Imaging cyclic AMP changes in pancreatic islets of transgenic reporter mice

Abstract

Cyclic AMP (cAMP) and Ca(2+) are two ubiquitous second messengers in transduction pathways downstream of receptors for hormones, neurotransmitters and local signals. The availability of fluorescent Ca(2+) reporter dyes that are easily introduced into cells and tissues has facilitated analysis of the dynamics and spatial patterns for Ca(2+) signaling pathways. A similar dissection of the role of cAMP has lagged because indicator dyes do not exist. Genetically encoded reporters for cAMP are available but they must be introduced by transient transfection in cell culture, which limits their utility. We report here that we have produced a strain of transgenic mice in which an enhanced cAMP reporter is integrated in the genome and can be expressed in any targeted tissue and with tetracycline induction. We have expressed the cAMP reporter in beta-cells of pancreatic islets and conducted an analysis of intracellular cAMP levels in relation to glucose stimulation, Ca(2+) levels, and membrane depolarization. Pancreatic function in transgenic mice was normal. In induced transgenic islets, glucose evoked an increase in cAMP in beta-cells in a dose-dependent manner. The cAMP response is independent of (in fact, precedes) the Ca(2+) influx that results from glucose stimulation of islets. Glucose-evoked cAMP responses are synchronous in cells throughout the islet and occur in 2 phases suggestive of the time course of insulin secretion. Insofar as cAMP in islets is known to potentiate insulin secretion, the novel transgenic mouse model will for the first time permit detailed analyses of cAMP signals in beta-cells within islets, i.e. in their native physiological context. Reporter expression in other tissues (such as the heart) where cAMP plays a critical regulatory role, will permit novel biomedical approaches.

Figure 1. cAMP reporter, enhanced by a mutation in YFP.

A, B. CHO cells, stably expressing rtTA, were cotransfected with RII-CFP and either the mutated (F46L, top) or original (lower) C-YFP. Note similar CFP fluorescence (left) but enhanced YFP fluorescence (right) for the F46L mutant. C. Transfected cells from A. and B., functionally imaged for FRET and stimulated with10 µM Fsk+100 µM IBMX to elevate cAMP levels. The F46L mutant cAMP reporter (•) yields a larger peak FRET signal (F470/F535) than the original reporter (○) (mean±s.e.m., n = 18 cells).

Figure 2. Enhanced cAMP reporter, expressed in tissues of double transgenic CMV-rtTA/pBI-cAMP mice.

A. Transgenic construct in pBI vector, with C-YFP and RII-CFP in opposing orientations around a bi-directional tetracycline-inducible promoter. Genotyping primers (▸,◂) and the resulting PCR products are indicated. B. Example of genotyping on genomic DNA (gDNA) from a mouse lacking (non-Tg) or possessing the integrated transgene (pBI-cAMP Tg). PCRs with each template tested for RII-CFP (lanes 1, 4, 7), C-YFP (lanes 2, 5, 8) and an endogenous gene, PLCβ2 (lanes 3, 6, 9). C–E. Tissues from double transgenic CMV-rtTA/pBI-cAMP mice were immunostained with anti-GFP (green) to visualize the reporter in skeletal myofibers (C), cardiac myocytes (D) and pancreas (E). In the pancreas, only acinar cells express the reporter, while islets of Langerhans (immunostained with anti-insulin, red) do not. In D., nuclei are counterstained red with TO-PRO-3. Scale bars, 50 µm.

Figure 3. Pancreatic islets function normally in double transgenic Ins2-rtTA/pBI-cAMP mice induced to express the cAMP reporter in pancreatic islet β-cells.

A. Cryosections of a pancreas, immunostained with anti-insulin to reveal islets of Langerhans (red) and anti-GFP (green). The overlay (right) shows that only β-cells express the transgenic cAMP reporter. We detected no gross changes in islet histology in transgenic mice. B. Reporter expression does not interfere with glucose homeostasis. Double transgenic mice, subjected to a Glucose Tolerance Test before (○), and 1 week after (•) induction of the cAMP reporter showed similar rise and fall in plasma glucose (mean±s.e.m.; n = 5 mice). C. β-cells from Ins2-rtTA/pBI-cAMP mice show normal glucose-stimulated Δ[Ca²⁺]_i (imaged with Fura-2). Glucose was elevated from 3 mM (basal) to 11 mM (grey bar, 11G). Intracellular [Ca²⁺] decreased transiently (arrow), then rapidly increased with a series of oscillations that continued for several minutes after glucose returned to the basal concentration. Similar responses were obtained in islets from wild-type mice (not shown).

