The Plasma Membrane Polarity Is Higher in the Neuronal Growth Cone than in the Cell Body of Hippocampal and Cerebellar Granule Neurons

Abstract

The polarity of the biological membrane, or lipid order, regulates many cellular events. It is generally believed that the plasma membrane polarity is regulated according to cell type and function, sometimes even within a cell. Neurons have a variety of functionally specialized subregions, each of which bears distinct proteins and lipids, and the membrane polarity of the subregions may differ accordingly. However, no direct experimental evidence of it has been presented to date. In the present study, we used a cell-impermeable solvatochromic membrane probe NR12A to investigate the local polarity of the plasma membrane of neurons. Both in hippocampal and cerebellar granule neurons, growth cones have higher membrane polarity than the cell body. In addition, the overall variation in the polarity value of each pixel was greater in the growth cone than in cell bodies, suggesting that the lateral diffusion and/or dynamics of the growth cone membrane are greater than other parts of the neuron. These tendencies were much less notably observed in the lamellipodia of a non-neuronal cell. Our results suggest that the membrane polarity of neuronal growth cones is unique and this characteristic may be important for its structure and function.

INTRODUCTION

In a typical biological membrane, hundreds of different lipid molecular species are present. The fluid mosaic model is considered a basic model for cellular membranes,^1,2) whereas more recent models suggest that this remarkable lipid variety contributes to lateral heterogeneity in the plasma membranes due to the formation of dynamic sub-micron-sized lipid structures—lipid microdomains.^3,4) The driving force of the lateral heterogeneity in biological membranes is the formation of two lipid membrane phases: a dense liquid-ordered phase (Lo phase) composed of saturated phospholipids, cholesterol, and sphingomyelin, and a liquid-disordered phase (Ld phase), rich in unsaturated phospholipids.⁵⁾ The structure and function of membranes can be studied by monitoring the biophysical properties, such as polarity, viscosity, tension, and lipid order.⁶⁾ These properties are crucial for various cellular processes, including membrane protein function, cell signaling, membrane trafficking, and cellular adaptation to environmental changes.^1,2,4) Monitoring membrane polarity is vital for comprehending lipid organization, cellular chemical reactions, and morphological changes. For example, the Lo and Ld phases in biological membranes are characterized by low and high membrane polarity, respectively.⁷⁾

Neurons are polarized cells and have diverse morphologies depending on their function. Neurons contain highly curved membrane structures such as growth cones, neurite branches, and spines, and the membranes at these locations are thought to have highly polar properties. In fact, phosphatidylcholine lipids with arachidonic acid, a polyunsaturated fatty acid, are distributed across a proximal-to-distal gradient.⁸⁾ However, to the best of our knowledge, no work has to date directly reported the distribution of polarity in the plasma membrane of neurons.

Recent developments in fluorescent probes have facilitated the gradual clarification of the subcellular localization of lipids and the understanding of their biophysical properties.^6,7) Thus, the polarity-sensitive (solvatochromic) fluorescent probes based on pyrene distinguish Lo and Ld phases by a change in the fluorescence wavelengths of the probe.^9,10) Studies using these fluorescent probes have shown that the intracellular membranes of organelles are more polar than the plasma membrane in HeLa cells.^9,11) The fluorescent probes based on solvatochromic dye Nile Red, NR12S and NR12A, exhibit different fluorescence wavelengths in Lo and Ld phases and are designed to be impermeant of the lipid bilayer, enabling specific assessment of polarity and lipid order at the outer leaflet of the plasma membrane.^10,12) Recently, NR12A was used to show the aberrant polarity of cells expressing mutant sphingomyelin synthase.¹³⁾ In this study, we applied NR12A to the cultured primary neurons from the mouse brain to investigate local polarity within a single neuron. Our results indicate that the growth cone is characterized by systematically lower membrane polarity.

MATERIALS AND METHODS

Reagents and Animals

NR12A was synthesized as described previously.¹⁰⁾ Human embryonic kidney (HEK) 293T cells were kindly provided by Dr. Katsuhiko Mikoshiba (the University of Tokyo). All experimental protocols were approved by the Animal Care and Use Committee of Nagoya City University and performed according to the guidelines of the National Institutes of Health of Japan. Slc:ICR mice were purchased from Japan SLC (Shizuoka, Japan).

