Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

In vivo monitoring of intracellular Ca2+ dynamics in the pancreatic β-cells of zebrafish embryos

In vivo monitoring of intracellular Ca2+ dynamics in the pancreatic β-cells of zebrafish embryos ISLETS 2018, VOL. 10, NO. 6, 221–238 https://doi.org/10.1080/19382014.2018.1540234 RESEARCH PAPER 2+ In vivo monitoring of intracellular Ca dynamics in the pancreatic β-cells of zebrafish embryos a b,c,d a a a Reka Lorincz , Christopher H. Emfinger , Andrea Walcher , Michael Giolai , Claudia Krautgasser , b,d b,c a Maria S. Remedi , Colin G. Nichols , and Dirk Meyer a b Institute of Molecular Biology/CMBI, University of Innsbruck, Innsbruck, Austria; Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, MO, USA; Center for the Investigation of Membrane Excitability Diseases (CIMED), Washington University School of Medicine, St. Louis, MO, USA; Department of Medicine, Washington University School of Medicine, St. Louis, MO, USA ABSTRACT ARTICLE HISTORY Received 24 April 2018 Assessing the response of pancreatic islet cells to glucose stimulation is important for under- Revised 8 October 2018 standing β-cell function. Zebrafish are a promising model for studies of metabolism in general, Accepted 20 October 2018 including stimulus-secretion coupling in the pancreas. We used transgenic zebrafish embryos 2+ expressing a genetically-encoded Ca sensor in pancreatic β-cells to monitor a key step in KEYWORDS 2+ glucose induced insulin secretion; the elevations of intracellular [Ca ] . In vivo and ex vivo analyses cacna1c; Cav1.2 channel; 2+ of [Ca ] demonstrate that β-cell responsiveness to glucose is well established in late embry- i early zebrafish development; ogenesis and that embryonic β-cells also respond to free fatty acid and amino acid challenges. In GCaMP6s; glucose-sensing of beta cells; in vivo imaging vivo imaging of whole embryos further shows that indirect glucose administration, for example by 2+ yolk injection, results in a slow and asynchronous induction of β-cell [Ca ] responses, while 2+ intravenous glucose injections cause immediate and islet-wide synchronized [Ca ] fluctuations. Finally, we demonstrate that embryos with disrupted mutation of the Ca 1.2 channel gene cacna1c are hyperglycemic and that this phenotype is associated with glucose-independent 2+ [Ca ] fluctuation in β-cells. The data reveal a novel central role of cacna1c in β-cell specific stimulus-secretion coupling in zebrafish and demonstrate that the novel approach we propose – 2+ to monitor the [Ca ] dynamics in embryonic β-cells in vivo – will help to expand the under- standing of β-cell physiological functions in healthy and diseased states. Introduction Ultimately, however, it is desirable to incorpo- Assessing the response of pancreatic islet cells rate in vivo imaging of native intracellular [Ca to glucose stimulation is important for under- ] without interfering with the complex para- standing β-cell function in healthy and diseased crine signalling networks regulating islet activ- states. Until now, pancreatic β-cell physiology )Here ity in native tissue (for a review, see ref. has been analyzed mainly in isolated cell and we tested transgenic zebrafish embryos expres- 1–5 2+ islet systems. Importantly, β-cells under sing a genetically encoded Ca sensor in their these conditions likely exhibit different physiol- β-cells as a potential model for corresponding ogy when compared to cells in their natural in non-invasive in vivo applications. Non-mam- vivo environment. A key step in mammalian malian vertebrates such as zebrafish (Danio glucose-stimulated insulin secretion is the ele- rerio) have become relevant alternative organ- 2+ ].Non-invasive in vation of intracellular [Ca i isms to study diabetes and other metabolic 2+ ] has recently been facili- vivo imaging of [Ca i diseases, due to ease of genetic manipulation, tated by transplantation of pancreatic islets into high reproduction rates, and ready access for in the anterior chamber of the eye or the kidney vivo imaging due to larval transparency. capsule of mice. Such real-time monitoring Importantly, pancreata in mammals and in zeb- facilitates the study of islet physiology and vas- rafish have conserved physiological endocrine cularization longitudinally, and enables in vivo and exocrine function, similar cellular architec- screening of novel drugs and treatments. ture, and conserved expression and function of CONTACT Dirk Meyer dirk.meyer@uibk.ac.at Institute of Molecular Biology/CMBI, University of Innsbruck, Technikerstrasse 25, Innsbruck 6020, Austria. Supplemental data for this article can be accessed on the publisher's website. © 2018 The Authors. Published with license by Tayloy & Francis. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 222 R. LORINCZ ET AL. 9 2+ 30–32 most developmental genes. Accordingly, the using Ca -sensitive dyes. More recently, 2+ zebrafish has proven highly productive for stu- genetically encoded Ca indicators have been 10–12 dies of pancreas development, and introduced as tools for non-invasive approaches to 13,14 regeneration. In mammals as well as in study excitable cells such as β-cells. The analysis of 2+ zebrafish the pancreas develops from the endo- transients in corresponding models systems Ca dermal germ layer and later compromises endo- will help in understanding underlying causes of β- crine and exocrine tissue. Within one day of cell dysfunction (for example in the context of 33,34 2 development, the zebrafish embryo forms a sin- diabetes risk factors. ) In vivo imaging of [Ca gle ‘primary’ pancreatic islet with ~60–70 ] dynamics in transplanted (intraocular) mouse mono-hormonal α-, β-, δ-, ε-cells. As develop- pancreatic islets showed reduction at a prediabetic 2+ ment proceeds, the primary islet increases in stage, suggesting the potential of [Ca ] as a func- size and additional secondary islets are formed, tional marker to evaluate β-cells in vivo in diseased to adopt to growth-related requirements. states. Nevertheless, the low litter numbers, com- Glucose metabolism in zebrafish is also very plicated methodologies, high costs, and large similar to mammalian glucose metabolism, and experiment variability limit the use of mammalian species for such studies. Since genetic manipulation overfed zebrafish displays obesity-related diabetes is easy and inexpensive in transparent zebrafish phenotypes including impaired glucose tolerance The molecular larvae, large-scale drug and genetic screens as well and increased insulin production. basis of glucose detection is well understood in as fluorescent imaging applications for studying β- 36–38 mammalian pancreatic β-cells (for a review, see cell physiology in living animals are feasible. 17 18 2+ ref. and. ) Glucose is taken up by the facilitative Monitoring of the glucose-stimulated Ca influx in glucose transporter (GLUT) GLUT2/SLC2A2, and isolated juvenile and larvae zebrafish islets has been We recently generated a is metabolized through glycolysis and oxidative previously described. phosphorylation, thereby generating adenosine tri- transgenic line expressing the genetically encoded 2+ phosphate (ATP) and increasing the ATP/ADP Ca -sensor GCaMP6s, and in this study we have 11,19 2 ratio. The altered [ATP/ADP] ratio in the β- established protocols for in vivo monitoring of [Ca + + cell then leads to the closure of ATP-sensitive K - ] dynamics in late embryonic animals. We found channels (K -channels), depolarization of the that β-cells in larvae show rare baseline sponta- ATP membrane, and consequent opening of voltage- neous activities, but after intravenous injections of dependent calcium channels (VDCCs). The glucose, virtually all β-cells showed very rapid ele- 2+ influx of Ca then triggers release of insulin vations in GCaMP6s fluorescence. from secretory granules. Orthologues of all Additionally, we have used the GCaMP6s major genes (GLUT, K -channels and VDCCs) reporter to assess β-cell associated function of ATP −/- involved in mammalian glucose sensing and insu- VDCCs. β-cell specific Ca 1.2 mice exhibit lin secretion are also expressed in zebrafish, and mild basal hyperglycemia, impaired glucose tol- 22–25 show functional similarities. Studies suggest erance and lack of first-phase insulin secretion,- 40,41 that glucose uptake in zebrafish, similar to mam- but this important role in mammalian mals, occurs through GLUT transporters, with glucose-induced insulin secretion has not been 40–43 GLUT2 expression found in the endocrine pan- assessed in zebrafish. 26,27 Furthermore, we and creas of zebrafish larvae. others recently demonstrated that zebrafish islet β- channels with con- cells express functional K Results ATP served structure and metabolic sensitivity to their 2+ Zebrafish β-cell [Ca ] responses depend on the mammalian counterparts, supporting the use of route of glucose administration in vivo zebrafish as a model animal in islet glucose sensing 28,29 2+ and diabetes research. To enable in vivo studies of β-cell [Ca ] in Excitability of β-cells has been investigated by zebrafish larva, we generated a transgenic line multiple strategies including monitoring of the expressing a genetically encoded membrane- 2+ membrane potential by electrical recordings, and tagged Ca -sensor together with a Histone- ISLETS 223 2+ tagged RFP, both under control of the insulin β-cell specific Ca changes either in the living promoter (Tg[ins:lynGCaMP6s; ins:H2B-RFP], embryo (Figure 1(C)) or in isolated islets referred to hereafter as ins:lynGCaMP6s; (Figure 1(D)). In the first set of experiments, Figure 1(A,B)). The combined expression of we recorded β-cell responses to different routes lynGCaMP6s and H2B-RFP in β-cells enabled of glucose administration. To simulate physio- rapid β-cell identification and measurements of logically relevant feeding-related gradual and 2+ Ca responses in selected β-cell areas. We continued increase of blood glucose level, we further established protocols for determining tested injections of glucose into the yolk, the Figure 1. Specific expression of GCaMP6s sensor in pancreatic β-cells enables the study of Ca2+ fluxes in live zebrafish. (A) Illustration of the construct used to generate Tg[ins:lynGCaMP6s ins:H2BtagRFP] (ins:lynGCaMP6s). The GCaMP6s sensor was linked N-terminal with a lyn membrane tag. The additional ins:H2BtagRFP sequence was added to the construct to ease the screening of transgenic embryos and to identify the β-cells due to the nuclear specific H2BtagRFP signal. (B) Antibody staining on 5 days post fertilization (dpf) ins:lynGCaMP6s embryos detecting >95% overlapping of endogenous insulin and GCaMP6s (using polyconal GFP antibody) and >80% overlapping of H2B:RFP and GCaMP6s (n = 12 larvae, 20 β-cells per larva were analysed). (C) Schematic outline of intravenous glucose injection of zebrafish larvae. Using α-bungarotoxin (BTX) protein injection, the zebrafish larvae are paralyzed. After 10–15 min, the larvae are embedded in low melt agarose. Using a micromanipulator assembled to the microscope, glucose (in combination with Rhodamine B isothiocyanate-Dextran) is injected to the dorsal aorta of the embryo while the imaging is running. (D) Schematic outline of perfusion system on extracted pancreatic islet of zebrafish embryos. Islets from ins: 2+ lynGCaMP6s fish are isolated and placed into glass-bottomed culture plates with agarose gel wells containing islet media for Ca imaging as solutions are changed. Scale bar = 10 µm 224 R. LORINCZ ET AL. responsive cells varying from 5–30% (Sup. pericardium or the hindbrain ventricle. In all 2+ Fig. 1A–C). In contrast, intravenous injection approaches, the glucose injection induced Ca of D-glucose (~10–20 mM final concentration) fluctuation in β-cells (Sup. Fig. 1D–F), but the in wild-type5 dpflarvaeevokedrapid and responses were very variable, with delays robust elevations in the GCaMP6s fluorescence between injection and response ranging from in almost all labelled cells (Figure 2(A), blue secondstominutes andthe numberof Figure 2. Zebrafish β-cells show rapid glucose response in vivo. (A) In vivo experiments in larva revealed rapid fluorescence signal alteration (peak within 5 sec) in response to intravenous injection of >10 mM D-glucose (blue line) (5 dpf) (n = 8 larvae). To confirm that the response is due to D-glucose metabolism, L-glucose (red line) using same concentration (>10 mM) was also injected and showed no response (n = 5 larvae). (B) Fluorescence image series of the islet showing β-cell specific GCaMP6 signal at the indicated time-points (sec) after intravenous injection of D-glucose (>10 mM final concentration) in 5 dpf larva. Scale bar = 10 µm (C) Quantitative GCaMP6 signal intensity in the larvae shown in (B), and two other individual larvae (gray dotted lines), showing calcium influx after sequential glucose injections (1.injection: >10 mM, 2. Injection: >20 mM, 3. Injection: >30 mM estimated final concentration). FIU intensities for the whole islet were determined using Image-J and normalized to the FIU baseline (intensity measured in the first time point). (D) Representative fluorescence traces of manually selected β-cell areas (n = 4, based on red RFP expression as indicated) in a 5 dpf old zebrafish larva, normalized to initial fluorescence intensity, while 1–2 nl of 0.5 M D-glucose (>10 mM final concentration) was intravenously injected in vivo (at dotted black line). After glucose injection β-cell areas show oscillation. (E) Cross-correlation matrix (determined by PeakCaller, MATLAB) of the 4, in (D) indicated, β-cell areas after intravenous glucose injection. The panel represent one islet with 4–4 cell comparison. Colour map key is given to the right of the panel. Further examples are shown in Sup. Fig. 2B. FIU = Fluorescence Intensity Units (AU, Arbitrary Units). ISLETS 225 line). Sequential intravenous glucose injections Zebrafish islets are uncoupled ex vivo resulted in repeated similar rapid responses, but To assess the concentration-dependence of glucose the normalized Fluorescence Intensity Unit and other nutrients on β-cell excitation, we per- (FIU) peaks decreased after the second and formed perfusion studies with isolated pancreatic third injection (Figure 2(B,(C)). Control injec- islets (ex vivo) from 5 dpf old zebrafish larvae tions with the same concentration of L-glucose (Figure 1(D)). Isolated larval islets were first incu- caused no GCaMP6s signals (Figure 2(A), red bated in 2 mM glucose solution, then stimulated line), confirming that the response is due to with 20 mM glucose solution and at the end of the glucose metabolism, rather than osmotic stress recording, 30 mM KCl solution was applied to or other potential injection artefacts. In intra- confirm excitability. In these experiments, higher venously D-glucose-injected embryos, the fluor- resolution imaging by spinning disc microscopy escence increased almost immediately after enabled single cell analysis of the isolated islets. injection, and reached its peak within 5 seconds, Consistent with the in vivo data, 20 mM glucose followed by oscillations across the islets (5 ± 1 increased GCaMP6s FIU/FIU baseline intensity peaks/minute; Figure 2(D), Sup.Video 1 and 2). (Figure 4(A,B)), almost all β-cells (83% ± 24%, Correlation analyses revealed that glucose- Figure 4(C)) responded to the glucose stimulus, induced responses are synchronized across the 2+ and the induced [Ca ] oscillated with similar islet when the glucose is injected intravenously frequencies (5 ± 1 and 6 ± 1 peaks/minute, respec- (Figure 2(E)andSup. Fig. 2B), which was not tively, Figure 4(D)). However, in contrast to what thecasefor yolk injection(Sup. Fig. 2C), indi- was observed after intravenous glucose injection, cating that the in vivo responsiveness of zebra- ex vivo β-cell responses were out of phase (correla- fish β-cells to glucose depends on the route of tion analyses shown Figure 4(E,C), Sup. Fig. 3 and glucose administration. 