SEARCH

SEARCH BY CITATION

Keywords:

  • α-synuclein;
  • clathrin-mediated endocytosis;
  • membrane fluidity;
  • membrane trafficking;
  • Parkinson’s disease;
  • polyunsaturated fatty acids;
  • synaptic vesicle recycling

Abstract

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgment
  7. References

α-Synuclein (αS) is an abundant neuronal cytoplasmic protein implicated in Parkinson’s disease (PD), but its physiological function remains unknown. Consistent with its having structural motifs shared with class A1 apolipoproteins, αS can reversibly associate with membranes and help regulate membrane fatty acid composition. We previously observed that variations in αS expression level in dopaminergic cultured cells or brains are associated with changes in polyunsaturated fatty acid (PUFA) levels and altered membrane fluidity. We now report that αS acts with PUFAs to enhance the internalization of the membrane-binding dye, FM 1-43. Specifically, αS expression coupled with exposure to physiological levels of certain PUFAs enhanced clathrin-mediated endocytosis in neuronal and non-neuronal cultured cells. Moreover, αS expression and PUFA-enhanced basal and -evoked synaptic vesicle (SV) endocytosis in primary hippocampal cultures of wild type (wt) and genetically depleted αS mouse brains. We suggest that αS and PUFAs normally function in endocytic mechanisms and are specifically involved in SV recycling upon neuronal stimulation.

The neuronal protein, α-synuclein (αS), has been implicated in the pathogenesis of Parkinson’s disease (PD) at both the genetic and the cytopathological levels (1–7). Despite the involvement of this abundant neuronal protein in sporadic and familial forms of PD and related α-synucleinopathies, both its normal function and the mechanism by which it gradually accumulates in dopaminergic and other neurons in disease remain unclear.

A portion of αS associates with membranes in vitro (8–16) and in vivo (17–22). These observations are consistent with its primary structure, which contains six imperfect apolipoprotein A1-like repeats in its N-terminal region that may mediate lipid binding (23,24). We obtained evidence that αS can associate with polyunsaturated fatty acids (PUFAs) in vitro and in neuronal cells and brain tissue (18). Importantly, we found that changes in αS expression can affect membrane and cytosolic PUFA composition and alter membrane fluidity. Specifically, we observed higher levels of certain long-chain PUFAs and higher fluidity in membranes of MES 23.5 dopaminergic cells overexpressing αS than in those of parental cells and lower levels of such PUFAs and lower fluidity in membranes of αS−/− than normal mouse brains (25). More recently, it was reported that αS can affect brain lipid metabolism and specifically PUFA metabolism (26–30). In agreement with our initial observation that αS expression affects membrane fatty acid (FA) composition (25), these studies in αS null mouse brains documented reduced incorporation of certain FAs into membrane phospholipids as well as decreases in FA uptake and turnover (26,27,29,30).

During endocytosis, a small region of the plasma membrane invaginates to form a new intracellular vesicle containing various cargo molecules. Clathrin-mediated endocytosis (CME) is the major entry route for extracellular molecules such as nutrients, hormones and signaling factors and serves to regulate the internalization of transmembrane receptors, including the recycling of pre- and postsynaptic neuronal membrane proteins (31–33). Although clathrin-coated vesicles are found in all eukaryotic cells, their components are particularly enriched in brain, where clathrin and its partner proteins are implicated in the biogenesis of presynaptic vesicles, the major secretory organelles within the nervous system (34,35). The cargo for endocytosis is usually recognized by a specific receptor on the cell surface. Most receptor-mediated endocytosis (RME) is mediated by clathrin-coated pits.

PUFAs have been found to play a role in the formation and/or maintenance of synaptic vesicles (SVs) (36,37), including dopaminergic vesicles (38–43). It has been hypothesized that because of their ‘cone shape’, PUFAs affect membrane curvature in a way that promotes vesicle budding and membrane trafficking (44,45). In this context, it is of interest that αS is localized in part to presynaptic neuronal terminals and has been found to be involved in the genesis and/or maintenance of the reserve, or resting, presynaptic vesicle pools (17,46,47).

In this study, we assess the effects of αS and PUFAs on plasma membrane trafficking. We provide evidence that αS and PUFAs affect endocytosis and vesicle recycling in both neuronal and non-neuronal cells and specifically activate SV recycling after neuronal stimulation by enhancing CME.

Results

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgment
  7. References

αS occurs in soluble oligomers and affects the cellular incorporation of 14C oleic acid

We previously reported that a portion of cellular αS can be detected as low-n soluble oligomers (dimers up to hexamers) (48). However, the visualization of these sodium dodecyl sulfate (SDS)-stable oligomers on western blots required preextraction of lipids, for example by heating cytosolic or membrane fractions to 65°C or else performing chloroform/methanol extraction. Such treatments could potentially induce protein denaturation and aggregation in the presence of SDS sample buffer. To address whether the lipid-associated soluble αS oligomers occur under non-denaturing conditions in vivo, we have incubated high-speed cytosols (post 100 000 × g) of mouse brains with Lipidex-1000 (4°C, 16 h), followed by native gel electrophoresis. This revealed αS forms migrating higher than the monomer in wt mouse brains and to a greater extent in A53T αS+/+ transgenic mouse brains (49) (Figure 1A). The signal was not present in cytosol from αS−/−(50) mouse brains, confirming its specificity (Figure 1A). We next analyzed the high-speed cytosols of untransfected and αS-overexpressing MES dopaminergic cells grown in serum-free medium supplemented with 35 μm14C oleic acid (OA) and 35 μm FA-free BSA (a known FA carrier protein) for 16 h, followed by native gel electrophoresis. The αS-overexpressing cells incubated under these conditions consistently showed greater amounts of 14C OA incorporation throughout the lane than did untransfected naive cells (Figure 1B). We conclude that cytosolic αS occurs natively in the form of soluble oligomers and affects the cellular FA content.

image

Figure 1. αS occurs in soluble oligomers in vivo and promotes incorporation of 14C OA. A) Samples of post 100 000 × g cytosols (15 μg) of brains from wt (16 month), αS +/+ (9 month) and αS −/− (15 month) mice were incubated with Lipidex-1000 at 4°C for 16 h followed by native gel electrophoresis (10% Tris-Glycine) and western blotting with syn-1 antibody. B) Samples of post 100 000 × g cytosols (15 μg) from naive or αS-overexpressing MES cells conditioned with 14C OA (35 μm) for 16 h and subjected to native gel electrophoresis (autoradiogram).