Cell Culture

HEK 293T cells were cultured using Dulbecco’s Modified Eagle’s Medium High Glucose (#044-29765, FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) supplemented with 10% fetal calf serum (#175012, Nichirei, Tokyo, Japan). The culturing on mouse primary hippocampal neurons was performed as previously described.^14,15) Briefly, embryonic day 18 pups were sacrificed, and hippocampi were obtained. They were then dissociated after trypsin (#25200056, ThermoFisher Scientific, Waltham, MA, U.S.A.) and DNase I (#043-26773, FUJIFILM Wako Pure Chemical Corporation, 0.1% in Hanks’ Balanced Salt Solution with 12 mM MgSO₄) treatment. The neurons were seeded on 35 mm glass-bottom dishes and cultured in Neurobasal medium (#21103049, ThermoFisher Scientific) supplemented with 2% B27 supplement (#17505044, ThermoFisher Scientific), 50 Units/mL penicillin/streptomycin, and 2 mM L-glutamine. Cerebellar granule neurons (CGNs) were prepared from postnatal day 3 mouse neonates and cultured essentially in the same way as the hippocampal neurons, the only difference being the addition of 20 mM KCl to the culturing medium.

Fluorescence Images and Calculation of the Polarity

The cells were incubated with NR12A (10 nM) for 8 min in a dark incubator. After washing with phosphate buffered saline briefly, the cells were imaged with a confocal microscope LSM800 with Airyscan (Carl Zeiss, Oberkochen, Germany). The samples were then excited at 488 nm wavelength and two fluorescence images, at 550–600 nm (I550–600) and 600- nm (I600>) wavelengths were obtained. The observation was completed within 40 min after the addition of NR12A. From each image, we selected the area corresponding to the cell body, the growth cone, and their boundary regions. For the cell body, the region with a smooth curve close to the nucleus was selected. For the growth cones, the region containing the filopodia at the tips of the neurite was selecetd. For the boundary region, the region around the base of the neurite without any filopodia was selected. The ratiometric image was created by dividing the I600> image by the I550–600 image, yielding I600 > /I550–600 ratio.

Statistical Analyses

In each set of experiment, multiple cells were analyzed and their means of ratio and coefficient of quartile deviation (CQD) were obtained. The specific values indicated in the bar or line graphs are the means ± standard error of the mean. The number of cells analyzed in each set of experiment is indicated in the figure legends. Microsoft Excel and Origin Pro were used for statistical analysis. A paired t-test was used to test between two paired groups. To test three paired groups, we applied Bonferroni’s correction: the corrected p-value threshold was determined by dividing the significance level by three. *, p < 0.05; **, p < 0.01. N.S., not significant. The p-values are given in Figure legends.

RESULTS

The Polarity of the Growth Cone Is Higher than That of the Cell Body in the Cultured Mouse Hippocampal Neurons

Mouse hippocampal neurons were cultured, and their membrane polarity was investigated with NR12A. The culture period was set less than 30 h so that the growth cones and cell bodies appear but still fit in a single picture (Fig. 1A). At this stage, the axon and dendrite are not determined but one or a few growth cones are clearly observed.¹⁶⁾ Each pixel in the calculated image has a ratiometric value that is shown in pseudocolor according to that value. The increase in the I600 > /I550–600 ratio corresponds to the pseudo-color change from blue to red, implying higher local polarity (Fig. 1E). We also carefully excluded the non-plasma-membrane (i.e., out of confocal plane) area because its ratio values may reflect the polarity of intracellular organelles membrane, whose polarity is known to be higher than that of the plasma membrane.⁹⁾ Roughly speaking, the red and blue pixels indicate the Ld and Lo phases, respectively. Because the value of a single pixel fluctuated considerably, we compared the average ratio of three areas: the cell body (Fig. 1B), the growth cone (Fig. 1D), and the boundary between them (Fig. 1C). Growth cones present more pixels in the orange-red range (Fig. 1D). On the other hand, the cell body has more pixels in the green-blue range. To determine how much the I600 > /I550-600 ratio values varied within the area, the CQD was also calculated from those values. The ratio values were systematically higher in the growth cone than in the cell body (Fig. 2A) and those of the boundary area were approximately between those of the cell body and growth cone, but closer to the cell body (Fig. 2A). Overall, it became evident that the polarity of the growth cone membrane was higher than that of the cell body or of the boundary region (Fig. 2B). The CQD was also higher in the growth cone than in the cell body or the boundary area (Figs. 2C, D). It was thus concluded that the growth cones exhibit higher membrane polarity than the cell bodies in hippocampal neurons, and the CQD of membrane polarity is also higher in hippocampal neurons.