2B). Notably, the time to peak in the ex vivo studies were highly variable (ranging from a few seconds to minutes) and on average were more Zebrafish islets are glucose responsive in early than 30 times longer than those seen after intra- developmental stages venous glucose injection (4.4 ± 0.5 sec in vivo and 140 ± 30 sec ex vivo, Figure 4(F)). We next assessed whether the glucose-responsive- ness of β-cells observed in 5 dpf larvae is also seen at earlier developmental stages. Studies in rodents L-glutamine and palmitic acid also stimulate have established that islet maturity and function 2+ [Ca ] influx in zebrafish pancreatic islet vary during development, with glucose responsive- ness occurring in later stages, particularly near the In mammalian β-cells, excitation occurs not only weaning transition. Previous studies showed that in response to glucose but also to various other 46–49 insulin-dependent glucose homeostasis in zebra- stimuli. To test whether amino acids and free 2+ fish starts at 3 dpf. Normalized FIU intensities fatty acids contribute to [Ca ] signalling of zeb- 2+ showed similar glucose-induced [Ca ] dynamics i rafish pancreatic islets, perfusion experiments with over time (Figure 3(a)), and quantification of the L-glutamine and palmitic acid on extracted islets GCaMP6s signal revealed significantly higher peak were carried out. Whole islets extracted from 5 dpf FIU intensities after glucose injection compared to zebrafish larvae showed increase in GCaMP6s baseline (before glucose), but no significant differ- FIU/FIU baseline intensity when 20 mM ences between developmental stages (3, 4 and L-glutamine or 1 mM palmitic acid alone and no 5 dpf) (Figure 3(b)). In addition, little or no spon- glucose was applied in the perfusion system taneous responses were observed in unstimulated (Figure 5(A,B)). To confirm that the responses larvae at 3–5 dpf (Figure 3(c,d)). This shows that were specific to the nutrients and were not due the embryonic onset of insulin-dependent glucose to osmotic stress, we performed perfusion experi- homeostasis correlates with the requirement for ments with 20 mM sucrose on the extracted pan- glucose-sensing mechanisms after 3 dpf. creatic islet; no significant increase in GCaMP6s 226 R. LORINCZ ET AL. Figure 3. Similar in vivo glucose responses in 3 dpf, 4 dpf and 5 dpf zebrafish larvae. (A) Quantitation GCaMP6 signal intensity show rapid fluorescence signal alteration after intravenous injection of D-glucose (>10 mM final concentration) in 3, 4 and 5 dpf larvae (n = 5, 3 and 8 larvae, respectively), (C) and no or little alteration in unstimulated larvae (n = 8, 8 and 6 larvae, respectively). FIU intensities for the whole islet were determined using Image-J and normalized to the FIU baseline (intensity measured in the first time point) (B) Peak FIU intensities (FIU ) of the whole islets before and after glucose peak injection (normalized to baseline FIU intensity measured in the first time point) (curves shown in (A) at different developmental stages (3, 4 and 5 dpf) (n = 5, 3 and 8 larvae, respectively). (D) Peak FIU intensities (FIU ) of the whole islets in unstimulated larvae peak (normalized to baseline FIU intensity measured in the first time point) (curves shown in (C)) at different developmental stages (3, 4 and 5 dpf) (n = 8, 8 and 6 larvae, respectively). FIU = Fluorescence Intensity Units (AU, Arbitrary Units).Data are shown as mean values ±s.d., *p < 0.05 in one-way ANOVA, Tukey’s Multiple Comparison test to compare the glucose-injected groups (between 3, 4 and 5 dpf old larvae after glucose) in (B) and the unstimulated larvae in (D), and two-tailed, paired t-test was done to compare before and after glucose groups in (B). 2+ fluorescence intensity was observed. In all cases, acid, without glucose, stimulates [Ca ] influx in islets were excitable, as revealed by the change in embryonic zebrafish pancreatic islets. fluorescence upon 30 mM KCl stimulation 2+ fluctuations occurred (Figure 5(C)). Again, Ca m458 Ca 1.2/isl mutant zebrafish are asynchronously across the whole islet in all ex vivo hyperglycemic conditions, indicating that β-cells were not coupled with one other (Figure 5(A,B), right Next we used the ins:lynGCaMP6s reporter to panels). These results revealed that short-term assess β-cell specific defects in L-type calcium exposure to L-glutamine alone or to palmitic channel mutant zebrafish embryos. Currently, the ISLETS 227 Figure 4. Comparison of ex vivo and in vivo glucose responsiveness. (A) Fluorescence image series of isolated pancreatic islet showing GCaMP6 signal alteration at the indicated time-points shown in (B) applying 2 mM, 20 mM glucose and 30 mM KCl. Single β-cell is defined as ˜3× 5 μm white rectangular selection drawn around one 228 R. LORINCZ ET AL. genetic roles of Ca 1.2 and Ca 1.3 in glucose of free glucose in control and mutant embryos V V homeostasis are not well defined. Physiological showed elevated glucose levels in 4–5 dpf Ca 1.2/ m458 studies indicate that Ca 1.2 and Ca 1.3 are key isl (Figure 6(A) and Sup. Fig. 5) but normal V V tc323d regulators of mammalian glucose-induced insulin glucose levels in 5 dpf Ca 1.3a/gem mutants ), but correspond- (Sup. Fig. 6). These data suggest impaired β-cell secretion (for a review, see ref. m458 1.2/isl mutant fish. ing mouse knockout models displayed only minor functionality in the Ca m458 effects on insulin secretion and glucose homeosta- To test whether Ca 1.2/isl mutants affect β- 2+ sis, perhaps due to compensation by other chan- cell specific [Ca ] dynamics, the mutation was nels. In particular, it was found that Ca 1.2 in β- crossed into the ins:lynGCaMP6s background. In cells is required for proper first-phase glucose- vivo fluorescence recordings of untreated 5 dpf induced insulin secretion, and that Ca 1.3 may transgenic larvae showed significantly higher 2+ modulate second-phase secretion, each of which basal islet [Ca ] dynamics in mutants than con- 40,43 is significant for normal glucose tolerance. trol animals (Figure 6(C), Sup. Video 3 and 4). The zebrafish has three Ca 1.2 and Ca 1.3 encod- Hyperglycemia in these animals raised the possi- V V ing gene orthologues, of which two – cacna1c bility that non-islet effects could underlie the 2+ m458 (encoding Ca 1.2) and cacna1da (Ca 1.3a) – increased [Ca ] dynamics in the Ca 1.2/isl V V i V show expression in the developing pancreatic mutants. To eliminate non-islet effects, we per- 50–52 islet. The third gene cacna1db (Ca 1.3b)is formed ex vivo recordings of isolated mutant and 2+ exclusively expressed in the central nervous control islets. Importantly, increased [Ca ] system. To test cell type specific expression of dynamics in mutants were maintained in isolated Ca 1.2 and Ca 1.3a, we performed whole mount islets and appeared to be glucose independent. In V V stains in different endocrine GFP-reporter stains particular, we found that in mutant islets the 1.2 and Ca 1.3a (Sup. Fig. 4). We found that Ca change from 2 mM to 20 mM glucose had no V V are expressed in all hormone-positive pancreatic significant effect on the peak frequency (2 mM: cells of embryos older than one day (Sup. Fig. 4.). 0.6 ± 0.3 and 2.6 ± 0.7 peaks/min; 20 mM: To determine requirements of these genes in glu- 6.0 ± 1.0 and 1.8 ± 0.5, control and mutant, cose homeostasis, we took advantage of previously respectively) (Figure 6(C)), or on the integrated m485 described mutants for Ca 1.2 (isl, isl , island signal intensity (2 mM: 0.018 ± 0.004 and 50,53 tc323d beat) and Ca 1.3a (gem, gem , gemini).- 0.065 ± 0.008 AUC; 20 mM: 0.057 ± 0.013 and 54,55 Both mutant alleles carry premature stop 0.074 ± 0.011, control and mutant, respectively) codons that result in the expression of truncated (Figure 6(D) and Sup. Fig 7A), while in control (and most likely inactive) proteins. Measurements islets these parameters were significantly increased piece of the GCamP6s-expressing membrane located between two individual β-cells. Scale bar = 10 µm. (B) Single cell GCaMP6s fluorescence traces for isolated islet from wild-type zebrafish larvae (5 dpf) after applying 2 mM, 20 mM glucose followed by 30 mM KCl. FIU intensities for the single cells (n = 6 cells) were determined using Image-J, normalized to the FIU baseline (intensities measured in the first 2 min when 2 mM glucose was applied) and displayed as FIU/FIU baseline (AU) for single cells (coloured dotted line) and for the average of the singles cells in the islet (black line). (C) Quantification showing the percentage of β-cells that respond upon stimulation with 20 mM glucose, ex vivo and the percentage of glucose-responsive β-cell area after intravenous injection of >10 mM glucose in vivo. (n = 8 and 9 larvae were analyzed, in vivo and ex vivo, respectively) (D) Average oscillation frequencies of 4 selected β-cell areas of the pancreatic islet (n = 8 islets, in each islet 4 β-cell areas were manually selected based on the red RFP expression) after intravenous injection (~1–2 nl of 0.5 M glucose, >10 mM final concentration) of 5 dpf old zebrafish larvae (in vivo, imaged by epifluorescence microscope), and average oscillation frequencies of individual β-cells (n = 5 islets, in each islet >6 cells were analyzed) of the isolated pancreatic islets after 20 mM glucose stimulation (ex vivo, imaged by spinning disc microscope). Oscillation peaks were identified by PeakCaller, MATLAB on the FIU/FIU baseline traces of individual β-cells/areas (see Sup. Fig. 2A). The peak number was divided by the time duration until oscillation was observed (˜30–60 sec, in vivo, ˜150–300 sec, ex vivo) (E) Cross-correlation matrix (determined by PeakCaller, MATLAB) of the islet between individual cells shown in (A) and (B), with color key to the right of the matrix graph. Numbers on the x and y axes represent individual cells that are marked in (A) and (B). (F) Time to peak calculated from the time point when addition of the new glucose solution was finished in isolated islets, ex vivo (n = 7 islets, the addition took in average ˜38 seconds), and when glucose was injected in living zebrafish larvae, in vivo (n = 8 islets from n = 8 larvae). Data are shown as mean values ±s.e.m., *p < 0.05 in unpaired, two-tailed t-test between ex vivo and in vivo groups in (C), (D) and (F). FIU = Fluorescence Intensity Units (AU, Arbitrary Units). ISLETS 229 Figure 5. Stimulation of β-cell by L-glutamine and palmitic acid. Representative single cell GCaMP6s fluorescence traces for isolated islet from wild-type zebrafish larvae (5 dpf) after applying 2 mM glucose followed by 20 mM L-glutamine (A), and 1 mM palmitic acid (B) stimulation and 30 mM KCl. FIU intensities for the single cells (n = 10–10 cells) were determined using Image-J, normalized to the FIU baseline (intensities measured in the first 2 min when 2 mM glucose was applied) and displayed as FIU/FIU baseline (AU) for single cells (coloured dotted line) and for the average of the singles cells in the islet (black line). Cross-correlation matrices (determined by PeakCaller, MATLAB) of the islets between individual cells shown in (A) left panel and (B) left panel, with color key to the right of the matrix graph. Numbers on the x and y axes represent individual cells. Cross-correlation analysis was done on the section where L-glutamine/palmitate stimulation was applied (excluding KCl section). (C) Calcium responses to 20 mM L-glutamine and 1 mM palmitic acid when compared to 2 mM glucose. The changes in GCaMP6s fluorescence intensity in β-cells of wild-type pancreatic islets were averaged for the whole islet and the FIU peak intensity upon different nutrient stimulation was normalized to the FIU peak intensity upon KCl stimulation (n = 23, 7, 5, 4 and 7 islets were analyzed for 2 mM glucose, 20 mM glucose, 20 mM L-glutamine, 20 mM sucrose, and 1 mM palmitic acid stimulation, respectively) (in each islet at least n = 5–8 cells were analyzed and the FIU peak intensity was measured in the average FIU trace for the whole islet). Data are shown as mean values ±s.e.m., *p < 0.05 in unpaired t-test between 2 mM glucose and 20 mM sucrose, 20 mM L-glutamine, and 1 mM palmitic acid stimulation. FIU = Fluorescence Intensity Units (AU, Arbitrary Units). 230 R. LORINCZ ET AL. Figure 6. Ex vivo imaging of β-cells Ca + dynamics in Ca 1.2 mutant zebrafish larvae. i V m458 (A) Ca 1.2/isl mutant shows elevated whole larval glucose levels at 4 and 5 dpf. Glucose levels measured at 4 and 5 dpf whole larval extracts (pool of 10–10 larvae), normalized to the protein level measured by Nanodrop, and showed as relative to the average control value. (n = 4 and 8 biological replicates, respectively) (at 5 dpf, results combined from 2 independent experiments) (***p < 0.0001, two-tailed t-test). (B) Fluorescence image series of an isolated islet from 5 dpf old Ca 1.2 mutant crossed with ins: lynGCaMP6s larva (upper panel). Quantitation of GCaMP signal intensity in the time series shown in the upper panel, indicating FIU/ FIU baseline changes after 20 mM glucose as well as after 30 mM KCl stimulation (lower panel). (C) Quantification of GCaMP6s fluorescence intensity peaks determined by PeakCaller, MATLAB in vivo, without any stimulation (n = 6 wild type and 8 Ca 1.2 mutant larva), and ex vivo, upon 2 and 20 mM glucose stimulation in wild-type and Ca 1.2 mutant larva (n = 7–7 and 6–6 larvae, −1 respectively), shown as oscillation frequency (min ) (peak number/cell number*minute). β-cell number was defined based on red nuclei RFP expression in one focal plane (4–10 cells per islet were analyzed) (**p < 0.005, two-tailed, nonparametric Mann-Whitney ISLETS 231 by increasing glucose levels. Thus, there is a coun- contrast to the ex vivo data, did not allow analyses at 2+ terintuitive increase in [Ca ] oscillations at basal single cell resolution. The combination of the deep glucose in the mutant islets. In mouse, Ca 1.3 tissue localization, the frequent presence of light subunit knockout was found to be compensated absorbing pigment cells, and permanent slow tissue 1.2 upregulation. To test whether upregu- by Ca movements due to gut peristaltic movement pre- 1.3a may account for the Ca 1.2/isl- lation of Ca vented more detailed analyses. However, our data V V m458 mutant phenotype, we performed RT-qPCR revealed that embryonic in vivo β-cell responses are (reverse transcriptase quantitative PCR) analyses. immediate and synchronized when glucose is However, transcripts levels for Ca 1.2 and directly injected into the bloodstream, and that gra- Ca 1.3a in whole embryos (Figure 6(E)) and iso- dually increasing blood glucose levels result in 2+ lated islets (Figure 6(F)) showed no significant ] fluctuations, delayed, non-synchronized [Ca differences between controls and mutants. These very similar to those found in glucose-stimulated data show that genetic loss of Ca 1.2 results in perfused islets. Currently, we can only speculate glucose-independent hyperactivation of β-cells, about the mechanism underlying the synchronized and suggest that this activation is independent of responses after intravenous glucose injection. cell-autonomous compensatory effects. Possibly, the synchronicity is caused by a cell-auton- omous but simultaneous response of all β-cells to the simultaneous prompt exposure to a high glucose Discussion challenge. Alternatively, synchronicity could be trig- 2+ We have successfully assessed [Ca ] dynamics in gered by a currently undefined paracrine signal. In 2+ β-cells of zebrafish pancreatic islets, both in vivo ] dynamics in this context, the much faster [Ca after injection of glucose and ex vivo using per- living animals, as compared to isolated islets, could fused isolated islet. We found that zebrafish β-cells indicate the involvement of neuronal driven triggers. are directly glucose-responsive, and that respon- Consistent with this possibility, it was recently siveness is established early in embryonic develop- shown that the embryonic zebrafish islet is inner- ment. These data are consistent with a conserved vated at the relevant time points, and that a neuro- glucose sensing and stimulus-secretion coupling transmitter, namely galanin expressed in these 59,60 mechanism in fish and mammals. However, our neurons, is involved in blood glucose regulation. analyses also showed that glucose-induced β-cell Notably, the delayed glucose response in the isolated responses in perfused embryonic zebrafish islets zebrafish islets is similar to that seen in mammalian are not well-synchronized, while β-cells in per- islets. While this delay is attributed to the time taken 61,62 fused mammalian islets show well-synchronized for glucose diffusion and metabolism, it might be 2+ 56– [Ca ] oscillations throughout the whole islet. further affected by the cleavage of neural connection 58 2+ 60 In agreement with very recent ex vivo Ca during the islet isolation procedure. Measurements recordings of perfused juvenile zebrafish islet, of real-time glucose response of pancreatic mouse the data suggest that zebrafish islets are uncoupled, islets by autofluorescence NAD(P)H signals also sug- at least in the developing zebrafish. gests that there might be differences in glucose meta- We also evaluated different approaches for glu- bolism after isolation. However, the time to reach cose administration in living zebrafish embryos. It is the initial GCaMP fluorescence peak after glucose important to note that our in vivo imaging data, in injection in mice in vivo (~1–2minutes)is still much test) (D) Quantification of GCaMP6s fluorescence intensity ex vivo in wild-type and Ca 1.2 mutant larvae shown as area under curve (AUC) upon 2 mM and 20 mM glucose stimulation normalized to the treatment duration (n = 9 and 10 larvae, respectively) (***p < 0.0005, two-tailed, unpaired t-test). (E) Relative normalized Ca 1.2 and Ca 1.3a mRNA expression levels as determined by RT- V V qPCR (reverse transcriptase quantitative PCR) in 5 dpf old wild type and Ca 1.2 mutant larvae (pools of 15 whole larvae) (n = 3 biological replicates, 2 technical replicates) (F) Relative normalized Ca 1.2, Ca 1.3a, insulin and neuroD mRNA expression levels V V determined by RT-qPCR in isolated pancreatic islets from 5 dpf old wild type and Ca 1.2 mutant larva (pools of 15 islets) (n = 3 biological replicates, 2 technical replicates). Data are shown as mean values ±s.e.m (*p < 0.05, two-tailed t-test). FIU = Fluorescence Intensity Units (AU, Arbitrary Units). 