Download figure to PowerPoint

αS enhances the internalization of FM1-43-labeled plasma membrane

To assess the possible involvement of αS in membrane trafficking, we initially used the well-characterized lipophilic dye, FM1-43 (reviewed in 51,52), to detect newly formed endocytic vesicles budding from the plasma membrane of MES dopaminergic neuronal cells maintained in standard serum-supplemented medium. FM1-43 (2.5 μm) was added to the conditioning medium for 4 min at 37°C, and then, free dye was removed by multiple washes. The labeled cells were then cultured at 37°C for an additional 10 min to allow endocytosis. Higher levels of FM1-43-labeled endocytic vesicles were invariably detected by confocal microscopy in cells stably overexpressing human wt αS than in non-transfected MES cells maintained under standard serum conditions (Figure 2A). Quantifying the FM1-43 signal revealed a ∼10-fold increase in wt αS-overexpressing cells than parental MES cells. This increase in FM1-43 internalization correlates with the ∼13-fold increase in αS expression versus naive untransfected cells. In quantitative experiments employing similar levels of αS expression and identical FM1-43-containing medium, stable expression of the PD-causing A53T αS mutant induced a more marked effect (two to three times wt levels) in the amount of internalized FM1-43-labeled membrane (Figure 2A).

image

Figure 2. FM1-43 internalization into MES 23.5 dopaminergic cells is induced by αS overexpression and 18:3 PUFA. A) Naive, wt and A53T αS-overexpressing MES cells conditioned in standard serum-containing medium and labeled with 2.5 μm FM1-43 fluorescent dye (Molecular Probes) for 4 min. Access of dye was removed, and 10 min later, the cells were fixed and tested for intracellular FM1-43 fluorescence. Pictures were taken in confocal, laser scanning microscope (Zeiss LSM 410) under nonsaturating conditions. B) Naive, wt and A53T αS-overexpressing MES cells were conditioned for 16–18 h in serum-free medium supplemented with 50 μm BSA, followed by labeling with FM1-43 as in (A). C) Cells as in (B) but conditioned in serum-free medium supplemented with BSA only (50 μm) plus 250 μm 18:3 PUFA. D) Quantitative presentation of the intracellular FM1-43 in BSA- versus PUFA-treated cultures. Quantification was performed on the captured images at selected image plane with the greatest intracellular FM1-43 signal, using ImageJ software, measuring the average fluorescence intensity (above threshold) across the entire cell divided by the area of the tested cell. Mean of 10–15 cells ± SD. *Significant over the corresponding treatment in naive cells, and **significant over the corresponding treatment in wt αS-overexpressing cells. t-test, p < 0.05. E) Cells were treated as in (B) and (C) but with 20 μm BSA with or without 100 μm 18:3, respectively. F) Western blot of cytosolic extract from naive, wt and A53T-overexpressing cells, treated for oligomers detection (48) and probed with anti αS H3C Ab. Bar represents 10 μm.

Download figure to PowerPoint

In view of our earlier findings that increased cellular PUFA levels enhance αS oligomerization (48) and that αS expression levels affect cellular PUFA composition (25), we next asked whether PUFAs play a role in the accumulation of αS-dependent, FM1-43-labeled endocytic vesicles. For this, we conditioned naive- and αS-overexpressing MES cells in serum-free medium supplemented with BSA only (50 μm) or with BSA plus the 18:3 PUFA (250 μm) (see Materials and Methods). The cells were conditioned in the specified medium for 16–18 h prior to labeling with FM1-43 (as described above), and the effects on endocytic vesicle formation were again observed by confocal microscopy. A consistent increase in FM1-43 fluorescence was detected in wt αS-overexpressing cells conditioned in serum-free medium supplemented with BSA only, specifically, an increase of approximately sevenfold over the parental, naive cells (Figure 2B). Thus, the effect of αS overexpression in enhancing endocytosis can be seen in the absence of any exogenous lipids in the conditioning medium. An increase in intracellular FM1-43 fluorescence was detected in parental cells conditioned in BSA + 250 μm 18:3 (increased approximately fivefold versus BSA alone) (Figure 2C). Thus, the effect of PUFA on endocytosis can occur independently of αS expression. Furthermore, a marked enhancement of plasma membrane internalization was observed by combining wt αS overexpression with 250 μm 18:3 (increased ∼30-fold versus parental MES cells conditioned in BSA alone, i.e., no 18:3 and no αS) (Figure 2D). The combined effect of αS stable expression and 18:3 was far higher than the addition of the effects of each factor alone, suggesting that PUFAs and αS act synergistically in this assay. Further enhancement in FM1-43 fluorescence was observed in MES cells stably expressing the A53T αS and conditioned in medium supplemented with BSA or BSA-PUFA (Figure 2B–D). Importantly, the fold increase (versus serum-free medium) of plasma membrane internalization in standard serum-containing medium or in medium supplemented with a PUFA (i.e., BSA + 250 μm 18:3) was closely similar. That is, an increase of ∼10- and ∼12-fold was detected with standard serum and with PUFA, respectively, in αS-overexpressing cells (Figure 2A,C). This closely similar effect of serum and a single PUFA indicates that the substantial response of the cells to a concentration of 250 μm 18:3 is in a physiological range. Nevertheless, we next performed this experiment at a lower concentration of 100 μm 18:3 and obtained similar results but with a lower fold induction; again, αS or PUFA alone each induced FM1-43 internalization, and they showed an apparent synergistic effect upon their combination (Figure 2E). Finally, we performed western blot analyses of wt and A53T αS-overexpressing cells and detected similar αS expression level in both clones (Figure 2F). Importantly, we tested FM1-43 internalization as a function of expression of the related family member, βS, and found no effect of βS expression alone or when combined with 18:3 treatment.

We next compared the effects of αS expression in neuronal and non-neuronal cell lines to determine whether the effect was restricted to neuronal cells. We examined HeLa, HEK 293, SK-N-SH, MN9D and MES cells, the two latter being dopaminergic neuronal lines. In all the five cell lines, αS expression consistently and markedly enhanced the levels of internalized FM1-43-labeled membrane (data not shown). We conclude that the robust effect of αS expression in enhancing the internalization of FM1-43-labeled plasma membrane is not specific to neuronal cells and can be observed in other cell types upon its overexpression.

PUFAs of greater carbon chain length and higher degree of unsaturation are more potent in inducing αS-dependent endocytosis

To confirm and extend the effects of 18:3 on the accumulation of endocytic vesicles, we compared the effects of different FAs on the formation of FM1-43-labeled vesicles in the presence of αS overexpression. Intriguingly, the saturated FAs (SFA), 18:0 and 20:0 (each at 250 μm with 50 μm BSA as the standard carrier protein) decreased the accumulation of FM1-43-labeled endocytic vesicles compared with BSA only. The monounsaturated fatty acid (MUFA), 18:1, produced no detectable alteration compared with BSA alone and longer chain MUFA, 20:1, slightly but significantly enhanced the accumulation of FM1-43. In contrast, the PUFAs, 18:2, 18:3, 20:2, 20:3 and 20:4, each consistently induced the accumulation of FM1-43 endocytic vesicles (Table 1). We found that the quantitative effect of PUFAs on the FM1-43 internalization signal varied as a function of both increasing carbon chain length and increasing degree of unsaturation. The longer and more polyunsaturated the FA, the more FM1-43 fluorescently labeled endocytic vesicles accumulated inside the cells (Figure 3 and Table 1). In general, the vesicles appeared larger and more coalesced after treatment with the longer, more unsaturated FAs versus with BSA alone.

Table 1.  FM1-43 internalization as a function of FA carbon chain length and level of unsaturation
FAFM1-43 internalizationa
  • a

    See Materials and Methods for quantification of FM1-43 signal.

  • b

    Mean of (15–20 representative cells) ± SE.

  • *

    p < 0.05, t-test versus BSA alone.

BSA6.86 ± 1.7b
18:0 Stearate0.33 ± 0.12*
18:1 Oleate7.3 ± 0.941
18:2 Linoleate18.7 ± 1.2*
18:3 Linolenate27.5 ± 2.7*
20:0 Arachidate0.4 ± 0.2*
20:1 Eicosenoate13.6 ± 1.67*
20:2 Eicosadienoate22.3 ± 3.87*
20:3 Eicosatrienoate30.6 ± 3.89*
20:4 Arachidonate45.9 ± 2.67*
image

Figure 3. FM1-43 endocytosis as a function of FA length and saturation. αS-overexpressing MES dopaminergic cells were grown on 12-well cover glass and conditioned for 16 h in serum-free medium supplemented with BSA only (50 μm) or with 250 μm of the indicated FA in sister cultures in parallel. Cells were then labeled with FM1-43 as in Figure 2A and fixed. Representative pictures are presented. Image acquisition by Zeiss LSM 410. Acquisition parameters were kept fixed throughout. Bar represents 10 μm.

