Mitochondrial malfunction mediates impaired cholinergic Ca 2 + signalling and submandibular salivary gland dysfunction in diabetes

Xerostomia (dry-mouth syndrome) is a painful and debilitating condition that frequently occurs in individuals with diabetes and is associated with impaired saliva production and salivary gland hypofunction. Saliva fluid production relies on Ca 2 + -coupled secretion driven by neurotransmitter stimulation of submandibular acinar cells. Although impairments in intracellular Ca 2 + signalling have been reported in various xerostomia models, the specific Ca 2 + -dependent mechanisms underlying saliva fluid hypofunction in diabetes remain unclear. In this study, we show that diabetic animals exhibit severe xerostomia, evident by reduced saliva flow rate, diminished total protein content, and decreased amylase activity in the saliva secreted by submandibular glands. These impairments remained resistant to exogenous cholinergic stimulation. In submandibular acinar cells, the intra-cellular Ca 2 + signals evoked by cholinergic stimulation were reduced and delayed in diabetes, caused by mal-functioning mitochondria. Upon initiation of cholinergic-evoked Ca 2 + signals, mitochondria accumulate higher Ca 2 + and fail to redistribute Ca 2 + influx and facilitate the store-operated Ca 2 + entry effectively. Structural damage to mitochondria was evident in the acinar cells in diabetes. These findings provide insights into the potential targeting of malfunctioning mitochondria for the treatment of diabetic xerostomia as an alternative strategy to the existing pharmacotherapeutic approaches. This article is part of the Special Issue on "Ukrainian Neuroscience"


Introduction
Diabetes affects a vast number of individuals globally, surpassing 420 million, and is a severe health condition responsible for approximately 1.5 million deaths annually.Among the various systemic complications that arise from diabetes, xerostomia is a chronic and debilitating disorder characterised by dry mouth syndrome.This condition is associated with persistent pathological thirst, leading to increased susceptibility to infections, heightened pain sensation, tooth decay, and other related issues (Borgnakke et al., 2015;López-Pintor et al., 2016).Xerostomia and hyposalivation are common early signs of diabetes and prevalent comorbidities found in over half of all diabetes cases (Malicka et al., 2014).However, treating xerostomia remains insufficient primarily due to the unclear mechanisms that mediate salivary gland hypofunction.Xerostomia originates from the inability of acinar cells to synthesise and secrete essential components of saliva (Ambudkar, 2018;Verstappen et al., 2021).Among the three pairs of major salivary glands, the submandibular glands are the primary source of secreted saliva fluid and electrolytes, responsible for maintaining basal saliva secretion to ensure oral moistening (Pedersen et al., 2018).
Human salivary glands produce approximately 1-1.5 L of saliva, predominantly daily through stimulus-driven salivation, a highly regulated process of neurotransmitter-evoked secretion (Melvin et al., 2005).Fluid secretion relies on cholinergic signalling driven by acetylcholine (ACh) released from the presynaptic terminals of parasympathetic nerves.Activation of muscarinic receptors in the acinar cells facilitates the secretion of fluid and electrolytes (Bymaster et al., 2003;Nakamura et al., 2004;Yamamoto et al., 1996).Therefore, submandibular salivary glands are among the most susceptible organs to the cholinergic (anticholinergic) medication, as they modulate saliva secretion and flow in patients (Arany et al., 2021).
At the cellular level, parasympathomimetics trigger a robust rise in cytosolic Ca 2+ , which global signal profile is essential for synchronised activation of Ca 2+ -dependent Cl − and K + channels in polarised acinar cells, ultimately driving saliva fluid secretion (for review, see Lee et al., 2012;Melvin et al., 2005).This intracellular Ca 2+ rise relies on sustained store-operated Ca 2+ entry (SOCE) induced by Ca 2+ release from intracellular Ca 2+ stores (Parekh and Putney, 2005).Our previous studies (Kopach et al., 2008(Kopach et al., , 2011) ) have demonstrated the vital role of mitochondria in sustaining SOCE and facilitating sufficient Ca 2+ refilling of the endoplasmic reticulum (ER) through the formation of 'Ca 2+ microdomains' -tight contact sites between mitochondria and the ER or plasma membrane.This mitochondrial role is critical for generating Ca 2+ signals during continuous saliva production following sustained cholinergic stimulation of the submandibular salivary glands, for example, during the tasting and chewing of food when ACh release from parasympathetic nerve terminals increases up to several-fold.
Cumulative evidence indicates that mitochondrial dysfunction represents an early hallmark of various pathologies associated with persistent salivary gland hypofunction.Damaged mitochondrial morphology, including mitochondrial swelling, membrane ruptures, loss and disorganisation of crests, and mitophagosomes, has been reported in patients with the autoimmune disease Sjögren's syndrome (Barrera et al., 2021;Katsiougiannis et al., 2023) and animal models of diabetes mellitus (Huang et al., 2020;Ittichaicharoen et al., 2017;Xiang et al., 2020).Additionally, these pathologies exhibit mitochondrial malfunction, such as membrane depolarisation and increased production of reactive oxygen species (ROS) within the salivary glands.Furthermore, detecting a mitochondrial stress protein in the saliva of type 2 diabetic patients (Yuan et al., 2011) indicates mitochondrial stress as a potential molecular biomarker in saliva secretion.
Mitochondrial malfunction has raised an important question regarding the regulation of intracellular Ca 2+ signals in the acinar cells of salivary glands in diabetes.Our previous studies have identified that diabetes-induced hyposalivation by the submandibular salivary glands is associated with altered Ca 2+ release from the ER and suppressed Ca 2+ -ATPase activity in the plasma membrane and ER in acinar cells (Fedirko et al., 2006).However, the specific connection between malfunctioning mitochondria and compromised SOCE as a potential mechanism for diabetic-induced xerostomia remains unknown.In the present study, we aimed to address this critical question by monitoring cholinergic-induced Ca 2+ signals in the cytosol and mitochondria using the designed experimental protocols to investigate SOCE upon various stimulation conditions (Kopach et al., 2011) in acinar cells obtained from healthy animals and those with diabetic xerostomia.Our findings indicate that mitochondrial malfunction is the primary cause of the failure of SOCE in diabetes.This failure results in the inability of acinar cells to secrete saliva, subsequently leading to the development of xerostomia.

