Transthyretin constitutes a functional component in pancreatic β-cell stimulus-secretion coupling

Abstract

Transthyretin (TTR) is a transport protein for thyroxine and, in association with retinol-binding protein, for retinol, mainly existing as a tetramer in vivo. We now demonstrate that TTR tetramer has a positive role in pancreatic β-cell stimulus-secretion coupling. TTR promoted glucose-induced increases in cytoplasmic free Ca²⁺ concentration ([Ca²⁺]_i) and insulin release. This resulted from a direct effect on glucose-induced electrical activity and voltage-gated Ca²⁺ channels. TTR also protected against β-cell apoptosis. The concentration of TTR tetramer was decreased, whereas that of a monomeric form was increased in sera from patients with type 1 diabetes. The monomer was without effect on glucose-induced insulin release and apoptosis. Thus, TTR tetramer constitutes a component in normal β-cell function. Conversion of TTR tetramer to monomer may be involved in the development of β-cell failure/destruction in type 1 diabetes.

Transthyretin (TTR) is a protein that is synthesized in the liver, the choroid plexus of the brain, and the endocrine pancreas (1, 2). It is a transport protein for thyroxine and, in association with retinol-binding protein, for retinol. It has been reported that 1-2% of plasma TTR circulates bound to high-density lipoprotein (HDL) and that the association to the HDL vesicle occurs through binding to apolipoprotein A1 (3). TTR has a complex equilibrium between different quaternary structures in serum (4),but exists mainly as a tetrameric protein of 14-kDa subunits (160-380 mg/liter) with only a small amount of TTR monomer present in vivo in normal individuals (5, 6). Consequently, measurements of TTR in serum by conventional methods mainly reflect the tetrameric form. The TTR amyloidoses are human diseases in which misfolded TTR protein aggregates in different tissues. Several point mutations in TTR have been related to familial amyloidotic polyneuropathy (FAP) (7). The fact that TTR is also produced within the pancreatic islet made us interested in evaluating a possible role of this protein in β-cell stimulus-secretion coupling.

Methods

Identification of TTR in Human Sera. Sera from type 1 diabetes (T1D) patients and control subjects were collected, identically sterile-processed, heat-inactivated by incubation at 56°C for 30 min, and stored frozen at -20°C until used. Changes in cytoplasmic free Ca²⁺ concentration ([Ca²⁺]_i) were tested in β-cells when they were depolarized with 25 mM KCl. Those diabetic sera that induced a higher increase in [Ca²⁺]_i than sera from controls were centrifuged, and the supernatant was passed through a 0.45-mm sterile filter. Samples were loaded on Sep-Pak C₁₈ preconditioned with 0.1% trifluoroacetic acid. After application, the sample proteins were eluted with 60% acetonitrile in 0.1% trifluoroacetic acid. This procedure was repeated twice. The two fractions were pooled, and the volume was reduced by lyophilization.

The lyophilized material was submitted to SDS/PAGE with a reference of control sera from healthy subjects. The gel was silver stained, and bands of the diabetic and nondiabetic sera were compared. The band at 14 kDa showed an up-regulation in diabetic sera.

Two SDS/PAGEs were run at the same time. One gel was stained with Coomassie brilliant blue, and one was blotted on a poly(vinylidene difluoride) membrane and lightly stained. The band at 14 kDa was cut out and digested in the gel with trypsin. After digestion, the material was run on a micro HPLC. Fractions were collected and mass fingerprinted by MALDI mass spectrometry, identifying TTR. This result was confirmed by amino acid sequencer analysis for 15 cycles.

Media. The medium used for isolation of pancreatic β-cells as well as in the experiments was a Hepes buffer (pH 7.4) containing 125 mM NaCl, 5.9 mM KCl, 1.2 mM MgCl₂, 1.28 or 2.56 mM CaCl₂, 25 mM Hepes, and 3 mM glucose. BSA was added at a concentration of 1 mg/ml. RPMI 1640 culture medium supplemented with 10% FBS was used for mouse cells, and CMRL 1066 medium was used for human pancreatic islets.

Preparation of Cells. Pancreatic islets from ob/ob mice were isolated by a collagenase technique and mechanically disrupted into cells (8, 9). Cells were either seeded onto glass coverslips or used as suspensions. The human islets were isolated and quantified by the procedure of Ricordi et al. (10).