Fig. 1. Ratiometric Imaging of the Plasma Membrane Polarity in a Cultured Hippocampal Neuron Using a Cell-Impermeable Solvatochromic Dye NR12A

A, whole picture of a culture hippocampal neuron. Scale bar, 5 µm. B–D, the enlarged image of the cell body (B), the boundary (C), and the growth cone (D), indicated by red, white, and blue boxes shown in A, respectively. The area used for the calculation is indicated by the white line. E, pseudocolor scale for the indication of the I600 > /I550–600 ratio.

Fig. 2. The Overall Polarity and Fluctuation Are Higher in the Growth Cone than in the Cell Body of the Cultured Hippocampal Neuron

A, the ratiometric values of cell body, boundary, and growth cone from individual neurons in a set of experiment are plotted and connected by a line. B, the summary of five independent experiments. The p-values for comparison between cell body/growth cone, cell body/boundary, and boundary/growth cone are 1.98 × 10⁻³, 4.72 × 10⁻², and 7.77 × 10⁻³, respectively. C, the CQD of the cell body, boundary, and growth cone from individual neurons in a set of experiment are plotted and connected by a line. D, the summary of five independent experiments. The p-values for comparison between cell body/growth cone, cell body/boundary, and boundary/growth cone are 1.33 × 10⁻³, 4.75 × 10⁻¹, and 4.58 × 10⁻³, respectively. Five to ten neurons are analyzed in each experiment.

The Polarity of the Growth Cone Is Higher than That of the Cell Body in the Cultured CGNs

We next checked whether membrane polarity was also heterogeneously distributed in neurons other than the hippocampus. We used mouse CGNs, which basically have a single axon and do not develop dendrites in primary culture conditions. As with hippocampal neurons, the growth cones present more pixels in the yellow to red range (Figs. 3A, D). On the other hand, the cell body has more pixels in the green to blue range (Fig. 3B). The cell body, boundary, and growth cone areas were enclosed to determine the averaged ratiometric values and CQD among these areas. The tendency was almost the same as that of hippocampal neurons: the polarity and CQD were higher in the growth cones than those in the cell body and the boundary areas (Figs. 3F–H).

Fig. 3. The Overall Polarity and Fluctuation Are Higher in the Growth Cone than in the Cell Body in the CGNs

A, a ratiometric fluorescence image of a CGN. Scale bar, 5 µm. B–D, the enlarged image of the cell body (B), the boundary (C), and the growth cone (D), indicated by red, white, and blue boxes shown in A, respectively. The area used for the calculation is indicated by the white line. E, pseudocolor scale for the indication of I600 > /I550–600 ratio. F, the ratiometric values of cell body, boundary, and growth cone from individual neurons in a set of experiment are plotted and connected by a line. G, the summary of polarity from three independent experiments. The p-values for comparison between cell body/growth cone, cell body/boundary, and boundary/growth cone are 3.08 × 10⁻³, 5.28 × 10⁻¹, and 2.81 × 10⁻³, respectively. H, the summary of CQD from three independent experiments. The p-values for comparison between cell body/growth cone, cell body/boundary, and boundary/growth cone are 2.72 × 10⁻³, 8.48 × 10⁻², and 1.25 × 10⁻³, respectively. Fourteen to twenty neurons are analyzed in each experiment.

The Polarity of Lamellipodia of Non-neuronal Cells Is Not Much Different from That of Other Area

The lamellipodia of HEK293T cells, which are non-neuronal, structurally resemble the growth cones of neurons (Fig. 4A). We examined the difference in the plasma membrane polarity between the lamellipodia and the non-lamellipodia part of HEK293T cells (Figs. 4A–C). The polarity of the lamellipodia was higher than that of the non-lamellipodia in some cells but the difference was not very obvious (Fig. 4E). Quantification revealed that the polarity of lamellipodia and non-lamellipodia membrane were only marginally different (Fig. 4F) and the same went with the CQD (Fig. 4G). These results suggested that the lamellipodia-like structure contributes to the high polarity of the plasma membrane of the growth cone in neurons.

