Sphingosine Kinase Activity and Sphingosine-1 Phosphate Production in Rat Pancreatic Islets and INS-1 Cells: Response to Cytokines

S1P and sphingosine were from Biomol Research Laboratory (Plymouth Meeting, PA). [γ-³²P]ATP (3,000 Ci/mmol) and [³²P]orthophosphate (8,500–9,120 Ci/mmol) were from Perkin Elmer (Boston, MA). Fetal bovine serum (FBS) was from Atlanta Biologicals (Norcross, GA). Human recombinant IL-1β, TNF-α, and interferon-γ were obtained from R&D Systems (Minneapolis, MN). Culture media and molecular biology reagents were from Invitrogen (Grand Island, NY). QuantumRNA Classic 18S standards (488 bp) were from Ambion (Austin, TX). BSA, essentially fatty acid free, fraction V (FAF-BSA), ATP, and other reagents were from Sigma Chemical (St. Louis, MO).

Rat pancreatic islets were isolated as described previously (31). All animal procedures were approved by the Institutional Animal Care and Use Committee. Islets were cultured in CMRL-1066 medium containing 9% FBS, penicillin (100 units/ml), and streptomycin (100 μg/ml), for up to 24 h in the absence or presence of various stimuli indicated in the text. Cells of the rat insulinoma cell line, INS-1e (a gift of Dr. Claes Wollheim, Geneva, Switzerland) were cultured in modified RPMI-1640 (32); for 24 h before and during the experimental treatments, the INS-1 cells were cultured in medium in the absence of serum and the presence of 0.1% FAF-BSA.

Levels of SPHK mRNA were determined by semiquantitative RT-PCR (11). Islet (15–20 islets per extract) RNA was extracted essentially as described previously (11). PCR was carried out using primers to SPHK1a: sense, 5′-AGCCACCTTCAAGGAGTGAC-3′; and antisense, 5′-CAGTCTGCTGGTTGCATAGC-3′ (319 bp). PCR was carried out using primers to SPHK2: sense, 5′-CCAGGCTGCTCCTATTGGTC-3′; and antisense, 5′-TTGAGCAACAGGTCAACACC-3′ (367 bp). PCR was carried out and quantitated as described previously (11) with 0.5 μl of Quantum RNA Classic 18S rRNA primers and 2 units of Platinum pfx DNA polymerase (SPHK1a) or Taq polymerase (SPHK2), 1.5 mmol/l MgSO₄ (SPHK1a), 1.2 mmol/l MgCl₂ (SPHK2), and 6 μl of enhancer solution (SPHK1a). Typical reaction conditions were as follows: 94°C for 2 min for initial denaturation, followed by 28–38 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 45 s, and extension at 68°C for 60 s, with a final extension at 68°C for 7 min. Cycle time to achieve product linearity was determined in preliminary experiments. The image densities of SPHK mRNA and 18S rRNA were compared to determine the ratio of expression in experimental samples. PCR product identity was confirmed by automated DNA sequencing (11) using appropriate primers described above.

Cells (2 × 10⁶ cells per 10- × 30-mm dish) were cultured in serum-free media for 24 h to deprive cells of S1P present in serum. Then cells were treated with cytokines for the times indicated. Islets and harvested cells were lysed by freeze-thawing in SPHK buffer (33), followed by brief sonication for islets. Lysates were centrifuged at 15,000g for 20 min at 4°C, and the protein concentration of the supernatant fraction was determined. SPHK activity assays included (unless otherwise indicated): 40–80 μg of protein per reaction, 50 μmol/l sphingosine complexed with FAF-BSA (4 mg/ml suspension), and [³²P]ATP (2.5 μCi, 1 mmol/l) containing MgCl₂ (10 mmol/l) (14). Reaction linearity was determined in preliminary assays. SPHK isozyme activities were determined in the presence of 400 mmol/l KCl or 1% Triton X-100, as described previously (12). Labeled S1P was extracted (33) and separated by thin-layer chromatography (TLC) on heat-activated silica gel G plates using a solvent system containing 1-butanol:methanol:acetic acid:water (80:20:10:20 [vol/vol]). [³²P]S1P was quantitated by phosphoimager densitometric analysis using Multi-Analyst software (Bio-Rad Laboratories, Hercules, CA). Sample densities were converted to counts per minute (cpm). The authenticity of the S1P in the samples was tested by treating lipid extracts with potassium hydroxide to remove lipids other than S1P (1) and by ninhydrin staining of TLC-resolved S1P standard. SPHK specific activity was expressed as picomoles of S1P formed per minute per milligram of total protein.