Download figure to PowerPoint

αS and PUFAs enhance endocytosis of transferrin

To elucidate the type of endocytic mechanism that is activated by αS and PUFAs, we studied the internalization of transferrin (Tf) by the transferrin receptor (TfR), the archetypical cargo for internalization by CME, as a model for RME. TfR binds its ligand (ferrotransferrin) at the cell surface and is internalized into early endosomes, where it releases the bound iron and recycles back to the plasma membrane. To measure the effects of αS and PUFAs on Tf endocytosis, we used fluorescently labeled human Tf (Alexa-488-Tf). MN9D cells were conditioned in serum-free DMEM supplemented with BSA +/− 18:3 or 18:1 (50 and 250 μm for BSA and 18:3, respectively) for 16 h, followed by the application of Alexa-488-Tf (50 μg/mL) at 4°C for 60 min to allow the binding of Tf to its receptor without internalization. Cells were then washed and transferred to 37°C to allow internalization, fixed in 4% paraformaldehyde (PFA) and processed to visualize and quantify Alexa-488-Tf by confocal microscopy (see Materials and Methods).

Using the naive, untransfected MN9D cells supplemented with BSA only (at 50 μm) as the control for basal endocytic activity, we found that 18:3 by itself (at 250 μm) induced Tf endocytosis into large endosomal vesicles (Figure 4A,B). αS overexpression alone also induced Tf endocytosis, resulting in a diffuse pattern of many smaller endocytic vesicles. The combination of αS overexpression and 18:3 in the conditioning medium further enhanced endocytosis of fluorescent Tf by the receptor (Figure 4A,B). Importantly, the corresponding 18:1 treatment had no effect on Tf endocytosis (data not shown). We performed the same experiments in MES dopaminergic cells and neuroblastoma SK-N-SH cells, yielding very similar results (data not shown). Additionally, we tested the effect of αS expression on Tf endocytosis in the presence of standard serum-supplemented conditioning medium and found that the PUFA activation of Tf endocytosis is within the physiologic range of serum alone (Figure 4B).

image

Figure 4. αS and PUFA act to induce endocytosis of Tf. A) Naive and wt αS-overexpressing MN9D cells were grown on 12-well cover glass and conditioned for 18 h in DMEM containing 250 μm 18:3 PUFA or DMEM with BSA (50 μm). Alexa-488 human Tf (Molecular Probes) (50 μg/mL) was then added to the cells in plain DMEM medium (without serum, antibiotics or nutrients) in cold to allow binding of Tf to the TfR. After the removal of unbound Tf, cells were incubated for 10 min at 37°C in serum-free medium supplemented with BSA or PUFA accordingly. Cells were then fixed and observed by Olympus FluoView FV300 confocal microscope. B) Quantification of Alexa-488-Tf endocytosis. Cells and treatments as in (A) plus naive and αS-overexpressing MN9D conditioned in standard serum-supplemented medium. All images were taken at the same settings, and the image plane containing maximal intracellular signal was selected. Threshold was set as to measure only endocytic vesicles, rather than including a more diffuse background. All measurements were of ‘selected region of interest’ normalized to cell area. Mean of 10 cells of each treatment ± SD. *, t-test p < 0.01. C) The endocytosis of Alexa-488-Tf is PUFA dose dependent in naive and αS-overexpressing cells. αS-overexpressing SK-N-SH cells were grown as in (A), conditioned with the indicated 18:3 concentration and quantified as in (B). Graph presenting the sum of signal in endocytic vesicles (above threshold) to 18:3 concentrations. Ten cells counted ± SD of αS-overexpressing cells (dashed line) and naive (filled line). Bar represents 10 μm.

Download figure to PowerPoint

We next measured Tf endocytosis as a function of increasing 18:3 concentrations in the SK-N-SH cells with or without αS overexpression. A dose-dependent effect of 18:3 concentrations on degree of Tf endocytosis was detected in naive and αS-overexpressing cells (Figure 4C). Higher degree of endocytosis was detected in αS-overexpressing cells throughout the 18:3 range of concentrations up to the highest concentration tested, that is 250 μm (Figure 4C). Importantly, the increased slope in αS-overexpressing (a = 0.13) versus naive cells (a = 0.06) suggests that αS and 18:3 may act synergistically. A significant increase in endocytosis of Tf was detected already at the lowest 18:3 concentration tested (i.e. 50 μm, t-test, p < 0.05), which is well within the physiological range of brain PUFA concentrations (53). Increased levels of Tf endocytosis were detected in a linear relationship to the increased 18:3 concentrations, with a correlation coefficient of 0.96 (p < 0.001, data not shown).

αS and PUFAs affect TfR cellular distribution and expression level

We examined the protein levels and cellular distribution of TfR in naive and αS-overexpressing cells. MN9D cells were incubated in serum-free DMEM (to obtain low iron levels) and supplemented with BSA and 18:3 for 16–18 h, as above. Following the addition of Tf (see above), cells were fixed in 4% PFA and processed for immunocytochemistry (ICC) with an antibody to TfR to determine receptor level and distribution. Using naive MN9D cells supplemented with BSA only (50 μm) as the control for basal TfR level and cellular distribution, we found that adding 18:3 (250 μm) or overexpressing αS resulted in quantitative increases in total receptor signal at the cytoplasm and the cell surface, with further enhancement of TfR signal by combining αS and PUFA (Figure 5A).

image

Figure 5. αS and PUFA affect TfR levels and cellular distribution. A) Naive and human wt αS-overexpressing MN9D neuronal cells were grown on 12-well cover glass, conditioned for 18 h in 250 μm 18:3 PUFA or in BSA only (50 μm). Tf (50 μm) was added as in Figure 4A, and cells were processed for ICC using anti-TfR Ab (Zymed Laboratories). B) αS and PUFA increase levels of cell surface TfR. Biotinylation of cell surface TfR, normalized to total TfR using anti-TfR Ab (Zymed Laboratories) and to actin (Sigma) and its densitometric analysis. C) αS and PUFA induce TfR expression levels. Quantitative real-time polymerase chain reaction of total RNA extracted from naive and αS-overexpressing MN9D cells treated in BSA ± 18:3 (at 50 and 250 μm for BSA and 18:3). The levels of TfR mRNA were normalized to the 18S mRNA. Mean of three experiments ± SE. Bar represents 10 μm. *Naive serum and αS serum, p<0.01.

Download figure to PowerPoint

We next performed surface biotinylation experiments to confirm the apparent increase in cell surface TfR levels. Naive and αS-overexpressing MN9D cells were conditioned in serum-free medium supplemented with BSA +/− 18:3 (at 250 μm) for 16 h. Cells were then transferred to 4°C to prevent internalization of receptors. Cell surface proteins were biotinylated with sulfo-NHS-LC-biotin (0.5 mg/mL). The amount of biotinylated TfR was normalized to the amounts of total TfR protein levels and to actin in the same sample using quantitative western blotting. A higher portion of surface TfR was detected upon αS expression or PUFA treatment, with highest surface TfR ratios detected in the αS-overexpressing PUFA-treated cells (Figure 5B). The effect of αS expression on surface TfR levels was also determined under standard serum-supplemented conditioning medium with very similar results (Figure 5B).