Measurements of Insulin Release. Mouse. Cells in suspension were incubated overnight with 150 mg/liter TTR tetramer (Sigma) and 1 or 2 μg/ml TTR monomer. Control cells were incubated with the vehicle (water) of TTR. Dynamics of insulin release were studied by perifusing islet cell aggregates mixed with Bio-Gel P4 polyacrylamide beads (Bio-Rad) in a 0.5-ml column at 37°C (11). The flow rate was 0.2 ml/min, and 2-min fractions were collected and analyzed for insulin by RIA using a rat insulin standard (Novo-Nordisk, Copenhagen).

Human. Islets were incubated overnight with 150 mg/liter TTR, 2 μg/ml monomer, or the vehicle. In these experiments, a 0.27-ml column was used. The flow rate was 0.1 ml/min, and 1-min fractions were collected, and insulin was analyzed with ELISA (Mercodia, Uppsala, Sweden). PicoGreen double-stranded DNA (dsDNA) quantification reagent kit (Molecular Probes) was used to quantify dsDNA in the human islets from each perifusion column. The amount of insulin released is presented as microunits/ng of dsDNA.

Measurements of [Ca²⁺]_i. Changes in [Ca²⁺]_i were recorded in cells preincubated with 150 mg/liter TTR in the presence or absence of anti-TTR (1:500, Dako Cytomation, Glostrup, Denmark) and 1 or 2 μg/ml TTR monomer. Cells were attached to coverslips, loaded with 2 μM fura-2 acetoxymethyl ester (Molecular Probes), and mounted on an inverted epifluorescence microscope (Zeiss, Axiovert 135) connected to a Spex Industries Fluorolog-2 system for dual-wavelength excitation fluorimetry. The emissions due to the two excitation wavelengths of 340 and 380 nm were used to calculate the fluorescence ratio 340/380, reflecting changes in [Ca²⁺]_i. To compensate for possible variations in output of light intensity from the two monochromators, each experiment also included measurement of a 360/360 ratio. Every experiment was normalized by dividing all fluorescence ratios by the corresponding 360/360 ratio.

Measurements of TTR. TTR was measured at the Department of Clinical Chemistry, Karolinska University Hospital, with kinetic nephelometry (Beckman Coulter Immage).

Purification of the TTR Monomer. TTR was run on reverse-phase HPLC with an Everest 238 EV 54-C₁₈ (0.46 × 25 cm) column. The separation was made by using a linear gradient of 20-60% acetonitrile in 0.1% trifluoroacetic acid for 30 min at 1 ml/min. Fractions were collected and lyophilized. They were then run on a Novex Tricine gel 10-20%.

Patch Clamp. Whole-cell Ca²⁺ currents were recorded in mouse pancreatic β-cells, incubated overnight with 150 mg/liter TTR or the vehicle by using the perforated-patch variant of the whole-cell patch-clamp recording technique. Electrodes were pulled from borosilicate glass capillaries (Hilgenberg, Malsfeld, Germany) on a horizontal programmable puller (DMZ Universal Puller, Zeitz-Instrumente, Augsburg, Germany) and filled with 76 mM Cs₂SO₄, 1 mM MgCl₂, 10 mM KCl, 10 mM NaCl, and 5 mM Hepes (pH 7.35), as well as amphotericin B (0.24 mg/ml) to permeabilize the cell membrane and allow low-resistance electrical access without breaking the patch. The cells were bathed in a solution containing 138 mM NaCl, 10 mM tetraethylammonium chloride, 10 mM CaCl₂, 5.6 mM KCl, 1.2 mM MgCl₂, 5 mM Hepes, and 3 mM d-glucose (pH 7.4). Whole-cell currents induced by voltage pulses (from a holding potential of -70 mV to several clamping potentials from -60 to 50 mV in 10-mV increments, 100 ms, 0.5 Hz) were filtered at 1 kHz and recorded. All recordings were made with an Axopatch 200 amplifier (Axon Instruments, Foster City, CA) at room temperature (≈22°C). Acquisition and analysis of data were done using the program pclamp6 (Axon Instruments). Perforated-patch whole-cell recordings were also employed to register membrane potential in the β-cells. Typical electrode resistance was 2-4 MΩ. Electrodes were filled with 76 mM K₂SO₄, 1 mM MgCl₂, 10 mM KCl, 10 mM NaCl, and 10 mM Hepes (pH 7.35), as well as amphotericin B (0.24 mg/ml). The cells were bathed in a solution containing 138 mM NaCl, 5.6 mM KCl, 2.6 mM CaCl₂, 1.2 mM MgCl₂, 5 mM Hepes, and 3 mM glucose (pH 7.4) for 1 h before membrane potential registration. Membrane potential was recorded during perifusion with 3 and 16.7 mM glucose in current-clamp mode with an Axopatch 200 amplifier. All recordings were made at 34°C.

