SEARCH

SEARCH BY CITATION

Keywords:

  • apical membrane;
  • cell polarity;
  • development;
  • glia;
  • myelin;
  • nerve injury

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

The proteolipid plasmolipin is member of the expanding group of tetraspan (4TM) myelin proteins. Initially, plasmolipin was isolated from kidney plasma membranes, but subsequent northern blot analysis revealed highest expression in the nervous system. To gain more insight into the functional roles of plasmolipin, we have generated a plasmolipin-specific polyclonal antibody. Immunohistochemical staining confirms our previous observation of glial plasmolipin expression and proves plasmolipin localization in the compact myelin of rat peripheral nerve and myelinated tracts of the CNS. Western blot analysis indicates a strong temporal correlation of plasmolipin expression and (re-) myelination in the PNS and CNS. However, following axotomy plasmolipin expression is also recovered in non-regenerating distal nerve stumps. In addition, we detected plasmolipin expression in distinct neuronal subpopulations of the CNS. The observed asymmetric distribution of plasmolipin in compact myelin, as well as in epithelial cells of kidney and stomach, indicates a polarized cellular localization. Therefore, we purified myelin from the CNS and PNS and demonstrated an enrichement of phosphorylated plasmolipin protein in detergent-insoluble lipid raft fractions, suggesting selective targeting of plasmolipin to the myelin membranes. The present data indicate that the proteolipid plasmolipin is a structural component of apical membranes of polarized cells and provides the basis for further functional analysis.

Abbreviations used
4-TM

four transmembrane

β-CD, methyl

β-cyclodextrin

ad

adult

BSA

bovine serum albumin

Cy-3

carboxyfluorescein

DIGs

detergent-insoluble glyco-sphingolipid/cholesterol enriched rafts

FITC

fluorescein-5-isothiocyanate

NGS

normal goat serum

P

postnatal day

PAGE

polyacrylamide gel electrophoresis

SDS

sodium dodecyl sulfate

TBS

tris-buffed saline

TGN

trans golgi network

TNE

tris-NaCl-EDTA-buffer

TX-100

triton X-100

The formation of highly specialized myelin by Schwann cells or oligodendrocytes requires the co-ordinate synthesis and integration of large quantities of specific proteins and lipids into the organized multi-lamellar structure (Morell and Ousley 1994). Many of the proteins embedded in compact myelin are structurally related and consist of four transmembrane domains (4-TM). Interestingly, mutations in their genes often result in severe hereditary demyelinating diseases of human and rodents (Müller 2000; Chance 2001; Woodward and Malcolm 2001; Gabreels-Festen 2002). Differential screening of rat sciatic nerve cDNA libraries performed in our laboratory resulted in the identification of the cDNA encoding plasmolipin, a 20-kDa protein with four putative transmembrane domains (Gillen et al. 1996). Plasmolipin was initially isolated from kidney plasma membranes (Tosteson and Sapirstein 1981) and classified as a proteolipid based on its solubility in organic solvents (Lees et al. 1979). Northern and in situ hybridization analysis revealed post-natal expression of plasmolipin by glial cells (Gillen et al. 1996). Plasmolipin mRNA is confined to myelinating Schwann cells in peripheral nerve and to myelinating oligodendrocytes in the CNS (Gillen et al. 1996). Following peripheral nerve lesions of the adult rat plasmolipin, mRNA is re-expressed during nerve regeneration (Gillen et al. 1996). Outside the nervous system, plasmolipin mRNA is most abundantly expressed in rat kidney (Gillen et al. 1996). Unfortunately, previous western blot analyses revealed highly variable molecular weight estimations ranging from 11.5 to 18.5 kDa (Shea et al. 1986; Sapirstein et al. 1988; Cochary et al. 1990; Fischer et al. 1991; Sapirstein et al. 1992a). Therefore, it is necessary to prove previous data concerning cellular expression and distribution of plasmolipin protein. Very recently, we observed a widespread plasmolipin expression in tissues with a high degree of complexity, most containing polarized cells or cell layers, and described a putative association of plasmolipin as a candidate gene or modifier of the human Bardet–Biedl syndrome (Hamacher et al. 2001).

Interestingly, plasmolipin shares some sequence similarities with MAL/VIP17 (Magyar et al. 1997; Perez et al. 1997), another 4-TM proteolipid protein of 17 kDa that is characterized as a myelin component of the PNS and CNS (Kim et al. 1995; Schaeren-Wiemers et al. 1995). MAL/VIP17 is associated with detergent-insoluble, glycosphingolipid/cholesterol-enriched rafts (DIGs) (Kim et al. 1995; Frank et al. 1998; Erne et al. 2002) and is involved in sorting and delivery of proteins from the trans golgi network (TGN) to the apical surface of epithelial cells in vitro (Cheong et al. 1999; Puertollano et al. 1999).