Densitometry of western blots probed with anti-TfR antibody showed elevated total TfR protein in response to PUFA treatment alone (∼60% increase), to αS overexpression alone (∼85% increase) or to both combined (∼200% increase) (Figure 5B). In accord, we detected an approximately threefold increase in TfR messenger RNA (mRNA) levels in the αS-overexpressing cells compared with the naive cells (Figure 5C) conditioned in standard serum-supplemented medium. Marked increases in TfR mRNA levels were detected in the BSA- and PUFA-treated cells. These may result from the effect of serum starvation and low iron levels; nevertheless, under these serum-free conditions, the levels of TfR mRNA were higher (but not significantly higher) in the PUFA-treated than BSA-treated cells (Figure 5C). Therefore, αS expression enhanced TfR expression, and combining αS expression and PUFA treatment resulted in a further increase in TfR expression.

αS and PUFA enhance the rate of TfR internalization and recycling

Using Alexa-488-Tf, we next measured αS and PUFAs effects on the rates of internalization and recycling of TfR. Naive and αS-overexpressing MN9D cells were conditioned in serum-free medium supplemented with BSA +/− 18:3 (at 50 μm BSA and 250 μm 18:3) for 16 h. We were concerned that comparing cellular membrane trafficking of cells treated in BSA only versus with PUFA was not ideal. We therefore performed the experiment comparing the effects of 18:1 and PUFA (18:3). Nevertheless, no difference was found between the effects of BSA and 18:1 on TfR recycling (data not shown). To measure internalization, cells were incubated with Alexa-488-Tf (25 μm) for the indicated times, surface-bound Alexa-488-Tf was then removed by acid wash and internalized Alexa-488-Tf was measured using flow cytometry [fluorescence-activated cell sorter (FACS) analyses] (Figure 6A). The results suggest that αS alone or PUFA alone increases the rate of internalization, with the highest rate of internalization observed in αS-overexpressing cells treated with PUFA. Internalization rate constants (calculated as internalized fluorescent per minute) are 2.05 (R= 0.92) and 3.05 (R2 = 0.91) for naive cells treated with BSA and PUFA, respectively, and 3.09 (R= 0.85) and 4.7 (R2 = 0.89) for αS-overexpressing cells treated with BSA and PUFA, respectively. Two-way anova indicated significant differences between the curves (Figure 6A), and the post hoc Student–Newman–Keuls test indicated a significant effect for αS expression, PUFA treatment and the combined αS and PUFA (p < 0.05). Interestingly, the internalization of Tf in naive cells with 18:3 and in αS cells with 18:1 was highly similar, suggesting that either αS expression or PUFA treatment on their own enhances Tf endocytosis to the same levels.

image

Figure 6. αS and PUFA activate the rate of membrane trafficking. Naive and wt αS-overexpressing MN9D cells were conditioned for 16–18 h in serum-free medium supplemented with BSA ± 18:3 PUFA (50 and 250 μm for BSA and PUFA, respectively). TfRs were loaded with Alexa-488-Tf. Recycling of Tf was measured by FACS analyses (see Materials and Methods). Graphs present the mean and SD of three experiments. A) Internalization, results are presented as percent of the sample with maximal internalized Tf in each test. B) Recycling, results are presented as the loss (percent) of cell associated fluorescence for each treatment.

Download figure to PowerPoint

To measure TfR recycling, Alexa-488-Tf (25 μm) was loaded for 60 min at 37°C to reach steady state. Excess unbound Alexa-488-Tf was then removed by washes, and cells were further incubated at 37°C for the indicated times (Figure 6B). Enhanced recycling is observed when combining αS expression and PUFA treatment and to a lesser extent with each treatment on its own. Recycling rate constants (calculated as internalized fluorescence per minute) are −0.032 (R= 0.86) and −0.035 (R2 = 0.85) for naive cells treated with BSA and PUFA, respectively, and −0.045 (R= 0.92) and −0.055 (R2 = 0.82) for αS-overexpressing cells treated with BSA and PUFA, respectively. Two-way anova indicated a significant effect for αS expression on the rate of recycling (p < 0.05). However, the effects of PUFA on recycling were not significant.

In accord with the increased TfR levels at the cell surface, Tf binding assays at 4°C (to allow binding but not internalization) showed increased Tf binding to the cell surface TfR (data not shown). Specifically, considering naive MN9D cells treated with BSA only as a control for basal Tf binding to the surface receptors, either 18:3 treatment or αS overexpression resulted in enhanced levels of bound Tf, with the maximal level of surface-bound Tf observed in αS-overexpressing cells treated with 18:3 PUFA (data not shown).

Taken together, the enhanced binding to TfR, enhanced levels of cell surface TfR and enhanced rate of internalization all act to increase Tf endocytosis in the presence of PUFA, αS or both.

αS- and PUFA-enhanced endocytosis is mediated by clathrin

In light of our recent findings suggesting αS involvement in affecting PUFA membrane composition and membrane fluidity (25), we sought to exclude the remote possibility that αS and PUFA induce Tf endocytosis through an alternative RME mechanism. To verify that αS- and PUFA-induced Tf endocytosis is mediated by CME, we treated naive and αS-overexpressing dopaminergic MN9D cells with BSA (50 μm) or BSA + 18:3 (250 μm) for 16 h and determined clathrin protein level by western blot probed with anti-clathrin or anti-actin antibodies. Densitometric analyses of the clathrin heavy chain normalized to actin on the same blot indicated increases in clathrin protein levels by αS expression, by PUFA and much further by combining them (∼40, 60 and 200% increases, respectively; Figure 7A). A similar result was observed by ICC with increases in clathrin signal and in accord, Tf signal, upon either αS over expression or PUFA treatment with highest signals observed by combining αS and PUFA (Figure 7B, left panel). Quantifying clathrin and Tf signals suggests a correlation between the increases in clathrin and Tf (Figure 7C). Considering naive cells treated with BSA as control, we found that Tf internalization and clathrin expression were significantly increased by either PUFA treatment or αS overexpression and further by αS and PUFA combined (anova< 0.001). These data support the conclusion that the increase in Tf internalization signal invariably observed with αS and PUFA (Figures 4 and 6A) results from activation of CME.

imageimageimage

Figure 7. αS- and PUFA-induced Tf endocytosis is clathrin dependent. Naive and wt αS-overexpressing MES cells were conditioned for 16–18 h in serum-free medium supplemented with BSA ± 18:3 PUFA (50 and 250 μm for BSA and PUFA, respectively). A) Whole cell extract was analyzed by western blot and probed with clathrin (heavy chain) antibody. Results are normalized to actin and presented in arbitrary units. B) Naive and wt αS-overexpressing MN9D cells were grown on 12-well cover glass and transfected with either mock (empty Plko.1 plasmid) or shRNA for clathrin heavy chain. Thirty-two hours post DNA transfection, the cells were transferred to serum-free conditioning medium, supplemented with BSA + 18:3 PUFA (50 and 250 μm for BSA and FA, respectively) for 16 h. Forty-eight hours post DNA transfection, 50 μg/mL Alexa-488 human Tf were added to the cells (as in Figure 4A). Cells were fixed and processed for detection of Alexa-488-Tf and ICC with anti-clathrin antibody (see Materials and Methods). C) Quantification of Alexa-488-Tf endocytosis and clathrin protein levels. Cells and treatments as in (B), images were taken at the same settings and the image plane containing maximal intracellular signal was selected. Threshold was set as to measure only endocytic vesicles rather than including a more diffuse background. All measurements were of selected region of interest normalized to cell area. Mean of 10 cells of each treatment ± SD. anova p < 0.001. D) Clathrin (heavy chain) protein levels in αS-overexpressing MES cells transfected with plko.1 or clathrin shRNA and treated with PUFA as in (B). Densitometric analyses of western blot probed with anti-clathrin antibody and normalized to tubulin signal on the same blot. E) Quantification of Alexa-488-Tf endocytosis and clathrin protein levels in MES cells transfected with clathrin siRNA (B). Images as in (C). Bar represents 10 μm.