Nuclei were prepared as described previously (34). Cells were washed, centrifuged briefly, and resuspended in buffer A (10 mmol/l HEPES, pH 7.5, 5 mmol/l MgCl₂, 15 mmol/l KCl, and 1 mmol/l phenylmethylsulfonyl fluoride) at 5 × 10⁶ cells/ml. The cells were then frozen in liquid nitrogen, thawed rapidly at 37°C, and passed through a 25-gauge needle 15 times. The homogenate was layered on top of a 500-μl sucrose cushion (50% sucrose in buffer A) and centrifuged for 5 min at 15,000g. Nuclei-free cytoplasm was removed from above the sucrose layer. Intact nuclei were resuspended in 200 μl of SPHK buffer (see above). The nuclei-free cytoplasm was then centrifuged at 100,000g for 60 min, and the pelleted membrane fraction was resuspended in 200 μl of SPHK buffer; the supernatant composed the cytosolic fraction.

INS-1 cells in serum- and antibiotic-free RPMI-1640 were transfected with 200 ng of pcDNA3.1-mSPHK1a vector (generous gift of Dr. Sarah Spiegel, Medical College of Virginia, Richmond, VA) or empty vector (generated by BSTX1 digestion of pcDNA3.1-mSPHK1a vector with subsequent religation), 1 μl of LipofectAMINE reagent, and 4 μl of LipofectAMINE PLUS reagent per well. Stable clones were established and assayed for SPHK activity as described above.

INS-1 cells (2 × 10⁶ cells per 10- × 30-mm dish) were prelabeled with [³²P]orthophosphate (40 μCi/ml) for 24 h and then treated with agents indicated in the text. After washing with PBS, cells were lysed with acidic methanol (1 ml of methanol + 10 μl of 3 N HCl). [³²P]S1P was extracted in chloroform:1 mol/l sodium chloride:3 N sodium hydroxide (2:1:0.1 [vol/vol/vol]) (9). [³²P]S1P was resolved by TLC as described above.

Values are means ± SE. Significant differences between treatment groups were determined by Student’s t test (paired) or one-way ANOVA with post hoc analysis using the Student-Newman-Keuls multiple-comparison test. Values of P ≤ 0.05 were accepted as significant.

Total SPHK activity, quantitated as the phosphorylation of sphingosine to S1P, was determined in INS-1 cell homogenate cytosolic fractions after treatment of intact cells with cytokines for 8 h. Cytosolic fractions were initially chosen for enzyme activity characterization because membrane-associated enzyme has been reported to be unstable (34). Control SPHK activity in samples containing sphingosine substrate was detectable above basal activity in samples lacking sphingosine (Fig. 1A, lanes 1 and 6). IL-1β, at a concentration (2 ng/ml) within the range previously reported to induce apoptosis in rat β-cells (35,36), significantly increased the SPHK activity in INS-1 cells by about 2.5-fold versus control with sphingosine (Fig. 1A, lanes 1 and 2, and B). At 2 ng/ml IL-1β, the stimulation of SPHK activity (335 ± 53% of control) was maximal, because SPHK activity at 0.1 ng/ml IL-1β (222 ± 59% of control) was not significantly different (P > 0.05). Treatment of cells with an apoptosis-inducing maximal concentration of TNF-α (37) resulted in a significant increase in SPHK activity to a level equivalent to that observed in IL-1β–treated cells (Fig. 1A, lanes 1 and 3, and B). There was no difference in INS-1 cell SPHK activity between 1 and 20 ng/ml TNF-α (data not shown). When IL-1β and TNF-α treatments were combined, SPHK activity increased significantly in the INS-1 cells versus control cells (Fig. 1A, lanes 1, 4, and 5, and B). This treatment also significantly increased activity versus either cytokine alone (Fig. 1A, lanes 2–5, and B).