Animals
Male Wistar rats (150-200 g) were group-housed with water and food available ad libitum.The animals were used in accordance with protocols approved by the Animal Care and Use Committee at the Bogomoletz Institute of Physiology.We made every effort to minimise animal suffering and reduce the number of animals used.

Experimental diabetes
Diabetes was induced by a single administration of streptozotocin (STZ) dissolved in saline (i.p. injection, 60 mg/kg), as described earlier (Fedirko et al., 2006).Saline administration was used as a control.Glucose concentration in blood serum was measured for each animal using a strip-operated blood glucose sensor (Roche Diagnostic, Mannheim, Baden-Württemberg, Germany).The average glucose concentration in blood serum was 6.0 ± 1.0 mM (n = 20) for the saline-treated and 22 ± 5 mM (n = 50) for the STZ-treated animals.The experiments were conducted 6 weeks after the onset of diabetes.

Collection of saliva from the submandibular glands; analysis of salivation
The saliva secreted by the submandibular salivary glands was collected using the cannulation method as described in our previous studies (Fedirko et al., 2006;Kopach et al., 2012) and by others (Kuriki et al., 2011).In brief, rats were anaesthetised with a mixture of ketamine (i.p., 100 mg/kg, CuraMed Pharma GmbH, Karlsruhe, Germany) and lysthenon (i.p., 0.05 ml/kg, Nycomed GmbH, Austria) and secured in a supine position.To ensure the collection of the saliva secreted by the submandibular glands, cannulation of the Wharton's duct was performed using a glass cannula tightly covering the duct papilla; the cannula was connected with a peristaltic pump to collect the secreted saliva into a small tube (Fig. 1A).To test drug effects on salivation by the submandibular glands, an agonist (or antagonist) was injected into the two glands.Saliva was collected for either 10 min (for an agonist) or 15-20 min (in the case of an antagonist) following the drug administration.For the parasympathetic stimulation of salivation, carbachol was chosen as a parasympathomimetic that binds to the same muscarinic receptors as acetylcholine but is resistant to hydrolysis by acetylcholinesterase.
Various parameters were measured, such as saliva flow rate, protein content and amylase activity.The saliva flow rate was defined as the volume of saliva secreted during 1 h normalized to the animal weight.We measured saliva protein content using a classical Lowry method with spectroscopy and saliva amylase activity via the Caraway method using an amylase-test kit (Biotron Diagnostics, Hemet, CA, USA).

Calcium imaging 2.5.1. Epifluorescence imaging of the cytosolic Ca 2+ level
Isolated acinar cells were bulk loaded with a ratiometric Ca 2+ dye fura-2/AM by incubating a suspension of freshly isolated cells with the dye (5 μM, 25-30 min, 35 • C) in the presence of 0.02% pluronic F-127 (Kopach et al., 2008(Kopach et al., , 2011(Kopach et al., , 2012)).After loading the dye, cells were allowed for de-etherification (15-20 min) and then plated on glass coverslips, pre-coated with poly-L-lysine (for 24 h).For imaging, the cells on a coverslip were transferred into a perfusion chamber mounted on an Olympus BX50WI microscope (Olympus Corporation, Japan) fitted with a 60 × (0.90 NA) water immersion objective and a highly sensitive 12-bit CCD camera (Sensicam, PCO, Germany).The dye was excited at 340 and 380 nm using a PolyChrome IV monochromator (Till Photonics, Germany) under the control of Imaging Workbench software (INDEC System, USA); emission was recorded at 510 ± 10 nm.

Calibration of cytosolic Ca 2+ levels
The classical calibration method was performed in the acinar cells using a calcium ionophore ionomycin to determine the intracellular O. Kopach et al. concentration of free Ca 2+ level ([Ca 2+ ] cyt ).The experiment involved beginning with zero calcium and gradually increasing the concentration until fura 2 became saturated.Collected images were processed for the background signal subtraction, and [Ca 2+ ] cyt values were calculated using the Grynkiewicz equation (Grynkiewicz et al., 1985).All experiments were performed at room temperature (21-23 • C).
Rhod-2 was excited at 543 nm; emission was recorded at 576 nm.[Ca 2+ ] mit was expressed as the changes in rhod-2 fluorescence at the maximum fluorescent signal over the baseline (ΔF/F 0 ).The background fluorescence was subtracted for all images.The changes in [Ca 2+ ] ER were expressed as the changes in the ratio of mag-fura-2 fluorescence at 340 nm to those at 380 nm (ΔF 340 /F 380 ) before and after drug application.For the experiments in Ca 2+ -free extracellular medium, CaCl 2 was omitted, MgCl 2 was increased to 4 mM, and 1 mM EGTA was added.