As our previous observations concerning plasmolipin mRNA expression suggest a putative function of plasmolipin related to myelination in the mammalian CNS and PNS, we have generated a plasmolipin-specific polyclonal antibody as a prerequisite for comprehensive western blot analyses and supplementary immunohistochemical localization studies. Taking into account that plasmolipin is expressed in various tissues with a high degree of complexity (Hamacher et al. 2001), we proved whether asymmetric protein distribution is a characteristic feature of plasmolipin expression and we further investigated the association of plasmolipin to lipid rafts.

Animals and surgery

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Adult male Wistar rats (180–220 g) were anaesthetized and sciatic nerves were either crushed with jeweller's forceps or transected with a fine pair of scissors at upper thigh level. In order to prevent spontaneous reanastomosis, both stumps of transected nerve were tied off with surgical silk. Prior to protein preparation from the distal nerve stumps, a segment of 2–3 mm adjacent to the site of injury was removed and discarded. Two to three independent pools, each with at least six injured nerves were collected. All animal experiments were performed according to the guidelines of German and European animal rights law, respectively.

Antibodies

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Polyclonal anti-plasmolipin antibody (Antibody Service Dr Pineda, Berlin, Germany), polyclonal anti-P0 antibody (Antibody Service Dr Pineda), monoclonal anti-MBP antibody (Chemicon, Hofheim, Germany), monoclonal anti-GFAP antibody (Roche, Mannheim, Germany), monoclonal anti-MAP-2 antibody (Sigma, Deisenhofen, Germany) and monoclonal anti-NeuN antibody (Chemicon, Hofheim, Germany). Secondary Antibodies: horseradish peroxidase coupled goat anti-mouse (DAKO, Hamburg, Germany), horseradish peroxidase coupled goat anti-rabbit (Southern Biotechnology, Birmingham, AL, USA), biotinylated goat anti-rabbit antibody (Vector, Burlingame, CA, USA), biotinylated goat anti-rabbit antibody (Jackson Immuno-Research, West Grove, PA, USA), biotinylated goat anti-mouse antibody (Vector), biotinylated horse anti-mouse (Vector).

Immunohistochemistry

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

De-paraffinized sections were treated with 3% H2O2 in methanol for 5 min and then with 2% normal goat serum (NGS) in Tris-buffered saline (TBS) for 30 min. Incubation with primary antibodies was performed for 1 h at 20°C. Sections were then washed several times and incubated with either (i) biotinylated secondary goat anti-rabbit antibody (1 : 200; Vector) for 30 min, then with avidin-biotinylated peroxidase complex (1 : 50; Vector) for 30 min and finally with the substrate diaminobenzidine/H2O2 for 5 min, or (ii) Alexa Green- or CY3-conjugated secondary antibodies (Jackson Immuno-Research) were used for signal visualisation. Finally, samples were extensively rinsed in phosphate-buffered saline (PBS).

Myelin preparation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Myelin was prepared from sciatic nerves or total brain of Wistar rats as described previously (Hasse et al. 2002). In brief, tissues were dissected and homogenized in ice-cold 0.9 m sucrose using an Ultra-Turrax T25. Homogenate was overlaid with 0.25 m sucrose and centrifuged at 100 000 g for 3 h at 4°C. Myelin was recovered from the 0.9- and 0.25-m sucrose interface and further purified by two rounds of hypo-osmotic shock through re-suspension in a large volume of ice-cold water, followed by a second round of centrifugation in a sucrose step gradient. Purified myelin was collected from the interface, washed twice with ice-cold water, re-suspended in a small volume of water and frozen in aliquots at − 20°C. Prior to our phophorylation experiments, all buffers were supplemented with 50 mm Na-fluoride, 5 mm pyrophosphate and phosphatase inhibitor cocktails 1 and 2.

Sucrose gradient centrifugation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

All detergent extracts were adjusted to 40% (w/v) sucrose by adding an equal volume (1 mL) of 80% (w/v) sucrose in TNE buffer to the myelin (see above) to a final volume of 2 mL (w/v) and placed at the bottom of an ultracentrifuge tube (Beckman, Palo Alto, CA, USA). For the step gradient, 5 mL of 30% (w/v) sucrose in TNE was layered over the lysate and an additional 5 mL of 5% (w/v) sucrose in TNE was placed on top of the 30% (w/v) sucrose layer. Gradients were centrifuged for 18 h at 160 000 g at 4°C in a Beckman SW41 rotor. Fractions of 1 mL were harvested from the top and analyzed by western blot analysis.