232 R. LORINCZ ET AL. 35 40 slower than in our study in zebrafish (˜5 sec). inhibition of first-phase insulin secretion. The Therefore, we also cannot exclude the possibility expression of Ca 1.2 not only in β-cells but rather that the fast response in zebrafish is caused by in all pancreatic endocrine cells leaves other pos- bypassing glucose metabolism, with glucose activat- sibilities for non-cell autonomous β-cell defects 2+ ] by other mechanisms, such as the swel- ing [Ca due to paracrine effects such as dysfunctional α- ling-sensing LRRC8 channels, for example, similar or δ-cells. A disturbed intra-islet communication 64 69 to mouse β-cells. Detailed analyses is required to (for a review see ref. ) could explain why β-cell m458 investigate the potential function of the LRRC8 pro- phenotypes in Ca 1.2/isl mutants appear to be teins in zebrafish β-cells, but expression profiling of completely different from those seen in β-cell spe- zebrafish islets cells suggests that the 5 lrrc8 genes are cific Ca 1.2 knock-out mice. 52,65 differentially expressed between these cell types. Our demonstration of glucose-dependence of 2+ Our finding that both palmitate and [Ca ] provides support for zebrafish as an appro- 2+ L-glutamine resulted in elevated [Ca ] correlates priate model to study stimulus-secretion coupling 2+ with increased Ca signalling and insulin secre- in the pancreas, both for fundamental understand- tion in mammalian islets exposed to free fatty ing of secretion control, and for development of acids and amino acids. However, and in contrast novel approaches to β-cell related diseases. to mammalian islets, zebrafish embryonic islets However, in defining distinct consequences of 2+ did not require the simultaneous presence of glu- manipulation of the voltage-gated L-type Ca 2+ 47,66,67 cose for elevated [Ca ] signals. channel, our work also highlights the need to Our studies further reveal a novel central role of determine underlying differences between mam- Ca 1.2 in β-cell specific stimulus-secretion cou- malian and zebrafish β-cell physiology to better pling in zebrafish. We find that genetic loss of understand this framework for applying zebrafish 1.2 but not of Ca 1.3a, results in elevated Ca as a diabetes model. V V glucose level during development. Further, we m458 show that β-cells in Ca 1.2/isl mutant zebra- 2+ Materials and methods fish display increased basal [Ca ] and a complete loss of glucose-responsiveness. Future studies will Zebrafish breeding and maintenance be required to unravel the underlying mechanism Zebrafish (Danio rerio) were maintained according of the increased basal activity of the Ca 1.2 trun- to standard protocols. Developing embryos were cated β-cells. The counterintuitive increase in [Ca staged by hours post fertilization (hpf) when incu- ] oscillations at basal glucose in the mutant cells bated at 28°C. All procedures were approved by could be caused by an overcompensation of other the Austrian Bundesministerium für Wissenchaſt VDCCs, as was seen in Ca 1.3 knockout mice. und Forschung (GZ BMWFW-66.008/0004-WF/ While our qPCR analyses revealed no evidence for II/3b/2014). up-regulation of Ca 1.2 and Ca 1.3a, this does V V not exclude the possibility of dysregulation of 2+ other Ca channels (N-, P/Q-type), that might Construction of Tg[ins:lynGCaMP6s,ins:H2B:RFP] be expressed and functional in the zebrafish islet. Importantly, our studies also suggest that β-cell The lyn membrane tag sequence was PCR-ampli- specific Ca 1.2 functions might be very different fied from pcDNA8D3cpv (Addgene plasmid in zebrafish and mice. Notably, Ca 1.2 mutant #36323; Primer1: 5′TGGGATCCCCACCATGGG mice are not viable and therefore comparable CTGCATCA-3′;Primer2:5′-TCTCTAGACCGG embryonic β-cell defects have not been determined TATCGGGTGGATCGATTCTCAGTTTGGGT- in the mouse mutants. However, β-cell specific GGGGTCTTCATGTCCACGC-3′)andligated 2+ loss of Ca 1.2 has only moderate effects on [Ca ] into the pME vector using the BamHI and XbaI V i handling and glucose responsiveness in mice, sites. Subsequently, the GCaMP6s sequence from despite significantly reduced Ca currents and Tol2 HuC GCaMP6s vector were added using V ISLETS 233 Age1 sites for ligation. The final construct was GFP rabbit (1:200, Abcam, ab6556); secondary assembled by 4-fragment Gateway cloning (LR antibody: Alexa Fluor 488 goat anti-rabbit Clonase II Enzyme mix; Invitrogen, 11–791-100) (1:1000, Roche, A32731). Immunostainings using pDestTol2CG2 ins:H2B-RFP (Addgene were performed as described above. plasmid), p5E insulin promoter, p3E polyA and the pME lynGCaMP6s plasmids. In vivo glucose injection and imaging Immunohistochemistry Larvae (3, 4 and 5 dpf) for live imaging were Immunostaining of whole embryos or larvae was anesthetized using cold-shock (embryos in egg performed as previously described. Larvae were water-containing dishes are set directly on ice) collected at 5 dpf, and fixed for 2 hours at room and immobilized by injection of ~10 nl of 1 mg/ temperature in 4% PFA/1% DMSO in PBS, then ml purified α-bungarotoxin (BTX, B137, Sigma- washed 3 × 5 minutes with 1 × PBS/0.1% Tween. Aldrich, T9641) either to the yolk or to the gut. Zebrafish embryos were incubated for one hour in After 10–15 minutes, embryos were embedded blocking solution (PBS/1%DMSO/1%BSA/1% in 1.2 % low melt agarose and imaged either on Triton), then incubated overnight at 4°C with pri- a Leica DM6000B microscope equipped with a mary antibodies diluted in blocking solution. SPOT-RT3 digital camera (Diagnostic Afterwards, washing was conducted with PBS+0.2% Instruments, Inc., Sterling Heights, MI) (intrave- Triton. Primary antibodies and dilutions were: gui- nous injection) or on a Zeiss Cell Observer SD nea pig anti-insulin (1:200, Dako, A0564), rabbit using a 25X water immersion objective (brain, anti-GFP (1:200, Torrey Pines Biolabs, yolk, pericardium injection). Time lapse images NC9589665). Secondary antibodies (1:1000 dilution) were collected with 500 ms–1000 ms time inter- were Alexa-conjugated (Invitrogen). Embryos were vals, 50–100 ms exposure for GCaMP6s, and washed and then reblocked, and incubated in sec- 150 ms forRFP,2×2binning. Foradjusting ondary antibody overnight at 4°C. the needle (GB120F-10, 69 × 1,20 × 100 mm with filament, Science Products GmbH) into the dorsal aorta, or into the hindbrain ventricle, Whole-mount in situ hybridization (WISH) yolk and pericardium, 5X or 10X dry objectives, WISH was performed as described previously. and for intravenous injection of ~1–2nlof Antisense probes were Digoxigenin (DIG RNA 0.5 M glucose (Carl Roth GmBH) (final concen- Labeling Mix, Roche, 11277073910) labelled and tration of ~10–20 mM, assuming the blood visualized with α-Digoxigenin-AP antibody volume of zebrafish embryo is 60 nl) (79) in (1:4000, Roche, 11093274910). The following combination with 1.25% Rhodamine B isothio- probes were used: cyanate-Dextran (R9379, Sigma-Aldrich) to con- Ca 1.2: Ca 1.2 Xba S: GGACAACTTTG firm intravenous injection, 40X water objective V V ACTATTTAACACGG; Ca 1.2_R: CTGTTCCC wasused(Sup.Video1). In vivo imaging of CCTACAAACTCG hindbrain ventricle, yolk, pericardium glucose Ca 1.3a: Ca 1.3a 5′ S: CTGAAAGGAAAGA injection (˜9 nl of 1 M glucose in combination V V CCTCATCAAAGAG, Ca 13a_R: ACCAGGTGC with 1.25% Rhodamine B isothiocyanate- TATAAAGTGGTAAT Dextran, final concentration of ˜30 mM, assum- All antisense RNA probes were generated by ing the total volume of a 5 dpf old larva is linearization of the corresponding plasmids and 300 nl), unstimulated larvae and Ca 1.2/isl transcription was performed using T3, T7 or SP6 mutants crossed with ins:lynGCaMP6s was car- RNA polymerase. The eGFP of transgenic ani- ried out with a Zeiss Cell Observer SD spinning mals was visualized by immunostaining after the disc microscope using a 25X water immersion in situ hybridisation. Primary antibody: anti- objective. 234 R. LORINCZ ET AL. Larvae islet isolation exposure for GCaMP6s) while indicated solutions were changed using a handmade perfusion sys- Zebrafish embryos at 5 dpf were euthanized using tem (5 ml syringes connected to plastic tubes cold-shock, rolled onto their left sides and, using through 26 G needles). Salts and glucose were 26 G sterile hypodermic-needles, yolk and visc- purchased from either Roth GmbH or eral organs were removed by applying gentle AppliChem GmbH. Solutions containing 2 mM pressure using needles until fully separated. The and20mMglucose,20mML-glutamine islets were identified and confirmed by either (Thermo Fischer, 25030081), 1 mM palmitic GCaMP6s green fluorescence or RFP red fluores- acid (sodium salt, 98%, Thermo Fischer, cence. Not all of the islet surrounding tissue was 416700050) and 30 mM KCl were used. removed in order to avoid the collapse of the islet. Freshly dissected islets were transferred by glass pipette to RPMI (ThermoFisher 11875–093) Image analysis supplemented with 1 mM HEPES, antibiotic solu- tion (Sigma A5955, 10 mll−1 solution), 10% fetal Visiview soſtware (Visitron Systems, Puchheim, bovine serum and diluted with glucose-free RPMI Germany) (for in vivo intravenous glucose injection to final glucose concentration of 6.67 mM (zebra- experiments) and ZEN software (for ex vivo experi- fish islet media) in a 12-wells culture plate. Islets ments and in vivo brain, yolk, pericardium injection) were incubated at 28°C for 1.5–2hoursbefore were used to capture time lapses, and fluorescence imaging. images were processed using ImageJ (Fiji) ROI Analyzer. StackReg plugin in the ImageJ/Fiji was used for the recursive alignment of time series, as Glucose level measurement previously described. The ROI manager in ImageJ/ Glucose levels in whole larvae were measured Fiji was used to manually draw either rectangular using Glucose Assay Kit (BioVision, K606-100) selection (~3 × 5 μm) around one piece of the as previously described. In brief, pools of 10 GCaMP6s-expressing membranes located between whole larvae were collected in 200 µl cold PBS, two individual β-cells (referred as single β-cell, in ex and lysed by glass bead (0.5 mm) homogenization vivo experiments done by Zeiss spinning disc micro- in Precellys24 bead homogenizer (one cycle of 20 s scopy), or polygon selections around the whole islet or at 5000 rpm) (Peqlab). After centrifugation (10, 4 non-overlapping single-cell sized β-cell areas per 000 rpm for 30 s at 4°C), supernatants were trans- islet that were defined based on red nuclei RFP ferred into fresh tubes, and glucose assay was expression (called as β-cell area) (in vivo experiments carried out according to manufacturer’s recom- done by Leica epifluorescence microscopy) to mea- mendation. In all experiments, at least 3 biological sure the mean GCaMP6s fluorescence intensity (FIU) replicates (10 whole larvae), and 2 technical repli- for each time frame. The mean GCaMP6s FIU per β- cates were measured. cell/β-cell area or per islet was normalized to the baseline (which is calculated as the mean intensity during incubation in 2 mM glucose for ex vivo experi- Ex vivo imaging ments, and as the intensity in the first time point for in Islets from ins:lynGCaMP6s were placed into vivo experiments) and displayed as FIU/FIU baseline glass-bottomed culture plates containing 2 mM trace (Arbitrary Unit, AU) over time (sec or min). To glucose solution (2 mM glucose in KRBH salt correct for fluorescence decay due to photo bleaching, solution: 114 mM NaCl, 4.7 mM KCl, 1.16 mM non-linear regression analysis (2 phase exponential MgSO heptahydrate, 1.2 mM KH PO ,2.5 mM decay, without weight data points and fixing of para- 4 2 4 CaCl dihydrate, 5 mM NaHCO , 20 mM HEPES meters) was carried out using GraphPad Prism ver- 2 3 and 0.1 g/100 ml BSA) with agarose gel wells. sion 5.00 for Windows (GraphPad Software, La Jolla, Islets were imaged on a Zeiss Cell Observer SD California USA, www.graphpad.com). Total β-cell 2+ using a 40X water immersion objective (time ] dynamics (ex vivo) was quantified by two [Ca lapse with 500 ms-1 s time interval, 50 ms different methods, either by calculating the area ISLETS 235 under the curve (AUC) as described previously, or qPCR was performed in a CFX Connect Real-Time for quantitative curves in Figure 5(C) and Sup. Fig. 7, System (BioRad) using the HOT Fire-Pol EvaGreen the glucose and other nutrient (sucrose, L-glutamine, qPCR MicPlus(SolisBioDyne).Datashown is the palmitic acid) responses are shown normalized to the average of 3 biological replicates, presented as gene change in fluorescence in response to KCl (showing expression level relative to controls, after normaliza- maximum islet depolarization), because the maxi- tion to the housekeeping gene ef1alpha.PCR primer mum excitability of an islet or β-cell can vary, and pairs for Ca 1.2 (5′ to 3′): TCTTCA the intensity of the islet’s fluorescence can also vary. GGTGTGCGACAGGA (F), CTTCTCACAAGGG For AUC analysis, we calculated the average of the CGGTTTG (R), Ca 1.3a: CCTCAGGCTGT single FIU/FIU baseline traces for 2 min before sti- GCTTCTGCT (F), CACAGCTTGCCTGGCATA 2+ mulation (basal [Ca ] dynamics), and 6 min after CA (R), insulin: GCCCAACAGGCTTCTTCTAC 2+ 20 mM glucose stimulation (stimulated [Ca ] AAC (F), GCAGATTTAGGAGGAAGGAAACCC dynamics), and normalized to the treatment duration (R), neuroD: CTTTCAACACACCCTAGAGTTCC (Figure 6(D)). For calculating the fraction of high G (F), GCATCATGCTTTCCTCGCTGTATG (R), 2+ glucose-responsive area (Figure 4(C)), the Ca influx ef1alpha: TCTCTACCTACCCTCCTCTTGGTC (F), response was defined as positive when the normalized TTGGTCTTGGCAGCCTTCTGTG (R). GCaMP6s FIU within the selected β-cell area (in vivo) or single β-cell (ex vivo) showed >7% increase in comparison to the time point before glucose injection Statistics or 20 mM glucose stimulation, and this increase in Graphical and statistical analyses were performed fluorescence intensity was higher than the peak fluor- using GraphPad Prism version 5.00 for Windows, escence intensity at baseline. In all ex vivo experi- GraphPad Software, La Jolla, California USA, www. ments, β-cells that failed to show fluorescence graphpad.com. Significance was tested using two- increase (>7% compared to time point before stimu- tailed distribution and either t-test or ONE-way lation) upon KCl stimulation were excluded from ANOVA analysis with p < 0.05 considered significant. quantification. Cross-correlation analysis (Fig. 4(E), Data presented are generally representative of at least 5(A,B), Sup. Fig. 2B, 3) and identification of the three independent experiments (except as indicated). peak numbers for oscillation frequencies (Figures 4 Error bars represent standard error (s.e.m) (except as (D)and 6(C)) on the FIU/FIU baseline traces of indicated). individual β-cells/areas were done by PeakCaller in (required rise: 1%, required fall: 3%, max. MATLAB lookback: 5 pts, max. lookahead: 5 pts) Cross-correla- Disclosure of potential conflicts of interest tion analysis was done on the section where glucose stimulation was applied (excluding KCl section). No potential conflicts of interest were disclosed. −1 Oscillation frequencies (min ) are average peak number per minute of selected β-cells (>6 cells, ex vivo)or β-cell areas (4 non-overlapping single-cell Acknowledgments sized selected region based on RFP red expression, We thank Robin Kimmel, Alexandra Koschak, Pidder Jansen- in vivo) normalized to the time duration until oscilla- Dürr, Petronel Tuluc, Thorsten Schwerte for helpful discussions, tion was observed in the time series (˜30–60 sec, in Sonja Töchterle and Dzenana Tufegdzic for expert zebrafish care, vivo, ˜150–300 sec, ex vivo). and all members of the Molecular Biology Institute for technical assistance and advice. qPCR Funding Total RNA was prepared from dissected islets of 5 dpf old zebrafish larvae (pool of 15 islets) and from whole This project was supported by the Universität Innsbruck/ zebrafish larvae (pool of 15 larvae) using Trizol University of Innsbruck and a travel fellowship to R.L. by (Ambion), cDNA was prepared using the First the Marshallplan-Jubiläumsstiftung/Austrian Marshall Plan Strand cDNA synthesis kit (Thermo Scientific). Foundation. 