As an alternative approach to confirm the involvement of clathrin in αS- and PUFA-induced endocytosis, we designed a small hairpin RNA (shRNA) to inhibit the expression of the clathrin heavy chain and thereby the assembly of clathrin (54). We selected an shRNA duplex that targeted the segment 4643–4663 of the mouse clathrin heavy chain open reading frame and inserted it into the Plko.1 plasmid (see Materials and Methods). Clathrin shRNA and unmodified Plko.1 plasmids were transfected into naive and αS-overexpressing dopaminergic lines MES and MN9D with indistinguishable results (data not shown). Thirty-two hours after transfection of the small interfering RNA (siRNA), the cells were transferred to conditioning medium (serum free) supplemented with BSA ± 18:3 (50 and 250 μm, respectively) and maintained under these conditions for 16 h. Then (48 h after transfection), the cells were assayed for clathrin levels and Alexa-488-Tf endocytosis by ICC or western blotting.

To quantify the reduction in clathrin heavy chain expression, the transfected cells were extracted and blotted for clathrin heavy chain (and tubulin as a loading control). A sharply reduced level of clathrin heavy chain was detected in the cells transfected with clathrin siRNA versus cells transfected with empty plasmid or left untransfected, whereas tubulin was unchanged. Quantification of αS-overexpressing MES cells transfected with the clathrin shRNA or empty Plko.1 and treated with 18:3 PUFA showed that the cellular concentration of clathrin heavy chain had dropped to ∼27% of control levels (Figure 7D). In accord, reduced clathrin expression throughout the cell was found by ICC in the siRNA versus mock-transfected cells (Figure 7B, right panel, and Figure 7E) accompanied with reduced Tf endocytosis in the clathrin siRNA-transfected cells but not in the mock-transfected cells (Figure 7B,E). The reduced clathrin signal and Tf endocytosis following transfection with clathrin siRNA was observed in all four combinations of αS expression and PUFA, that is, naive cells treated with BSA or PUFA and in αS-overexpressing cells treated with BSA or PUFA (with ∼60–90% reduced signal throughout the treatments).

We next confirmed these results with a second shRNA that targeted the segment of 3387–3408 of the clathrin heavy chain versus its scrambled sequence. The results indicated that the clathrin siRNA but not its scrambled sequence reduced clathrin expression throughout the cell and reduced Tf endocytosis by ∼65% in αS-overexpressing cells treated with PUFAs (data not shown).

αS expression and PUFA treatment enhance endocytic vesicle formation in primary neurons

αS is a presynaptic protein believed to be involved in SV generation and/or maintenance (46,47). To determine whether the striking effects of αS plus PUFA exposure pertain to synaptic activity, we again used the lipophilic dye, FM1-43. This dye has been used extensively in diverse experimental systems to study mechanisms involved in SV recycling (reviewed in 52,55,56). The concept behind using FM1-43 to assess synaptic and endocytic vesicles is similar. In both cases, FM1-43 is a marker for plasma membrane internalization. Furthermore, it is well established that SV formation is mediated by CME (reviewed in 34,57).

Primary hippocampal neurons (18 days in vitro) from normal and αS null (−/−) mouse brains were grown in serum-free medium supplemented with BSA +/− 18:3 (10 and 50 μm, respectively) for 16 h prior to membrane depolarization with KCl (70 mm for 60 seconds), and SV formation was analyzed in the presence of FM1-43 (10 μm). Neurons were rinsed of excess FM1-43, fixed and processed for ICC with anti-synaptophysin antibody as a marker for synapses (Figure 8). The chemical stimulation resulted in uploading the fluorescent dye into the hippocampal neurons. We also performed this experiment with an even lower 18:3 concentration (25 μm) with very similar results (data not shown).

image

Figure 8. Reduced FM1-43 internalization into SVs and reduced synaptic pool in αS−/− primary hippocampal neurons. Hippocampal primary cultures (at 18 days) of normal and αS−/− mouse brains were conditioned for 16 h in serum-free medium supplemented with BSA ± 50 μm 18:3. A) αS expression and PUFA treatment activate SV endocytosis. Fluorescence image after exposure to 10 μm FM1-43 in 90 mm K+ solution for 60 seconds, followed by washout. Bar represents 50 μm. B) ICC with anti-synaptophysin Ab (DAKO) performed on slides from (A). Arrowheads represent the area of the inset. C) Fluorescence image of cultures treated as in (A) after a second stimulation with 70 mm K+ (FM1-43 download). D) αS expression and PUFA treatment induced constitutive internalization of FM1-43 into neuronal cell bodies. Bar represents 25 μm. Fluorescent images as in (A and B) but without chemical stimulation.

Download figure to PowerPoint

Overall, we detected higher levels of FM1-43 dye in synaptic terminals in the neuropils of wt than αS−/− mouse hippocampal cultures exposed to BSA alone and then saw a substantially more pronounced induction upon exposure of the cultures to PUFA (Table 2 and Figure 8A). Importantly, each factor alone, that is, αS expression or 18:3 treatment, induced FM1-43 uptake into SV, indicating that they can act independently of each other to enhance new SV formation (Figure 8A). However, the number and size of synaptic puncta were higher in the hippocampal neurons from wt mouse brain treated with 18:3, again indicating at least an additive effect between αS expression and PUFA levels. Furthermore, these results suggest that either αS−/− synapses maintain a smaller pool of SV or that the recycling of the vesicles in αS−/− synapses is delayed.

Table 2.  Quantification of synaptophysin-positive synapses and FM1-43 uptake in primary hippocampal neurons from wt and αS−/− mouse brains exposed to PUFA or not
 wt-BSAwt-PUFAαS −/−BSAαS −/−PUFA
  • a

    Quantification was done on microscopic fields containing highly similar neuritic densities and measured as signal above threshold and presented as % area.

  • b

    Means ± SD of seven to nine fields for each treatment.

FM1-43a,b9.13 ± 1.78b19.75 ± 7.12.25 ± 1.26.31 ± 2.45
Synaptophysina,b15.71 ± 2.7818.53 ± 7.646.33 ± 2.36.83 ± 3.41

In accord with the reduced FM1-43 signal, we detected reduced signal for synaptophysin in the αS−/− synaptic terminals (Table 2 and Figure 8B). Specifically, we detected reduced synapse size and reduced intensity of synaptophysin signal in the αS−/− neurons. Quantifying the total synaptophysin signal per cell revealed that the signal in αS−/− neurons was ∼58% lower than the signal detected in the wt neurons. However, no significant effect for PUFA on synaptophysin signal was detected by ICC (Table 2) or quantitative western blotting (data not shown). Furthermore, to determine FM1-43 uptake specifically into boutons, we calculated the percent of synaptophysin-positive boutons that were also positive for FM1-43. We found that ∼40 and ∼75% of synaptophysin-positive boutons were also positive for FM1-43 in wt cultures treated with BSA and BSA + PUFA, respectively, indicating that PUFA-induced FM1-43 internalization occurs into synaptic pools at the synapses. Nevertheless, a substantial amount of the PUFA-induced FM1-43 internalization signal appeared along the neurite and not within synaptic terminals or boutons (i.e., was not colocalized with synaptophysin). This result indicates that PUFA-induced internalization is not restricted to SV at synaptic pools. In αS−/− neurites, we observed that PUFA alone induced FM1-43 internalization into synaptic terminals as well as into SV along the neurite. However, because of the lower number of synaptophysin-positive boutons in the αS−/− neurites, we found that ∼60 and ∼95% of the synaptophysin-positive boutons were positive for FM1-43 in the BSA- and BSA + PUFA-treated cultures, respectively.