Figure 4. Glucose stimulation results in dynamic and dose-dependent changes of cAMP concentration in β-cells.

Islets were harvested from double transgenic Ins2-rtTA/pBI-cAMP mice after induction. A. When cAMP was elevated (grey bar, 10 µM Forskolin) in islets excited at 430 nm, FRET emission (upper, 535 nm, orange symbols) dropped while CFP fluorescence (upper, 470 nm emission, cyan symbols) increased slightly. The ratio of these emission (lower, F470/F535; black symbols) is a monitor of changing cAMP concentration. Example shown is mean±s.e.m. (n = 14 cells in different regions of 1 islet). B. Repeated stimulation with 11 mM glucose (grey bars, 11G) produced consistent cAMP responses in β-cells (mean±s.e.m.; n = 6 cells). C. Stimulating islets with increasing concentrations of glucose (5.5 to 35 mM, grey bars) evoked increasing cAMP responses in β-cells. Upper trace is ratio measurements across a single islet, lower trace is mean±s.e.m. for 4 islets. D. The concentration-response function (mean±s.e.m., n = 4 islets) for cAMP, with EC50 = 9 mM glucose, corresponds well with glucose-stimulated insulin release in isolated islets , .

Figure 5. Prolonged stimulation with glucose results in a biphasic pattern of cAMP accumulation in β-cells.

A. A living islet from an induced Ins2-rtTA/pBI-cAMP mouse, viewed for YFP fluorescence. Dotted circles are ROIs analyzed functionally, and correspond to numbered traces in B. Scale bar, 20 µm. B. Prolonged glucose stimulation of this islet (grey bar, 11G) resulted in a nearly synchronous, biphasic elevation of intracellular cAMP in β-cells across the surface of the islet (black traces correspond to numbered ROIs shown in A; red symbols are mean±s.e.m. for the 8 ROIs).

Figure 6. Glucose-stimulated cAMP does not require Ca²⁺ elevation.

A. Islets expressing cAMP reporter were loaded with Fura-2 to measure Ca²⁺ and cAMP concurrently. In response to glucose (11G, grey bar), the increase in cAMP (black trace) precedes Ca²⁺ elevation and oscillations (blue trace). B–D. Glucose-stimulated cAMP in β-cells is independent of [Ca²⁺]_i. B. Glucose–evoked Ca²⁺ oscillations are completely eliminated in the absence of extracellular Ca²⁺, as widely reported for β-cells. Trace shows Fura 2 responses (F340/F380) to 11 mM glucose (grey bars, 11G) before, during and after the depletion of extracellular Ca²⁺. C. In contrast, glucose-evoked cAMP responses persist when extracellular Ca²⁺ is removed. D. Mean responses from Ca²⁺ - and cAMP imaging to 11 mM glucose before, during, and after Ca²⁺ is removed from bath (±s.e.m., n = 6 islets each). Ca²⁺- and cAMP-imaging were conducted independently to prevent any spectral overlap.

Figure 7. Elevation of cAMP causes PKA catalytic subunit translocation to the nucleus in β-cells.

A. Islets from induced Ins2-rtTA/pBI-cAMP mice were incubated for 30 min in control media or with added glucose (25 mM), forskolin (10 µM) or IBMX (100 µM). Nuclear C-YFP fluorescence is visible after prolonged elevation of cAMP (especially with fsk) in contrast to cytoplasmic localization in control islets. Scale bar, 20 µm. B. Z-stacks of confocal images to illustrate cytoplasmic C-YFP (i.e. with dark nucleus, left) and nuclear translocated (i.e. with bright nucleus, right) following IBMX for 30 min. C. Fluorescence intensity was quantified in Regions Of Interest (dotted circles in inset) over the nucleus and cytoplasm of cells treated as in A. The ratio of nuclear to cytoplasmic fluorescence was significantly higher relative to control when cAMP levels were elevated (* p≤0.05; ** p≤0.01; Dunnett's multiple comparisons test; n = 24–47 cells in 3 experiments for each treatment).