Electron microscopy
Isolated acinar cells were fixed with 2% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 12 h at 4 • C, subsequently rinsed in a cold phosphate buffer and post-fixed in a solution containing 1% of osmium tetroxide.Cells were dehydrated in the increasing series of ethanol, pre-embedded with propylene oxide, and embedded in Epoxide resin (Plano, Wetzlar, Germany).Ultrathin sections were cut (60-70 nm) and processed for post-staining with lead citrate and uranyl acetate.Images were collected with a JEM electron microscope (JEOL, Japan) at 80 kV.

Data analysis
All datasets were tested for normality using the Shapiro-Wilk test using Origin Pro software.The datasets distributed normally were presented as mean ± S.E.M with n referring to the number of cells analysed or animals tested.Statistical significance was calculated using paired and unpaired two-tailed Student's t-test as appropriate or one-way analysis of variance (ANOVA) followed with Bonferroni's post-hoc test.The datasets with normality rejected were presented as median values, and the nonparametric Mann-Whitney test was performed for statistical comparison.A p value of ≤0.05 was considered statistically significant for either test.

Hyposalivation and impaired saliva content secreted by the submandibular salivary glands in diabetic xerostomia are resistant to parasympathomimetics
Salivary gland hypofunction is an early indicator of developing diabetes, commonly associated with ongoing thirst, a symptom of xerostomia.To confirm that xerostomia in diabetes is mediated by reduced saliva secretion by the submandibular glands, we analysed various parameters of saliva collected from the gland ducts in both control (healthy) rats and STZ-treated rats.Our analysis showed significant suppression of all tested parameters in STZ-treated rats compared to control animals under basal conditions (no stimulation).Specifically, in diabetes, the saliva flow rate decreased by 7-fold (median value, 8.4 ml/kg/h, n = 17 in control vs 1.2 ml/kg/h, n = 23 in STZanimals, p = 0.0001 − 3 , Mann-Whitney test), total saliva protein content decreased by approximately 2-fold (2.5 mg/ml vs 0.3 mg/ml, respectively, p = 0.00008, Mann-Whitney test), and total amylase activity decreased by about 3-fold (152 mg/s/L vs 49 mg/s/L, respectively, p = 0.0009, Mann-Whitney test; Fig. 1B).These findings indicate a severe reduction in salivation by the submandibular glands, resulting in diabetic xerostomia.
Notably, atropine, an anticholinergic agent, suppressed salivation in healthy animals to a level similar to that observed in diabetic rats.Specifically, the salivary flow rate decreased to 2.7 ± 0.3 ml/kg/h (n = 17, p = 0.00087 − 8 compared to unstimulated control, the two-tailed paired t-test) and amylase activity dropped to 80 ± 10 mg/s/L (n = 13, p = 0.0002, the two-tailed paired t-test; Fig. 1C).These results demonstrate the submandibular salivary gland hyposalivation that mediates xerostomia in diabetes and cannot be improved by salivary cell activation with cholinomimetics.

Impaired cholinergic-evoked [Ca 2+ ] cyt signals in submandibular acinar cells in diabetes
Saliva fluid secretion relies on increased intracellular Ca 2+ levels in acinar cells.Because fluid secretion is primarily driven by cholinergic stimulation of salivary cells, we investigated changes in [Ca 2+ ] cyt signals evoked by the principal neurotransmitter of the parasympathetic nervous system ACh in acinar cells in diabetes.We applied exogenous ACh (5 μM, 15-30 s) to acinar cells isolated from the submandibular salivary glands of age-matched control and diabetic animals.Our analysis of ACh-induced [Ca 2+ ] cyt signals revealed significant differences in the amplitude and kinetics of the signals between the two groups (Fig. 2A).The peak amplitude was decreased in diabetes (by 27 ± 2%, n = 33 in control and n = 19 in diabetes, p = 0.0027, the two-tailed unpaired ttest), and the kinetics was slower (the half-rise time prolonged by 113 ± 13%, p = 0.0006, and the half-decay time extended by 249 ± 32%, p = 0.0001, the two-tailed unpaired t-test; Fig. 2B).
The observed changes were independent of extracellular Ca 2+ .Even when cells were stimulated with ACh in the absence of extracellular Ca 2+ , the amplitude remained lower in acinar cells in the diabetic group (248 ± 26 nM, n = 21 in control and 187 ± 23 nM, n = 12 in diabetes), and the amplitude decrease was similar between both cohorts (by 22 ± 2%, n = 21 and 20 ± 2%, n = 12, respectively).The decay kinetics also remained prolonged in Ca 2+ -free medium in the diabetic group (by ~238%, n = 11, p = 0.0025 compared with control, the two-tailed unpaired t-test).The half-rise time was not significantly different (2.9 ± 0.3 s in control and 3.5 ± 0.5 s in diabetes, p = 0.371, the two-tailed unpaired t-test; Fig. 2C).These findings suggest impaired handling of intracellular Ca 2+ during the signal generation.
Consistent with our previous observations (Fedirko et al., 2006), acinar cells displayed higher levels of resting [Ca 2+ ] cyt under diabetic conditions, regardless of whether in standard or Ca 2+ -free medium (Fig. 2A).This further supports impaired cytosolic Ca 2+ handling in the submandibular acinar cells from diabetic animals.