Flow Cytometric Analysis of Cell Death. Cells from ob/ob mice were cultured for 48 h in the presence of 2 μg/ml TTR monomer, 150 mg/liter total TTR, 40 μg/ml apolipoprotein CIII (apoCIII), TTR monomer or total TTR, and apoCIII or the vehicle. The whole cell population was collected and stained with propidium iodide (PI) (BD Pharmingen) and analyzed on a FACscan using CELLQuest acquisition software (Becton Dickinson, Immunocytometry System). FACS gating, based on forward and side scatter, was used to exclude cellular debris and autofluorescence. Typically, 10,000 cells were selected for analysis.

Statistical Analysis. Statistical significance was evaluated by Student's paired and unpaired t test, and P values <0.05 were considered to be significant. Data are presented as means ± SEM.

Results and Discussion

The stimulus-secretion coupling in the pancreatic β-cell is a complex process in which changes in [Ca²⁺]_i serve as key regulators of insulin release (12). We have measured changes in [Ca²⁺]_i in β-cells pretreated with commercial TTR, which constitutes both monomeric and tetrameric forms, and stimulated with glucose and KCl, the latter to directly open the voltage-gated L-type Ca²⁺-channel by depolarizing the cell. The addition of 25 mM KCl will give an approximate increase in membrane potential from -60 to -30 mV (13). TTR at a concentration of 150 mg/liter induced a more pronounced increase in [Ca²⁺]_i, subsequent to stimulation with glucose and KCl (Fig. 1 A, C, and F), compared with control. This effect was abolished when cells were incubated with TTR and antibody against TTR (Fig. 1E).

Fig. 1.

Effects of TTR on [Ca²⁺]_i, voltage-gated Ca²⁺ currents, and membrane potential. (A and C) Cells incubated with total TTR. (B and D) Cells incubated with the monomer. (A and B) The Δ increase in [Ca²⁺]_i is shown when cells were stimulated with glucose. (A) n = 64 for control cells and 82 for TTR-treated cells. (B) n = 50 for control cells and 73 for cells treated with 1 μg/ml TTR monomer, and n = 50 for control cells and 47 for cells treated with 2 μg/ml TTR monomer. (C and D) The Δ increase in [Ca²⁺]_i is shown when cells were depolarized with KCl. (C) n = 71 for control cells and 90 for TTR treated cells. (D) n = 83 for control cells and 83 for cells treated with 1 μg/ml TTR, and n = 83 for control cells and 60 for cells treated with 2 μg/ml TTR. (E) Cells incubated with total TTR and anti-TTR. Δ increases in [Ca²⁺]_i are shown when cells are stimulated with glucose (n = 37 for control cells, n = 22 for TTR-treated cells, and n = 52 for TTR and anti-TTR-treated cells) and depolarized with KCl (n = 89 for control cells, n = 24 for TTR-treated cells, and n = 42 for TTR and anti-TTR-treated cells). (F) Representative traces showing the effects of glucose and KCl on [Ca²⁺]_i in cells incubated with TTR (upper trace) and control cells (lower trace). (G) Sample whole-cell Ca²⁺ current traces generated by a depolarizing voltage pulse (100 ms) to -20 mV from a holding potential of -70 mV from a control cell (upper trace) and a cell incubated with TTR (lower trace). (H) Whole-cell Ca²⁺ current-voltage relationships in pancreatic β-cells in the presence and absence of TTR. The TTR-treated β-cells (n = 45) displayed larger Ca²⁺ currents than control cells (n = 50) when they were depolarized to -20 mV. (I) No significant difference in the resting membrane potential was detected between control cells (n = 21) and cells treated with TTR (n = 17). (J) TTR-treated cells (n = 17) exhibited a significantly faster action potential frequency than control cells (n = 21). **, P < 0.01 versus control. (K) Samples of membrane potential traces recorded from a control cell (Upper) and a cell incubated with TTR overnight (Lower). There was a more frequent firing of action potentials in cells incubated with TTR compared with control cells. **, P < 0.01 and ***, P < 0.001.