Fig. 4. The Polarity of Lamellipodium Is Not Significantly Different from That of Other Regions in HEK293T Cells

A, a ratiometric fluorescence image of a HEK293T cell. Scale bar, 5 µm. B, C, the enlarged image of the lamellipodium (B) and non-lamellipodium (C) area, indicated by blue and red boxes shown in A, respectively. The area used for the calculation is indicated by the white line. D, pseudocolor scale for the indication of I600 > /I550–600 ratio. E, the ratiometric values of non-lamellipodium and lamellipodium from individual cells in a set of experiment are plotted and connected by a line. F, the summary of three independent experiments. The p-value is 1.68 × 10⁻². G, the summary of CQD from three independent experiments. The p-value is 1.67 × 10⁻². Ten to twenty-one cells are analyzed in each experiment.

DISCUSSION

Membrane polarity influences the flexibility and mobility of enzymes, which is crucial for conformational changes necessary for substrate binding, catalysis, and product release.^1,17) Proper membrane polarity ensures the dynamic behavior of enzymes, facilitating their functional conformational transitions. Membrane polarity can impact the lateral organization and distribution of both enzymes and substrates within the membrane.^1,17) The brain contains a huge repertoire of lipids,^18,19) and neurons utilize many of them to fulfill their versatile and critical functions.^20,21) Some secreted factors are known to affect the lipid composition of the brain.^22–24) Among those factors, a secreted glycoprotein Reelin contributes to increased levels of polyunsaturated fatty acids in neurons,²³⁾ which may partly explain why Reelin induces dendrite elongation.^25–27) However, how neurons regulate plasma membrane polarity is not understood. To the best of our knowledge, no experimental evidence on the heterogeneous distribution of plasma membrane polarity within a neutron has been reported to date. Our objective was to tackle this question.

In recent years, technological developments have begun to reveal the localization of lipid molecules in cells and the local membrane polarity of cells. One such technique is the utilization of fluorescent probes. The fluorescent probe NR12A has different fluorescence wavelengths in the Lo and Ld phases and does not enter the cells,^6,10) making it possible to specifically evaluate the polarity of the plasma membrane.

When we first started to obtain images using NR12A, we noticed that the signal of each pixel fluctuated considerably. This may be due to the lateral diffusion of NR12A or lipid molecules. In any case, it was difficult (and probably meaningless) to determine the exact value of the polarity of a specific pixel. We instead took the average of each region in the neurons, which was much more stable. Then, we found that the polarity of the growth cone was generally higher than that of the cell body (Figs. 2, 3). To the best of our knowledge, this is the first demonstration of the unique membrane polarity of the neuronal growth cone.