Higher mitochondrial Ca 2+ levels in diabetes prevent mitochondria to shape cholinergic-evoked [Ca 2+ ] cyt signals
Mitochondrial Ca 2+ uptake represents a high-capacitive mechanism for rapid Ca 2+ clearance from the cytoplasm (Duchen et al., 2008).To evaluate how mitochondria contribute to the rearrangement of intracellular Ca 2+ upon generation of cholinergic-induced [Ca 2+ ] cyt signals in acinar cells, we inhibited Ca 2+ uptake by mitochondria.For the inhibition of mitochondrial Ca 2+ uniporter, we used a protonophore FCCP (2 μM) or a mixture of rotenone (10 μM) and oligomycin (10 μM), which inhibits mitochondrial respiratory complex I and mitochondrial ATP synthase, depleting cellular ATP level to 64% and 55% of control, respectively (Li et al., 2003).Firstly, we have confirmed that inhibition of mitochondrial Ca 2+ uptake caused a steady increase of [Ca 2+ ] cyt in acinar cells in the control group (Fig. 3A).This increase reflects Ca 2+ release from mitochondria and is consistent with our previous findings (Kopach et al., 2008(Kopach et al., , 2011)).However, in the diabetic group, such increase was significantly higher (by 58 ± 8%, n = 18, p = 0.032 compared with control, Mann-Whitney test).This indicates that mitochondria accumulate higher levels of Ca 2+ in diabetic conditions.Notably, in the Ca 2+ -free extracellular medium, [Ca 2+ ] cyt rise was lower in acinar cells from the STZ-treated group than in the control (by 58 ± 8%, n = 3, p < 0.01, the two-tailed unpaired t-test; Fig. 3A), indicating that increased mitochondrial Ca 2+ accumulation in diabetes is attributed to Ca 2+ influx.
To assess the role of mitochondria in the generation of cholinergicevoked [Ca 2+ ] cyt signals, we designed the experimental protocol to inhibit mitochondrial Ca 2+ uptake once the ACh-induced [Ca 2+ ] cyt signal reached its peak magnitude.This approach enabled us to visualise an amount of Ca 2+ rapidly accumulated by mitochondria during the signal generation.We observed a robust [Ca 2+ ] cyt elevation, characterised by a slow decay, which recovered back to basal level within 5-10 min (Fig. 3B).In the STZ-treated group, such [Ca 2+ ] cyt rise had the higher magnitude than in control (by ~57%, n = 4 in diabetes and n = 12 in control, p < 0.01, the two-tailed unpaired t-test); its integrated function was also larger (by ~59% compared with control).The rise remained larger in the STZ-treated group when the cells were stimulated in the Ca 2+ -free extracellular medium (by ~17%, n = 5 in control and ~30%, n = 3 in the diabetic group; Fig. 3B).
Having confirmed that mitochondria accumulate higher Ca 2+ levels upon acinar cell stimulation in diabetes, we next examined how it shapes the Ca 2+ signal profile.We applied the experimental protocol of stimulating acinar cells with ACh before and after inhibiting mitochondrial Ca 2+ uptake for paired comparison.The analysis revealed reduced AChevoked [Ca 2+ ] cyt signals upon inhibited mitochondrial Ca 2+ uptake in control cells (by 29 ± 6%, n = 18, p < 0.05, the two-tailed paired t-test); this reduction was, however, more substantial in diabetes (by 69 ± 11%, n = 11, p < 0.001, the two-tailed paired t-test; Fig. 3C and D).Also, the signal kinetics was substantially decelerated upon mitochondrial Ca 2+ uptake inhibition in diabetes: a decrease in the half-rise time by ~119% (n = 18, p < 0.01) in control but ~140% (n = 12, p < 0.01) in the diabetic group.Notably, preventing mitochondrial Ca 2+ uptake in control cells decelerated the half-rise time to a similar value as in diabetic cells (7.9 ± 1.5 s and 7.7 ± 0.9 s, respectively, Fig. 3D).

