Analysis of sphingomyelin in plasma membrane at single cells using luminol electrochemiluminescence

Abstract

Sphingomyelin in plasma membrane at single cells was analyzed using luminol electrochemiluminescence, which involved serial reactions to generate hydrogen peroxide from membrane sphingomyelin for luminescence detection. A large deviation of the luminescence ratio before and after the introduction of the enzyme mixture was obtained from single cell analysis indicating the high cellular heterogeneity.

---

Sphingomyelin is one of the most important sphingolipids in plasma membrane at cells incorporating a phosphocholine head group, a sphingosine and a fatty acid. The synthesis or the degradation of sphingomyelin generates diacylglycerol or ceramide, which play significant roles in cell signaling pathways.^1,2 The recent study revealed that cholesterol, the other main component in plasma membrane, had different activity (escape tendency) due to the different molecular ratio between cholesterol and sphingomyelin in micro-domains, which led to the intracellular cholesterol trafficking.³ Therefore, the analysis of sphingomyelin at cellular membrane is significant for the understanding of many cellular signaling and the interaction of cholesterol and lipids.⁴ Currently, the classic detection strategy for the sphingomyelin at cellular plasma membrane utilized the sequential biological reactions including the hydrolysis of sphingomyelin by sphingomyelinase to generate ceramide and phosphorylcholine. The phosphorylcholine generated was reacted with alkaline phosphatase to create choline, which was oxidized by choline oxidase for the production of hydrogen peroxide. Using the classic colorimetric kit, the determination of hydrogen peroxide could offer the quantitative information of sphingomyelin.^5,6 This strategy was simple, however, this method worked on the cell population and could not be used for single cell analysis of sphingomyelin. Due to the high cellular heterogeneity, the accomplishment of single cell sphingomyelin analysis is still important for the sphingomyelin-related studies.^7,8

Previously, our group utilized luminol electrochemiluminescence for the analysis of active membrane cholesterol at single cell level.^9,10 Active cholesterol in plasma membrane at cells was reacted with cholesterol oxidase to generate hydrogen peroxide. In presence of luminol, hydrogen peroxide and luminol were electrochemically oxidized into diazaquinone and oxygen radicals, respectively, which were further reacted together to generate the excited 3-aminophthalate species for the emission light.^11,12 Since the luminescence intensity was correlated with the concentration of hydrogen peroxide, active membrane cholesterol at cells were semi-quantitatively measured. This strategy was further developed to realize sing cell cholesterol analysis by inserting a cell-sized pinhole between the electrode cultured with cells and PMT. When all the cells generated luminescence through hydrogen peroxide produced from active membrane cholesterol, only the luminescence from the target cell above the pinhole was captured for single cell analysis. Therefore, active membrane cholesterol at the target cell could be determined by this luminescence. A relative high deviation on the amount of active membrane cholesterol at individual cells was observed exhibiting the cellular heterogeneity on the plasma membrane. As compared with colorimetric assay and fluorescence assay, our electrochemiluminescence method had lower detection limit that facilitated single cell analysis.

Considering that the colorimetric assay of sphingomyelin generated hydrogen peroxide, our luminol electrochemiluminescence should be adapted for the analysis of sphingomyelin at single cell level. The cells cultured on the electrode were exposed to the luminol and enzyme mixture, including sphingomyelinase, alkaline phosphatase and choline oxidase. As proposed in Fig. 1, sphingomyelin at cellular membrane were reacted with sphingomyelinase, alkaline phosphatase and choline oxidase in the solution in serial to generate hydrogen peroxide, which induced the luminescence with luminol under the positive potential. In this communication, the luminescence difference before and after the introduction of enzyme mixture was measured at single cell level to exhibit single cell analysis of membrane sphingomyelin.

Fig. 1 The schematic process for the analysis of sphingomyelin in plasma membrane at cells.

The aqueous sphingomyelin was detected firstly to validate our electrochemiluminescence assay. The background luminescence from 200 μM L012 (luminol analog) and 200 nM sphingomyelin in 100 mM phosphate buffer saline (PBS, pH 7.4) was recorded with the potential, as shown in Fig. 2A (black curve). A maximal luminescence was observed at the potential of 1.0 V. After the addition of 5 U ml⁻¹ sphingomyelinase, alkaline phosphatase and choline oxidase into the solution, more luminescence was observed (Fig. 2A, red curve). The luminescence increase indicated the reaction of sphingomyelin with the enzymes to generate hydrogen peroxide, which induced the additional luminescence from L012. The luminescence reached the steady state after the reaction for 5 min, which was determined as the time for the conversion of aqueous sphingomyelin into hydrogen peroxide. Under this condition, sphingomyelin in the range of 100 nM to 5 μM was measured. At each concentration, the ratio of luminescence intensity before and after the introduction of enzymes was calculated to correlate with the concentration. More ratio of luminescence intensity in Fig. 2B with high concentration of sphingomyelin revealed that our electrochemiluminescence assay could determine the concentration of sphingomyelin in the solution.