236 R. LORINCZ ET AL. 12. Heimberg H, De Vos A, Vandercammen A, Van ORCID Schaftingen E, Pipeleers D, Schuit F. Heterogeneity in Reka Lorincz http://orcid.org/0000-0002-9077-3300 glucose sensitivity among pancreatic beta-cells is cor- Christopher H. Emfinger http://orcid.org/0000-0002-9130- related to differences in glucose phosphorylation rather than glucose transport. Embo J. 1993;12(7):2873–2879. Andrea Walcher http://orcid.org/0000-0001-6624-4284 13. Moss LG, Caplan TV, Moss JB. Imaging beta cell Michael Giolai http://orcid.org/0000-0003-4166-8202 regeneration and interactions with islet vasculature in Maria S. Remedi http://orcid.org/0000-0003-2048-472X transparent adult zebrafish. Zebrafish. 2013;10(2):249– Colin G. Nichols http://orcid.org/0000-0002-4929-2134 257. doi:10.1089/zeb.2012.0813. Dirk Meyer http://orcid.org/0000-0002-4020-4484 14. Andersson O, Adams BA, Yoo D, Ellis GC, Gut P, Anderson RM, German MS, Stainier DY. NIH Public Access. 2013;15(6):885–894. doi:10.1016/j. cmet.2012.04.018.Adenosine. References 15. Argenton F, Zecchin E, Bortolussi M. Early appearance of pancreatic hormone-expressing cells in the zebrafish 1. Krishnan K, Ma Z, Björklund A, Islam MS. Calcium embryo. Mech Dev. 1999;87(1–2):217–221. doi:10.1016/ signaling in a genetically engineered human pancreatic S0925-4773(99)00151-3. β-cell line. Pancreas. 2015;44(5):773–777. doi:10.1097/ 16. Zang L, Shimada Y, Nishimura N. Development of a MPA.0000000000000318. novel zebrafish model for type 2 diabetes mellitus. Sci 2. Almaça J, Liang T, Gaisano HY, Nam HG, Berggren Rep. 2017;7(1). doi:10.1038/s41598-017-01432-w. PO, Caicedo A. Spatial and temporal coordination of 17. Marty N, Dallaporta M, Thorens B. Brain glucose sen- insulin granule exocytosis in intact human pancreatic sing, counterregulation, and energy homeostasis. islets. Diabetologia. 2015;58(12):2810–2818. Physiology (Bethesda). 2007;22:241–251. doi:10.1152/ doi:10.1007/s00125-015-3747-9. physiol.00010.2007. 3. Dolenšek J, Špelič D, Klemen MS, Žalik B, Gosak M, 18. Polakof S, Mommsen TP, Soengas JL. Glucosensing Rupnik M, Stožer A. Membrane potential and calcium and glucose homeostasis: from fish to mammals. dynamics in beta cells from mouse pancreas tissue Comp Biochem Physiol - B Biochem Mol Biol. slices: theory, experimentation, and analysis. Sensors 2011;160(4):123–149. doi:10.1016/j.cbpb.2011.07.006. (Switzerland). 2015;15(11):27393–27419. doi:10.3390/ 19. Röder PV, Wu B, Liu Y, Han W. Pancreatic regulation s151127393. of glucose homeostasis. Exp Mol Med. 2016;48(3):e219. 4. Kenty JHR, Melton DA. Testing pancreatic islet func- doi:10.1038/emm.2016.6. tion at the single cell level by calcium influx with 20. Ro der PV, Wong X, Hong W, Han W. Molecular regula- associated marker expression. PLoS One. 2015;10 tion of insulin granule biogenesis and exocytosis. Biochem (4):1–17. doi:10.1371/journal.pone.0122044. J. 2016;473(18):2737–2756. doi:10.1042/BCJ20160291. 5. Stožer A, Dolenšek J, Rupnik MS. Glucose-stimulated 21. Gilon P, Henquin JC. Influence of membrane potential calcium dynamics in islets of langerhans in acute changes on cytoplasmic CA2+ concentration in an mouse pancreas tissue slices. PLoS One. 2013;8(1). electrically excitable cell, the insulin-secreting pancrea- doi:10.1371/journal.pone.0054638. tic B-cell. J Biol Chem. 1992;267(29):20713–20720. 6. Leibiger IB, Berggren P-O. Intraocular in vivo imaging of 22. Eames SC, Philipson LH, Prince VE, Kinkel MD. Blood pancreatic islet cell physiology/pathology. Mol Metab. sugar measurement in zebrafish reveals dynamics of 2017;6(9):1002–1009. doi:10.1016/j.molmet.2017.03.014. glucose homeostasis. Zebrafish. 2010;7(2):205–213. 7. Chandra R, Liddle RA. Recent advances in the regula- doi:10.1089/zeb.2009.0640. tion of pancreatic secretion. Curr Opin Gastroenterol. 23. Remedi MS, Emfinger C. Pancreatic β-cell identity in dia- 2014. doi:10.1097/MOG.0000000000000099. betes. Diabetes, Obes Metab. 2016 September;18:110–116. 8. Seth A, Stemple DL, Barroso I. The emerging use of doi:10.1111/dom.12727. zebrafish to model metabolic disease. Dis Model Mech. 24. JurczykA,Roy N,BajwaR,Gut P,Lipson K,Yang C, 2013;6(5):1080–1088. doi:10.1242/dmm.011346. Covassin L, Racki WJ, Rossini AA, Phillips N, et al. 9. Pan FC, Wright C. Pancreas organogenesis: from bud Dynamic glucoregulation and mammalian-like responses to plexus to gland. Dev Dyn. 2011;240(3):530–565. to metabolic and developmentaldisruptioninzebrafish. doi:10.1002/dvdy.22584. Gen Comp Endocrinol. 2011;170(2):334–345. doi:10.1016/ j.ygcen.2010.10.010. 10. Fernandez J, Valdeolmillos M. Synchronous glucose- 25. Tseng Y, Chen R, Lee J, Liu S, Lee S, Hwang P. Specific dependent [Ca2+](i) oscillations in mouse pancreatic islets expression and regulation of glucose transporters in of Langerhans recorded in vivo. FEBS Lett. 2000;477(1– zebrafish ionocytes. Am J Physiol-Reg I. 2009;297 2):33–36. doi:10.1016/S0014-5793(00)01631-8. (2):275–290. doi:10.1152/ajpregu.00180.2009. 11. Henquin JC. Triggering and amplifying pathways of reg- 26. Marín-Juez R, Rovira M, Crespo D, van der Vaart M, ulation of insulin secretion by glucose71. Diabetes. 2000;49 Spaink HP, Planas JV. GLUT2-mediated glucose uptake (11):1751–1760. doi:10.2337/diabetes.49.11.1751. ISLETS 237 and availability are required for embryonic brain develop- growth. Nat Commun. 2017;8(1). doi:10.1038/s41467- ment in zebrafish. J Cereb Blood Flow Metab. 2015;35 017-00461-3. (1):74–85. doi:10.1038/jcbfm.2014.171. 40. Schulla V, Renström E, Feil R, Feil S, Franklin I, Gjinovci A, 27. CohenM,KitsbergD,TsytkinS,Shulman M, Aroeti B, Jing XJ, Laux D, Lundquist I, Magnuson MA, et al. NahmiasY.Liveimaging of GLUT2glucose-dependent Impaired insulin secretion and glucose tolerance in beta trafficking and its inhibition in polarized epithelial cysts. cell-selective Ca(v)1.2 Ca2+ channel null mice. Embo J. Open Biol. 2014;4(7):140091. doi:10.1098/rsob.140091. 2003;22(15): 3844–3854.doi:10.1093/emboj/cdg389. 28. Emfinger CH, Welscher A, Yan Z, Wang Y, Conway H, 41. U.K. Prospective Diabetes Study Group. Overview of 6 Moss JB, Moss LG, Remedi MS, Nichols CG. years’ therapy of type II diabetes: a progressive disease. Expression and function of ATP-dependent potassium Diabetes. 1995;44:1249–1258. doi:10.2337/db06-0880. channels in zebrafish islet β-cells. R Soc Open Sci. CHS. 2017;4(2):160808. doi:10.1098/rsos.160808. 42. Namkung Y, Skrypnyk N, Jeong MJ, Lee T, Lee MS, 29. Li M, Maddison LA, Page-McCaw P, Chen W. Kim HL, Chin H, Suh PG, Kim SS, Shin HS. Overnutrition induces -cell differentiation through pro- Requirement for the L-type Ca(2+) channel alpha(1D) longed activation of -cells in zebrafish larvae. AJP subunit in postnatal pancreatic beta cell generation. Jci. Endocrinol Metab. 2014.doi:10.1152/ajpendo.00686.2013. 2001;108(7):1015–1022. doi:10.1172/JCI13310. 30. Pace CS, Price S. Electrical responses of pancreatic islet 43. Yang S-N, Berggren P-O. The role of voltage-gated cells to secretory stimuli. Biochem Biophys Res calcium channels in pancreatic beta-cell physiology Commun. 1972;46(4):1557–1563. doi:10.1016/0006- and pathophysiology. Endocr Rev. 2006;27(6):621– 291X(72)90785-1. 676. doi:10.1210/er.2005-0888. 31. Grynkiewicz G, Poenie M, Tsien RY. A new generation 44. Stolovich-Rain M, Enk J, Vikesa J, Nielsen FC, Saada of Ca2+ indicators with greatly improved fluorescence A, Glaser B, Dor Y. weaning triggers a maturation step properties. J Biol Chem. 1985;260(6):3440–3450. of pancreatic β cells. Dev Cell. 2015;32(5):535–545. 32. Bootman MD, Rietdorf K, Collins T, Walker S, doi:10.1016/j.devcel.2015.01.002. Sanderson M. Ca2+-sensitive fluorescent dyes and 45. Musson MC, Jepeal LI, Mabray PD, Zhdanova IV, intracellular Ca2+ imaging. Cold Spring Harb Protoc. Cardoso WV, Wolfe MM. Expression of glucose- 2013;8(2):83–99. doi:10.1101/pdb.top066050. dependent insulinotropic polypeptide in the zebrafish. 33. Matveyenko AV, Butler PC. Relationship between β- Am J Physiol Regul Integr Comp Physiol. 2009;297: cell mass and diabetes onset. Diabetes Obes Metab. R1803–R1812. doi:10.1152/ajpregu.00288.2009. 2008;10:23–31. doi:10.1111/j.1463-1326.2008.00939.x. 46. Newsholme P, Brennan L, Bender K. Amino acid meta- 34. Gotthardt M, Eizirik DL, Cnop M, Brom M. Beta cell bolism, β-cell function, and diabetes. Diabetes. 2006;55 imaging - a key tool in optimized diabetes prevention (Suppl. 2). doi:10.2337/db06-S006. and treatment. Trends Endocrinol Metab. 2014;25 47. Gravena C, Mathias PC, Ashcroft SJH. Acute effects of (8):375–377. doi:10.1016/j.tem.2014.02.002. fatty acids on insulin secretion from rat and human 35. Chen C, Chmelova H, Cohrs CM, Chouinard JA, Jahn islets of Langerhans. J Endocrinol. 2002;173(1):73–80. SR, Stertmann J, Uphues I, Speier S. Alterations in β- doi:10.1677/JOE.0.1730073. cell calcium dynamics and efficacy outweigh islet mass 48. MalaisseWJ,BestL,KawazuS,Malaisse-LagaeF,SenerA. adaptation in compensation of insulin resistance and The stimulus-secretion coupling of glucose-induced insu- prediabetes onset. Diabetes. 2016;65(9):2676–2685. lin release: fuel metabolism in islets deprived of exogenous doi:10.2337/db15-1718. nutrient. Arch Biochem Biophys. 1983;224(1):102–110. 36. Barros TP, Alderton WK, Reynolds HM, Roach AG, doi:10.1016/0003-9861(83)90193-5. Berghmans S. Zebrafish: an emerging technology for in 49. Zhang Y, Louchami K, Carpentier YA, Malaisse WJ, vivo pharmacological assessment to identify potential Sener A. L-glutamine and palmitate catabolism in pan- safety liabilities in early drug discovery. Br J Pharmacol. creatic islets from rats depleted in long-chain polyun- 2008;154(7):1400–1413. doi:10.1038/bjp.2008.249. saturated ω3 fatty acids. Cell Biochem Funct. 2008;26 37. Henke K, Bowen ME, Harris MP. Perspectives for (1):82–86. doi:10.1002/cbf.1403. identification of mutations in the zebrafish: making 50. Rottbauer W, Baker K, Wo ZG, Mohideen MA, Cantiello use of next-generation sequencing technologies for for- HF,Fishman MC.Growthand function of theembryonic ward genetic approaches. Methods. 2013;62(3):185– heart depend upon the cardiac- specific L-type calcium 196. doi:10.1016/j.ymeth.2013.05.015. channel alpha1 subunit. Dev Cell. 2001;1(2):265–275. 38. Kamel M, Ninov N. Catching new targets in metabolic 51. Sidi S, Busch-Nentwich E, Friedrich R, Schoenberger U, disease with a zebrafish. Curr Opin Pharmacol. Nicolson T. gemini encodes a zebrafish L-type calcium 2017;37:41–50. doi:10.1016/j.coph.2017.08.007. channel that localizes at sensory hair cell ribbon synapses. 39. Singh SP, Janjuha S, Hartmann T, Kayisoglu Ö, Konantz J, JNeurosci. 2004;24(17):4213–4223. doi:10.1523/ BirkeS,MurawalaP,AlfarEA,MurataK,EugsterA,etal. JNEUROSCI.0223-04.2004. Different developmental histories of beta-cells generate 52. Tarifeño-Saldivia E, Lavergne A, Bernard A, Padamata K, functional and proliferative heterogeneity during islet BergemannD,VozML,ManfroidI,Peers B. 238 R. LORINCZ ET AL. Transcriptome analysis of pancreatic cells across distant essential for swelling-activated Cl − currents and embryo- species highlights novel important regulator genes. BMC nic development in zebrafish. Physiol Rep. 2016;4(19): Biol. 2017;15(1):21. doi:10.1186/s12915-017-0362-x. e12940. doi:10.14814/phy2.12940. 53. Stainier DY, Fouquet B, Chen JN, Warren KS, Weinstein 66. Rorsman P, Braun M, Zhang Q. Regulation of calcium BM, Meiler SE, Mohideen MA, Neuhauss SC, Solnica- in pancreatic α- and β-cells in health and disease. Cell Krezel L, Schier AF, et al. Mutations affecting the forma- Calcium. 2012. doi:10.1016/j.ceca.2011.11.006. tion and function of the cardiovascular system in the 67. Liu Z, Jeppesen PB, Gregersen S, Chen X, Hermansen zebrafish embryo. Development. 1996;123:285–292. K. Dose- and glucose-dependent effects of amino acids 54. Granato M, van Eeden FJ, Schach U, Trowe T, Brand on insulin secretion from isolated mouse islets and M, Furutani-Seiki M, Haffter P, Hammerschmidt M, clonal INS-1E beta-cells. Rev Diabet Stud. 2008;5 Heisenberg CP, Jiang YJ, et al. Genes controlling and (4):232–244. doi:10.1900/RDS.2008.5.232. mediating locomotion behavior of the zebrafish 68. Seisenberger C, Specht V, Welling A, Platzer J, Pfeifer embryo and larva. Development. 1996;123(1):399–413. A, Kühbandner S, Striessnig J, Klugbauer N, Feil R, 55. Nicolson T, Rüsch A, Friedrich RW, Granato M, Hofmann F. Functional embryonic cardiomyocytes Ruppersberg JP, Nüsslein-Volhard C. Genetic analysis after disruption of the L-type α1C (Cav1.2) calcium of vertebrate sensory hair cell mechanosensation: the channel gene in the mouse. J Biol Chem. 2000. zebrafish circler mutants. Neuron. 1998;20(2):271–283. doi:10.1074/jbc.M006467200. doi:10.1016/S0896-6273(00)80455-9. 69. Brereton MF,Vergari E,ZhangQ,Clark A.Alpha-,delta- 56. GylfeE,GrapengiesserE,Hellman B. Propagationofcyto- and PP-cells: are they the architectural cornerstones of islet plasmic Ca2+ oscillations inclustersof pancreatic β-cells structure and co-ordination? J Histochem Cytochem. exposed to glucose. Cell Calcium. 1991;12(2–3):229–240. 2015;63(8):575–591. doi:10.1369/0022155415583535. 57. Martín F, Soria B. Glucose-induced [Ca2+]i oscillations in 70. Westerfield M. The zebrafish book. a guide for the single human pancreatic islets. Cell Calcium. 1996;20 laboratory use of zebrafish (Danio rerio). 5th ed.: (5):409–414. doi:10.1016/S0143-4160(96)90003-2. Univ Oregon Press, Eugene; 2007. 58. Jonkers FC, Jonas JC, Gilon P, Henquin JC. Influence 71. Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, of cell number on the characteristics and synchrony of Schilling TF. Stages of embryonic development of the Ca2+ oscillations in clusters of mouse pancreatic islet zebrafish. Dev Dyn. 1995;203(3):253–310. doi:10.1002/ cells. J Physiol. 1999;520(Pt 3):839–849. aja.1002030302. 59. Podlasz P, Jakimiuk A, Chmielewska-Krzesinska M, 72. Kimmel RA, Dobler S, Schmitner N, Walsen T, Kasica N, Nowik N, Kaleczyc J. Galanin regulates Freudenblum J, Meyer D. Diabetic pdx1-mutant zebra- blood glucose level in the zebrafish: a morphological fish show conserved responses to nutrient overload and and functional study. Histochem Cell Biol. 2016. anti-glycemic treatment. Sci Rep. 2015 August;5:14241. doi:10.1007/s00418-015-1376-5. doi:10.1038/srep14241. 60. Yang YHC, Kawakami K, Stainier DY. A new mode of 73. Thisse C, Thisse B. High-resolution in situ hybridiza- pancreatic islet innervation revealed by live imaging in tion to whole-mount zebrafish embryos. Nat Protoc. zebrafish. Elife. 2018;7:e34519. doi:10.7554/eLife.34519. 2008;3(1):59–69. doi:10.1038/nprot.2007.514. 61. Yamada K, Nakata M, Horimoto N, Saito M, Matsuoka 74. Guo Y, Veneman WJ, Spaink HP, Verbeek FJ. Three- H, Inagaki N. Measurement of glucose uptake and dimensional reconstruction and measurements of zebra- intracellular calcium concentration in single, living fish larvae from high-throughput axial-view in vivo ima- pancreatic ??-cells. J Biol Chem. 2000;275(29):22278– ging. Biomed Opt Express. 2017;8(5):2611. doi:10.1364/ 22283. doi:10.1074/jbc.M908048199. BOE.8.002611. 62. Kaminski MT, Lenzen S, Baltrusch S. Real-time analy- 75. Kimmel RA, Meyer D. Zebrafish pancreas as a model sis of intracellular glucose and calcium in pancreatic for development and disease. Elsevier Ltd; 2016. Vol. beta cells by fluorescence microscopy. Biochim Biophys 134. doi:10.1016/bs.mcb.2016.02.009. Acta. 2012;1823(10):1697–1707. doi:10.1016/j. 76. Schneider CA, Rasband WS, Eliceiri KW. NIH Image bbamcr.2012.06.022. to ImageJ: 25 years of image analysis. Nat Methods. 63. Li G, Wu B, Ward MG, Chong ACN, Mukherjee S, 2012;9(7):671–675. doi:10.1038/nmeth.2089. Chen S, Hao M. Multifunctional in vivo imaging of 77. Thévenaz P, Ruttimann UE, Unser M. A pyramid pancreatic islets during diabetes development. J Cell approach to subpixel registration based on intensity. IEEE Sci. 2016;129(14). doi:10.1242/jcs.190843. Trans Image Process. 1998;7(1):27–41. doi:10.1109/ 64. Kang C, Xie L, Gunasekar SK, Mishra A, Zhang Y, Pai 83.650848. S, Gao Y, Kumar A, Norris AW, Stephens SB, et al. SWELL1 is a glucose sensor regulating β-cell excitabil- 78. Artimovich E, Jackson RK, Kilander MBC, Lin YC, ity and systemic glycaemia. Nat Commun. 2018;9(1). Nestor MW. PeakCaller: an automated graphical inter- doi:10.1038/s41467-017-02664-0. face for the quantification of intracellular calcium 65. Yamada T,WondergemR,MorrisonR,YinVP,StrangeK. obtained by high-content screening. BMC Neurosci. Leucine-rich repeat containing protein LRRC8A is 2017;18(1). doi:10.1186/s12868-017-0391-y. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Islets Taylor & Francis

In vivo monitoring of intracellular Ca2+ dynamics in the pancreatic β-cells of zebrafish embryos

Loading next page...
 
/lp/taylor-francis/in-vivo-monitoring-of-intracellular-ca2-dynamics-in-the-pancreatic-RmuHsei4Cw

References (83)

Publisher
Taylor & Francis
Copyright
© 2018 The Authors. Published with license by Tayloy & Francis.