Impaired SOCE and Ca 2+ refilling of the ER after cholinergic stimulation of acinar cells in diabetes
Ca 2+ influx is the major source of the intracellular Ca 2+ rise upon the agonist-induced [Ca 2+ ] cyt signal generation.To investigate whether and how diabetes changes SOCE, we implemented an experimental protocol to visualise SOCE-mediated [Ca 2+ ] cyt rise (Fig. 4A), as in our previous studies (Kopach et al., 2011).In particular, we triggered the ER depletion in a Ca 2+ -free extracellular medium and subsequently added a Ca 2+ -containing medium (2 mM) to produce SOCE.Different modes of cell stimulation were tested to mimic either brief or prolonged agonist action, thus, the level of ER depletion.In control cells, the SOCE-mediated [Ca 2+ ] cyt rise depended on the cell stimulation mode, hence more potent stimulation produced larger SOCE (p < 0.001, one-way ANOVA with Bonferroni post-hoc test; Fig. 4B).Such a correlation is entirely consistent with our previous findings (Kopach et al., 2011).Although there was no significant difference in the SOCE amplitude between control and diabetic cells in the case of brief ACh stimulation (p = 0.184), more potent stimulation did not result in larger SOCE in diabetes (Fig. 4B).We observed a similar SOCE magnitude in acinar cells between different protocols tested in diabetes (p = 0.101, one-way ANOVA with Bonferroni post-hoc test).In particular, the amplitude was ~91 nM (n = 28) in the case of 30-s ACh, ~102 nM (n = 11) following SERCA inhibition with thapsigargin (TG, 3 μM), and ~134 nM in the case of 5-min ACh (n = 22; Fig. 4B).
In addition to the amplitude, we measured SOCE kinetics, which is thought to more accurately reflect the number of open store-operated Ca 2+ channels (Glitsch et al., 2002a(Glitsch et al., , 2002b)).Again, in contrast to control cells, where the SOCE kinetics depended on the stimulation mode (p < 0.001, one-way ANOVA with Bonferroni post-hoc test), in diabetic cells, there was no significant difference in the SOCE kinetics between stimulation protocols (2.3 ± 0.6 s − 1 , n = 12 for 30-s ACh, 2.3 ± 0.5 s − 1 , n = 10 for TG, and 3.6 ± 0.4 s − 1 , n = 18 for 5-min ACh, p = 0.110, one-way ANOVA with Bonferroni post-hoc test; Fig. 4B).The similar kinetics suggests a similar number of activated channels in diabetic cells regardless of the stimulation mode.
To validate Ca 2+ release from the ER following acinar cell stimulation, we visualised the intraluminal Ca 2+ dynamics in acinar cells of control and diabetic groups using a low-affinity Ca 2+ dye mag-fura-2/ AM.We observed a decrease of [Ca 2+ ] ER in response to ACh stimulation (5 μM, 30 s) in both groups (Fig. 4C).Although ACh stimulation resulted in a drop of [Ca 2+ ] ER of a similar magnitude in control and diabetic cells (0.036 ± 0.004, n = 11 and 0.040 ± 0.004, n = 6, respectively, p = 0.525, the two-tailed unpaired t-test), the kinetics of [Ca 2+ ] ER recovery was dramatically prolonged in diabetes (by ~72% compared with control, p = 0.010; Fig. 4D).Sustained ACh stimulation (5 min) produced even more prominent [Ca 2+ ] ER response in diabetic cells than in control (0.039 ± 0.005, n = 4 in control vs 0.069 ± 0.007, n = 5 in diabetes, p = 0.009, the two-tailed unpaired t-test), and the [Ca 2+ ] ER recovery after 5-min stimulation was even more dramatically prolonged in diabetes (by 7-fold compared with that in control, p = 0.005, the two-tailed unpaired t-test; Fig. 4D).These data demonstrate severely impaired Ca 2+ refilling of the ER after cholinergic stimulation of acinar cells in diabetes.

Malfunctioning mitochondria fail to maintain SOCE
Mitochondrial Ca 2+ uptake is critical for shaping SOCE and adequate Ca 2+ refiling of the ER upon sustained stimulation of acinar cells (Arnaudeau et al., 2001;Kopach et al., 2011;Liu et al., 2007).We, therefore, further investigated the contribution of malfunctioning mitochondria to impaired SOCE in submandibular acinar cells in diabetes.We applied the same experimental protocols to activate SOCE as described above but inhibited mitochondrial Ca 2+ uptake after depleting the ER stores (Fig. 5A).Our data revealed that inhibition of mitochondrial Ca 2+ uptake in the control cells resulted in a similar SOCE magnitude between different protocols tested (p = 0.120, one-way ANOVA with Bonferroni post-hoc test).Eventually, mitochondrial inhibition equalised SOCE between control and diabetic cells, so there were no significant differences between the two groups (p > 0.05, the two-tailed unpaired t-test) across different stimulation protocols tested (Fig. 5B).Also, the rise slope decelerated to a similar level in control between different SOCE protocols (p = 0.961, one-way ANOVA with Bonferroni post-hoc test).The kinetics became similar between control and diabetic cells regardless of the protocol tested (30-s ACh: 0.7 ± 0.2 s − 1 and 0.5 ± 0.1 s − 1 ; TG: 0.8 ± 0.2 s − 1 and 0.8 ± 0.3 s − 1 ; 5-min ACh: 0.9 ± 0.2 s − 1 and 0.9 ± 0.3 s − 1 , respectively).These data indicate that mitochondria determine SOCE in submandibular acinar cells depending on cell stimulation mode and play a key role in SOCE maintenance.The results also confirm that mitochondria are solely responsible for impaired SOCE in diabetes.
To directly visualise the contribution of mitochondrial Ca 2+ uptake to SOCE maintenance in submandibular acinar cells, we performed monitoring of Ca 2+ dynamics inside mitochondria during SOCE activation following sustained cholinergic stimulation of the cells (5 min ACh).We visualised [Ca 2+ ] mit rise during SOCE activation in control and diabetic cells, which was dramatically lower in diabetic cells (Fig. 5C).The median amplitude of [Ca 2+ ] mit rise was 76% (n = 9) in the control, but 23% (n = 11) in the diabetic group (p = 0.0031, Mann-Whitney test; Fig. 5D).The kinetics was also decreased in diabetic conditions (by ~72%, p = 0.001 compared with control, Mann-Whitney test; Fig. 5D).This further confirms malfunction of mitochondria in diabetic acinar cells, resulting in diminished SOCE in diabetes.