The polarity of the plasma membrane of neuronal cells is controlled by several factors. First, the types of lipids present in the plasma membrane greatly impact its polarity. Phospholipids have different fatty acid chains a higher proportion of unsaturated phospholipids increases membrane polarity.^1,2,4) Polyunsaturated fatty acids such as docosahexaenoic acid have been shown to increase neurite length and branching.^28,29) Arachidonic acid incorporated into phosphatidylinositol is essential for neuronal migration and neurite extension.³⁰⁾ Cholesterol is another essential component of the plasma membrane that influences its polarity.^1,2,4) The roles of cholesterol and cholesterol-rich microdomain have been investigated in detail,^31–33) although there has been no universal consensus on their roles. Integral membrane proteins also affect membrane polarity.^1,2,4) As it is conceivable that neurons require an optimal level of membrane polarity for proper morphology and functioning, the balance between these factors ensures an appropriate level of polarity that allows membrane dynamics and facilitates functional integrity. The mechanism underlying these events remains largely unknown, but our results in this study will pave the way for further research in this direction.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Nicolson GL, Ferreira de Mattos G. The fluid-mosaic model of cell membranes: a brief introduction, historical features, some general principles, and its adaptation to current information. Biochim. Biophys. Acta. Biomembr., 1865, 184135 (2023).
• 2) Suzuki KGN, Kusumi A. Refinement of Singer–Nicolson fluid-mosaic model by microscopy imaging: lipid rafts and actin-induced membrane compartmentalization. Biochim. Biophys. Acta. Biomembr., 1865, 184093 (2023).
• 3) Lingwood D, Simons K. Lipid rafts as a membrane-organizing principle. Science, 327, 46–50 (2010).
• 4) Sezgin E, Levental I, Mayor S, Eggeling C. The mystery of membrane organization: composition, regulation and roles of lipid rafts. Nat. Rev. Mol. Cell Biol., 18, 361–374 (2017).
• 5) Feigenson GW, Enoki TA. Nano-scale domains in the plasma membrane are like macroscopic domains in asymmetric bilayers. Biophys. J., 122, 925–930 (2023).
• 6) Klymchenko AS. Fluorescent probes for lipid membranes: from the cell surface to organelles. Acc. Chem. Res., 56, 1–12 (2023).
• 7) Niko Y, Klymchenko AS. Emerging solvatochromic push-pull dyes for monitoring the lipid order of biomembranes in live cells. J. Biochem., 170, 163–174 (2021).
• 8) Yang H-J, Sugiura Y, Ikegami K, Konishi Y, Setou M. Axonal gradient of arachidonic acid-containing phosphatidylcholine and its dependence on actin dynamics. J. Biol. Chem., 287, 5290–5300 (2012).
• 9) Niko Y, Didier P, Mely Y, Konishi G, Klymchenko AS. Bright and photostable push-pull pyrene dye visualizes lipid order variation between plasma and intracellular membranes. Sci. Rep., 6, 18870 (2016).
• 10) Danylchuk DI, Moon S, Xu K, Klymchenko AS. Switchable solvatochromic probes for live-cell super-resolution imaging of plasma membrane organization. Angew. Chem. Int. Ed. Engl., 58, 14920–14924 (2019).
• 11) Valanciunaite J, Kempf E, Seki H, Danylchuk DI, Peyriéras N, Niko Y, Klymchenko AS. Polarity mapping of cells and embryos by improved fluorescent solvatochromic pyrene probe. Anal. Chem., 92, 6512–6520 (2020).
• 12) Kucherak OA, Oncul S, Darwich Z, Yushchenko DA, Arntz Y, Didier P, Mély Y, Klymchenko AS. Switchable nile red-based probe for cholesterol and lipid order at the outer leaflet of biomembranes. J. Am. Chem. Soc., 132, 4907–4916 (2010).
• 13) Sokoya T, Parolek J, Foged MM, Danylchuk DI, Bozan M, Sarkar B, Hilderink A, Philippi M, Botto LD, Terhal PA, Mäkitie O, Piehler J, Kim Y, Burd CG, Klymchenko AS, Maeda K, Holthuis JCM. Pathogenic variants of sphingomyelin synthase SMS2 disrupt lipid landscapes in the secretory pathway. eLife, 11, e79278 (2022).
• 14) Nakao Y, Yokawa S, Kohno T, Suzuki T, Hattori M. Visualization of reelin secretion from primary cultured neurons by bioluminescence imaging. J. Biochem., 171, 591–598 (2022).