Western blot analyses

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Proteins were separated on 12 or 15% (w/v) sodium dodecyl sulfate (SDS)-polyacrylamide gels under reducing conditions and transferred to Hybond ECL nitrocellulose membranes (Amersham, Freiburg, Germany) by semi-dry electroblotting. Blots were blocked overnight at 4°C with Tris-buffered saline (TBS, 50 mm Tris-HCl, ph 7.4, 0.9% w/v NaCl) containing 3% (w/v) non-fat milk and 2% (w/v) bovine serum albumin (BSA) and incubated for 1 h at room temperature with the indicated primary antibody. After several washings, blots were incubated for 1 h with goat anti-mouse (Dako) or goat anti-rabbit (Cappel, Turnhout, Germany) secondary antibody coupled to horseradish peroxidase (both 1 : 2500), washed extensively and developed using a chemiluminescent detection system (ECL, Amersham). The data for each analyses were confirmed by at least two independent protein preparations and subsequent repeated western-analyses.

Immunoprecipitation and phosphorylation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Sciatic nerves from adult Wistar rats and 15-day-old rat pups were dissected and sliced into 0.5-cm pieces. Segments from four adult nerves were pooled and incubated, essentially according to the procedure of Iyer et al. (1996) in Krebs–Ringer bicarbonate buffer containing 430 μm freshly prepared vanadyl hydroperoxide (pervanadate), 5 mm sodium pyrophosphate and the protease inhibitors (Hofmann–La Roche, Mannheim, Germany) leupeptin, aprotinin, pepstatin and phenylmethylsulfonyl fluoride at 25 μg/mL, for 1 h at 37°C. Additionally, we supplemented the buffer with 50 mm sodium fluoride and phosphatase inhibitor cocktail 1 and 2 (Sigma) to retard phosphatase activity. Segments of six nerves from 15-day-old rats were incubated under the same conditions. Nerve segments were then homogenized to prepare myelin and rafts (see above) or were extracted in lysis buffer containing 1.2% NP-40, 1.2% TX-100, 0.1% SDS, 50 mm Tris (pH 7.8) and 150 mm NaCl for immunoprecipitation experiments. Extracts were cleared by centrifugation and then subjected to immunoprecipitation with the anti-phosphoserine antibody (Biomol, Hamburg, Germany) or an irrelevant antibody (polyclonal anti-rMDC15; Antibody Service Dr Pineda) as negative control, pre-bound to protein G-Sepharose, overnight at 4°C. The beads were washed three times with precipitation buffer supplemented with 0.5 m NaCl and resuspended in 1 × SDS sample buffer. Immunoprecipitated proteins were boiled for 5 min, fractionated by SDS–polyacrylamide gel electrophoresis (PAGE) and electro-transferred onto nitrocellulose membranes for immunoblotting.

Generation and characterisation of plasmolipin-specific polyclonal antibody

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

The peptide RGVGSNAATSQMAGGYS, comprising amino acids 168–182 of rat plasmolipin, was synthesized, coupled to keyhole limpet hemocyanin and used to immunize rabbits. The selected peptide sequence represents the C-terminus of the plasmolipin protein (Fig. 1a) and is predicted to be intracellularly localized according to the membrane topological model previously proposed for plasmolipin (Gillen et al. 1996).

image

Figure 1. An epitope-specific antibody was raised against the C-terminal amino acids 168–182 of plasmolipin protein. (a) Schematic representation of the proposed membrane topology of plasmolipin. The peptide recognized by the antibody Pla-4 is indicated by a striped line. The putative phosphorylation sites Ser9 and Ser130 are marked by stars. (b) Pla-4 recognizes a single band of approximately 20 kDa in western blots of total protein extracted from rat sciatic nerve (SN), brain (Br) and kidney (Kd). Cultivated HeLa-cells show slight endogenous plasmolipin expression (WT), which is significantly increased after recombinant plasmolipin overexpression (rec.). No signal is obtained by pre-adsorption of the Pla-4 antibody with the epitope-specific C-terminal peptide. +, peptide preadsorption; –, no peptide pre-adsorption.

Download figure to PowerPoint

The rabbit antiserum was affinity-purified on a peptide column (Pineda Antibody Service, Berlin, Germany) and its specificity was verified by detection of both, endogenous plasmolipin protein (see Fig. 1b) and recombinant plasmolipin protein expressed by transfected cultured cells (Fig. 1b) or transformed Escherichia coli bacteria (data not shown). The polyclonal antibody (Pla-4) recognized a protein band of the predicted size of 20 kDa in sciatic nerve, brain and kidney (Fig. 1b). In addition, the C-terminal peptide used for immunization was able to completely block the recognition of plasmolipin by the antiserum when Pla-4 was pre-incubated for 1 h with 500 ng/mL of the peptide prior to immunoblotting (Fig. 1b). Likewise, no staining was observed when using pre-immune serum (data not shown).