Islets also showed significant and sustained increases in SPHK activity after culture with IL-1β for 15 min, 1 h, or 8 h (Fig. 2A). Similar responses were observed in islets cultured with TNF-α for 1 or 8 h (Fig. 2A). SPHK activation appeared maximal after 1 h of stimulation with IL-1β or TNF-α because activity was not increased further after 8 h (Fig. 2A). SPHK activity in islets treated with IL-1β plus TNF-α for 15 min was significantly lower than activity in islets treated with either cytokine alone; however, the activity increased in a time-dependent manner for as long as 8 h (Fig. 2A) and was significantly higher than control activity levels after 1 and 8 h (Fig. 2A).

INS-1 cells also showed rapid SPHK activation after culture with either IL-1β, TNF-α, or the combination of these cytokines (Fig. 2B). SPHK activation by the cytokines appeared to be maximal within 15 min because activity was not significantly higher after 1 h with either or both cytokines or after 8 h with the combined stimulus (Fig. 2B). No additivity in the responses to IL-1β or TNF-α was apparent at 15-min and 1-h treatment times (Fig. 2B).

The effects of cytokine stimulation on SPHK activity in subcellular fractions of INS-1 cells was also determined. In control cells, basal SPHK activity was approximately two- to threefold enriched in the particulate membrane and nuclear fractions of cells compared with the high-speed cytosolic fraction and the activity in whole-cell homogenates (Fig. 3). SPHK activity in the cytosolic fractions was not significantly different from those observed in the paired whole-cell homogenate preparations. After treatment of INS-1 cells with IL-1β and TNF-α for 18 h, there was a significant increase in SPHK activity in the cytosolic fraction compared with basal control cytosolic activity (Fig. 3). However, whereas SPHK activity in the particulate membrane and nuclear fractions of cytokine-stimulated cells exhibited a trend toward higher activity, the activity in these fractions was more variable than the cytosolic or whole-cell preparations and was not significantly different from the basal activity in comparable fractions isolated from control cells (Fig. 3). Moreover, no significant differences in SPHK activity were observed in whole-cell homogenates of cytokine-treated and untreated cells (Fig. 3).

The contribution of isozymes SPHK1 and SPHK2 to total SPHK activity in β-cells was determined by comparison of their activities in the presence of high salt concentration (KCl) and detergent (Triton X-100) (12). SPHK1 activity is inhibited by high salt concentration and activated by Triton X-100, whereas SPHK2 behaves in a reciprocal manner. INS-1 cell SPHK activity was activated in KCl-containing buffer by approximately sixfold, whereas the activity was almost undetectable in the presence of Triton X-100 (Fig. 4A and B), which is consistent with SPHK2 activity. In comparison, the effects of KCl and Triton X-100 on SPHK1 activity were assessed in INS-1 cells transfected with recombinant murine SPHK1a. Transfected INS-1 cells showed a significant activation of activity after treatment with Triton X-100 (Fig. 4A and C). Unexpectedly, the recombinant SPHK1a activity was not inhibited by KCl (Fig. 4C). It is unlikely that the lack of an inhibitory effect of KCl in transfected cells was due to activation of endogenous SPHK2 activity because the relative SPHK2 activity in nontransfected cells was only increased by ∼5 pmol S1P · min⁻¹ · mg protein⁻¹ with KCl treatment (Fig. 4B), and in transfected cells, SPHK total activity was 331 ± 81 pmol · min⁻¹ · mg protein⁻¹ (Fig. 4C). Molar concentrations of KCl may be required for maximal enzyme inhibition (12).

SPHK1 has been described as a predominantly cytosolic enzyme that can translocate to the nucleus and plasma membrane (34,38), whereas SPHK2 has a functional nuclear localization signal (26). In INS-1 subcellular fractionation studies, KCl-stimulated SPHK2 activity was identified in the membrane-free cytosolic, membrane, and nuclear fractions (Fig. 5). Consistent with data in Fig. 4, SPHK2 cytosolic activity was significantly activated by KCl treatment. SPHK activity in each subcellular fraction was inhibited by treatment with Triton X-100 (Fig. 5).

