression pattern and interaction of IRAG2 with different IP 3 receptors in the murine pancreas as well as its role in Ca 2+ signaling and exocrine pancreatic function.
Western blot analysis showed expression of IRAG2 in whole pancreas tissue of IRAG2-WT mice but not in IRAG2-KO mice ( Figure 1A,B). To investigate the cell types, which express IRAG2 in the murine pancreas, X-Gal staining for β-galactosidase activity was performed on pancreata from lacZ×IRAG2-KO mice and IRAG2-WT mice, where βgalactosidase acts as a reporter for IRAG2. Staining of β-galactosidase was detected in pancreatic acinar cells of lacZ×IRAG2-KO mice but not in IRAG2-WT mice ( Figure 1F). No β-galactosidase activity was detected in islets of Langerhans ( Figure 1F) or pancreatic ducts ( Figure S2). To confirm these data, Western blot analysis of isolated pancreatic acinar cells was performed, and IRAG2 was detected in acinar cells of IRAG2-WT but not in IRAG2-KO ( Figure 1C,D). Furthermore, expression of β-galactosidase was analyzed in lysates using an anti-β-galactosidase-antibody and was detected in whole pancreas tissue and pancreatic acinar cells from lacZ×IRAG2-KO mice, but not in IRAG2-WT mice ( Figure 1E). Taken together, these data revealed IRAG2 expression in murine pancreas and localization in pancreatic acinar cells. . (E) Western blot analysis indicated expression of β-galactosidase as a reporter for IRAG2 in whole pancreas and isolated pancreatic acinar cells of lacZ×IRAG2-KO mice (KO) but not in IRAG2-WT mice (WT). (F) Localization of β-galactosidase by X-Gal staining on pancreas cryosections from lacZ×IRAG2-KO and IRAG2-WT mice. Expression was detected in pancreatic acinar cells (indicated by arrow) but not in islets of Langerhans from lacZ×IRAG2-KO pancreata. No expression of β-galactosidase was detected in IRAG2-WT pancreata. In the graphs, mean ± SEM are shown. Images of total protein (TP) are shown in Figure  S4.

Interaction of IRAG2 with IP3R Types 1, 2, and 3 and Regulation of IP3R Expression
Because of its homology to IRAG1, especially in its coiled-coil domain [8][9][10], it might be possible that IRAG2 has similar interaction partners like IRAG1. To test whether IRAG2 . (E) Western blot analysis indicated expression of β-galactosidase as a reporter for IRAG2 in whole pancreas and isolated pancreatic acinar cells of lacZ×IRAG2-KO mice (KO) but not in IRAG2-WT mice (WT). (F) Localization of β-galactosidase by X-Gal staining on pancreas cryosections from lacZ×IRAG2-KO and IRAG2-WT mice. Expression was detected in pancreatic acinar cells (indicated by arrow) but not in islets of Langerhans from lacZ×IRAG2-KO pancreata. No expression of β-galactosidase was detected in IRAG2-WT pancreata. In the graphs, mean ± SEM are shown. Images of total protein (TP) are shown in Figure S4.
As IRAG2 has been previously described as a homolog of IRAG1 [8][9][10], we investigated if IRAG1 is also expressed in whole pancreas tissue or pancreatic acinar cells.
Detection of IRAG1 was neither seen in pancreatic acinar cells nor in whole pancreas tissue of IRAG2-WT or IRAG2-KO mice ( Figure S3). As a positive control for IRAG1 expression, platelets of IRAG1-WT and as a negative control platelets of IRAG1-KO mice were applied since expression of IRAG1 in platelets has been described before [15].