In trying to elucidate the cellular mechanism underlying the increase in [Ca²⁺]_i, the activity of voltage-gated Ca²⁺ channels was analyzed in cells incubated with 150 mg/liter TTR (Fig. 1 G and H). Ca²⁺ currents registered from TTR-treated cells, depolarized to -20 mV, were significantly enhanced as compared with those from control cells (Fig. 1 G and H). It is slightly surprising that TTR is effective only at -20 mV. However, mouse β-cells are equipped with multiple types of voltage-gated Ca²⁺ channels (14). These channels become activated at different potentials. The significant effect of TTR on voltage-gated Ca²⁺ currents appeared at -20 mV and disappeared gradually with larger depolarizations. It is noteworthy that there is a tendency of an increase in voltage-gated Ca²⁺ currents by TTR also at -10 mV. Most likely, this finding reflects that TTR, in addition to up-regulating the L_D-type Ca²⁺ current, also up-regulates the R-type Ca²⁺ current (15, 16). The disappearance of TTR effect at the larger depolarizations (-10 to 50 mV) used in our experiments is probably due to the activation of higher threshold voltage-gated Ca²⁺ channels that are unresponsive to TTR. These data support the idea that TTR acts on β-cell voltage-gated Ca²⁺ channels because the effects occurred in the range of physiological depolarizations (17).

To clarify the extent to which the TTR-induced changes in [Ca²⁺]_i were paralleled by changes in insulin release, mouse β-cells were incubated overnight with commercial TTR and thereafter stimulated with glucose and KCl (Fig. 2A) in a perifusion system. At a concentration of 150 mg/liter, TTR enhanced both basal (P < 0.05) and glucose-stimulated insulin release (P < 0.05) evaluated as area under the curve (Fig. 2 E and F). This concentration of TTR did not influence KCl-stimulated insulin release. The reason is not clear at the moment, but may reflect an additional interaction of TTR with the metabolic steps involved in glucose-stimulated insulin release. In this context, it was therefore interesting to analyze changes in β-cell membrane potential, a parameter determined by glucose metabolism. β-cell resting membrane potential, mainly determined by K_ATP channels, did not differ between control and TTR-treated cells during perifusion with 3 mM glucose (Fig. 1I), suggesting that exposure to TTR did not affect K_ATP channel activity. Interestingly, when stimulated with 16.7 mM glucose, TTR-treated β-cells displayed a significantly higher frequency of action potentials than control cells (Fig. 1 J and K). These data suggest that overnight treatment with TTR indeed led to enhanced glucose metabolism and thereby increased glucose-induced electrical activity. We have also studied insulin release from human islets obtained from three healthy subjects (donors of islets for transplantation) incubated with 150 mg/liter TTR, and the results were similar to those obtained in mouse β-cells (Fig. 2B).

Fig. 2.

Effects of total TTR and TTR monomer on insulin secretion. Insulin release was measured in mouse β-cells preincubated with total TTR at 150 mg/liter (n = 5) (A), 1 μg/ml TTR monomer (n = 3) (C), and 2 μg/ml TTR monomer (n = 3) (D). (B) A representative trace (n = 3) for insulin release measured in human islets exposed to 150 mg/liter TTR. μU, microunits. (E-H) The insulin release from A, C, and D expressed as area under the curve for basal (E and G) and glucose-stimulated (F and H) secretion. *, P < 0.05.

In T1D there is a specific destruction of the insulin-secreting pancreatic β-cell. Although the exact molecular mechanisms underlying β-cell destruction are not known, 25-40% of sera (L.J.-B., unpublished data.) from newly diagnosed T1D patients have been shown to increase the activity of voltage-gated L-type Ca²⁺ channels in β-cells (18). This increase in activity results in unphysiological increases in [Ca²⁺]_i upon depolarization and thereby β-cell apoptosis effects that can be prevented in vitro by a Ca²⁺-channel blocker (18). Ca²⁺-induced apoptosis may thus aggravate the complex autoimmune reaction associated with T1D. The key question has been what factor in T1D serum is responsible for the unphysiological changes in [Ca²⁺]_i. We have recently identified such a factor, namely apoCIII (19). T1D serum that affects [Ca²⁺]_i contains increased levels of apoCIII, the latter inducing both increased [Ca²⁺]_i and β-cell death. The effects of T1D sera and apoCIII on [Ca²⁺]_i and β-cell apoptosis are abolished in the presence of an antibody against apoCIII.

We have used kinetic nephelometry to measure the concentration of TTR in serum from 12 children with T1D and 10 healthy controls (Table 1). The levels were 116 ± 11 mg/liter in diabetic sera and 220 ± 18 mg/liter in control sera (P < 0.001), which is in accordance with previous studies (2, 20-22).