Although the predominant basal SPHK isozyme in INS-1 cells appeared to be SPHK2, there remained the possibility that the activation of SPHK activity by cytokines shown in Figs. 1 and 2 could be explained by an augmentation of SPHK1a and/or SPHK2 activities. To address this question, INS-1 cells were treated with IL-1β and TNF-α for 1, 8, and 24 h, and cytosolic SPHK activity was assayed in the presence of either KCl or Triton X-100. SPHK activity was significantly increased by cytokine treatment at each time point, and SPHK2 activity appeared to predominate, as indicated by complete inhibition by Triton X-100 (Fig. 6). Although basal SPHK activity was stimulated by KCl treatment, indicative of SPHK2, the stimulated level of SPHK activity induced by cytokine treatment at the various time points was not further activated by KCl (Fig. 6).

Endogenous S1P production in INS-1 cells was determined after overnight equilibration with ³²P_i. When prelabeled INS-1 cells were cultured with sphingosine for 20 min to act as an exogenous substrate for SPHK, S1P levels were significantly increased relative to untreated cells (Fig. 7A and B). The presence of sphingosine did not affect secreted [³²P]S1P levels in media (Fig. 7A), which were largely undetected. Treatment of ³²P_i-prelabeled intact INS-1 cells with IL-1β plus TNF-α for 8 h induced a significant increase in intracellular [³²P]S1P levels of 152 ± 13% over control (P < 0.02; Fig. 8). In contrast, ³²P-labeled total phospholipid levels in IL-1β–plus TNF-α–treated cells were only 120 ± 8% of control (P < 0.05) after 8 h (data not shown).

To determine whether new protein synthesis in part mediated the cytokine-induced change in S1P levels, INS-1 cells were cultured in the presence of cycloheximide during treatment with IL-1β and TNF-α. The cytokine-induced increase in [³²P]S1P levels in cycloheximide-treated cells was reduced by ∼40% compared with cells not treated with the inhibitor (Fig. 8). The levels of [³²P]S1P in cycloheximide-treated cells were similar to those recovered from cells cultured under basal conditions (Fig. 8).

The regulation of SPHK genes during cytokine challenge to β-cells was also investigated. Levels of SPHK1a and SPHK2 mRNAs were normalized to 18S rRNA in INS-1 cells cultured in the absence or presence of IL-1β, TNF-α, and a combination of these cytokines. Culture of INS-1 cells with IL-1β for 1 h significantly increased the ratio of SPHK1a mRNA/18S rRNA by nearly 70% versus basal values (Fig. 9A). However, the increase in relative SPHK1a mRNA levels was not sustained when the cells were continuously treated with IL-1β for 8 h (Fig. 9A). TNF-α also stimulated an early increase in SPHK1a mRNA within 1 h, but the effect was not apparent after 8 h (Fig. 9A). A combination of the cytokines did not raise relative SPHK1a mRNA levels above the levels observed with either cytokine alone after 1 h, and there was no apparent stimulatory effect of the cytokines after 8 h (Fig. 9A). In contrast to SPHK1a mRNA, there were no significant changes in the relative levels of INS-1 cell SPHK2 mRNA after stimulation by cytokines (Fig. 9B).

Isolated rat pancreatic islets also responded to IL-1β treatment with changes in SPHK mRNA expression. Within 1 h, relative SPHK1a mRNA levels in IL-1β–treated islets increased to more than 2.5-fold the basal values; similar results were observed with a combination treatment of IL-1β and TNF-α. However, TNF-α alone failed to elicit any changes in SPHK1a mRNA levels in islets (Fig. 10). Moreover, relative SPHK2 mRNA levels remained unchanged in the face of a 1-h islet challenge by IL-1β or TNF-α or a combination of these cytokines (Fig. 10). After a longer period (8 h) of islet exposure to IL-1β or a combination of IL-1β and TNF-α, the relative levels of SPHK1a mRNA were not significantly different (P > 0.05) from basal values (142 ± 15% and 116 ± 40% of basal, respectively); moreover, relative SPHK2 mRNA levels were not significantly different from basal values (131 ± 9% and 94 ± 27% of basal, respectively).