• 15) Ishii K, Kohno T, Sakai K, Hattori M. Reelin regulates the migration of late-born hippocampal CA1 neurons via cofilin phosphorylation. Mol. Cell. Neurosci., 124, 103794 (2023).
• 16) Dotti CG, Sullivan CA, Banker GA. The establishment of polarity by hippocampal neurons in culture. J. Neurosci., 8, 1454–1468 (1988).
• 17) de Santis A, Scoppola E, Ottaviani MF, Koutsioubas A, Barnsley LC, Paduano L, D’Errico G, Russo Krauss I. Order vs. disorder: cholesterol and omega-3 phospholipids determine biomembrane organization. Int. J. Mol. Sci., 23, 5322 (2022).
• 18) Yoon JH, Seo Y, Jo YS, Lee S, Cho E, Cazenave-Gassiot A, Shin Y-S, Moon MH, An HJ, Wenk MR, Suh P-G. Brain lipidomics: from functional landscape to clinical significance. Sci. Adv., 8, eadc9317 (2022).
• 19) Aureli M, Grassi S, Prioni S, Sonnino S, Prinetti A. Lipid membrane domains in the brain. Biochim. Biophys. Acta, 1851, 1006–1016 (2015).
• 20) Igarashi M, Honda A, Kawasaki A, Nozumi M. Neuronal signaling involved in neuronal polarization and growth: lipid rafts and phosphorylation. Front. Mol. Neurosci., 13, 150 (2020).
• 21) Chauhan MZ, Arcuri J, Park KK, Zafar MK, Fatmi R, Hackam AS, Yin Y, Benowitz L, Goldberg JL, Samarah M, Bhattacharya SK. Multi-omic analyses of growth cones at different developmental stages provides insight into pathways in adult neuroregeneration. Iscience, 23, 100836 (2020).
• 22) Asante I, Louie S, Yassine HN. Uncovering mechanisms of brain inflammation in Alzheimer’s disease with APOE4: Application of single cell-type lipidomics. Ann. N. Y. Acad. Sci., 1518, 84–105 (2022).
• 23) Mizukami T, Ikeda K, Shimanaka Y, Korogi K, Zhou C, Takase H, Tsuiji H, Kono N, Kohno T, Arai H, Arita M, Hattori M. Reelin deficiency leads to aberrant lipid composition in mouse brain. Biochem. Biophys. Res. Commun., 505, 81–86 (2018).
• 24) Mu P, Liu Y, Jiang S, Gao J, Sun S, Li L, Gao D. Glial cell line-derived neurotrophic factor alters lipid composition and protein distribution in MPP + -injured differentiated SH-SY5Y cells. J. Cell. Physiol., 235, 9347–9360 (2020).
• 25) Niu S, Renfro A, Quattrocchi CC, Sheldon M, D’Arcangelo G. Reelin promotes hippocampal dendrite development through the VLDLR/ApoER2-Dab1 pathway. Neuron, 41, 71–84 (2004).
• 26) Leemhuis J, Bouche E, Frotscher M, Henle F, Hein L, Herz J, Meyer DK, Pichler M, Roth G, Schwan C, Bock HH. Reelin signals through apolipoprotein E receptor 2 and Cdc42 to increase growth cone motility and filopodia formation. J. Neurosci., 30, 14759–14772 (2010).
• 27) Kohno T, Honda T, Kubo K-I, Nakano Y, Tsuchiya A, Murakami T, Banno H, Nakajima K, Hattori M. Importance of Reelin C-terminal region in the development and maintenance of the postnatal cerebral cortex and its regulation by specific proteolysis. J. Neurosci., 35, 4776–4787 (2015).
• 28) Calderon F, Kim H-Y. Docosahexaenoic acid promotes neurite growth in hippocampal neurons. J. Neurochem., 90, 979–988 (2004).
• 29) Cao D, Kevala K, Kim J, Moon H-S, Jun SB, Lovinger D, Kim H-Y. Docosahexaenoic acid promotes hippocampal neuronal development and synaptic function. J. Neurochem., 111, 510–521 (2009).
• 30) Lee H-C, Inoue T, Sasaki J, Kubo T, Matsuda S, Nakasaki Y, Hattori M, Tanaka F, Udagawa O, Kono N, Itoh T, Ogiso H, Taguchi R, Arita M, Sasaki T, Arai H. LPIAT1 regulates arachidonic acid content in phosphatidylinositol and is required for cortical lamination in mice. Mol. Biol. Cell, 23, 4689–4700 (2012).
• 31) Tassew NG, Mothe AJ, Shabanzadeh AP, Banerjee P, Koeberle PD, Bremner R, Tator CH, Monnier PP. Modifying lipid rafts promotes regeneration and functional recovery. Cell Reports, 8, 1146–1159 (2014).
• 32) Roselló-Busquets C, de la Oliva N, Martínez-Mármol R, Hernaiz-Llorens M, Pascual M, Muhaisen A, Navarro X, Del Valle J, Soriano E. Cholesterol depletion regulates axonal growth and enhances central and peripheral nerve regeneration. Front. Cell. Neurosci., 13, 40 (2019).
• 33) Galvan C, Camoletto PG, Dotti CG, Aguzzi A, Ledesma MD. Proper axonal distribution of PrP^C depends on cholesterol-sphingomyelin-enriched membrane domains and is developmentally regulated in hippocampal neurons. Mol. Cell. Neurosci., 30, 304–315 (2005).