Table 1. Concentration of TTR in sera from T1D patients and healthy controls.

	TTR, mg/liter
No.	Diabetic sera	Control sera
1	103	227
2	112	248
3	90	169
4	76	203
5	134	205
6	151	172
7	70	178
8	130	318
9	151	317
10	196	165
11	78
12	105
Mean ± SEM	116 ± 11	220 ± 18

When trying to differentiate the monomeric form of TTR from the tetrameric form, it should be noted that it is possible to estimate the amount of monomeric TTR both by SDS/PAGE and by reverse-phase HPLC. As shown in Fig. 3, the monomer eluted before the tetramer on reverse-phase HPLC (peak 2 and peak 3, respectively; peak 1 being non-TTR material). The monomeric form was verified by total mass determination, 14 kDa by MALDI mass spectrometry, and a monomer band on SDS/PAGE. The amount of TTR monomer was analyzed in both T1D and control sera. In adults (Fig. 4A) as well as in children (Fig. 4B) with T1D, the band representing the TTR monomer has a higher intensity on silver-stained SDS/PAGE compared with sera from healthy controls. Of 10 sera tested from children with T1D, 6 showed an up-regulation of the monomeric band. The fact that the band on SDS/PAGE representing TTR monomer was of a higher intensity in T1D patients, whereas the serum level of the tetramer was decreased, made us test the effect of the monomer on β-cell [Ca²⁺]_i. The TTR monomer was purified from commercially available TTR by HPLC, and [Ca²⁺]_i measurements were made in β-cells incubated with 1 and 2 μg/ml monomer. These concentrations were chosen based on the only study, to our knowledge, in which the levels of monomer have been measured in serum from patients with familial amyloid polyneuropathy and normal subjects (6). In familial amyloid polyneuropathy patients, the concentration was 0.3 ± 0.1 μg/ml and in the controls 0.6 ± 0.2 μg/ml. Because our diabetic patients had a stronger band than the controls, we decided to test one concentration in the upper normal range and, in addition, one higher concentration. In cells exposed to 1 and 2 μg/ml monomer, glucose and depolarization with high K⁺ concentration gave a more pronounced increase in [Ca²⁺]_i compared with control cells (Fig. 1 B and D). Insulin secretion, stimulated by glucose and KCl, was not affected by the monomer in either of the chosen concentrations (Fig. 2 C, D, and H), although basal insulin secretion was significantly higher in cells treated with 2 μg/ml monomer (Fig. 2G). Interestingly, whereas the TTR tetramer (150 mg/liter) partly protected against apoCIII-induced β-cell apoptosis, the monomer (2 μg/ml) was without effect (Fig. 4 C and D).

Fig. 3.

SDS/PAGE pattern of TTR. Commercial TTR run on HPLC with a Vydac (Hesperia, CA) C₁₈ column, acetonitrile in 0.1% trifluoroacetic acid; gradient 20-60% in 30 min. AU, absorbance units at wavelength 214 nm. (Inset) Fractions 1-3 were collected and run on SDS/PAGE, where std refers to the commercial TTR before HPLC run.

Fig. 4.

SDS/PAGE of sera from adults and children with T1D and healthy controls and effect of total TTR and TTR monomer in the presence or absence of apoCIII on cell death. (A) Five T1D sera (d) and one control serum (c) from adults. (B) Sera from two children with TID (d) and two control children (c). The band for the monomeric form of TTR is marked with an arrow. In both A and B, samples with commercial TTR and a broad-range marker (m) were run. (C and D) FACS analysis of pancreatic β-cells exposed to total TTR (C) or TTR monomer (D) in the presence or absence of apoCIII or apoCIII alone (n = 5). Cell death is expressed as the percentage of 10,000 counted cells. **, P < 0.01 and ***, P < 0.001.

We have now demonstrated that TTR tetramer not only constitutes a functional component in normal pancreatic β-cell stimulus-secretion coupling but also preserves β-cell integrity. We have also demonstrated that the concentration of total TTR is decreased in T1D, whereas that of the monomeric form is increased. An interesting observation was that the monomer neither constituted a functional component in normal pancreatic β-cell stimulus-secretion coupling nor preserved β-cell integrity. Hence, it is likely that conversion of the tetrameric form of TTR to the monomeric form may serve as an early step associated with the development of β-cell failure/destruction in T1D.