2.2.
Interaction of IRAG2 with IP 3 R Types 1, 2, and 3 and Regulation of IP 3

R Expression
Because of its homology to IRAG1, especially in its coiled-coil domain [8][9][10], it might be possible that IRAG2 has similar interaction partners like IRAG1. To test whether IRAG2 interacts with different IP 3 -receptor subtypes in murine pancreas, IRAG2 was immunoprecipitated by IRAG2 antibody, and interaction partners were analyzed via Western blot. All three subtypes of IP 3 receptors, IP 3 R types 1, 2, and 3, were detected as coimmunoprecipitated interaction partners (Figure 2A). To see whether a knockdown of IRAG2 leads to different expressions of IP 3 R subtypes, we analyzed the expression of IP 3 R1, IP 3 R2, and IP 3 R3 in pancreata from IRAG2-WT and IRAG2-KO mice. Expression was normalized to total protein ( Figure S3) and calculated as x-fold change of IRAG2-KO expression normalized to IRAG2-WT. Expression of IP 3 R3 was significantly higher in IRAG2-KO mice compared to IRAG2-WT ( Figure 2B,C), and expression of IP 3 R2 ( Figure 2D,E) was significantly lower in pancreata from IRAG2-KO. IP 3 R1 expression ( Figure 2F,G) was not altered between IRAG2-KO and IRAG2-WT pancreas.

Intracellular Ca 2+ Regulation in IRAG2-KO Pancreatic Acinar Cells
IRAG1 interacts with IP 3 receptors and modulates intracellular Ca 2+ release in smooth muscle cells [11,12]. Since IRAG2, as a homolog of IRAG1 [8][9][10], interacted with different subtypes of IP 3 receptors in the murine pancreas (2.2.) and due to localization of IRAG2 in pancreatic acinar cells (2.1.), we isolated the acinar cells of the exocrine pancreas to see whether knockdown of IRAG2 impacts Ca 2+ signaling. Isolated, Fura2-AM loaded acinar cells were equilibrated with Ringer's solution for 2 min. Following that, cells were first stimulated with 1 µM carbachol, and after 2 min recovery time under perfusion with Ringer s solution, cells were stimulated with 10 µM carbachol to evoke Ca 2+ release from the endoplasmic reticulum (ER). The Ca 2+ transients were monitored by the ratio of Fura-2 fluorescence (R 340/380 ). Basal, unstimulated ratio (R 340/380 ) was highly significantly lower in IRAG2-KO cells compared to IRAG2-WT cells, indicating that basal intracellular Ca 2+ was decreased in IRAG2-KO ( Figure 3A,B). Upon stimulation with 1 µM carbachol and 10 µM carbachol, there was a tendency for a slightly decreased Ca 2+ release in IRAG2-KO, indicated by a decreased R 340/380 , which was not significant ( Figure 3B). To normalize the quantity of released Ca 2+ , the maximum ratio upon stimulation with 1 µM or 10 µM carbachol was calculated as x-fold of basal ratio (∆ R/R 0 ). IRAG2-KO showed a higher ∆ R/R 0 than IRAG2-WT after stimulation with 1 µM, whereas 10 µM carbachol had no effect. This implied that relatively more Ca 2+ was released upon stimulation with 1 µM carbachol in IRAG2-KO compared to IRAG2-WT ( Figure 3C). The area under the curve (AUC) was calculated over a stimulation period of 2 min when cells were stimulated with 1 µM or 10 µM carbachol. This value allows the comparison of the amount of released Ca 2+ between IRAG2-WT and IRAG2-KO cells. Calculated AUC was significantly higher in IRAG2-KO than in IRAG2-WT when stimulating with 1µM, but not with 10 µM carbachol, indicating that higher amounts of Ca 2+ were released upon stimulation in IRAG2-KO ( Figure 3D). The slope of the release curve was calculated to assess velocity of Ca 2+ release. Slope was not altered between IRAG2-WT and IRAG2-KO neither on stimulation with 1 µM nor on 10 µM carbachol ( Figure S6A). To evaluate stimulatability of cells with 10 µM carbachol after treating them with 1 µM carbachol before, maximal ratio R 340/380 on stimulation with 10 µM carbachol was calculated x-fold of maximal ratio R 340/380 on stimulation with 1 µM carbachol. Between IRAG2-WT and IRAG2-KO, no differences in stimulatability of cells upon treatment with higher carbachol concentrations were seen when stimulating with lower concentrations before ( Figure S6B). to different expressions of IP3R subtypes, we analyzed the expression of IP3R1, IP3R2, and IP3R3 in pancreata from IRAG2-WT and IRAG2-KO mice. Expression was normalized to total protein ( Figure S3) and calculated as x-fold change of IRAG2-KO expression normalized to IRAG2-WT. Expression of IP3R3 was significantly higher in IRAG2-KO mice compared to IRAG2-WT ( Figure 2B,C), and expression of IP3R2 ( Figure 2D,E) was significantly lower in pancreata from IRAG2-KO. IP3R1 expression ( Figure 2F,G) was not altered between IRAG2-KO and IRAG2-WT pancreas.   Figure S5. Numbers of analyzed lysates are indicated in the graphs, and the mean ± SEM is shown by bars. Significant differences are indicated by (*) (p < 0.05) and (**) (p < 0.01).  At lower concentrations of carbachol, pancreatic acinar cells show Ca 2+ oscillations. This phenomenon is achieved through the release of Ca 2+ from intracellular stores and ATP-dependent Ca 2+ reuptake, leading to oscillating signals [16,17]. We observed Ca 2+ oscillations in IRAG2-WT ( Figure 3E) and IRAG2-KO ( Figure 3F) when treating the cells with 1 µM carbachol. Oscillations per min were counted and are significantly higher in IRAG2-KO cells than in IRAG2-WT cells ( Figure 3G), showing that the frequency of oscillations was higher in IRAG2-KO than in IRAG2-WT cells.