Next, the primary cultures were depolarized in KCl one more time to download the FM1-43 dye from the SV (see Materials and Methods). This second cycle of depolarization with KCl now evoked exocytosis of the FM1-43-loaded SV and therefore resulted in a markedly reduced fluorescent signal, confirming FM1-43 downloading by evoked SV recycling (Figure 8C). We conclude from these results that the FM1-43 dye is internalized specifically into SV.

A similar increase in FM1-43 uptake into vesicles was also observed in the neuronal cell bodies. Consistent with the increased FM1-43 internalization into SV, this increase is dependent on αS expression and the presence of PUFA (Figure 8D). However, this internalization of FM1-43 dye into neuronal cell bodies occurred without chemical depolarization of neurons. Furthermore, this FM1-43 signal was not washed out with a subsequent depolarization and did not colocalize with synaptophysin. This result suggests the possible involvement of αS expression and PUFA in constitutive mechanisms of neuronal endocytosis.

Very importantly, no induction of FM1-43 uptake was observed at either the somata or the neurites when hippocampal neurons from wt and αS−/− mice were conditioned in the presence of 18:1 (50 μm), performed in parallel to the 18:3 experiments (data not shown). This control indicates that PUFAs but not MUFAs of identical carbon chain length are capable of activating these endocytic mechanisms.

Discussion

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgment
  7. References

Taken together, these experiments provide new evidence that αS and PUFAs can act in at least an additive fashion to modulate membrane trafficking, with a specific focus on their individual and combined effects on endocytosis mediated by clathrin. Overall, we found that αS expression or PUFA treatment induced endocytosis, with further induction upon their combination. In general, at the same carbon chain length and concentration, PUFAs induced and SFAs inhibited endocytosis. MUFAs effect on induced endocytosis was milder and was observed only with the longer carbon-chained FAs. Using the canonical example of ligand internalization through RME, we show that αS and PUFA induce the endocytosis and recycling of Tf by the TfR. αS- and PUFA-induced Tf endocytosis is molecularly specific as it is inhibited by clathrin siRNA. Relevant to the physiological function of αS in the brain, our findings indicate that αS and PUFAs can act together to stimulate SV formation in primary hippocampal neurons. Specifically, αS is implicated in the endocytic recycling of SV following neuronal stimulation. Importantly, primary hippocampal neurons genetically deleted of αS showed reduced number and size of synaptophysin-positive boutons and reduced FM1-43 endocytosis in SV, suggesting a reduced recycling pool. Based on these invariant results in primary neuronal cultures, we suggest that αS and PUFA are normally involved in the endocytic machinery leading to SV formation and replenishment and thus synaptic strengthening.

Early reports that αS is localized to nerve terminals (17,58,59) and associates with SV (60) suggested the possible involvement of αS in the formation/maintenance of SV and/or neurotransmitter release. Several subsequent studies have shown that αS expression has a role in the formation or maintenance of distinct synaptic pools. In particular, reduced sizes of distinct synaptic pools in primary hippocampal neurons were observed upon partially silencing αS expression (47) and also in hippocampal neurons from αS−/− brains (46). In accord, increased accumulation of docked vesicles was observed in PC12 cells upon αS overexpression (61). Furthermore, αS has been postulated to be involved in the replenishment of synaptic pools after a depleting stimulation (46). Nevertheless, these various observations were not consistent with certain findings in other αS−/− mouse models. No alterations in synaptic pool size or replenishment of recycling SV were observed in αS−/− mice or in the double knockout αS−/− and αS−/− mice (62), and αS was suggested to negatively regulate the readily releasable pool of dopamine-containing vesicles (63). These variations in observations may result from the different mouse models and experimental designs employed and highlight the complexity of synuclein’s physiological role. The new results presented herein strongly support a role for αS in the formation/maintenance of SV and/or neurotransmitter release. Furthermore, the results indicate that αS is involved in mechanisms leading to synaptic strengthening. That is, the reduced number and size of synaptophysin-positive boutons observed in αS−/− neurons indicate lower synaptic activity. This, in turn, could affect the plasticity of neuronal networks that is known to underlie cognitive functions such as learning and memory.

Recent studies have suggested that αS is involved in other aspects of membrane trafficking, possibly in exocytosis or in the secretory pathway. It has been reported that αS expression specifically inhibits ER-to-Golgi trafficking, resulting in cytotoxicity that was prevented by Rab1 expression (64). It was further shown that αS expression affected vesicle docking or fusion to the Golgi apparatus after an efficient budding from the ER (65). In PC12 cells, αS overexpression inhibited evoked catecholamine release and increased the ‘docked’ vesicle pool (61). Other studies have suggested an indirect role for αS in promoting the assembly of the SNARE complex (62,66). SNAREs catalyze the fusion of vesicles with their target membranes to enable the release of cargo from the vesicle (reviewed in 67,68). Collectively, growing evidence implicate αS in membrane trafficking, including endocytosis and exocytosis.

In addition to our observed induction of FM1-43 uptake into SV, αS plus PUFA additively induced dye uptake into neuronal cell bodies. Despite extensive investigation, it remains unclear whether the endocytic pathways at nerve terminals are specializations of the pathways that exist in all cells or rather represent a neuron-specific mechanism for rapid membrane recycling (69). In this regard, whether neurons have classical endosomes is also unresolved. Therefore, the nature of the induced vesicles we observed in the neuronal cell bodies is yet to be determined. Nevertheless, because we found that αS can activate endocytosis at both neuronal and non-neuronal vesicles, these pathways most likely share a high degree of similarity.

We previously reported that αS expression in cells and brain is associated with enriched levels of certain PUFAs in both the cytosol and the membranes, and this was associated with apparent changes in membrane biophysical properties reflected by increased membrane fluidity (25). Our initial working hypothesis after we obtained these results was that the αS effects on membrane fluidity may relate directly to its effects on membrane trafficking. Our rationale was that αS may act to enrich the cytoplasmic membrane leaflet with PUFAs. PUFAs in the membranes act as cone-shaped lipids that induce a positive curvature of the membrane leaflet (44,70), thus inducing invagination and fission of the membrane and thereby enhancing membrane trafficking. The results presented herein support this initial working hypothesis. It is supported by the specific inducing effect of PUFAs (but never MUFAs or SFAs of identical carbon chain length and concentration) – either with or without αS overexpression – on endocytosis and SV formation. Therefore, we speculate that αS and PUFA mechanically alter membrane curvature as a result of enrichment of the plasma membrane with PUFAs, thereby facilitating endocytosis, including, among other mechanisms, CME.