Fig. 2 (A) The typical luminescence-potential trace collected from 100 mM PBS (pH 7.4) with 200 nM sphingomyelin before (black curve) and after the introduction of 5 U ml⁻¹ sphingomyelinase, alkaline phosphatase and choline oxidase. (B) The correlation of luminescence ratio-concentration of sphingomyelin. The error bar presented the standard deviation of three independent experiments.

After the establishment of our assay for aqueous sphingomyelin, the cell population were used to confirm the analysis of sphingomyelin in plasma membrane using luminol electrochemiluminescence. 20 000 cells were cultured on ITO slides and PBS solution with 200 μM luminol was used as the extracellular buffer. The introduction of sphingomyelinase, alkaline phosphatase and choline oxidase was proposed to react with sphingomyelin at membrane for the generation of hydrogen peroxide, which induced the luminescence increase under positive potential. Our previous work demonstrated that the application of potential did not affect the cellular activity.⁹ As shown in Fig. 2A, the background luminescence by the application of the potential cycle from −1.0 to 1.0 V was recorded in PBS solution with luminol. After the introduction of sphingomyelinase, alkaline phosphatase and choline oxidase for 1 min, the luminescence increase indicated that sphingomyelin at cellular membrane was reactive with all the enzymes to generate hydrogen peroxide. To confirm our assay of membrane sphingomyelin, sphingomyelinase, alkaline phosphatase or choline oxidase was added into the solution, individually, which should not generate hydrogen peroxide from PCs. As compared with background luminescence, no more luminescence increase was observed in these three control experiments. The results exhibited that the detection of membrane sphingomyelin followed the proposed procedure, which needed the cooperation of sphingomyelinase, alkaline phosphatase and choline oxidase. Meanwhile, the specificity of the assay was critical because some other phospholipid, mainly phosphatidylcholine (PCs), was the key component of cholesterol-rich micro-domain, as call rafts.¹³ As the result, the distinguishment of PCs and sphingomyelin was important for the future study of lipid–cholesterol interaction. 1 mM dipalmitoylphosphatidylcholine (DPPC) was added into the extracellular solution and no further luminescence increase was observed confirming that membrane PCs were not interfere the assay of sphingomyelin.

The reaction kinetics between the enzymes and membrane sphingomyelin was critical for the following single cell analysis because the slow consumption of sphingomyelin by the enzymes induced the significant diffusion of hydrogen peroxide into the bulk solution, and thus, lowered the concentration of hydrogen peroxide on the electrode surface to offer smaller luminescence change. For the maximum signal from membrane sphingomyelin, the overdose of sphingomyelinase, alkaline phosphatase and choline oxidase was required to achieve full consumption of sphingomyelin shortly. To investigate the consumption process of membrane sphingomyelin, a constant potential of 1.0 V was applied on the electrode and the steady state luminescence was exhibited as the background signal. When the enzyme mixture was added into the solution, a jump in luminescence should be observed and the time to reach the peak luminescence should offer the information about the sphingomyelin consumption. As shown in Fig. 2B, a minimal time of 13 ± 4 s was obtained for the peak luminescence intensity when the doses of enzymes increased up to 5 U l⁻¹. Therefore, the holding time of 17 s was set as the waiting time between the addition of the enzymes and the luminescence recording. Under the condition, the luminescence ratio before and after the introduction of the enzymes oxidase was 2.15 ± 0.31 from five independent measurements. The relative standard deviation was calculated to be 14.4% (Fig. 3).

Fig. 3 (A) The typical luminescence trace collected on 20 000 cells on ITO electrode before and after the introduction of sphingomyelinase, alkaline phosphatase and choline oxidase in 10 mM PBS with 200 μM luminol. The potential cycle was from −1.0 to 1.0 V with a scan rate of 1 V s⁻¹. (B) The luminescence trace collected from the cell population with a constant potential of 1.0 V. The PMT voltage was 600 V.

To estimate the amount of sphingomyelin at cellular membrane, the luminescence increase was correlated with the amount of hydrogen peroxide on the electrode. Experimentally, the luminescence increase after the introduction of enzymes was recorded after the reaction for 1 min, in which hydrogen peroxide generated diffused into the whole bulk solution. Afterwards, the solution in the cell chamber was replaced by the fresh buffer including luminol and aqueous hydrogen peroxide. The calibration curve of the luminescence increase on the concentration of hydrogen peroxide was established. From the curve, ∼200 nM hydrogen peroxide was produced from 20 000 cells in 20 μl solution, which suggested 0.12 × 10⁹ PCs molecules on one cell. Considering 10⁹ lipid (mainly phospholipid) molecules were at the plasma membrane of one cell and 10% of them were sphingomyelin,¹ our estimated number was close to the literature value, which supported that our assay could determine the amount of sphingomyelin from cellular membrane at different disease states.