Mitochondrial structural damage in acinar cells in diabetes
To find whether diabetes causes structural changes in acinar cells, especially damage to mitochondria, we next performed EM studies.We observed swelling of submandibular acinar cells from diabetic rats (Fig. 6).This was confirmed by counting the length of basolateral plasma membrane (PM), using 11 separated cuttings obtained from isolated acinar cells of submandibular salivary glands in 3 control and 3 STZ-treated animals.Mean PM was 540 ± 20 μm in 13 cells of the control group but 957 ± 33 μm in 14 cells of the STZ-treated group (p = 0.0089, the two-tailed unpaired t-test; Fig. 6B).In the basolateral region, polarised acinar cells contain no secretory granules, which are densely packed in the apical region; therefore, the increased PM in diabetes is due to the cell swelling, regardless of any possible changes in the pool of Given that the initiation of [Ca 2+ ] cyt signals driving exocytosis in polarised acinar cells occurs in the basolateral region, we focused on mitochondria and internal mitochondrial structure within the basolateral region.We counted 112 mitochondria in the control (n = 13) and 141 mitochondria in the diabetic group (n = 14) in the basolateral region of acinar cells.The average area of the mitochondrial membrane located close to the basal PM was 56 ± 1.5 μm in the control and 77 ± 3 μm in the diabetic group (p = 0.074, the two-tailed unpaired t-test; Fig. 6B).This indicates mitochondrial swelling in diabetes mellitus.Furthermore, individual mitochondria displayed structural damage diabetic conditions, including membrane ruptures, numerous vacuolization and considerable damage to mitochondrial cristae (Fig. 6B).

