Two-Photon Excitation Imaging of Pancreatic Islets With Various Fluorescent Probes

RESEARCH DESIGN AND METHODS

Preparation of mouse pancreatic islets.

Pancreatic islets were isolated from ICR mice aged >8 weeks using collagenase and cultured for between 1 h and 3 days (7). The islets cultured for 1 day were immersed in a solution (solution A) containing 140 mmol/l NaCl, 5 mmol/l KCl, 2 mmol/l CaCl₂, 1 mmol/l MgCl₂, 10 mmol/l Hepes-NaOH (pH 7.3), and 2 mmol/l glucose. The solution also contained lucifer yellow (0.4 mmol/l), sulforhodamine B (0.7 mmol/l), FM1-43 (4 μmol/l), BODIPY-forskolin (10 μmol/l), or tetramethylrhodamine-ethylester (TMRE) (10 μmol/l). Staining with fura-2 was performed by placing the islets in a serum-free culture medium (Dulbecco’s modified Eagle’s medium [DMEM]) containing 10 μmol/l fura-2-AM, 0.1% pluronic F127, and 0.1% bovine serum albumin for 30 min at 37°C and washing with solution A. These fluorescence indicators were purchased from Molecular Probes (Eugene, OR). BODIPY-forskolin and fura-2-acetoxymethyl were first dissolved in DMSO at high concentrations. An adenovirus vector for yellow cameleon 2.1 was constructed as described elsewhere (8–10). Adenovirus infection was performed on islets cultured for 1 day. The islets were placed in a serum-free culture medium (DMEM) containing the virus (5.7 × 10⁸ pfu/ml) for 1 h, washed, and subsequently cultured for 2 days. Imaging experiments were then performed at room temperature (24.5–25.5°C).

Multiphoton excitation imaging.

Two-photon excitation imaging of islets was performed with an inverted microscope (IX70; Olympus, Tokyo) and a laser-scanning microscope (FluoView; Olympus) using a water immersion objective lens (UApo 60× water/IR) as described elsewhere (4). The laser power at the specimen was 3–10 mW. Lucifer yellow was excited at 720 nm, whereas sulforhodamine, yellow cameleon 2.1, FM1-43, BODIPY-forskolin, fura-2, and TMRE were excited at 830 nm. The fluorescence of indicators other than yellow cameleon 2.1 was measured with a photomultiplier of the laser-scanning microscope at wavelengths of 400–620 nm. The fluorescence of yellow cameleon 2.1 was measured at 400–510 nm (F480) and 530–650 nm (F530), and the ratio between F530 and F480 was calculated. Twelve-bit images were color-coded with either the autumn or green color codes of FluoView.

RESULTS

Hydrophilic probes.

When acutely isolated islets were perifused with a solution containing hydrophilic tracers, such as lucifer yellow (Fig. 1A) or sulforhodamine (data not shown), highly fluorescent areas (Fig. 1A, the white arrow and arrowheads) and puncta (Fig. 1A, the blue arrow and arrowheads) quickly appeared within 30 s (Fig. 1B), which apparently reflected the major blood vessels and microvessels in the islets, respectively. In addition, we noted all the interstitial spaces of islets were subsequently filled with tracers (Fig. 1A, the green arrow and arrowheads; Fig. 1B). These structures were evenly stained between the inner as well as outer layers of islets, when images were obtained at planes within 20–50 μm of the bottom of an islet on a glass coverslip. Thus, two-photon excitation imaging can visualize deep cell layers (down to the third or fifth layers) in the islet. These staining patterns were preserved in islet preparations cultured for >6 days.

The adenovirus vector can be transfected into deep cell layers of an islet. Figure 1C shows an islet transfected with an adenovector carrying the gene encoding yellow cameleon 2.1, a Ca²⁺ indicator protein based on green fluorescent protein (GFP) and calmodulin (11). Fluorescent expression was detected in the cells deep within the islet (Fig. 1C). The fact that expression levels of cameleon 2.1 greatly differ from cell to cell indicates that cameleon cannot penetrate the gap junctions (12). Furthermore, synchronous increases in the fluorescence ratio at 480 nm/530 nm were induced in many cells in the islet by application of a high-concentration potassium solution from a glass pipette, indicating that Ca²⁺ signaling is preserved after the virus infection. Thus, GFP-based Ca²⁺ indicator can monitor the cytosolic Ca²⁺ concentration of cells in deep layers of islets, unlike acetoxymethyl esters of organic Ca²⁺ indicators (Fig. 2E).

Lipophilic probes.

FM1-43 is a lipophilic yet membrane-impermeable tracer (13). We found that FM1-43 also brightly stained the islet blood vessels and the plasma membrane facing the interstitial space throughout the islet preparations within 3 min (Fig. 2A). The fluorescent intensity of blood vessels in the middle of the islets was as strong as those in the superficial layers (Fig. 2B), and there was little cytosolic staining within 20 min after application.

In contrast, BODIPY-forskolin and fura-2-acetoxymethyl, which are lipophilic and membrane-permeable tracers, were accumulated in the superficial cell layers of the islets (Fig. 2C and E) and did not penetrate into deeper cell layers even 30 min to 2 h after the application (data not shown) (14). This result indicates that, even though these dyes are dissolved into solution by DMSO, they probably stick to the plasma membrane of cells in the superficial cell layers, are internalized, and do not readily leave the superficial cell layers. Therefore, the concentrations of substances that can reach the deeper cell layers can be very low. Lipophilic probes that have unusually high membrane permeability, however, were found to penetrate into deeper cell layers but over a longer period of time. For example, TMRE stained mitochondria throughout the islet 30 min after application.

DISCUSSION

We have demonstrated that various fluorescence probes can be successfully applied for two-photon excitation imaging of the whole-islet preparations. The rapid penetration of hydrophilic substances via blood vessels is consistent with nearly synchronous increases in NADH autofluorescence throughout the islet during high glucose stimulation (5). Direct visualization of islet interstitial space with either polar tracers or FM1-43 will provide a promising methodology for the direct imaging of insulin exocytosis in the whole-islet preparations.

On the other hand, membrane-permeable fluorescence probes tend to get trapped in the superficial layer of the islets and exhibited steep concentration gradients along the depth axis of a single islet (Fig. 2D and F). Thus, the most efficient method for cytosolic loading of the cells in the islet seems to involve virus-mediated gene transfer. In fact, we have shown that an adenovirus carrying the cameleon 2.1 gene can induce expression of cameleon proteins in deep cell layers of islets and enable the detection of increase in cytosolic Ca²⁺ by depolarization (Fig. 1C), which was impossible with commonly used acetoxymethyl esters of organic Ca²⁺ indicators (Fig. 2E).

It has been reported that insulin exocytosis from single β-cells shows high sensitivity to cytosolic cAMP (12,15). In contrast, cytosolic cAMP was claimed to play a less significant role in the whole-islet preparation because the membrane-permeable antagonist of cAMP, Rp-cAMP (16,17), or inhibitors (protein kinase inhibitors [PKIs]) of cAMP-dependent protein kinase (PKA) (9,10) showed relatively weak effects on glucose-induced insulin secretion. It needs to be noted, however, that Rp-cAMP did not sufficiently reach the deep cell layers of islets (18,19) and that the myristoylated PKI peptide (18) or PKI peptide in liposome (19,20) was trapped in the most superficial layers of islets. Direct evidence for the involvement of cAMP/PKA in insulin exocytosis will be gained either by using adenoviruses carrying cAMP/PKA inhibitory peptides or proteins or by using two-photon excitation imaging of insulin exocytosis in whole-islet preparations.