Amylase Content and Release in IRAG2-WT and IRAG2-KO Pancreatic Acinar Cells
The release of Ca 2+ from intracellular stores in response to secretagogues triggers the exocytosis of zymogen granules and consequently the release of amylase [1, 18,19]. We stimulated dispersed acini with 1 µM or 10 µM carbachol for 30 min and measured secreted pancreatic amylase to evaluate the exocrine pancreatic function since IRAG2 appears to regulate Ca 2+ signaling in pancreatic acinar cells (2.3.). IRAG2-KO pancreatic acinar cells showed a reduced basal, unstimulated amylase release compared to IRAG2-WT cells over a stimulation period of 30 min ( Figure 4A). When stimulating the cells with 1 µM or 10 µM carbachol, the maximum amount of secreted amylase during 30 min was lower in IRAG2-KO than in IRAG2-WT, but not significantly ( Figure 4A). Release of amylase calculated as x-fold of basal amylase release revealed a not significant tendency for an enhanced amylase secretion in IRAG2-KO compared to IRAG2-WT upon stimulation with 1 µM carbachol, but not for 10 µM carbachol ( Figure 4B). The activity of amylase was examined in pancreas lysates and serum probes from IRAG2-WT and IRAG2-KO and showed no altered amylase activity between IRAG2-KO and IRAG2-WT neither in pancreas lysates ( Figure S7A) nor in serum probes ( Figure S7B).
To compare the total amount of amylase in exocrine pancreas tissue, immunohistochemical analysis of amylase in whole pancreas tissue was performed. IRAG2-KO showed an increased amount of amylase in pancreas tissue compared to IRAG2-WT ( Figure 4C). The amount of amylase was quantified by detecting the fluorescence in IRAG2-KO and IRAG2-WT pancreata. Fluorescence and therefore quantity of expressed amylase in IRAG2-KO was normalized to IRAG2-WT, and results were given as x-fold change of amylase expression in IRAG2-KO normalized to IRAG2-WT ( Figure 4D).
Since the amount of amylase was increased in IRAG2-KO pancreas tissue, hematoxylineosin (HE) staining was performed to compare the morphology of pancreata and the amount of zymogen granules in IRAG2-KO and IRAG2-WT mice. HE staining revealed no differences in morphology or zymogen granules between IRAG2-WT and IRAG2-KO exocrine pancreatic tissue ( Figure S8). Images of dispersed pancreatic acinar cells also did not show any differences between isolated cells from IRAG2-WT and IRAG2-KO ( Figure S9).  Deficiency in exocrine pancreatic function can lead to underdigestion of food and reduced body weights [3]. To analyze if IRAG2-KO mice show any abnormalities in body weight, mice were weighed starting day 9 after birth until day 47 after birth. Before weaning, no differences in body weights between IRAG2-WT and IRAG-KO mice were observed ( Figure S10A,B,C). Some days after weaning (day 33-35), IRAG2-KO mice showed a slightly reduced body weight, which normalized until day 45 after birth ( Figure S10A,D,E). After postnatal day 45, no differences in body weights between IRAG2-WT and IRAG2-KO were seen ( Figure S10A,E).