Discussion
Salivary gland dysfunction and hyposalivation are the primary causes of xerostomia, a chronic debilitating condition associated with oral infections, periodontal diseases, tooth loss, and various soft tissue pathologies.Xerostomia (dry mouth syndrome) is one the earliest symptoms of diabetes mellitus, characterised by the acinar cells' inability to produce and secrete saliva.Despite the high prevalence of xerostomia among individuals with diabetes, effective treatments to restore salivary gland function are lacking, and progress in this area is limited due to the unclear mechanisms underlying the development and maintenance of xerostomia at the cellular/subcellular levels.Here, we present evidence highlighting the crucial role of mitochondria in generating intracellular Ca 2+ signals in acinar cells during cholinergic stimulation of the submandibular salivary glands.Our research shows that mitochondrial Ca 2+ uptake is essential for sustained SOCE and adequate Ca 2+ refilling of the intracellular stores to shape the profile of cholinergic-induced [Ca 2+ ] cyt signals, which are vital for driving Ca 2+ -dependent secretion.
Among three major salivary glands, the submandibular salivary glands play a key role in saliva fluid secretion, providing continuous O. Kopach et al. saliva outflow.Using an experimental model of STZ-induced diabetes mellitus, we observed severe xerostomia in STZ-treated rats, characterised by a dramatic decrease of saliva secretion by the submandibular salivary glands.This was evident by severely diminished saliva flow rate (a 7-fold decrease) and impaired saliva content (decreased total proteins and enzymatic activity), pointing at the primary impact of submandibular gland dysfunction to diabetes-induced xerostomia.It is important to note that anaesthetics may affect salivation and bias findings from animal studies where sedation is essential.In clinical practice, ketamine has been known for decades to stimulate oral secretions, hence is often coadministered with anticholinergics to mitigate hypersalivation during sedation (Gingrich, 1994;Heinz et al., 2006).Taking this stimulating effect of ketamine into account, xerostomia observed in diabetic animals under ketamine-induced sedation could be even more severe.Interestingly, hyposalivation in diabetes persisted despite treatment with carbachol, a cholinomimetic resistant to hydrolysis by acetylcholinesterase, indicating no efficacy in alleviating diabetic-induced xerostomia by cholinergic treatment.This is consistent with our previous findings using other muscarinic receptor agonist pilocarpine (Fedirko et al., 2006).Similar observations have been reported in other xerostomia models, such as Sjögren's syndrome, a systemic autoimmune pathology (Barrera et al., 2021;Katsiougiannis et al., 2023;Verstappen et al., 2021) and irradiation-induced salivary gland hypofunction (Liu et al., 2013(Liu et al., , 2021)).
Saliva secretion is a Ca 2+ -coupled process, where the generation of neurotransmitter-induced [Ca 2+ ] cyt signals is essential for the sequential activation of spatially localised ion channels and transporters in polarised acinar cells to trigger secretion (Lee et al., 2012;Melvin et al., 2005).Studies have consistently reported impairments in Ca 2+ signalling in acinar cells of salivary glands in Sjogren's syndrome (Cortés et al., 2019;Dawson et al., 2006;Sun et al., 2023;Teos et al., 2015;Zeng et al., 2017).Similarly, impaired Ca 2+ signalling in acinar cells of the pancreas is known to cause necrosis, a hallmark of all forms of pancreatitis (Petersen et al., 2021;Voronina and Tepikin, 2012).However, there are not as many reports of salivary cell dysfunction in diabetes (Biswas et al., 2018;Ittichaicharoen et al., 2017).Our previous studies have revealed impaired intracellular Ca 2+ homeostasis in submandibular acinar cells for the first time in diabetes mellitus (Fedirko et al., 2006).The present research expands upon our understanding of diabetic-induced impairments in cholinergic-induced [Ca 2+ ] cyt signals by uncovering the mechanism of impaired SOCE and [Ca 2+ ] ER refilling following cholinergic stimulation of acinar cells due to malfunctioning mitochondria.Mitochondrial Ca 2+ uptake is vital to maintain SOCE upon sustained cholinergic stimulation of submandibular acinar cells (Kopach et al., 2011), preventing Ca 2+ -dependent inactivation of store-operated Ca 2+ channels (Glitsch et al., 2002b;Rizzuto et al., 2004).In acinar cells, SOCE represents the crucial mechanism of replenishing depleted intracellular Ca 2+ stores.Following the uptake of Ca 2+ entering through the store-operated channels, mitochondria subsequently release Ca 2+ via the mitochondrial Na + /Ca 2+ exchanger, enabling Ca 2+ refilling of the depleted storesthe vital mechanism for acinar cells during sustained stimulation (Kopach et al., 2011).Inhibition of Ca 2+ entry in salivary gland cells has been proposed as a critical factor contributing to ER stress in Sjogren's syndrome (Sun et al., 2023).Our study has revealed that malfunctioning mitochondria are solely responsible for impaired SOCE in diabetes, resulting in decreased [Ca 2+ ] ER recovery after cholinergic stimulation of acinar cells.Diabetes does not apparently alter the number of store-operated channels but impairs SOCE maintenance due to impaired mitochondria.On the other hand, mitochondrial respiration and ATP production can modulate ion channel conductance.
Previous electron microscopy studies have shown the strategic localization of mitochondria beneath the basolateral PM in acinar cells, allowing mitochondria to rapidly uptake Ca 2+ entering through storeoperated channels, thereby facilitating SOCE (Kopach et al., 2011).Our present data demonstrate that mitochondria accumulate higher Ca 2+ levels in acinar cells in diabetes, both under resting conditions and upon agonist stimulation.The elevated mitochondrial Ca 2+ accumulation may be associated with the suppressed activity of Ca 2+ ATPases in the PM and ER of acinar cells in submandibular salivary glands in diabetes (Fedirko et al., 2006).Mitochondrial Ca 2+ overload can cause mitochondrial damage, including changes in respiratory capacity and ATP production, thus significantly inhibiting saliva secretiona process highly dependent on ATP availability.Sustained Ca 2+ overload can damage mitochondrial membranes and internal structures.Indeed, our EM microscopy data confirmed damaged mitochondrial structure in acinar cells from diabetic rats, consistent with other reports demonstrating swollen and ruptured mitochondria, vacuolation and dissolution of mitochondria and Golgi elements, and increased lipid accumulation in intracellular spaces, progressing with the development of diabetes (Huang et al., 2020;Xiang et al., 2020).
Mitochondrial dysfunction is commonly associated with heightened oxidative stress and lipid peroxidation; it also leads to imbalanced production of reactive oxygen species (ROS), reported in salivary glands in STZ-induced diabetes (Biswas et al., 2018;Zalewska et al., 2015), pre-diabetic (obesity) conditions (Ittichaicharoen et al., 2017) and in patients with diabetes mellitus (Al-Rawi, 2011;Su et al., 2012).Mitochondrial membrane depolarisation, accompanied by increased ROS production, leads to impaired intracellular Ca 2+ signalling in salivary acinar cells, heightened inflammation and apoptosis in obese rats (Ittichaicharoen et al., 2017) and mitophagy in type 2 diabetes (Yamamoto et al., 1996).In addition to impaired Ca 2+ signals, studies have suggested the link between high levels of ROS and the expression of aquaporins, water channels critical for maintaining water flow and salivary gland fluid secretion (D'Agostino et al., 2020;Soyfoo et al., 2012).Specifically, the suppressed expression of aquaporins 1 and 5 in hyperglycemia (Biswas et al., 2018;Sada et al., 2016) can represent a part of the complex molecular cascades underlying hyposalivation in diabetes.In support of this, experimental aquaporin gene therapy has effectively restored salary gland function in Sjögren's syndrome (Lai et al., 2016) and irradiation-induced xerostomia (Teos et al., 2016).
Dry mouth syndrome is a common and troublesome condition of diverse etiologies, including radiotherapy, chemotherapies, treatments with neuroactive drugs, various antidepressants, and others; it is often associated with medications administered to treat different systemic conditions (Tanasiewicz et al., 2016).Numerous medications with anticholinergic effects can lead to the anticholinergic burden.For instance, muscarinic antagonists, commonly used to treat hyperactivity of visceral smooth muscles, i.e., overactive bladder, cause dry mouth syndrome as a side effect, making it difficult for patients to speak or swallow.Therefore, our experimental findings in diabetic xerostomia may have broader implications for other conditions where similar salivary gland dysfunction occurs via similar impairments of Ca 2+ signalling.In clinics, targeting muscarinic receptors, i.e. with pilocarpine, remains one of only a few options available for treating dry mouth syndrome.Despite side effects, this approach is contraindicated in patients taking medications for hypertension, asthma, glaucoma, hyperthyroidism, arrhythmias, some other conditions.Treating xerostomia takes long and new strategies are needed.Our findings provide insights into targeting malfunctioning mitochondria as a potential novel pharmacotherapeutic approach to treating xerostomia via the mechanism of SOCE-dependent Ca 2+ -driven saliva fluid secretion.The described mechanism by which impaired mitochondria cause decreased secretion and ultimately result in xerostomia is important for developing targeted strategies that enhance mitochondrial function.This, in turn, will protect Ca 2+ signalling in systemic metabolic syndromes, such as diabetes or other conditions that cause dry mouse syndrome.Previous studies have reported improved salivary gland function in obese rats by targeting oxidative stress and mitochondrial dysfunction (Ittichaicharoen et al., 2018).Similar approaches have also shown promising results by administering mitochondrial scavengers in a model of radiation-induced xerostomia, linking mitochondrial ROS and salivary gland hypofunction (Liu et al., 2021).Altogether, it highlights the potential of therapeutic O. Kopach et al. targeting of mitochondria for improving persistent salivary gland funcin systemic metabolic disorders associated with xerostomia or dry mouth syndrome.