After the establishment of sphingomyelin assay on cellular population, single cell analysis was attempted by inserting a cell-sized pinhole between ITO electrode and PMT. The cell density was adjusted so that the cell-to-cell distance was larger than 100 μm. Under this condition, only one cell appeared above the pinhole was detected by PMT and no interruption of the cells nearby on the luminescence had been confirmed.⁹ Fig. 4A showed the luminescence change before and after the introduction of enzymes for 17 s. The luminescence increase exhibited that membrane sphingomyelin were detectable at single cell levels. To investigate the cellular heterogeneity on membrane sphingomyelin, 12 cells were analyzed individually. The luminescence ratio before and after the reaction of enzymes with sphingomyelin were shown in Fig. 4B. The statistical luminescence ratio was 2.54 ± 0.96. The average number was close to the value from the cell population, however, the relative standard deviation was 37.8%. The deviation from single cell analysis was larger than that from cell population exhibited high cellular heterogeneity on membrane sphingomyelin. Previously, high heterogeneity had been observed on membrane cholesterol,^9,10 while, no explanation could be reached on the co-existence of these cells with different amount in active cholesterol. Considering the cholesterol equilibrium was controlled by the composition of cholesterol in membrane lipids, the correlation of the differences in membrane cholesterol and sphingomyelin might unveil the mystery between the cellular heterogeneity and cellular equilibrium.

Fig. 4 (A) The typical luminescence trace collected on one cells on ITO electrode before and after the introduction of sphingomyelinase, alkaline phosphatase and choline oxidase in 10 mM PBS with 200 μM luminol. The potential cycle was from −1.0 to 1.0 V with a scan rate of 1 V s⁻¹. (B) The luminescence ratio collected on ten individual cells. The PMT voltage was 900 V.

In conclusion, sphingomyelin at single cells were firstly achieved using luminol electrochemiluminescence. Membrane sphingomyelin reacted with sphingomyelinase, alkaline phosphatase and choline oxidase in the solution to generate hydrogen peroxide, which induced the luminescence under the positive potential for the detection. A large relative standard deviation was observed on individual cells suggested the cellular heterogeneity on the amount of sphingomyelin at cellular membrane. The further work will focus on the correlation of the heterogeneity on membrane cholesterol and sphingomyelin to understand the cell difference. Also, some other important lipids in plasma membrane at single cells will be analyzed using the similar strategy to elucidate the lipid–cholesterol interaction.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 81200028) and Foundation of Outstanding Young Talent in The Second Affiliated Hospital of Chongqing Medical University.

Notes and references

1. G. Van Meer, D. R. Voelker and G. W. Feigenson, Nat. Rev. Mol. Cell Biol., 2008, 9, 112–124.
2. R. Kolesnick, Mol. Chem. Neuropathol., 1994, 21, 287–297.
3. T. L. Steck and Y. Lange, Trends Cell Biol., 2010, 20, 680–687.
4. E. Ikonen, Nat. Rev. Mol. Cell Biol., 2008, 9, 125–138.
5. M. R. Hojjati, Z. Li, H. Zhou, S. Tang, C. Huan, E. Ooi, S. Lu and X. C. Jiang, J. Biol. Chem., 2005, 280, 10284–10289.
6. M. R. Hojjati and X. C. Jiang, J. Lipid Res., 2006, 47, 673–676.
7. C. Schubert, Nature, 2011, 480, 133–137.
8. L. N. Zheng, M. Wang, B. Wang, H. Q. Chen, H. Ouyang, Y. L. Zhao, Z. F. Chai and W. Y. Feng, Talanta, 2013, 116, 782–787.
9. G. Z. Ma, J. Y. Zhou, C. X. Tian, D. C. Jiang, D. J. Fang and H. Y. Chen, Anal. Chem., 2013, 85, 3912–3917.
10. C. X. Tian, J. Y. Zhou, Z. Q. Wu, D. J. Fang and D. C. Jiang, Anal. Chem., 2014, 86, 678–684.
11. C. A. Marquette and L. Blum, Anal. Chim. Acta, 1999, 381, 1–10.
12. M. Chen, X. Wei and Y. Tu, Talanta, 2011, 85, 1304–1309.
13. R. T. Dobrowsky, Cell. Signalling, 2000, 12, 81–90.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental detail. See DOI: 10.1039/c5ra24275d

---