In addition to the general role of membrane curvature in endocytosis discussed above, the findings herein that αS+ PUFAs effects can alter actual protein levels of specific key factors in CME and RME, that is, TfR and clathrin, and also redirect TfR to the cell surface membrane indicate activation of regulatory mechanisms that may secondarily lead to the overexpression or stabilization of these specific endocytic participants. A hypothetical biological target for αS/PUFA activation is certain classes of membrane phospholipids, and specifically phosphoinositides, that are known to regulate CME (reviewed in 71). Further investigation will be needed to elucidate the specific molecular mechanisms by which αS and PUFAs activate endocytosis.

We view the data we have obtained to date as addressing a normal physiological role of αS. However, it is likely that there are pathophysiological implications of our findings. For example, we observed that the accumulation of αS into high MW assemblies, including soluble cytosolic dimers and low-n oligomers that could well serve as the nidus for αS aggregation into Lewy-type fibrillar deposits, is associated with alterations in neuronal PUFA composition. That is, we examined FA profiles in mesencephalic neuronal cells that stably express αS and found accumulation of PUFAs in cytosols as well as membrane fractions in such cells compared with parental cells that express low levels of endogenous αS. These changes could be relevant to the overexpression of wt αS in PD families with duplication or triplication of the αS locus and also to idiopathic PD cases that invariably accumulate wt αS in the neuronal cytoplasm. In this regard, it is interesting to consider the correlation between PUFA-induced αS oligomerization (48) and PUFA-induced FM1-43 uptake (herein). In both cases, the longer and more unsaturated the FA, the more αS oligomerization and the more FM1-43 internalization one observes. This correlation may suggest that the soluble, cytosolic oligomers are the active forms.

Collectively, our earlier and current results suggest that αS normally interacts with PUFAs to carry out its physiological functions, but under certain potentially pathogenic conditions, this interaction may lead to neuronal membrane dysfunction and ultimately αS aggregates and cell death. We propose that this shift on a continuum between normal and pathogenic αS–lipid interactions may be driven by transient increases in either the levels of certain PUFAs or the levels of αS monomers in the cytoplasm.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgment
  7. References

Cells

The mesencephalic neuronal cell lines MES 23.5 and MN9D, which have dopaminergic properties (72,73), were stably transfected with wt human αS complementary DNA (cDNA) in the pCDNA 3.1 vector using Lipofectamine 2000 (Invitrogen). A reported feature of the αS-transfected MES cells was that all stable clones gradually lost αS expression after being continuously passaged for 2–3 months or more (18). To overcome this technical problem and increase consistency of results, we kept frozen aliquots of MES and MN9D αS-overexpressing clones. The clones were frozen at 55–65 days post DNA transfection, and fresh aliquots were thawed routinely every 4–8 weeks. Experiments performed in SK-N-SH neuronal lines involved viral infection with αS-expressing adenovirus (see below).

Detection of αS oligomers and native gel electrophoresis

Normal, αS+/+(49) and αS−/−(50) mouse brains were homogenized and fractionated as described before (18). Samples (15 μg) of high speed supernatant (post 100 000 × g) were incubated with Lipidex 1000 (w/v) for 16 h. Samples were spun down to remove Lipidex-1000, and the supernatant was loaded on a 10% Tris-Glycine polyacrylamide gel (without SDS in gel, running or loading buffers). The proteins were transferred to polyvinylidene flouride (PVDF) membrane and probed with syn-1 antibody (Transduction Laboratories). For the in vivo association of αS with 14C OA, cells were incubated with the indicated mixture of FA and BSA, fractionated as above and samples of high speed supernatant was subjected to native gel electrophoresis as above. The gel was dried and autoradiogram obtained.

FM1-43 internalization into endocytic vesicles

To visualize endocytic membrane trafficking, cells (growing on cover glass of 12 wells) were incubated at the presence of 2.5 μm FM1-43 (Fixable; Molecular Probes) for 4 min at room temperature. Unbound probe was removed, and the cells were washed twice and reincubated for 10 min at 37°C with the indicated conditioning medium. The cells were then washed and fixed with 4% PFA. Cover glass was then either prepared for microscopic analysis to observe FM1-43 fluorescence or else subjected to ICC.

Tf binding and endocytosis

Alexa-488-human Tf was purchased from Molecular Probes. Cells (SK-N-SH, MES or MN9D) were plated on cover glass coated with 0.01 mg/mL poly-D-lysine the day before experiment. Alexa-488 human Tf (or the equivalent non-fluorescent Tf for ICC with TfR antibody) was added to the cells at 50 μg/mL in plain DMEM medium without any supplements in the presence of 0.1 mg/L of Fe(NO3)3*9H2O at 4°C for 60 min. Unbound Tf was removed and cells were transferred to 37°C for additional 10 min at the indicated conditioning medium to allow endocytosis. Cells were then fixed with 4% PFA for 10 min on ice and either processed for quantitative measurements of internalized Tf or processed for ICC with the indicated antibody (Ab) using secondary cy5-conjugated (emission at 633).

Immunocytochemistry

Following fixation with 4% PFA for 10 min on ice, cells were washed and permeabilized with 0.2% Triton-X-100 in PBS and 1% goat serum for 5 min at room temperature. Slides were then blocked with 1.5% goat serum in PBS. Primary abs: monoclonal mouse anti-clathrin heavy chain (1:250; BD Biosciences), monoclonal mouse anti-TfR (1:100; Zymed Laboratories), monoclonal mouse anti-synaptophysin Ab (1:200; DAKO) and secondary Ab at 1:200, anti-mouse-cy5 (Molecular probes), anti-mouse-Cy5 (Jackson). Slides were sealed with mounting medium (cat# M1289; Sigma) and analyzed by confocal microscopy.

Clathrin shRNA construction

The shRNA for clathrin heavy chain (mouse) was designed according to the Broad Institute, TRC, The RNA consortium website. The shRNA targeting sequence is GCGAACATCAATAGATGCTTA, total of 21 nucleotides at position 4643–4663. The sequence of the forward primer is CCGGGCGAACATCAATAGATGCTTACTCGAGTAAGCATCTATTGATGTTCGCTTTTTG and the reverse primer is AATTCAAAAAGCGAACATCAATAGATGCTTACTCGAGTAAGCATCTATTGATGTTCGC. The primers were annealed and cloned into the plko.1 vector (Sigma). The vector was transfected into cells using Lipofectamine 2000 (Invitrogen). Thirty-two hours after DNA transfection, cells were treated with the indicated medium for additional 16–18 h, followed by Tf endocytosis protocol and ICC with clathrin Ab (BD Biosciences) or collected and processed for western blotting with clathrin Ab (BD Biosciences). A second shRNA for clathrin heavy chain with the specific sequence TGAGCTGTTTGAAGAAGCA, total of 21 nucleotides at position 3387—3408, was kindly provided by Drs Bacharach E. and Ehrlich M. (Tel Aviv University, Israel).

Quantitative real-time polymerase chain reaction

Total RNA was extracted from naive and αS-overexpressing MN9D cells treated with BSA ± 18:3 at 250 μm) using Tri Reagent (Sigma). RNA samples were treated with DNAse I (RNAse free) and reverse transcribed with random hexamers using Reverse-iT (ABgene) to generate cDNA. Primers were generated according to (74) with 5′ primer sequence: 5-TCATGAGGGAAATCAATGATCGTA-3 and 3′ primer sequence 5-GCCCCAGAAGATATGTCGGAA-3. Results were normalized against the expression levels of 18S. Polymerase chain reaction analyses were performed using Applied Biosystems Prism 7000 sequence detection system with CYBR Green Master Mix (Applied Biosystems).