Discussion
IRAG2 is a type II membrane protein that has been firstly described by Behrens et al. [6]. It is expressed particularly in lymphoid tissues [6,14], but expression can also be found in intestinal tuft cells [13], in sinoatrial nodes [8], and in sweet, bitter, and umami taste-responsive cells [9]. According to the Human Protein Atlas (www.proteinatlas.org), protein expression, as well as mRNA expression of IRAG2, is detectable in the human exocrine pancreas [20]. In our study, we demonstrate that IRAG2 is expressed in the murine pancreas and localized in acinar cells of the exocrine pancreas. This localization suggests that IRAG2 might be involved in exocrine pancreatic function, e.g., in exocrine secretion. These findings correlate with the data from the Human Protein Atlas for expression of IRAG2 in the human pancreas, mainly in the exocrine glandular cells. The data from the Human Protein Atlas also indicate a weak expression of IRAG2 in human pancreatic duct cells [20]. However, we could not detect the expression of IRAG2 in murine pancreatic duct cells. A possible function of IRAG2 in the human pancreatic duct cells and in human pancreatic acinar cells is unknown.
IRAG2 reveals a homology of 44% with IRAG1 in its coiled-coil domain [8][9][10]. The coiled-coil domain of IRAG1 is essential for its interaction with the IP 3 R1 [11]. The homology of IRAG1 and IRAG2 raises the possibility that IRAG2 also interacts with IP 3 R in different murine tissues. It is already described that IRAG2 interacts with IP 3 R3 via its coiled-coil domain in the COS7 heterologous expression system and is co-expressed with IP 3 R3 in sweet, bitter, and umami taste-responsive cells [9]. In intestinal tuft cells, the interaction of IP 3 R2 and IRAG2 was reported [13]. In the pancreas, all subtypes of IP 3 receptors are expressed, and in acinar cells, all IP 3 receptor subtypes are localized to the secretory pole of the acinus [4]. Our data demonstrate that IRAG2 interacts with IP 3 receptor types 1, 2, and 3 in the murine pancreas. This leads to the assumption that IRAG2 is also expressed at the secretory pole of the acinus. In the Human Protein Atlas, immunohistochemical analysis of IRAG2 expression in human pancreata shows the distribution of IRAG2, particularly at the apical pole of the acinus [20], which would confirm the thesis that IRAG2 is also expressed at the secretory pole of the acinar cells in the murine pancreas. Moreover, to our knowledge, this is the first time that an interaction of IRAG2 with the IP 3 R subtype 1 is also shown. IP 3 R type 1 is highly expressed in brain tissue and plays a role in higher brain functions such as behavior, learning, and memory [21][22][23][24]. Its role in pancreatic acinar cells is widely unknown. Expression of IP 3 R type 1 is reported to be rather low compared to IP 3 R2 and IP 3 R3 in the pancreas [25]. IP 3 receptor types 2 and 3 have been shown to play an important role in exocrine secretion. Double knockouts of IP 3 R2 and IP 3 R3 reveal exocrine dysfunction, and Ca 2+ signaling is severely decreased in pancreatic acinar cells. Therefore, IP 3 R2 and IP 3 R3 are the major Ca 2+ release channels in pancreatic acinar cells that mediate secretagogue-induced exocrine pancreatic secretion [3]. In our study, we observed an upregulation of IP 3 R3 and a downregulation of IP 3 R2 expression in pancreata from IRAG2-KO compared to IRAG2-WT, whereas expression of IP 3 R1 is not altered between IRAG2-WT and IRAG2-KO. Our data show a significantly decreased basal Ca 2+ release in pancreatic acinar cells from IRAG2-KO mice. Despite the lower basal Ca 2+ levels in IRAG2-KO pancreatic acinar cells, upon stimulation with 1 µM of the muscarinic M3 receptor agonist carbachol, a higher amount of Ca 2+ normalized to basal Ca 2+ levels is released in IRAG2-KO pancreatic acinar cells compared to IRAG2-WT. However, maximum Ca 2+ levels in these cells in response to stimulation with 1 µM carbachol are slightly lower in IRAG2-KO than in IRAG2-WT, though this effect is not significant. These effects are not observed for higher concentrations of carbachol. Therefore, our data suggest that IRAG2 leads to an activation of IP 3 receptors under unstimulated, basal conditions. Loss of IRAG2 reduces the activity of the IP 3 receptors and consequently diminishes basal intracellular Ca 2+ levels and might therefore lead to a downregulation of IP 3 R2. This leads to the assumption that IP 3 R2 could be involved in basal Ca 2+ release in pancreatic acinar cells. A reason for the enhanced Ca 2+ release upon stimulation with 1 µM carbachol could be the upregulation of IP 3 R3 in the pancreas as a compensatory mechanism. It is shown that cytosolic Ca 2+ concentration affects the activity of IP 3 receptors in many cell types in a biphasic manner [26][27][28]. Therefore, another explanation could be that the lower basal Ca 2+ levels in IRAG2-KO cells lead to enhanced activation of IP 3 receptors upon stimulation and, therefore, to an increased amount of released Ca 2+ normalized to basal Ca 2+ levels. In IRAG2-KO