Fig. 1 .
Fig. 1.Diabetic-induced xerostomia is resistant to the treatment with cholinomimetics.(A) A cartoon showing the cannulation approach for collecting the saliva secreted by the submandibular glands from Wharton's ducts.(B) Statistical summary of the saliva parameters in control and diabetic rats: the saliva flow rate, reflecting the secreted saliva volume (left), total saliva protein content (middle), and total amylase activity (right bars).Boxes show median values; ***p < 0.001, nonparametric Mann-Whitney test.Number of animals tested are indicated.(C) Statistical summary of the tested saliva parameters in control rats following cholinergic stimulation with carbachol or anticholinergic treatment with atropine.(D) Summary of the saliva parameters in diabetic rats after cholinergic stimulation.Data are mean with s.e.m. **P < 0.01; ***p < 0.001, the two-tailed paired t-test; # p < 0.05; ### p < 0.001, the two-tailed unpaired t-test.

Fig. 3 .
Fig. 3. Mitochondrial Ca 2+ uptake shapes the cholinergic-induced [Ca 2+ ] cyt signals and maintains [Ca 2+ ] cyt level in acinar cells.(A) Representative recordings (left traces) and statistical summary of [Ca 2+ ] cyt rise in acinar cells in the conditions of inhibited mitochondrial Ca 2+ uptake.Middle panel: bars show median values; # p < 0.05, nonparametric Mann-Whitney test.Right panel: data are mean with s.e.m; *p < 0.05, the two-tailed unpaired t-test.(B) Representative ACh-induced [Ca 2+ ] cyt signals before and after inhibition of mitochondrial Ca 2+ uptake at the peak of [Ca 2+ ] cyt rise in control (left) and diabetic cells (right).Grey traces are recordings made in Ca 2+ -free extracellular medium.(C) Representative ACh-induced [Ca 2+ ] cyt signals before and following inhibition of mitochondrial Ca 2+ uptake.Insert shows superimposed ACh-induced [Ca 2+ ] cyt signals in paired recordings (recorded before and following treatment with rotenone and oligomycin (blue line) in control and diabetic cells.(D) Statistics of the inhibitory effect on the ACh-evoked [Ca 2+ ] cyt signal amplitude (left) and kinetics (right) in control and diabetes.Data are mean with s.e.m.Number of cells tested are indicated.*P < 0.05; **p < 0.01, the two-tailed paired t-test; n.s.non-significant, the two-tailed unpaired t-test.

Fig. 4 .
Fig. 4. Impaired SOCE in acinar cells from diabetic animals.(A).An experimental protocol of SOCE activation in response to cell stimulation with ACh (5 μM, 30 s or 5 min) or thapsigargin (TG, 3 μM).(B) Statistical summary of the amplitudes and kinetics of SOCE-mediated [Ca 2+ ] cyt rise in acinar cells of control and diabetic groups.*P < 0.05; ***p < 0.001, the two-tailed unpaired t-test; n.s, non-significant, one-way ANOVA with Bonferroni post-hoc test.(C) Representative recordings of [Ca 2+ ] ER responses in mag-fura-2-loaded acinar cells (images on the top panel) to a brief (30-s) ACh application (upper traces) and sustained agonist stimulation (5min, lower traces) in the control and diabetic groups.(D) Statistical summary of the amplitude of [Ca 2+ ] ER response and the kinetics of [Ca 2+ ] ER recovery (ER refiling) following brief (30 s, upper panel) and sustained ACh stimulations (5 min, lower panel) in control and diabetic cells.All data are mean with s.e.m.Number of cells tested are indicated.**P < 0.01, n.s, non-significant, the two-tailed unpaired t-test.

Fig. 5 .
Fig. 5. Malfunctioning mitochondria mediate impaired SOCE in acinar cells in diabetes.(A).Experimental protocol for activating SOCE in the conditions of inhibited mitochondrial Ca 2+ uptake (blue trace).Examples taken for an acinar cell from SZT-treated rat.(B) Statistical summary of SOCE amplitude in the conditions of inhibited mitochondrial Ca 2+ uptake in the control and diabetic groups for a brief ACh application (left graphs), administration of thapsigargin (TG, middle graphs) and sustained cholinergic stimulation (5-min ACh, right graphs).Data are mean with s.e.m.Number of cells tested are indicated.(C).Top panel, images of acinar cells loaded with Rhod-2 for monitoring changes in [Ca 2+ ] mit during SOCE activation.Dotted line highlights a patched acinar cell before and following SOCE activation.Traces show representative [Ca 2+ ] mit recordings during SOCE activation by sustained stimulation (ACh, 5 min) of acinar cells in control and diabetes.(D).Statistical summary of the [Ca 2+ ] mit amplitude and kinetics in acinar cells from control and diabetic animals.Boxes show median values; **p < 0.01, ***p < 0.001, nonparametric Mann-Whitney test.

Fig. 6 .
Fig. 6.Diabetes causes structural damage to acinar cells of submandibular salivary glands.(A) Representative EM images showing subcellular structure of acinar сells from submandibular salivary glands in control (left) and STZ-treated rats (right).Insert on A shows an enlarged view of mitochondria (mit) located close to the basolateral plasma membrane (PM) and surrounded by the ER in control.Nuc, nucleus.Scale bars, 1 μm.(B) Left bars, statistical summary of the basolateral PM length (upper) and mitochondrial area (lower) in control and diabetes.Right images show mitochondrial damage and swelling crests (arrows) in acinar cells from diabetic rats.Scale bars 0.5 μm; for inserts, 100 nm.Data are mean with s.e.m.Number of cells are indicated.**P < 0.01, the two-tailed unpaired t-test.