FACS measurements of endocytosis and recycling

MN9D cells were plated 2 days before experiment and treated with the specified medium 16–18 h before the labeling with AlexaFluor-488-Tf (Invitrogen). For internalization, cells were conditioned in 25 μm 488-Tf at 37°C for the times indicated. Internalization was stopped by chilling the cells on ice. Access of 488-Tf was removed by three washes in ice-cold serum-free medium and PBS. Surface-bound Tf was removed by an acid wash with DMEM at pH 4.5 (pH adjusted with acetic acid), followed by neutralization with DMEM at pH 7.5 and a wash in PBS. Results are calculated relative to a sample treated in parallel (including acid wash) without incubation at 37°C (designated 0%) and relative to the sample with the highest internalized Tf in each test (100%). To measure recycling, cells were plated and treated with the specified medium as for internalization. Cells were then incubated for 1 h at 37°C in the presence of 25 μg/mL 488-Tf to completely label the receptor population. Unbound 488-Tf was removed by washes with prewarmed serum-free medium (the amount of Tf at this point was referred to as 100%) and the cells were now conditioned for the indicated time at 37°C in serum-free medium supplemented with BSA ± 18:3 (50 and 250 μm, respectively), 200 μg/mL deferoxamine and 600 μg/mL unlabeled human Tf. The release of 488-Tf was stopped by chilling the cells on ice. The fluorescence intensity of 488-Tf was measured for 10 000 cells by flow cytometry using an FACSCalibur (BD Biosciences), and the average intensity of the cell population was recorded for each time-point.

Biotinylation of cell surface TfR

Surface levels of TfR were determined using the NHS-LC-biotin (Pierce Chemical Co.) as previously described (75). Briefly, cells were plated in 60-mm dishes 2 days before the biotinylation and treated with the specified medium for 16–18 h before cooling by three quick washes with ice-cold PBS containing Ca2+ and Mg2+ (PBS++). Selective biotinylation of the cell surface proteins was performed at 4°C with two rounds of 20 min incubations of 0.5 mg/mL NHS-LC-biotin in 10 mm boric acid and 150 mm NaCl, pH 8.0. Cells were washed and free sulfhydryl groups were quenched for 15 min at 4°C with 50 mm glycine in PBS++. Whole cell lysate was prepared in 0.5 mL Triton lysis buffer [100 mm NaCl, 5 mm ethylenediaminetetraacetic acid, pH 8.0, 100 mm triethanolamine Cl, pH 8.6, 2.5% (v/v) Triton-X-100, 0.02% (v/v) sodium azide and protein inhibitor cocktail 1:100 (Sigma)]. Cells were scraped and vortex at 4°C for 20 min. Cell lysate was centrifuged for 30 min at 12 000 × g at 4°C. Supernatant was transferred to a fresh tube. Protein concentration was determined using BCA kit (Pierce). Aliquots of equal protein concentration (30 μg) were analyzed by western blotting for total TfR, and aliquots of 45 μg protein were used to immunoprecipitate TfR on streptavidine agarose beads (sigma S1638).

Primary neuronal culture

Hippocampal cell cultures were prepared as described previously (76). Briefly, hippocampal CA1-CA3 regions were dissected from 1- to 2-day-old C57BL/6 mice, dissociated by trypsin treatment, followed by trituration with a siliconized Pasteur pipette and then plated onto coverslips coated with poly-d-lysine (Sigma) inside 24-well dish. Culture medium consisted of MEM (Invitrogen), 0.6% glucose, 0.1 gm/L bovine Tf (Calbiochem), 0.25 gm/L insulin (Sigma), 0.3 gm/L glutamine, 5–10% fetal calf serum (Sigma) and 2% B-27 supplement (Invitrogen). To eliminate the glia cells, 8 μm cytosine b-d-arabinofuranoside (Sigma) was added to the culture 3 days after preparation and removed after additional 3–4 days. Cultures were maintained at 37°C in a 95% air/5% CO2 humidified incubator, and culture medium was replaced every 4–7 days. Experiments were performed on cultures grown for 18 days. Normal mice are C57BL/6 obtained from Jackson laboratories, and αS−/− mice are C57BL/6 obtained from Harlan (73).

FM1-43 internalization into SVs

Neurons were loaded with 10 μm FM1-43 by high K+ application or without chemical stimulation. Specifically, neurons were chemically induced in solution containing 90 mm K+ in buffer containing 31.5 mm NaCl, 2 mm KCl, 2 mm CaCl2, 2 mm MgCl2, 30 mm HEPES and 10 mm glucose buffered to pH 7.4 for 60 seconds or maintained in PBS (calcium and magnesium free) in the presence of FM1-43. Neurons were than washed (two to three times) in low KCl solution containing 119 mm NaCl, 2.5 mm KCl, 2 mm CaCl2, 2 mm MgCl2, 25 mm HEPES and 30 mm glucose buffered to pH 7.4 for 5 min. A second stimulation with 90 mm KCl solution without FM1-43 for 60 seconds was applied for destaining, followed by washes and fixing the neurons in 4% PFA. Neurons were than processed for ICC with anti-synaptophysin Ab (DAKO cytomation) as described above.

Imaging and quantitative analysis of endocytosis

Most experiments were documented with a Zeiss LSM 410 confocal laser scanning system. The system is equipped with an argon laser 488 excitation with 510 nm pass barrier filter and two helium–neon lasers, excitation at 543 nm (570 nm emission) and excitation at 643 nm (665 nm emission). Fluorescence was collected simultaneously with differential interference contrast (DIC) images according to Nomarski using a transmitted light detector. The fluorescence was collected by employing a 63× 1.4 plan Apochromate oil immersion lens (Zeiss). In each experiment, exciting laser, intensity, background levels, photo multiplier tube (PMT) gain, contrast and electronic zoom size were maintained at the same level. Alternatively, we used (as indicated) an Olympus confocal laser scanning system FluoView FV300 equipped with an UPlanSApo 60×/1.35 (oil) lens. Lasers used are Melles Griot 488 Ion Argon laser and 633 He/Ne laser.

Image series were analyzed with Pro/Image J softwares (Media Cybernetics Inc.). The FM1-43 heavily labeled endosomes increased the probability that multiple endosomes would be clustered into an individual fluorescent point. Therefore, for the majority of experiments (FM1-43/Alexa-488-Tf/Cy5-clathrin and synaptophysin; Figures 2–4, 7 and 8), endocytosis was measured as the sum of fluorescent signal above threshold divided by the area of the cell. Fluorescence intensity in a single cell was obtained by choosing the image plane with the greatest number of stained endosomes, measuring the average fluorescence intensity across the entire cell area and subtracting the non-specific background fluorescence. Background was subtracted using imageJ rolling ball algorithm (Rolling Ball Radius = 50). The presented values are average integrated density of 5–20 cells. For Tf dose–response experiment (Figure 4) – a computer-controlled microscope stage was used for acquiring vertical series of images through the entire depth of a cell at 1 μm increments. The projections were analyzed using ‘particle analysis’ (to define the vesicle we used limitation in size of 10 pixels and above and circularity of 0–1 of imageJ definitions). The values represented are average properties of all labeled vesicles of six different cells for each treatment. Image acquisition parameters were kept fixed for each experiment. All intensity values are presented in arbitrary units.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgment
  7. References

This study was supported by the National Institute of Neurological Disorders and Stroke (NINDS) R01 NS051318 (D. J. S. and R. S.).

References

  1. Top of page
  2. Abstract
  3. Results
  4. Discussion
  5. Materials and Methods
  6. Acknowledgment
  7. References