pancreatic acinar cells, the frequency of Ca 2+ oscillations is higher compared to IRAG2-WT cells. This may also be a mechanism to compensate for the lower intracellular Ca 2+ levels in IRAG2-KO due to decreased activity of the IP 3 R to ensure an appropriate Ca 2+ release after muscarinic stimulation. The effect of the different IP 3 receptor subtypes and the extent of how they contribute to Ca 2+ signaling and Ca 2+ oscillations are still widely unknown [2]. In Hela cells and COS7 cells, it is reported that knockdown of IP 3 R1 leads to decreased total Ca 2+ levels and termination of Ca 2+ oscillations. In contrast, the knockdown of IP 3 R3 leads to more long-lasting Ca 2+ oscillations, and it is shown that IP 3 R3 functions as an anti-oscillatory unit [2]. In B cells, which are genetically engineered to express only IP 3 R2, Ca 2+ oscillations are more regular and long-lasting, whereas B cells that either express IP 3 R1 or IP 3 R3 showed rapidly damped Ca 2+ oscillations [29]. These results are in contrast to the oscillatory signals we observe in pancreatic acinar cells of IRAG2-KO, where a knockdown of IRAG2 leads to an upregulation of IP 3 R3 and results in a higher frequency of oscillations. An explanation could be that IRAG2 modulates the IP 3 receptor subtypes and, therefore, Ca 2+ oscillations in pancreatic acinar cells, independent of the amount of expressed IP 3 R3. Hence, it is possible that IRAG2 leads to a stabilization of IP 3 R, leading to a decreased frequency of oscillations. In pancreatic acinar cells, it is suggested that IP 3 R3 may play a major role in the generation of agonist-induced Ca 2+ oscillations [30][31][32]. This raises the possibility that modulation of IP 3 R3 by IRAG2 leads to a lower frequency of Ca 2+ oscillations in WT cells. The fact that cytosolic Ca 2+ concentration affects the activity of IP 3 receptors in a biphasic manner [26][27][28] could be another reason for the increased frequency of Ca 2+ oscillations in IRAG2-KO. It is possible that the lower basal Ca 2+ levels of IRAG2-KO leads to an activation of IP 3 receptors and followed by a higher frequency of Ca 2+ oscillations in IRAG2-KO cells, maybe through IP 3 R3 but maybe also through other IP 3 receptor subtypes with a higher affinity to cytosolic Ca 2+ , like IP 3 R1 or IP 3 R2 [29]. However, the fact that the expression of IP 3 R1 is rather low in the pancreas makes it unlikely that modulation of IP 3 R1 through IRAG2 contributes to Ca 2+ oscillations in pancreatic acinar cells [25,32]. The downregulation of IP 3 R2 in the pancreas of IRAG2-KO may also contribute to the higher frequency of oscillations in IRAG2-KO acinar cells, as IP 3 R2 leads to more long-lasting and regular oscillations in B cells [29].
Release of Ca 2+ and secretion of digestive enzymes like amylase are strongly linked, as the release of Ca 2+ triggers secretion of amylase [1, 18,19]. In our study, we also observe that reduced basal Ca 2+ release leads to lower basal amylase secretion. Secretion of amylase upon stimulation with carbachol shows a tendency for a higher amount of secreted amylase normalized to basal secretion. This effect is not significant, but the tendency of increased amylase secretion is in accordance with the enhanced Ca 2+ release upon stimulation when normalized to basal Ca 2+ release. In our study, immunohistochemical investigation of amylase in the pancreas reveals a higher amount of amylase in pancreatic acinar cells of IRAG2-KO compared to IRAG2-WT, suggesting that more amylase remains in the cells, due to lower basal Ca 2+ release, and therefore lower basal amylase secretion in IRAG2-KO acinar cells. Even if the amount of amylase is higher in IRAG2-KO cells compared to IRAG2-WT, we detect no accumulation of zymogen granules in IRAG2-KO cells compared to IRAG2-WT cells. It is reported that the double knockout of IP 3 R2 and IP 3 R3 causes exocrine dysfunction, leading to difficulties in nutrient digestion and decreased body weights of the double knockout mice [3]. IRAG2-KO mice show no differences in body weights until weaning; after weaning, IRAG2-KO mice reveal a tendency for decreased body weights. However, this effect is only temporary as body weight normalizes until day 45 after birth. This leads to the assumption that IRAG2-KO animals show almost normal nutrient digestion despite their lower basal amylase secretion. However, the accumulation of amylase in pancreata from IRAG2-KO mice could contribute to pancreatic diseases like pancreatitis. Therefore, IRAG2 could have a protective effect against pancreatic diseases. To determine the effect of IRAG2 on pancreatic diseases, further investigations using specific animal disease models would be needed.
In summary, with this work, we present new results about expression, molecular and physiological function of IRAG2: 1.
IRAG2 is expressed in acinar cells of the murine exocrine pancreas and interacts with IP 3 receptor types 1, 2, and 3 in the murine pancreas. Loss of IRAG2 leads to an upregulation of IP 3 R3 and to downregulation of IP 3 R2 expression in the pancreas.