ISSN
1938-2022
eISSN
1938-2014
DOI
10.1080/19382014.2018.1540234
Publisher site
See Article on Publisher Site

Abstract

ISLETS 2018, VOL. 10, NO. 6, 221–238 https://doi.org/10.1080/19382014.2018.1540234 RESEARCH PAPER 2+ In vivo monitoring of intracellular Ca dynamics in the pancreatic β-cells of zebrafish embryos a b,c,d a a a Reka Lorincz , Christopher H. Emfinger , Andrea Walcher , Michael Giolai , Claudia Krautgasser , b,d b,c a Maria S. Remedi , Colin G. Nichols , and Dirk Meyer a b Institute of Molecular Biology/CMBI, University of Innsbruck, Innsbruck, Austria; Department of Cell Biology and Physiology, Washington University School of Medicine, St. Louis, MO, USA; Center for the Investigation of Membrane Excitability Diseases (CIMED), Washington University School of Medicine, St. Louis, MO, USA; Department of Medicine, Washington University School of Medicine, St. Louis, MO, USA ABSTRACT ARTICLE HISTORY Received 24 April 2018 Assessing the response of pancreatic islet cells to glucose stimulation is important for under- Revised 8 October 2018 standing β-cell function. Zebrafish are a promising model for studies of metabolism in general, Accepted 20 October 2018 including stimulus-secretion coupling in the pancreas. We used transgenic zebrafish embryos 2+ expressing a genetically-encoded Ca sensor in pancreatic β-cells to monitor a key step in KEYWORDS 2+ glucose induced insulin secretion; the elevations of intracellular [Ca ] . In vivo and ex vivo analyses cacna1c; Cav1.2 channel; 2+ of [Ca ] demonstrate that β-cell responsiveness to glucose is well established in late embry- i early zebrafish development; ogenesis and that embryonic β-cells also respond to free fatty acid and amino acid challenges. In GCaMP6s; glucose-sensing of beta cells; in vivo imaging vivo imaging of whole embryos further shows that indirect glucose administration, for example by 2+ yolk injection, results in a slow and asynchronous induction of β-cell [Ca ] responses, while 2+ intravenous glucose injections cause immediate and islet-wide synchronized [Ca ] fluctuations. Finally, we demonstrate that embryos with disrupted mutation of the Ca 1.2 channel gene cacna1c are hyperglycemic and that this phenotype is associated with glucose-independent 2+ [Ca ] fluctuation in β-cells. The data reveal a novel central role of cacna1c in β-cell specific stimulus-secretion coupling in zebrafish and demonstrate that the novel approach we propose – 2+ to monitor the [Ca ] dynamics in embryonic β-cells in vivo – will help to expand the under- standing of β-cell physiological functions in healthy and diseased states. Introduction Ultimately, however, it is desirable to incorpo- Assessing the response of pancreatic islet cells rate in vivo imaging of native intracellular [Ca to glucose stimulation is important for under- ] without interfering with the complex para- standing β-cell function in healthy and diseased crine signalling networks regulating islet activ- states. Until now, pancreatic β-cell physiology )Here ity in native tissue (for a review, see ref. has been analyzed mainly in isolated cell and we tested transgenic zebrafish embryos expres- 1–5 2+ islet systems. Importantly, β-cells under sing a genetically encoded Ca sensor in their these conditions likely exhibit different physiol- β-cells as a potential model for corresponding ogy when compared to cells in their natural in non-invasive in vivo applications. Non-mam- vivo environment. A key step in mammalian malian vertebrates such as zebrafish (Danio glucose-stimulated insulin secretion is the ele- rerio) have become relevant alternative organ- 2+ ].Non-invasive in vation of intracellular [Ca i isms to study diabetes and other metabolic 2+ ] has recently been facili- vivo imaging of [Ca i diseases, due to ease of genetic manipulation, tated by transplantation of pancreatic islets into high reproduction rates, and ready access for in the anterior chamber of the eye or the kidney vivo imaging due to larval transparency. capsule of mice. Such real-time monitoring Importantly, pancreata in mammals and in zeb- facilitates the study of islet physiology and vas- rafish have conserved physiological endocrine cularization longitudinally, and enables in vivo and exocrine function, similar cellular architec- screening of novel drugs and treatments. ture, and conserved expression and function of CONTACT Dirk Meyer dirk.meyer@uibk.ac.at Institute of Molecular Biology/CMBI, University of Innsbruck, Technikerstrasse 25, Innsbruck 6020, Austria. Supplemental data for this article can be accessed on the publisher's website. © 2018 The Authors. Published with license by Tayloy & Francis. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 222 R. LORINCZ ET AL. 9 2+ 30–32 most developmental genes. Accordingly, the using Ca -sensitive dyes. More recently, 2+ zebrafish has proven highly productive for stu- genetically encoded Ca indicators have been 10–12 dies of pancreas development, and introduced as tools for non-invasive approaches to 13,14 regeneration. In mammals as well as in study excitable cells such as β-cells. The analysis of 2+ zebrafish the pancreas develops from the endo- transients in corresponding models systems Ca dermal germ layer and later compromises endo- will help in understanding underlying causes of β- crine and exocrine tissue. Within one day of cell dysfunction (for example in the context of 33,34 2 development, the zebrafish embryo forms a sin- diabetes risk factors. ) In vivo imaging of [Ca gle ‘primary’ pancreatic islet with ~60–70 ] dynamics in transplanted (intraocular) mouse mono-hormonal α-, β-, δ-, ε-cells. As develop- pancreatic islets showed reduction at a prediabetic 2+ ment proceeds, the primary islet increases in stage, suggesting the potential of [Ca ] as a func- size and additional secondary islets are formed, tional marker to evaluate β-cells in vivo in diseased to adopt to growth-related requirements. states. Nevertheless, the low litter numbers, com- Glucose metabolism in zebrafish is also very plicated methodologies, high costs, and large similar to mammalian glucose metabolism, and experiment variability limit the use of mammalian species for such studies. Since genetic manipulation overfed zebrafish displays obesity-related diabetes is easy and inexpensive in transparent zebrafish phenotypes including impaired glucose tolerance The molecular larvae, large-scale drug and genetic screens as well and increased insulin production. basis of glucose detection is well understood in as fluorescent imaging applications for studying β- 36–38 mammalian pancreatic β-cells (for a review, see cell physiology in living animals are feasible. 17 18 2+ ref. and. ) Glucose is taken up by the facilitative Monitoring of the glucose-stimulated Ca influx in glucose transporter (GLUT) GLUT2/SLC2A2, and isolated juvenile and larvae zebrafish islets has been We recently generated a is metabolized through glycolysis and oxidative previously described. phosphorylation, thereby generating adenosine tri- transgenic line expressing the genetically encoded 2+ phosphate (ATP) and increasing the ATP/ADP Ca -sensor GCaMP6s, and in this study we have 11,19 2 ratio. The altered [ATP/ADP] ratio in the β- established protocols for in vivo monitoring of [Ca + + cell then leads to the closure of ATP-sensitive K - ] dynamics in late embryonic animals. We found channels (K -channels), depolarization of the that β-cells in larvae show rare baseline sponta- ATP membrane, and consequent opening of voltage- neous activities, but after intravenous injections of dependent calcium channels (VDCCs). The glucose, virtually all β-cells showed very rapid ele- 2+ influx of Ca then triggers release of insulin vations in GCaMP6s fluorescence. from secretory granules. Orthologues of all Additionally, we have used the GCaMP6s major genes (GLUT, K -channels and VDCCs) reporter to assess β-cell associated function of ATP −/- involved in mammalian glucose sensing and insu- VDCCs. β-cell specific Ca 1.2 mice exhibit lin secretion are also expressed in zebrafish, and mild basal hyperglycemia, impaired glucose tol- 22–25 show functional similarities. Studies suggest erance and lack of first-phase insulin secretion,- 40,41 that glucose uptake in zebrafish, similar to mam- but this important role in mammalian mals, occurs through GLUT transporters, with glucose-induced insulin secretion has not been 40–43 GLUT2 expression found in the endocrine pan- assessed in zebrafish. 26,27 Furthermore, we and creas of zebrafish larvae. others recently demonstrated that zebrafish islet β- channels with con- cells express functional K Results ATP served structure and metabolic sensitivity to their 2+ Zebrafish β-cell [Ca ] responses depend on the mammalian counterparts, supporting the use of route of glucose administration in vivo zebrafish as a model animal in islet glucose sensing 28,29 2+ and diabetes research. To enable in vivo studies of β-cell [Ca ] in Excitability of β-cells has been investigated by zebrafish larva, we generated a transgenic line multiple strategies including monitoring of the expressing a genetically encoded membrane- 2+ membrane potential by electrical recordings, and tagged Ca -sensor together with a Histone- ISLETS 223 2+ tagged RFP, both under control of the insulin β-cell specific Ca changes either in the living promoter (Tg[ins:lynGCaMP6s; ins:H2B-RFP], embryo (Figure 1(C)) or in isolated islets referred to hereafter as ins:lynGCaMP6s; (Figure 1(D)). In the first set of experiments, Figure 1(A,B)). The combined expression of we recorded β-cell responses to different routes lynGCaMP6s and H2B-RFP in β-cells enabled of glucose administration. To simulate physio- rapid β-cell identification and measurements of logically relevant feeding-related gradual and 2+ Ca responses in selected β-cell areas. We continued increase of blood glucose level, we further established protocols for determining tested injections of glucose into the yolk, the Figure 1. Specific expression of GCaMP6s sensor in pancreatic β-cells enables the study of Ca2+ fluxes in live zebrafish. (A) Illustration of the construct used to generate Tg[ins:lynGCaMP6s ins:H2BtagRFP] (ins:lynGCaMP6s). The GCaMP6s sensor was linked N-terminal with a lyn membrane tag. The additional ins:H2BtagRFP sequence was added to the construct to ease the screening of transgenic embryos and to identify the β-cells due to the nuclear specific H2BtagRFP signal. (B) Antibody staining on 5 days post fertilization (dpf) ins:lynGCaMP6s embryos detecting >95% overlapping of endogenous insulin and GCaMP6s (using polyconal GFP antibody) and >80% overlapping of H2B:RFP and GCaMP6s (n = 12 larvae, 20 β-cells per larva were analysed). (C) Schematic outline of intravenous glucose injection of zebrafish larvae. Using α-bungarotoxin (BTX) protein injection, the zebrafish larvae are paralyzed. After 10–15 min, the larvae are embedded in low melt agarose. Using a micromanipulator assembled to the microscope, glucose (in combination with Rhodamine B isothiocyanate-Dextran) is injected to the dorsal aorta of the embryo while the imaging is running. (D) Schematic outline of perfusion system on extracted pancreatic islet of zebrafish embryos. Islets from ins: 2+ lynGCaMP6s fish are isolated and placed into glass-bottomed culture plates with agarose gel wells containing islet media for Ca imaging as solutions are changed. Scale bar = 10 µm 224 R. LORINCZ ET AL. responsive cells varying from 5–30% (Sup. pericardium or the hindbrain ventricle. In all 2+ Fig. 1A–C). In contrast, intravenous injection approaches, the glucose injection induced Ca of D-glucose (~10–20 mM final concentration) fluctuation in β-cells (Sup. Fig. 1D–F), but the in wild-type5 dpflarvaeevokedrapid and responses were very variable, with delays robust elevations in the GCaMP6s fluorescence between injection and response ranging from in almost all labelled cells (Figure 2(A), blue secondstominutes andthe numberof Figure 2. Zebrafish β-cells show rapid glucose response in vivo. (A) In vivo experiments in larva revealed rapid fluorescence signal alteration (peak within 5 sec) in response to intravenous injection of >10 mM D-glucose (blue line) (5 dpf) (n = 8 larvae). To confirm that the response is due to D-glucose metabolism, L-glucose (red line) using same concentration (>10 mM) was also injected and showed no response (n = 5 larvae). (B) Fluorescence image series of the islet showing β-cell specific GCaMP6 signal at the indicated time-points (sec) after intravenous injection of D-glucose (>10 mM final concentration) in 5 dpf larva. Scale bar = 10 µm (C) Quantitative GCaMP6 signal intensity in the larvae shown in (B), and two other individual larvae (gray dotted lines), showing calcium influx after sequential glucose injections (1.injection: >10 mM, 2. Injection: >20 mM, 3. Injection: >30 mM estimated final concentration). FIU intensities for the whole islet were determined using Image-J and normalized to the FIU baseline (intensity measured in the first time point). (D) Representative fluorescence traces of manually selected β-cell areas (n = 4, based on red RFP expression as indicated) in a 5 dpf old zebrafish larva, normalized to initial fluorescence intensity, while 1–2 nl of 0.5 M D-glucose (>10 mM final concentration) was intravenously injected in vivo (at dotted black line). After glucose injection β-cell areas show oscillation. (E) Cross-correlation matrix (determined by PeakCaller, MATLAB) of the 4, in (D) indicated, β-cell areas after intravenous glucose injection. The panel represent one islet with 4–4 cell comparison. Colour map key is given to the right of the panel. Further examples are shown in Sup. Fig. 2B. FIU = Fluorescence Intensity Units (AU, Arbitrary Units). ISLETS 225 line). Sequential intravenous glucose injections Zebrafish islets are uncoupled ex vivo resulted in repeated similar rapid responses, but To assess the concentration-dependence of glucose the normalized Fluorescence Intensity Unit and other nutrients on β-cell excitation, we per- (FIU) peaks decreased after the second and formed perfusion studies with isolated pancreatic third injection (Figure 2(B,(C)). Control injec- islets (ex vivo) from 5 dpf old zebrafish larvae tions with the same concentration of L-glucose (Figure 1(D)). Isolated larval islets were first incu- caused no GCaMP6s signals (Figure 2(A), red bated in 2 mM glucose solution, then stimulated line), confirming that the response is due to with 20 mM glucose solution and at the end of the glucose metabolism, rather than osmotic stress recording, 30 mM KCl solution was applied to or other potential injection artefacts. In intra- confirm excitability. In these experiments, higher venously D-glucose-injected embryos, the fluor- resolution imaging by spinning disc microscopy escence increased almost immediately after enabled single cell analysis of the isolated islets. injection, and reached its peak within 5 seconds, Consistent with the in vivo data, 20 mM glucose followed by oscillations across the islets (5 ± 1 increased GCaMP6s FIU/FIU baseline intensity peaks/minute; Figure 2(D), Sup.Video 1 and 2). (Figure 4(A,B)), almost all β-cells (83% ± 24%, Correlation analyses revealed that glucose- Figure 4(C)) responded to the glucose stimulus, induced responses are synchronized across the 2+ and the induced [Ca ] oscillated with similar islet when the glucose is injected intravenously frequencies (5 ± 1 and 6 ± 1 peaks/minute, respec- (Figure 2(E)andSup. Fig. 2B), which was not tively, Figure 4(D)). However, in contrast to what thecasefor yolk injection(Sup. Fig. 2C), indi- was observed after intravenous glucose injection, cating that the in vivo responsiveness of zebra- ex vivo β-cell responses were out of phase (correla- fish β-cells to glucose depends on the route of tion analyses shown Figure 4(E,C), Sup. Fig. 3 and glucose administration. 