S1P is an extracellular receptor ligand and an intracellular signaling molecule. A physiological role for extracellular S1P in rat islets and INS-1 cells mediated by EDG receptors for S1P was recently reported by this laboratory (11). Endogenous S1P has been reported to regulate intracellular Ca²⁺ mobilization and to promote cell growth and survival (9,39,40). SPHK uses sphingosine as a substrate to generate S1P, and activity has been demonstrated in numerous cell types (9,20,23). This study is the first to characterize SPHK expression and activity in rat pancreatic β-cells and demonstrates the intracellular regulation of INS-1 cell and islet SPHK expression, activity, and S1P biosynthesis. Total SPHK activity in INS-1 cell (5.9 ± 1.4 pmol S1P formed · min⁻¹ · mg protein⁻¹) and islet (3.0 ± 1.5 pmol S1P formed · min⁻¹ · mg protein⁻¹) cytosolic fractions was similar to the total activity reported for PC12 cell cytosolic fractions (28 ± 4.9 pmol S1P · min⁻¹ · mg protein⁻¹) (24). However, various tissues express widely different SPHK activities (33). In this study, rat INS-1 cells serve as a homogeneous β-cell model (32) in comparison with isolated rat islets, which are an organ composed of at least three types of endocrine cells and other cell types.

Several cloned SPHKs have been reported for species including human, rat, mouse, yeast, and plant (13). These data illustrate that mRNAs for SPHK1 and SPHK2 isozymes are expressed in rat pancreatic islets and INS-1 insulinoma cells. SPHK used sphingosine as substrate in the production of S1P in these cells. Moreover, the results illustrate that endogenous β-cell SPHK activity is acutely regulated by the cytokines IL-1β and/or TNF-α at concentrations that induce apoptosis. This cytokine-regulated SPHK activation was rapid and sustained and correlated with increased levels of intracellular S1P in INS-1 cells. Cytokines, including IL-1β and TNF-α, are major stress inducers in β-cells and have been implicated in the development of diabetes (27,35–37). Unexpectedly, islet SPHK activation by the combination of IL-1β and TNF-α was not as high as the enzyme activity induced by either cytokine alone within 1 h of treatment, although in INS-1 cells, both cytokines elicited comparable early increases in SPHK activation. However, after 8 h, the SPHK activities of cytokine-stimulated islets and INS-1 cells were comparable. INS-1 cells showed a higher responsivity after 15 min to cytokines than did islets. Although SPHK activity was enriched in the membrane and nuclear subcellular fractions of INS-1 cells, it was the SPHK activity in the cytosolic fraction that was significantly activated by treatment of cells with IL-1β and TNF-α. Whether this can be attributed to greater instability of the enzyme associated with membranes (34) is not known.

SPHK activation appeared to occur in at least two phases: an early phase (∼15 min) that is expected to be largely independent of new protein synthesis, and a late phase that is dependent on new protein synthesis. After 8 h, cycloheximide blocked most, if not all, of the cytokine-induced increase in S1P levels, suggesting that activation of the SPHK pathway involved activation of endogenous enzyme as well as induction of new enzyme. Kinases and signaling proteins might contribute to early activation of SPHK. SPHK is a largely cytosolic enzyme (39) that has been postulated to migrate to membranes where it can phosphorylate sphingosine (41). PKC-activating phorbol esters have been reported to activate SPHK (19). In preliminary studies, phorbol 12-myristate 13-acetate activated INS-1 cell cytosolic SPHK activity (L.D.M., unpublished data). Putative phosphorylation sites in SPHK may regulate activity. Phorbol esters can induce phosphorylation of SPHK1 (38), presumably via PKC and a mito-gen-activated serine/threonine protein kinase (42). Interestingly, IL-1β in islets acutely increases PKC activation (43), and TNF-α can potentiate IL-1β–induced PKC activity (44). Recently, phorbol ester–induced HEK-293 cell SPHK activation was reported to involve translocation of SPHK to the plasma membrane (38). Ca²⁺/calmodulin may also activate SPHK (40). Paralleling an increase in PKC activation by IL-1β is an increase in intracellular Ca²⁺ (45). Thus, the activation of islet SPHK may be modulated by Ca²⁺ and PKC. In addition, TNF-α receptor–associated factor-2, nuclear coactivator 62-kDa/Ski-interacting protein, and a novel SPHK-1-binding protein, RPK118, bind to SPHK and may regulate SPHK activation (46–48).