2.
Deletion of IRAG2 leads to lower basal Ca 2+ release in murine pancreatic acinar cells, suggesting that IRAG2 leads to activation of IP 3 R under unstimulated basal conditions. However, IRAG2-KO reveals a higher amount of released Ca 2+ normalized to basal release upon stimulation with 1 µM carbachol compared to IRAG2-WT. The frequency of Ca 2+ oscillations is higher in IRAG2-KO pancreatic acinar cells compared to IRAG2-WT, maybe due to modulation of IP 3 receptors through IRAG2.

3.
Decreased basal Ca 2+ release in IRAG2-KO acinar cells leads to lower basal amylase secretion. Lower release of amylase may cause a higher amount of amylase that remains in the zymogen granules of IRAG2-KO.

Animals
Global Irag2 deficient mice were generated as described before [13]. The targeting vector contained the exon 8 of murine Irag2 gene, which was flanked by a loxP-Neo cassette and loxP sites, sequences of the Flipase/FRT system (FRT) and the lacZ gene and was constructed by EUCOMM (Helmholtz Zentrum Munich, Neuherberg, Germany). "Floxed" mice containing this targeting vector were either mated with CMV Cre-mice (B6.C-Tg(CMV-cre)1Cgn/J, Jackson Laboratories), resulting in lacZ × Irag2 −/− mice expressing β-galactosidase, or mated with FLP-mice (TGM/hACTB-FLPe, kindly provided by Prof. Ralph Witzgall, University of Regensburg), followed by mating with CMV Cre mice, leading to Irag2 −/− mice not expressing β-galactosidase ( Figure S1). Animals were bred and maintained in the animal facilities of the University of Regensburg according to the Guidelines of the Federation of European Laboratory Animal Science Associations (Bavaria, Germany; Regierung von Unterfranken: DMS 2532-2-326) with free access to food and water ad libitum, following to all guidelines according to the German animal protection law.

Statistical Analysis
All data are shown with mean ± SEM. The Shapiro-Wilk test was used to test the samples for normality. For normally distributed parameters, an unpaired Student s ttest was used to calculate significant differences between two means when variances between the groups were not different. Non-parametric data were analyzed using the Wilcoxon-Mann-Whitney test. Statistical analysis was performed using "GraphPad Prism version 5.01". Significant differences in the graphs are shown by asterisks (*) (p < 0.05), (**) (p < 0.01), and (***) (p < 0.001).

Informed Consent Statement: Not applicable.
Data Availability Statement: The datasets for this manuscript are not publicly available because the raw data supporting the conclusions of this manuscript will be made available by the authors, without undue reservation, to any qualified researcher. Requests to access the datasets should be directed to the corresponding author jens.schlossmann@chemie.uni-regensburg.de.