2B). Notably, the time to peak in the ex vivo studies were highly variable (ranging from a few seconds to minutes) and on average were more Zebrafish islets are glucose responsive in early than 30 times longer than those seen after intra- developmental stages venous glucose injection (4.4 ± 0.5 sec in vivo and 140 ± 30 sec ex vivo, Figure 4(F)). We next assessed whether the glucose-responsive- ness of β-cells observed in 5 dpf larvae is also seen at earlier developmental stages. Studies in rodents L-glutamine and palmitic acid also stimulate have established that islet maturity and function 2+ [Ca ] influx in zebrafish pancreatic islet vary during development, with glucose responsive- ness occurring in later stages, particularly near the In mammalian β-cells, excitation occurs not only weaning transition. Previous studies showed that in response to glucose but also to various other 46–49 insulin-dependent glucose homeostasis in zebra- stimuli. To test whether amino acids and free 2+ fish starts at 3 dpf. Normalized FIU intensities fatty acids contribute to [Ca ] signalling of zeb- 2+ showed similar glucose-induced [Ca ] dynamics i rafish pancreatic islets, perfusion experiments with over time (Figure 3(a)), and quantification of the L-glutamine and palmitic acid on extracted islets GCaMP6s signal revealed significantly higher peak were carried out. Whole islets extracted from 5 dpf FIU intensities after glucose injection compared to zebrafish larvae showed increase in GCaMP6s baseline (before glucose), but no significant differ- FIU/FIU baseline intensity when 20 mM ences between developmental stages (3, 4 and L-glutamine or 1 mM palmitic acid alone and no 5 dpf) (Figure 3(b)). In addition, little or no spon- glucose was applied in the perfusion system taneous responses were observed in unstimulated (Figure 5(A,B)). To confirm that the responses larvae at 3–5 dpf (Figure 3(c,d)). This shows that were specific to the nutrients and were not due the embryonic onset of insulin-dependent glucose to osmotic stress, we performed perfusion experi- homeostasis correlates with the requirement for ments with 20 mM sucrose on the extracted pan- glucose-sensing mechanisms after 3 dpf. creatic islet; no significant increase in GCaMP6s 226 R. LORINCZ ET AL. Figure 3. Similar in vivo glucose responses in 3 dpf, 4 dpf and 5 dpf zebrafish larvae. (A) Quantitation GCaMP6 signal intensity show rapid fluorescence signal alteration after intravenous injection of D-glucose (>10 mM final concentration) in 3, 4 and 5 dpf larvae (n = 5, 3 and 8 larvae, respectively), (C) and no or little alteration in unstimulated larvae (n = 8, 8 and 6 larvae, respectively). FIU intensities for the whole islet were determined using Image-J and normalized to the FIU baseline (intensity measured in the first time point) (B) Peak FIU intensities (FIU ) of the whole islets before and after glucose peak injection (normalized to baseline FIU intensity measured in the first time point) (curves shown in (A) at different developmental stages (3, 4 and 5 dpf) (n = 5, 3 and 8 larvae, respectively). (D) Peak FIU intensities (FIU ) of the whole islets in unstimulated larvae peak (normalized to baseline FIU intensity measured in the first time point) (curves shown in (C)) at different developmental stages (3, 4 and 5 dpf) (n = 8, 8 and 6 larvae, respectively). FIU = Fluorescence Intensity Units (AU, Arbitrary Units).Data are shown as mean values ±s.d., *p < 0.05 in one-way ANOVA, Tukey’s Multiple Comparison test to compare the glucose-injected groups (between 3, 4 and 5 dpf old larvae after glucose) in (B) and the unstimulated larvae in (D), and two-tailed, paired t-test was done to compare before and after glucose groups in (B). 2+ fluorescence intensity was observed. In all cases, acid, without glucose, stimulates [Ca ] influx in islets were excitable, as revealed by the change in embryonic zebrafish pancreatic islets. fluorescence upon 30 mM KCl stimulation 2+ fluctuations occurred (Figure 5(C)). Again, Ca m458 Ca 1.2/isl mutant zebrafish are asynchronously across the whole islet in all ex vivo hyperglycemic conditions, indicating that β-cells were not coupled with one other (Figure 5(A,B), right Next we used the ins:lynGCaMP6s reporter to panels). These results revealed that short-term assess β-cell specific defects in L-type calcium exposure to L-glutamine alone or to palmitic channel mutant zebrafish embryos. Currently, the ISLETS 227 Figure 4. Comparison of ex vivo and in vivo glucose responsiveness. (A) Fluorescence image series of isolated pancreatic islet showing GCaMP6 signal alteration at the indicated time-points shown in (B) applying 2 mM, 20 mM glucose and 30 mM KCl. Single β-cell is defined as ˜3× 5 μm white rectangular selection drawn around one 228 R. LORINCZ ET AL. genetic roles of Ca 1.2 and Ca 1.3 in glucose of free glucose in control and mutant embryos V V homeostasis are not well defined. Physiological showed elevated glucose levels in 4–5 dpf Ca 1.2/ m458 studies indicate that Ca 1.2 and Ca 1.3 are key isl (Figure 6(A) and Sup. Fig. 5) but normal V V tc323d regulators of mammalian glucose-induced insulin glucose levels in 5 dpf Ca 1.3a/gem mutants ), but correspond- (Sup. Fig. 6). These data suggest impaired β-cell secretion (for a review, see ref. m458 1.2/isl mutant fish. ing mouse knockout models displayed only minor functionality in the Ca m458 effects on insulin secretion and glucose homeosta- To test whether Ca 1.2/isl mutants affect β- 2+ sis, perhaps due to compensation by other chan- cell specific [Ca ] dynamics, the mutation was nels. In particular, it was found that Ca 1.2 in β- crossed into the ins:lynGCaMP6s background. In cells is required for proper first-phase glucose- vivo fluorescence recordings of untreated 5 dpf induced insulin secretion, and that Ca 1.3 may transgenic larvae showed significantly higher 2+ modulate second-phase secretion, each of which basal islet [Ca ] dynamics in mutants than con- 40,43 is significant for normal glucose tolerance. trol animals (Figure 6(C), Sup. Video 3 and 4). The zebrafish has three Ca 1.2 and Ca 1.3 encod- Hyperglycemia in these animals raised the possi- V V ing gene orthologues, of which two – cacna1c bility that non-islet effects could underlie the 2+ m458 (encoding Ca 1.2) and cacna1da (Ca 1.3a) – increased [Ca ] dynamics in the Ca 1.2/isl V V i V show expression in the developing pancreatic mutants. To eliminate non-islet effects, we per- 50–52 islet. The third gene cacna1db (Ca 1.3b)is formed ex vivo recordings of isolated mutant and 2+ exclusively expressed in the central nervous control islets. Importantly, increased [Ca ] system. To test cell type specific expression of dynamics in mutants were maintained in isolated Ca 1.2 and Ca 1.3a, we performed whole mount islets and appeared to be glucose independent. In V V stains in different endocrine GFP-reporter stains particular, we found that in mutant islets the 1.2 and Ca 1.3a (Sup. Fig. 4). We found that Ca change from 2 mM to 20 mM glucose had no V V are expressed in all hormone-positive pancreatic significant effect on the peak frequency (2 mM: cells of embryos older than one day (Sup. Fig. 4.). 0.6 ± 0.3 and 2.6 ± 0.7 peaks/min; 20 mM: To determine requirements of these genes in glu- 6.0 ± 1.0 and 1.8 ± 0.5, control and mutant, cose homeostasis, we took advantage of previously respectively) (Figure 6(C)), or on the integrated m485 described mutants for Ca 1.2 (isl, isl , island signal intensity (2 mM: 0.018 ± 0.004 and 50,53 tc323d beat) and Ca 1.3a (gem, gem , gemini).- 0.065 ± 0.008 AUC; 20 mM: 0.057 ± 0.013 and 54,55 Both mutant alleles carry premature stop 0.074 ± 0.011, control and mutant, respectively) codons that result in the expression of truncated (Figure 6(D) and Sup. Fig 7A), while in control (and most likely inactive) proteins. Measurements islets these parameters were significantly increased piece of the GCamP6s-expressing membrane located between two individual β-cells. Scale bar = 10 µm. (B) Single cell GCaMP6s fluorescence traces for isolated islet from wild-type zebrafish larvae (5 dpf) after applying 2 mM, 20 mM glucose followed by 30 mM KCl. FIU intensities for the single cells (n = 6 cells) were determined using Image-J, normalized to the FIU baseline (intensities measured in the first 2 min when 2 mM glucose was applied) and displayed as FIU/FIU baseline (AU) for single cells (coloured dotted line) and for the average of the singles cells in the islet (black line). (C) Quantification showing the percentage of β-cells that respond upon stimulation with 20 mM glucose, ex vivo and the percentage of glucose-responsive β-cell area after intravenous injection of >10 mM glucose in vivo. (n = 8 and 9 larvae were analyzed, in vivo and ex vivo, respectively) (D) Average oscillation frequencies of 4 selected β-cell areas of the pancreatic islet (n = 8 islets, in each islet 4 β-cell areas were manually selected based on the red RFP expression) after intravenous injection (~1–2 nl of 0.5 M glucose, >10 mM final concentration) of 5 dpf old zebrafish larvae (in vivo, imaged by epifluorescence microscope), and average oscillation frequencies of individual β-cells (n = 5 islets, in each islet >6 cells were analyzed) of the isolated pancreatic islets after 20 mM glucose stimulation (ex vivo, imaged by spinning disc microscope). Oscillation peaks were identified by PeakCaller, MATLAB on the FIU/FIU baseline traces of individual β-cells/areas (see Sup. Fig. 2A). The peak number was divided by the time duration until oscillation was observed (˜30–60 sec, in vivo, ˜150–300 sec, ex vivo) (E) Cross-correlation matrix (determined by PeakCaller, MATLAB) of the islet between individual cells shown in (A) and (B), with color key to the right of the matrix graph. Numbers on the x and y axes represent individual cells that are marked in (A) and (B). (F) Time to peak calculated from the time point when addition of the new glucose solution was finished in isolated islets, ex vivo (n = 7 islets, the addition took in average ˜38 seconds), and when glucose was injected in living zebrafish larvae, in vivo (n = 8 islets from n = 8 larvae). Data are shown as mean values ±s.e.m., *p < 0.05 in unpaired, two-tailed t-test between ex vivo and in vivo groups in (C), (D) and (F). FIU = Fluorescence Intensity Units (AU, Arbitrary Units). ISLETS 229 Figure 5. Stimulation of β-cell by L-glutamine and palmitic acid. Representative single cell GCaMP6s fluorescence traces for isolated islet from wild-type zebrafish larvae (5 dpf) after applying 2 mM glucose followed by 20 mM L-glutamine (A), and 1 mM palmitic acid (B) stimulation and 30 mM KCl. FIU intensities for the single cells (n = 10–10 cells) were determined using Image-J, normalized to the FIU baseline (intensities measured in the first 2 min when 2 mM glucose was applied) and displayed as FIU/FIU baseline (AU) for single cells (coloured dotted line) and for the average of the singles cells in the islet (black line). Cross-correlation matrices (determined by PeakCaller, MATLAB) of the islets between individual cells shown in (A) left panel and (B) left panel, with color key to the right of the matrix graph. Numbers on the x and y axes represent individual cells. Cross-correlation analysis was done on the section where L-glutamine/palmitate stimulation was applied (excluding KCl section). (C) Calcium responses to 20 mM L-glutamine and 1 mM palmitic acid when compared to 2 mM glucose. The changes in GCaMP6s fluorescence intensity in β-cells of wild-type pancreatic islets were averaged for the whole islet and the FIU peak intensity upon different nutrient stimulation was normalized to the FIU peak intensity upon KCl stimulation (n = 23, 7, 5, 4 and 7 islets were analyzed for 2 mM glucose, 20 mM glucose, 20 mM L-glutamine, 20 mM sucrose, and 1 mM palmitic acid stimulation, respectively) (in each islet at least n = 5–8 cells were analyzed and the FIU peak intensity was measured in the average FIU trace for the whole islet). Data are shown as mean values ±s.e.m., *p < 0.05 in unpaired t-test between 2 mM glucose and 20 mM sucrose, 20 mM L-glutamine, and 1 mM palmitic acid stimulation. FIU = Fluorescence Intensity Units (AU, Arbitrary Units). 230 R. LORINCZ ET AL. Figure 6. Ex vivo imaging of β-cells Ca + dynamics in Ca 1.2 mutant zebrafish larvae. i V m458 (A) Ca 1.2/isl mutant shows elevated whole larval glucose levels at 4 and 5 dpf. Glucose levels measured at 4 and 5 dpf whole larval extracts (pool of 10–10 larvae), normalized to the protein level measured by Nanodrop, and showed as relative to the average control value. (n = 4 and 8 biological replicates, respectively) (at 5 dpf, results combined from 2 independent experiments) (***p < 0.0001, two-tailed t-test). (B) Fluorescence image series of an isolated islet from 5 dpf old Ca 1.2 mutant crossed with ins: lynGCaMP6s larva (upper panel). Quantitation of GCaMP signal intensity in the time series shown in the upper panel, indicating FIU/ FIU baseline changes after 20 mM glucose as well as after 30 mM KCl stimulation (lower panel). (C) Quantification of GCaMP6s fluorescence intensity peaks determined by PeakCaller, MATLAB in vivo, without any stimulation (n = 6 wild type and 8 Ca 1.2 mutant larva), and ex vivo, upon 2 and 20 mM glucose stimulation in wild-type and Ca 1.2 mutant larva (n = 7–7 and 6–6 larvae, −1 respectively), shown as oscillation frequency (min ) (peak number/cell number*minute). β-cell number was defined based on red nuclei RFP expression in one focal plane (4–10 cells per islet were analyzed) (**p < 0.005, two-tailed, nonparametric Mann-Whitney ISLETS 231 by increasing glucose levels. Thus, there is a coun- contrast to the ex vivo data, did not allow analyses at 2+ terintuitive increase in [Ca ] oscillations at basal single cell resolution. The combination of the deep glucose in the mutant islets. In mouse, Ca 1.3 tissue localization, the frequent presence of light subunit knockout was found to be compensated absorbing pigment cells, and permanent slow tissue 1.2 upregulation. To test whether upregu- by Ca movements due to gut peristaltic movement pre- 1.3a may account for the Ca 1.2/isl- lation of Ca vented more detailed analyses. However, our data V V m458 mutant phenotype, we performed RT-qPCR revealed that embryonic in vivo β-cell responses are (reverse transcriptase quantitative PCR) analyses. immediate and synchronized when glucose is However, transcripts levels for Ca 1.2 and directly injected into the bloodstream, and that gra- Ca 1.3a in whole embryos (Figure 6(E)) and iso- dually increasing blood glucose levels result in 2+ lated islets (Figure 6(F)) showed no significant ] fluctuations, delayed, non-synchronized [Ca differences between controls and mutants. These very similar to those found in glucose-stimulated data show that genetic loss of Ca 1.2 results in perfused islets. Currently, we can only speculate glucose-independent hyperactivation of β-cells, about the mechanism underlying the synchronized and suggest that this activation is independent of responses after intravenous glucose injection. cell-autonomous compensatory effects. Possibly, the synchronicity is caused by a cell-auton- omous but simultaneous response of all β-cells to the simultaneous prompt exposure to a high glucose Discussion challenge. Alternatively, synchronicity could be trig- 2+ We have successfully assessed [Ca ] dynamics in gered by a currently undefined paracrine signal. In 2+ β-cells of zebrafish pancreatic islets, both in vivo ] dynamics in this context, the much faster [Ca after injection of glucose and ex vivo using per- living animals, as compared to isolated islets, could fused isolated islet. We found that zebrafish β-cells indicate the involvement of neuronal driven triggers. are directly glucose-responsive, and that respon- Consistent with this possibility, it was recently siveness is established early in embryonic develop- shown that the embryonic zebrafish islet is inner- ment. These data are consistent with a conserved vated at the relevant time points, and that a neuro- glucose sensing and stimulus-secretion coupling transmitter, namely galanin expressed in these 59,60 mechanism in fish and mammals. However, our neurons, is involved in blood glucose regulation. analyses also showed that glucose-induced β-cell Notably, the delayed glucose response in the isolated responses in perfused embryonic zebrafish islets zebrafish islets is similar to that seen in mammalian are not well-synchronized, while β-cells in per- islets. While this delay is attributed to the time taken 61,62 fused mammalian islets show well-synchronized for glucose diffusion and metabolism, it might be 2+ 56– [Ca ] oscillations throughout the whole islet. further affected by the cleavage of neural connection 58 2+ 60 In agreement with very recent ex vivo Ca during the islet isolation procedure. Measurements recordings of perfused juvenile zebrafish islet, of real-time glucose response of pancreatic mouse the data suggest that zebrafish islets are uncoupled, islets by autofluorescence NAD(P)H signals also sug- at least in the developing zebrafish. gests that there might be differences in glucose meta- We also evaluated different approaches for glu- bolism after isolation. However, the time to reach cose administration in living zebrafish embryos. It is the initial GCaMP fluorescence peak after glucose important to note that our in vivo imaging data, in injection in mice in vivo (~1–2minutes)is still much test) (D) Quantification of GCaMP6s fluorescence intensity ex vivo in wild-type and Ca 1.2 mutant larvae shown as area under curve (AUC) upon 2 mM and 20 mM glucose stimulation normalized to the treatment duration (n = 9 and 10 larvae, respectively) (***p < 0.0005, two-tailed, unpaired t-test). (E) Relative normalized Ca 1.2 and Ca 1.3a mRNA expression levels as determined by RT- V V qPCR (reverse transcriptase quantitative PCR) in 5 dpf old wild type and Ca 1.2 mutant larvae (pools of 15 whole larvae) (n = 3 biological replicates, 2 technical replicates) (F) Relative normalized Ca 1.2, Ca 1.3a, insulin and neuroD mRNA expression levels V V determined by RT-qPCR in isolated pancreatic islets from 5 dpf old wild type and Ca 1.2 mutant larva (pools of 15 islets) (n = 3 biological replicates, 2 technical replicates). Data are shown as mean values ±s.e.m (*p < 0.05, two-tailed t-test). FIU = Fluorescence Intensity Units (AU, Arbitrary Units). 