A role for SPHK and the phosphorylated lipid product, S1P, in β-cells may be related to S1P having actions on apoptosis and cell growth opposing those of ceramide and sphingosine (14,39). Exposure of β-cells to SPHK-activating agents converts sphingosine to S1P, which may reduce the apoptotic potential. In addition, in some cells, S1P can regulate intracellular Ca²⁺ mobilization and promote cell growth and survival (9,39,40). On the other hand, it has been reported that SPHK2 promotes apoptosis (25). Thus, SPHK2 may be a heretofore unrecognized modulator of cytokine-induced apoptosis in β-cells.

IL-1β and TNF-α, alone or in combination, also stimulated an early increase in SPHK1a mRNA expression in INS-1 cells, whereas in rat islets, IL-1β but not TNF-α stimulated an early (1 h) increase in SPHK1a mRNA expression. The differences between islets and INS-1 cells could be related to the interaction of cytokine and glucose stimulation (11 mmol/l) in the INS-1 cell cultures versus 5.5 mmol/l glucose in the islet cultures or to inherent differences between transformed and primary cells. In INS-1 cells or islets, cytokine effects on SPHK1a mRNA levels were not sustained over 8 h, suggesting that mRNA stability and/or transcription may be affected.

Surprisingly, cytokines failed to affect the expression of SPHK2 mRNA in islets or INS-1 cells. These data are in contrast to the cytokine effects on SPHK activity levels in INS-1 cells, where it appeared that SPHK2 was the predominant isoform and that cytokines elicited maximal stimulation that was not further affected by high salt concentration. In contrast, SPHK activity in the presence of Triton X-100 was low in cytokine-treated INS-1 cell extracts. The residual SPHK activity observed in the presence of Triton X-100 may be SPHK1; however, there was no cytokine regulation of this activity over a 24-h period. The tight regulation of SPHK activity in the presence of cycloheximide suggests that new protein synthesis does account for long-term SPHK activation. If this late chronic S1P-synthesizing activity is accounted for by SPHK2, then increased translation of SPHK2 or enzyme activation must explain the results. There is no ready explanation for why SPHK1a isoform mRNA appears to be highly regulated in these cells but is not paralleled by a change in enzyme activity. It is possible that SPHK1 was exported from the cells (49) or is unstable or that levels of the enzyme defy biochemical detection.

The subcellular distribution of SPHK2 in INS-1 cells and islets in nuclei and cytosol is typical (27). Whereas SPHK1 was expected to be localized to the cytosol, little, if any, SPHK1 activity was observed in the presence of Triton X-100. This is in contrast to the Triton X-100–induced activation of a recombinant murine SPHK1a transfected into INS-1 cells. Localization of SPHK2 activity and S1P production in the β-cell nucleus may modulate changes in gene transcription or even mitogenesis (34). A role for the regulated expression of SPHK1a mRNA, however, remains to be determined.

In summary, SPHK is active in isolated islets and the β-cell line INS-1, and activity is responsive to cytokines, resulting in enhanced S1P levels. The failure to recover S1P in culture media suggests that islet S1P production would not contribute in a meaningful way to blood levels of S1P or EDG receptor stimulation by autocrine mechanisms in vivo even in the islet microenvironment. The levels of S1P in plasma can be in the 400- to 500-nmol/l range (6). Thus, S1P produced in β-cells is likely to have an intracellular site(s) of action. The rapid response of SPHK to cytokine treatment suggests that it potentially either modulates or offers protection against stress responses induced by cytokines.

S.G.L. was supported by a grant from the Juvenile Diabetes Research Foundation International (1-2002-613). L.D.M. was supported by fellowships from the Endocrine Fellows Foundation and the Juvenile Diabetes Research Foundation International.

We thank Dr. Sarah Spiegel for her generous provision of recombinant mouse SPHK1a plasmid.