232 R. LORINCZ ET AL. 35 40 slower than in our study in zebrafish (˜5 sec). inhibition of first-phase insulin secretion. The Therefore, we also cannot exclude the possibility expression of Ca 1.2 not only in β-cells but rather that the fast response in zebrafish is caused by in all pancreatic endocrine cells leaves other pos- bypassing glucose metabolism, with glucose activat- sibilities for non-cell autonomous β-cell defects 2+ ] by other mechanisms, such as the swel- ing [Ca due to paracrine effects such as dysfunctional α- ling-sensing LRRC8 channels, for example, similar or δ-cells. A disturbed intra-islet communication 64 69 to mouse β-cells. Detailed analyses is required to (for a review see ref. ) could explain why β-cell m458 investigate the potential function of the LRRC8 pro- phenotypes in Ca 1.2/isl mutants appear to be teins in zebrafish β-cells, but expression profiling of completely different from those seen in β-cell spe- zebrafish islets cells suggests that the 5 lrrc8 genes are cific Ca 1.2 knock-out mice. 52,65 differentially expressed between these cell types. Our demonstration of glucose-dependence of 2+ Our finding that both palmitate and [Ca ] provides support for zebrafish as an appro- 2+ L-glutamine resulted in elevated [Ca ] correlates priate model to study stimulus-secretion coupling 2+ with increased Ca signalling and insulin secre- in the pancreas, both for fundamental understand- tion in mammalian islets exposed to free fatty ing of secretion control, and for development of acids and amino acids. However, and in contrast novel approaches to β-cell related diseases. to mammalian islets, zebrafish embryonic islets However, in defining distinct consequences of 2+ did not require the simultaneous presence of glu- manipulation of the voltage-gated L-type Ca 2+ 47,66,67 cose for elevated [Ca ] signals. channel, our work also highlights the need to Our studies further reveal a novel central role of determine underlying differences between mam- Ca 1.2 in β-cell specific stimulus-secretion cou- malian and zebrafish β-cell physiology to better pling in zebrafish. We find that genetic loss of understand this framework for applying zebrafish 1.2 but not of Ca 1.3a, results in elevated Ca as a diabetes model. V V glucose level during development. Further, we m458 show that β-cells in Ca 1.2/isl mutant zebra- 2+ Materials and methods fish display increased basal [Ca ] and a complete loss of glucose-responsiveness. Future studies will Zebrafish breeding and maintenance be required to unravel the underlying mechanism Zebrafish (Danio rerio) were maintained according of the increased basal activity of the Ca 1.2 trun- to standard protocols. Developing embryos were cated β-cells. The counterintuitive increase in [Ca staged by hours post fertilization (hpf) when incu- ] oscillations at basal glucose in the mutant cells bated at 28°C. All procedures were approved by could be caused by an overcompensation of other the Austrian Bundesministerium für Wissenchaſt VDCCs, as was seen in Ca 1.3 knockout mice. und Forschung (GZ BMWFW-66.008/0004-WF/ While our qPCR analyses revealed no evidence for II/3b/2014). up-regulation of Ca 1.2 and Ca 1.3a, this does V V not exclude the possibility of dysregulation of 2+ other Ca channels (N-, P/Q-type), that might Construction of Tg[ins:lynGCaMP6s,ins:H2B:RFP] be expressed and functional in the zebrafish islet. Importantly, our studies also suggest that β-cell The lyn membrane tag sequence was PCR-ampli- specific Ca 1.2 functions might be very different fied from pcDNA8D3cpv (Addgene plasmid in zebrafish and mice. Notably, Ca 1.2 mutant #36323; Primer1: 5′TGGGATCCCCACCATGGG mice are not viable and therefore comparable CTGCATCA-3′;Primer2:5′-TCTCTAGACCGG embryonic β-cell defects have not been determined TATCGGGTGGATCGATTCTCAGTTTGGGT- in the mouse mutants. However, β-cell specific GGGGTCTTCATGTCCACGC-3′)andligated 2+ loss of Ca 1.2 has only moderate effects on [Ca ] into the pME vector using the BamHI and XbaI V i handling and glucose responsiveness in mice, sites. Subsequently, the GCaMP6s sequence from despite significantly reduced Ca currents and Tol2 HuC GCaMP6s vector were added using V ISLETS 233 Age1 sites for ligation. The final construct was GFP rabbit (1:200, Abcam, ab6556); secondary assembled by 4-fragment Gateway cloning (LR antibody: Alexa Fluor 488 goat anti-rabbit Clonase II Enzyme mix; Invitrogen, 11–791-100) (1:1000, Roche, A32731). Immunostainings using pDestTol2CG2 ins:H2B-RFP (Addgene were performed as described above. plasmid), p5E insulin promoter, p3E polyA and the pME lynGCaMP6s plasmids. In vivo glucose injection and imaging Immunohistochemistry Larvae (3, 4 and 5 dpf) for live imaging were Immunostaining of whole embryos or larvae was anesthetized using cold-shock (embryos in egg performed as previously described. Larvae were water-containing dishes are set directly on ice) collected at 5 dpf, and fixed for 2 hours at room and immobilized by injection of ~10 nl of 1 mg/ temperature in 4% PFA/1% DMSO in PBS, then ml purified α-bungarotoxin (BTX, B137, Sigma- washed 3 × 5 minutes with 1 × PBS/0.1% Tween. Aldrich, T9641) either to the yolk or to the gut. Zebrafish embryos were incubated for one hour in After 10–15 minutes, embryos were embedded blocking solution (PBS/1%DMSO/1%BSA/1% in 1.2 % low melt agarose and imaged either on Triton), then incubated overnight at 4°C with pri- a Leica DM6000B microscope equipped with a mary antibodies diluted in blocking solution. SPOT-RT3 digital camera (Diagnostic Afterwards, washing was conducted with PBS+0.2% Instruments, Inc., Sterling Heights, MI) (intrave- Triton. Primary antibodies and dilutions were: gui- nous injection) or on a Zeiss Cell Observer SD nea pig anti-insulin (1:200, Dako, A0564), rabbit using a 25X water immersion objective (brain, anti-GFP (1:200, Torrey Pines Biolabs, yolk, pericardium injection). Time lapse images NC9589665). Secondary antibodies (1:1000 dilution) were collected with 500 ms–1000 ms time inter- were Alexa-conjugated (Invitrogen). Embryos were vals, 50–100 ms exposure for GCaMP6s, and washed and then reblocked, and incubated in sec- 150 ms forRFP,2×2binning. Foradjusting ondary antibody overnight at 4°C. the needle (GB120F-10, 69 × 1,20 × 100 mm with filament, Science Products GmbH) into the dorsal aorta, or into the hindbrain ventricle, Whole-mount in situ hybridization (WISH) yolk and pericardium, 5X or 10X dry objectives, WISH was performed as described previously. and for intravenous injection of ~1–2nlof Antisense probes were Digoxigenin (DIG RNA 0.5 M glucose (Carl Roth GmBH) (final concen- Labeling Mix, Roche, 11277073910) labelled and tration of ~10–20 mM, assuming the blood visualized with α-Digoxigenin-AP antibody volume of zebrafish embryo is 60 nl) (79) in (1:4000, Roche, 11093274910). The following combination with 1.25% Rhodamine B isothio- probes were used: cyanate-Dextran (R9379, Sigma-Aldrich) to con- Ca 1.2: Ca 1.2 Xba S: GGACAACTTTG firm intravenous injection, 40X water objective V V ACTATTTAACACGG; Ca 1.2_R: CTGTTCCC wasused(Sup.Video1). In vivo imaging of CCTACAAACTCG hindbrain ventricle, yolk, pericardium glucose Ca 1.3a: Ca 1.3a 5′ S: CTGAAAGGAAAGA injection (˜9 nl of 1 M glucose in combination V V CCTCATCAAAGAG, Ca 13a_R: ACCAGGTGC with 1.25% Rhodamine B isothiocyanate- TATAAAGTGGTAAT Dextran, final concentration of ˜30 mM, assum- All antisense RNA probes were generated by ing the total volume of a 5 dpf old larva is linearization of the corresponding plasmids and 300 nl), unstimulated larvae and Ca 1.2/isl transcription was performed using T3, T7 or SP6 mutants crossed with ins:lynGCaMP6s was car- RNA polymerase. The eGFP of transgenic ani- ried out with a Zeiss Cell Observer SD spinning mals was visualized by immunostaining after the disc microscope using a 25X water immersion in situ hybridisation. Primary antibody: anti- objective. 234 R. LORINCZ ET AL. Larvae islet isolation exposure for GCaMP6s) while indicated solutions were changed using a handmade perfusion sys- Zebrafish embryos at 5 dpf were euthanized using tem (5 ml syringes connected to plastic tubes cold-shock, rolled onto their left sides and, using through 26 G needles). Salts and glucose were 26 G sterile hypodermic-needles, yolk and visc- purchased from either Roth GmbH or eral organs were removed by applying gentle AppliChem GmbH. Solutions containing 2 mM pressure using needles until fully separated. The and20mMglucose,20mML-glutamine islets were identified and confirmed by either (Thermo Fischer, 25030081), 1 mM palmitic GCaMP6s green fluorescence or RFP red fluores- acid (sodium salt, 98%, Thermo Fischer, cence. Not all of the islet surrounding tissue was 416700050) and 30 mM KCl were used. removed in order to avoid the collapse of the islet. Freshly dissected islets were transferred by glass pipette to RPMI (ThermoFisher 11875–093) Image analysis supplemented with 1 mM HEPES, antibiotic solu- tion (Sigma A5955, 10 mll−1 solution), 10% fetal Visiview soſtware (Visitron Systems, Puchheim, bovine serum and diluted with glucose-free RPMI Germany) (for in vivo intravenous glucose injection to final glucose concentration of 6.67 mM (zebra- experiments) and ZEN software (for ex vivo experi- fish islet media) in a 12-wells culture plate. Islets ments and in vivo brain, yolk, pericardium injection) were incubated at 28°C for 1.5–2hoursbefore were used to capture time lapses, and fluorescence imaging. images were processed using ImageJ (Fiji) ROI Analyzer. StackReg plugin in the ImageJ/Fiji was used for the recursive alignment of time series, as Glucose level measurement previously described. The ROI manager in ImageJ/ Glucose levels in whole larvae were measured Fiji was used to manually draw either rectangular using Glucose Assay Kit (BioVision, K606-100) selection (~3 × 5 μm) around one piece of the as previously described. In brief, pools of 10 GCaMP6s-expressing membranes located between whole larvae were collected in 200 µl cold PBS, two individual β-cells (referred as single β-cell, in ex and lysed by glass bead (0.5 mm) homogenization vivo experiments done by Zeiss spinning disc micro- in Precellys24 bead homogenizer (one cycle of 20 s scopy), or polygon selections around the whole islet or at 5000 rpm) (Peqlab). After centrifugation (10, 4 non-overlapping single-cell sized β-cell areas per 000 rpm for 30 s at 4°C), supernatants were trans- islet that were defined based on red nuclei RFP ferred into fresh tubes, and glucose assay was expression (called as β-cell area) (in vivo experiments carried out according to manufacturer’s recom- done by Leica epifluorescence microscopy) to mea- mendation. In all experiments, at least 3 biological sure the mean GCaMP6s fluorescence intensity (FIU) replicates (10 whole larvae), and 2 technical repli- for each time frame. The mean GCaMP6s FIU per β- cates were measured. cell/β-cell area or per islet was normalized to the baseline (which is calculated as the mean intensity during incubation in 2 mM glucose for ex vivo experi- Ex vivo imaging ments, and as the intensity in the first time point for in Islets from ins:lynGCaMP6s were placed into vivo experiments) and displayed as FIU/FIU baseline glass-bottomed culture plates containing 2 mM trace (Arbitrary Unit, AU) over time (sec or min). To glucose solution (2 mM glucose in KRBH salt correct for fluorescence decay due to photo bleaching, solution: 114 mM NaCl, 4.7 mM KCl, 1.16 mM non-linear regression analysis (2 phase exponential MgSO heptahydrate, 1.2 mM KH PO ,2.5 mM decay, without weight data points and fixing of para- 4 2 4 CaCl dihydrate, 5 mM NaHCO , 20 mM HEPES meters) was carried out using GraphPad Prism ver- 2 3 and 0.1 g/100 ml BSA) with agarose gel wells. sion 5.00 for Windows (GraphPad Software, La Jolla, Islets were imaged on a Zeiss Cell Observer SD California USA, www.graphpad.com). Total β-cell 2+ using a 40X water immersion objective (time ] dynamics (ex vivo) was quantified by two [Ca lapse with 500 ms-1 s time interval, 50 ms different methods, either by calculating the area ISLETS 235 under the curve (AUC) as described previously, or qPCR was performed in a CFX Connect Real-Time for quantitative curves in Figure 5(C) and Sup. Fig. 7, System (BioRad) using the HOT Fire-Pol EvaGreen the glucose and other nutrient (sucrose, L-glutamine, qPCR MicPlus(SolisBioDyne).Datashown is the palmitic acid) responses are shown normalized to the average of 3 biological replicates, presented as gene change in fluorescence in response to KCl (showing expression level relative to controls, after normaliza- maximum islet depolarization), because the maxi- tion to the housekeeping gene ef1alpha.PCR primer mum excitability of an islet or β-cell can vary, and pairs for Ca 1.2 (5′ to 3′): TCTTCA the intensity of the islet’s fluorescence can also vary. GGTGTGCGACAGGA (F), CTTCTCACAAGGG For AUC analysis, we calculated the average of the CGGTTTG (R), Ca 1.3a: CCTCAGGCTGT single FIU/FIU baseline traces for 2 min before sti- GCTTCTGCT (F), CACAGCTTGCCTGGCATA 2+ mulation (basal [Ca ] dynamics), and 6 min after CA (R), insulin: GCCCAACAGGCTTCTTCTAC 2+ 20 mM glucose stimulation (stimulated [Ca ] AAC (F), GCAGATTTAGGAGGAAGGAAACCC dynamics), and normalized to the treatment duration (R), neuroD: CTTTCAACACACCCTAGAGTTCC (Figure 6(D)). For calculating the fraction of high G (F), GCATCATGCTTTCCTCGCTGTATG (R), 2+ glucose-responsive area (Figure 4(C)), the Ca influx ef1alpha: TCTCTACCTACCCTCCTCTTGGTC (F), response was defined as positive when the normalized TTGGTCTTGGCAGCCTTCTGTG (R). GCaMP6s FIU within the selected β-cell area (in vivo) or single β-cell (ex vivo) showed >7% increase in comparison to the time point before glucose injection Statistics or 20 mM glucose stimulation, and this increase in Graphical and statistical analyses were performed fluorescence intensity was higher than the peak fluor- using GraphPad Prism version 5.00 for Windows, escence intensity at baseline. In all ex vivo experi- GraphPad Software, La Jolla, California USA, www. ments, β-cells that failed to show fluorescence graphpad.com. Significance was tested using two- increase (>7% compared to time point before stimu- tailed distribution and either t-test or ONE-way lation) upon KCl stimulation were excluded from ANOVA analysis with p < 0.05 considered significant. quantification. Cross-correlation analysis (Fig. 4(E), Data presented are generally representative of at least 5(A,B), Sup. Fig. 2B, 3) and identification of the three independent experiments (except as indicated). peak numbers for oscillation frequencies (Figures 4 Error bars represent standard error (s.e.m) (except as (D)and 6(C)) on the FIU/FIU baseline traces of indicated). individual β-cells/areas were done by PeakCaller in (required rise: 1%, required fall: 3%, max. MATLAB lookback: 5 pts, max. lookahead: 5 pts) Cross-correla- Disclosure of potential conflicts of interest tion analysis was done on the section where glucose stimulation was applied (excluding KCl section). No potential conflicts of interest were disclosed. −1 Oscillation frequencies (min ) are average peak number per minute of selected β-cells (>6 cells, ex vivo)or β-cell areas (4 non-overlapping single-cell Acknowledgments sized selected region based on RFP red expression, We thank Robin Kimmel, Alexandra Koschak, Pidder Jansen- in vivo) normalized to the time duration until oscilla- Dürr, Petronel Tuluc, Thorsten Schwerte for helpful discussions, tion was observed in the time series (˜30–60 sec, in Sonja Töchterle and Dzenana Tufegdzic for expert zebrafish care, vivo, ˜150–300 sec, ex vivo). and all members of the Molecular Biology Institute for technical assistance and advice. qPCR Funding Total RNA was prepared from dissected islets of 5 dpf old zebrafish larvae (pool of 15 islets) and from whole This project was supported by the Universität Innsbruck/ zebrafish larvae (pool of 15 larvae) using Trizol University of Innsbruck and a travel fellowship to R.L. by (Ambion), cDNA was prepared using the First the Marshallplan-Jubiläumsstiftung/Austrian Marshall Plan Strand cDNA synthesis kit (Thermo Scientific). Foundation. 236 R. LORINCZ ET AL. 12. Heimberg H, De Vos A, Vandercammen A, Van ORCID Schaftingen E, Pipeleers D, Schuit F. Heterogeneity in Reka Lorincz http://orcid.org/0000-0002-9077-3300 glucose sensitivity among pancreatic beta-cells is cor- Christopher H. Emfinger http://orcid.org/0000-0002-9130- related to differences in glucose phosphorylation rather than glucose transport. Embo J. 1993;12(7):2873–2879. Andrea Walcher http://orcid.org/0000-0001-6624-4284 13. Moss LG, Caplan TV, Moss JB. Imaging beta cell Michael Giolai http://orcid.org/0000-0003-4166-8202 regeneration and interactions with islet vasculature in Maria S. Remedi http://orcid.org/0000-0003-2048-472X transparent adult zebrafish. Zebrafish. 2013;10(2):249– Colin G. Nichols http://orcid.org/0000-0002-4929-2134 257. doi:10.1089/zeb.2012.0813. Dirk Meyer http://orcid.org/0000-0002-4020-4484 14. Andersson O, Adams BA, Yoo D, Ellis GC, Gut P, Anderson RM, German MS, Stainier DY. NIH Public Access. 2013;15(6):885–894. doi:10.1016/j. cmet.2012.04.018.Adenosine. References 15. Argenton F, Zecchin E, Bortolussi M. Early appearance of pancreatic hormone-expressing cells in the zebrafish 1. Krishnan K, Ma Z, Björklund A, Islam MS. Calcium embryo. Mech Dev. 1999;87(1–2):217–221. doi:10.1016/ signaling in a genetically engineered human pancreatic S0925-4773(99)00151-3. β-cell line. Pancreas. 2015;44(5):773–777. doi:10.1097/ 16. Zang L, Shimada Y, Nishimura N. Development of a MPA.0000000000000318. novel zebrafish model for type 2 diabetes mellitus. Sci 2. Almaça J, Liang T, Gaisano HY, Nam HG, Berggren Rep. 2017;7(1). doi:10.1038/s41598-017-01432-w. PO, Caicedo A. Spatial and temporal coordination of 17. Marty N, Dallaporta M, Thorens B. Brain glucose sen- insulin granule exocytosis in intact human pancreatic sing, counterregulation, and energy homeostasis. islets. Diabetologia. 2015;58(12):2810–2818. Physiology (Bethesda). 2007;22:241–251. doi:10.1152/ doi:10.1007/s00125-015-3747-9. physiol.00010.2007. 3. Dolenšek J, Špelič D, Klemen MS, Žalik B, Gosak M, 18. Polakof S, Mommsen TP, Soengas JL. Glucosensing Rupnik M, Stožer A. Membrane potential and calcium and glucose homeostasis: from fish to mammals. dynamics in beta cells from mouse pancreas tissue Comp Biochem Physiol - B Biochem Mol Biol. slices: theory, experimentation, and analysis. Sensors 2011;160(4):123–149. doi:10.1016/j.cbpb.2011.07.006. (Switzerland). 2015;15(11):27393–27419. doi:10.3390/ 19. Röder PV, Wu B, Liu Y, Han W. Pancreatic regulation s151127393. of glucose homeostasis. Exp Mol Med. 2016;48(3):e219. 4. Kenty JHR, Melton DA. Testing pancreatic islet func- doi:10.1038/emm.2016.6. tion at the single cell level by calcium influx with 20. Ro der PV, Wong X, Hong W, Han W. Molecular regula- associated marker expression. PLoS One. 2015;10 tion of insulin granule biogenesis and exocytosis. Biochem (4):1–17. doi:10.1371/journal.pone.0122044. J. 2016;473(18):2737–2756. doi:10.1042/BCJ20160291. 5. Stožer A, Dolenšek J, Rupnik MS. Glucose-stimulated 21. Gilon P, Henquin JC. Influence of membrane potential calcium dynamics in islets of langerhans in acute changes on cytoplasmic CA2+ concentration in an mouse pancreas tissue slices. PLoS One. 2013;8(1). electrically excitable cell, the insulin-secreting pancrea- doi:10.1371/journal.pone.0054638. tic B-cell. J Biol Chem. 1992;267(29):20713–20720. 6. Leibiger IB, Berggren P-O. Intraocular in vivo imaging of 22. Eames SC, Philipson LH, Prince VE, Kinkel MD. Blood pancreatic islet cell physiology/pathology. Mol Metab. sugar measurement in zebrafish reveals dynamics of 2017;6(9):1002–1009. doi:10.1016/j.molmet.2017.03.014. glucose homeostasis. Zebrafish. 2010;7(2):205–213. 7. Chandra R, Liddle RA. Recent advances in the regula- doi:10.1089/zeb.2009.0640. tion of pancreatic secretion. Curr Opin Gastroenterol. 23. Remedi MS, Emfinger C. Pancreatic β-cell identity in dia- 2014. doi:10.1097/MOG.0000000000000099. betes. Diabetes, Obes Metab. 2016 September;18:110–116. 8. Seth A, Stemple DL, Barroso I. The emerging use of doi:10.1111/dom.12727. zebrafish to model metabolic disease. Dis Model Mech. 24. JurczykA,Roy N,BajwaR,Gut P,Lipson K,Yang C, 2013;6(5):1080–1088. doi:10.1242/dmm.011346. Covassin L, Racki WJ, Rossini AA, Phillips N, et al. 9. Pan FC, Wright C. Pancreas organogenesis: from bud Dynamic glucoregulation and mammalian-like responses to plexus to gland. Dev Dyn. 2011;240(3):530–565. to metabolic and developmentaldisruptioninzebrafish. doi:10.1002/dvdy.22584. Gen Comp Endocrinol. 2011;170(2):334–345. doi:10.1016/ j.ygcen.2010.10.010. 10. Fernandez J, Valdeolmillos M. Synchronous glucose- 25. Tseng Y, Chen R, Lee J, Liu S, Lee S, Hwang P. Specific dependent [Ca2+](i) oscillations in mouse pancreatic islets expression and regulation of glucose transporters in of Langerhans recorded in vivo. FEBS Lett. 2000;477(1– zebrafish ionocytes. Am J Physiol-Reg I. 2009;297 2):33–36. doi:10.1016/S0014-5793(00)01631-8. (2):275–290. doi:10.1152/ajpregu.00180.2009. 11. Henquin JC. Triggering and amplifying pathways of reg- 26. Marín-Juez R, Rovira M, Crespo D, van der Vaart M, ulation of insulin secretion by glucose71. Diabetes. 2000;49 Spaink HP, Planas JV. GLUT2-mediated glucose uptake (11):1751–1760. doi:10.2337/diabetes.49.11.1751. ISLETS 237 and availability are required for embryonic brain develop- growth. Nat Commun. 2017;8(1). doi:10.1038/s41467- ment in zebrafish. J Cereb Blood Flow Metab. 2015;35 017-00461-3. (1):74–85. doi:10.1038/jcbfm.2014.171. 40. Schulla V, Renström E, Feil R, Feil S, Franklin I, Gjinovci A, 27. CohenM,KitsbergD,TsytkinS,Shulman M, Aroeti B, Jing XJ, Laux D, Lundquist I, Magnuson MA, et al. NahmiasY.Liveimaging of GLUT2glucose-dependent Impaired insulin secretion and glucose tolerance in beta trafficking and its inhibition in polarized epithelial cysts. cell-selective Ca(v)1.2 Ca2+ channel null mice. Embo J. Open Biol. 2014;4(7):140091. doi:10.1098/rsob.140091. 2003;22(15): 3844–3854.doi:10.1093/emboj/cdg389. 28. Emfinger CH, Welscher A, Yan Z, Wang Y, Conway H, 41. U.K. Prospective Diabetes Study Group. Overview of 6 Moss JB, Moss LG, Remedi MS, Nichols CG. years’ therapy of type II diabetes: a progressive disease. Expression and function of ATP-dependent potassium Diabetes. 1995;44:1249–1258. doi:10.2337/db06-0880. channels in zebrafish islet β-cells. R Soc Open Sci. CHS. 2017;4(2):160808. doi:10.1098/rsos.160808. 42. Namkung Y, Skrypnyk N, Jeong MJ, Lee T, Lee MS, 29. Li M, Maddison LA, Page-McCaw P, Chen W. Kim HL, Chin H, Suh PG, Kim SS, Shin HS. Overnutrition induces -cell differentiation through pro- Requirement for the L-type Ca(2+) channel alpha(1D) longed activation of -cells in zebrafish larvae. AJP subunit in postnatal pancreatic beta cell generation. Jci. Endocrinol Metab. 2014.doi:10.1152/ajpendo.00686.2013. 2001;108(7):1015–1022. doi:10.1172/JCI13310. 30. Pace CS, Price S. Electrical responses of pancreatic islet 43. Yang S-N, Berggren P-O. The role of voltage-gated cells to secretory stimuli. Biochem Biophys Res calcium channels in pancreatic beta-cell physiology Commun. 1972;46(4):1557–1563. doi:10.1016/0006- and pathophysiology. Endocr Rev. 2006;27(6):621– 291X(72)90785-1. 676. doi:10.1210/er.2005-0888. 31. Grynkiewicz G, Poenie M, Tsien RY. A new generation 44. Stolovich-Rain M, Enk J, Vikesa J, Nielsen FC, Saada of Ca2+ indicators with greatly improved fluorescence A, Glaser B, Dor Y. weaning triggers a maturation step properties. J Biol Chem. 1985;260(6):3440–3450. of pancreatic β cells. Dev Cell. 2015;32(5):535–545. 32. Bootman MD, Rietdorf K, Collins T, Walker S, doi:10.1016/j.devcel.2015.01.002. Sanderson M. Ca2+-sensitive fluorescent dyes and 45. Musson MC, Jepeal LI, Mabray PD, Zhdanova IV, intracellular Ca2+ imaging. Cold Spring Harb Protoc. Cardoso WV, Wolfe MM. Expression of glucose- 2013;8(2):83–99. doi:10.1101/pdb.top066050. dependent insulinotropic polypeptide in the zebrafish. 33. Matveyenko AV, Butler PC. Relationship between β- Am J Physiol Regul Integr Comp Physiol. 2009;297: cell mass and diabetes onset. Diabetes Obes Metab. R1803–R1812. doi:10.1152/ajpregu.00288.2009. 2008;10:23–31. doi:10.1111/j.1463-1326.2008.00939.x. 46. Newsholme P, Brennan L, Bender K. Amino acid meta- 34. Gotthardt M, Eizirik DL, Cnop M, Brom M. Beta cell bolism, β-cell function, and diabetes. Diabetes. 2006;55 imaging - a key tool in optimized diabetes prevention (Suppl. 2). doi:10.2337/db06-S006. and treatment. Trends Endocrinol Metab. 2014;25 47. Gravena C, Mathias PC, Ashcroft SJH. Acute effects of (8):375–377. doi:10.1016/j.tem.2014.02.002. fatty acids on insulin secretion from rat and human 35. Chen C, Chmelova H, Cohrs CM, Chouinard JA, Jahn islets of Langerhans. J Endocrinol. 2002;173(1):73–80. SR, Stertmann J, Uphues I, Speier S. Alterations in β- doi:10.1677/JOE.0.1730073. cell calcium dynamics and efficacy outweigh islet mass 48. MalaisseWJ,BestL,KawazuS,Malaisse-LagaeF,SenerA. adaptation in compensation of insulin resistance and The stimulus-secretion coupling of glucose-induced insu- prediabetes onset. Diabetes. 2016;65(9):2676–2685. lin release: fuel metabolism in islets deprived of exogenous doi:10.2337/db15-1718. nutrient. Arch Biochem Biophys. 1983;224(1):102–110. 36. Barros TP, Alderton WK, Reynolds HM, Roach AG, doi:10.1016/0003-9861(83)90193-5. Berghmans S. Zebrafish: an emerging technology for in 49. Zhang Y, Louchami K, Carpentier YA, Malaisse WJ, vivo pharmacological assessment to identify potential Sener A. L-glutamine and palmitate catabolism in pan- safety liabilities in early drug discovery. Br J Pharmacol. creatic islets from rats depleted in long-chain polyun- 2008;154(7):1400–1413. doi:10.1038/bjp.2008.249. saturated ω3 fatty acids. Cell Biochem Funct. 2008;26 37. Henke K, Bowen ME, Harris MP. Perspectives for (1):82–86. doi:10.1002/cbf.1403. identification of mutations in the zebrafish: making 50. Rottbauer W, Baker K, Wo ZG, Mohideen MA, Cantiello use of next-generation sequencing technologies for for- HF,Fishman MC.Growthand function of theembryonic ward genetic approaches. Methods. 2013;62(3):185– heart depend upon the cardiac- specific L-type calcium 196. doi:10.1016/j.ymeth.2013.05.015. channel alpha1 subunit. Dev Cell. 2001;1(2):265–275. 38. Kamel M, Ninov N. Catching new targets in metabolic 51. Sidi S, Busch-Nentwich E, Friedrich R, Schoenberger U, disease with a zebrafish. Curr Opin Pharmacol. Nicolson T. gemini encodes a zebrafish L-type calcium 2017;37:41–50. doi:10.1016/j.coph.2017.08.007. channel that localizes at sensory hair cell ribbon synapses. 39. Singh SP, Janjuha S, Hartmann T, Kayisoglu Ö, Konantz J, JNeurosci. 2004;24(17):4213–4223. doi:10.1523/ BirkeS,MurawalaP,AlfarEA,MurataK,EugsterA,etal. JNEUROSCI.0223-04.2004. Different developmental histories of beta-cells generate 52. Tarifeño-Saldivia E, Lavergne A, Bernard A, Padamata K, functional and proliferative heterogeneity during islet BergemannD,VozML,ManfroidI,Peers B. 238 R. LORINCZ ET AL. Transcriptome analysis of pancreatic cells across distant essential for swelling-activated Cl − currents and embryo- species highlights novel important regulator genes. BMC nic development in zebrafish. Physiol Rep. 2016;4(19): Biol. 2017;15(1):21. doi:10.1186/s12915-017-0362-x. e12940. doi:10.14814/phy2.12940. 53. Stainier DY, Fouquet B, Chen JN, Warren KS, Weinstein 66. Rorsman P, Braun M, Zhang Q. Regulation of calcium BM, Meiler SE, Mohideen MA, Neuhauss SC, Solnica- in pancreatic α- and β-cells in health and disease. Cell Krezel L, Schier AF, et al. Mutations affecting the forma- Calcium. 2012. doi:10.1016/j.ceca.2011.11.006. tion and function of the cardiovascular system in the 67. Liu Z, Jeppesen PB, Gregersen S, Chen X, Hermansen zebrafish embryo. Development. 1996;123:285–292. K. Dose- and glucose-dependent effects of amino acids 54. Granato M, van Eeden FJ, Schach U, Trowe T, Brand on insulin secretion from isolated mouse islets and M, Furutani-Seiki M, Haffter P, Hammerschmidt M, clonal INS-1E beta-cells. Rev Diabet Stud. 2008;5 Heisenberg CP, Jiang YJ, et al. Genes controlling and (4):232–244. doi:10.1900/RDS.2008.5.232. mediating locomotion behavior of the zebrafish 68. Seisenberger C, Specht V, Welling A, Platzer J, Pfeifer embryo and larva. Development. 1996;123(1):399–413. A, Kühbandner S, Striessnig J, Klugbauer N, Feil R, 55. Nicolson T, Rüsch A, Friedrich RW, Granato M, Hofmann F. Functional embryonic cardiomyocytes Ruppersberg JP, Nüsslein-Volhard C. Genetic analysis after disruption of the L-type α1C (Cav1.2) calcium of vertebrate sensory hair cell mechanosensation: the channel gene in the mouse. J Biol Chem. 2000. zebrafish circler mutants. Neuron. 1998;20(2):271–283. doi:10.1074/jbc.M006467200. doi:10.1016/S0896-6273(00)80455-9. 69. Brereton MF,Vergari E,ZhangQ,Clark A.Alpha-,delta- 56. GylfeE,GrapengiesserE,Hellman B. Propagationofcyto- and PP-cells: are they the architectural cornerstones of islet plasmic Ca2+ oscillations inclustersof pancreatic β-cells structure and co-ordination? J Histochem Cytochem. exposed to glucose. Cell Calcium. 1991;12(2–3):229–240. 2015;63(8):575–591. doi:10.1369/0022155415583535. 57. Martín F, Soria B. Glucose-induced [Ca2+]i oscillations in 70. Westerfield M. The zebrafish book. a guide for the single human pancreatic islets. Cell Calcium. 1996;20 laboratory use of zebrafish (Danio rerio). 5th ed.: (5):409–414. doi:10.1016/S0143-4160(96)90003-2. Univ Oregon Press, Eugene; 2007. 58. Jonkers FC, Jonas JC, Gilon P, Henquin JC. Influence 71. Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, of cell number on the characteristics and synchrony of Schilling TF. Stages of embryonic development of the Ca2+ oscillations in clusters of mouse pancreatic islet zebrafish. Dev Dyn. 1995;203(3):253–310. doi:10.1002/ cells. J Physiol. 1999;520(Pt 3):839–849. aja.1002030302. 59. Podlasz P, Jakimiuk A, Chmielewska-Krzesinska M, 72. Kimmel RA, Dobler S, Schmitner N, Walsen T, Kasica N, Nowik N, Kaleczyc J. Galanin regulates Freudenblum J, Meyer D. Diabetic pdx1-mutant zebra- blood glucose level in the zebrafish: a morphological fish show conserved responses to nutrient overload and and functional study. Histochem Cell Biol. 2016. anti-glycemic treatment. Sci Rep. 2015 August;5:14241. doi:10.1007/s00418-015-1376-5. doi:10.1038/srep14241. 60. Yang YHC, Kawakami K, Stainier DY. A new mode of 73. Thisse C, Thisse B. High-resolution in situ hybridiza- pancreatic islet innervation revealed by live imaging in tion to whole-mount zebrafish embryos. Nat Protoc. zebrafish. Elife. 2018;7:e34519. doi:10.7554/eLife.34519. 2008;3(1):59–69. doi:10.1038/nprot.2007.514. 61. Yamada K, Nakata M, Horimoto N, Saito M, Matsuoka 74. Guo Y, Veneman WJ, Spaink HP, Verbeek FJ. Three- H, Inagaki N. Measurement of glucose uptake and dimensional reconstruction and measurements of zebra- intracellular calcium concentration in single, living fish larvae from high-throughput axial-view in vivo ima- pancreatic ??-cells. J Biol Chem. 2000;275(29):22278– ging. Biomed Opt Express. 2017;8(5):2611. doi:10.1364/ 22283. doi:10.1074/jbc.M908048199. BOE.8.002611. 62. Kaminski MT, Lenzen S, Baltrusch S. Real-time analy- 75. Kimmel RA, Meyer D. Zebrafish pancreas as a model sis of intracellular glucose and calcium in pancreatic for development and disease. Elsevier Ltd; 2016. Vol. beta cells by fluorescence microscopy. Biochim Biophys 134. doi:10.1016/bs.mcb.2016.02.009. Acta. 2012;1823(10):1697–1707. doi:10.1016/j. 76. Schneider CA, Rasband WS, Eliceiri KW. NIH Image bbamcr.2012.06.022. to ImageJ: 25 years of image analysis. Nat Methods. 63. Li G, Wu B, Ward MG, Chong ACN, Mukherjee S, 2012;9(7):671–675. doi:10.1038/nmeth.2089. Chen S, Hao M. Multifunctional in vivo imaging of 77. Thévenaz P, Ruttimann UE, Unser M. A pyramid pancreatic islets during diabetes development. J Cell approach to subpixel registration based on intensity. IEEE Sci. 2016;129(14). doi:10.1242/jcs.190843. Trans Image Process. 1998;7(1):27–41. doi:10.1109/ 64. Kang C, Xie L, Gunasekar SK, Mishra A, Zhang Y, Pai 83.650848. S, Gao Y, Kumar A, Norris AW, Stephens SB, et al. SWELL1 is a glucose sensor regulating β-cell excitabil- 78. Artimovich E, Jackson RK, Kilander MBC, Lin YC, ity and systemic glycaemia. Nat Commun. 2018;9(1). Nestor MW. PeakCaller: an automated graphical inter- doi:10.1038/s41467-017-02664-0. face for the quantification of intracellular calcium 65. Yamada T,WondergemR,MorrisonR,YinVP,StrangeK. obtained by high-content screening. BMC Neurosci. Leucine-rich repeat containing protein LRRC8A is 2017;18(1). doi:10.1186/s12868-017-0391-y.

Journal

IsletsTaylor & Francis

Published: Nov 2, 2018

Keywords: cacna1c; Cav1.2 channel; early zebrafish development; GCaMP6s; glucose-sensing of beta cells; in vivo imaging

There are no references for this article.