Plasmolipin expression during peripheral (re)myelination

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

In order to compare the expression profile of plasmolipin with the time course of post-natal myelination, we have analyzed protein fractions from rat sciatic nerve at different time-points from birth (P0) through adulthood (ad). Already at post-natal day 4, plasmolipin protein could be detected and during the first three weeks the expression increased and then the plasmolipin protein level remained almost constant in the adult peripheral nerve (Fig. 2a).

image

Figure 2. Western blot analyses showing the plasmolipin expression pattern during (a) post-natal development of rat sciatic nerve (SN) and (b) after sciatic nerve lesion. (a) Protein was prepared from rat sciatic nerve at each time point indicated (P, post-natal day; ad, adult). Total protein (30 μg) was separated on a 15% SDS–PAGE and then probed with the polyclonal plasmolipin antibody Pla-4. (b) Protein was prepared from distal nerve fragments at each time point indicated (d, days after transection, Ctrl: unlesioned control nerve). Total protein (20 μg) was separated on a 15% SDS–PAGE and then probed either with polyclonal anti-plasmolipin-antibody (Pla-4) or with an antibody directed against myelin protein P0.

Download figure to PowerPoint

We further analysed plasmolipin expression following two types of traumatic nerve lesions. Sciatic nerves of adult rats were either crushed (data not shown) or transected (Fig. 2b) and distal nerve stumps were collected at different time points after lesion. Immunoblots were performed using membrane protein fractions. The observed rapid and distinct decrease in plasmolipin protein levels (Fig. 2b) of injured sciatic nerve fragments is characteristic for various established myelin genes after both types of lesion. In contrast to our expectation, a subsequent recovery to control levels was observed not only in the distal stumps of crushed (data not shown), but also in the non-regenerating transected nerves within 4 weeks after injury (Fig. 2b). For internal control, we compared the expression profiles of the established myelin genes P0, PMP22 and MBP, whose protein expression after nerve axotomy is closely linked to re-myelinating Schwann cells. As expected, re-expression of the latter three proteins P0 (Fig. 2b), PMP22 and MBP (data not shown) failed after 4 weeks within the distal fragments of non-regenerating transected sciatic nerves.

Localization of plasmolipin protein

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Previous studies have shown endogenous expression of plasmolipin in the nervous system (Shea et al. 1986; Gillen et al. 1996), as well as in various non-neural tissues (Fischer and Sapirstein 1994; Hamacher et al. 2001). However, the term plasmolipin was initially given to a protein doublet of variable and changing molecular weights identified by SDS–PAGE (Tosteson and Sapirstein 1981; Shea et al. 1986; Sapirstein et al. 1988; Cochary et al. 1990; Fischer et al. 1991; Sapirstein et al. 1992a). For this reason, a comprehensive and distinct characterization of the distribution and cellular localization of the 20 kDa proteolipid protein plasmolipin in vivo presently does not exist. Based on our previous observation that plasmolipin mRNA is expressed by Schwann cells in vivo and in vitro (Gillen et al. 1996) we examined plasmolipin protein distribution by immunocytochemistry in sciatic nerves of 6-month-old rats. Intense plasmolipin immunoreactivity was observed in the ring-like structures of compact peripheral myelin around the axons (Fig. 3a). Immunohistochemistry on adjacent nerve sections using a myelin basic protein (MBP)-specific antibody showed widespread co-localization of both proteins within the myelin compartment (data not shown). Plasmolipin staining of cells within the epineurium, perineurium or blood vessels was not observed.

image

Figure 3. Immunostaining showing expression of plasmolipin protein in cross sections of rat sciatic nerve (a), spinal cord (b), and coronal sections of rat adult brain (c, d). Plasmolipin protein is predominantly localized in the myelin sheath of the peripheral (a) and myelinated fibre tracts of the central nervous sytem (b, c, d). aC, anterior part of the anterior commisure; BV, blood vessel; Cpu, caudate putamen; fpCM, frontoparietal cortex motor area; ed, endoneurium; ep, epineurium; fpCss, frontoparietal cortex somatosensory area; GCC, genu of the corpus callosum; LO, lateral olfactory tract; mCP, middle cerebellar peduncle; prNc, pontine reticular nucleus; PY, ventral pyramidal tract; sCP, superior cerebellar peduncle; wm, white matter of the cerebellum; 4 V, fourth ventricle. Scale bars: 100 μm in a and b; 1 mm in c and d.

Download figure to PowerPoint

Likewise, cross sections of the spinal cord showed a distinct staining of white matter (Fig. 3b, close up). Double immunolabelling experiments in the spinal cord with antibodies against myelin basic protein MBP, microtubule-associated protein 2 (MAP-2), neuron-specific nuclear protein (NeuN) and glial fibrillary acidic protein (GFAP) proved co-localization of plasmolipin only with the structural myelin protein MBP, but not with the neuronal or astrocytic protein markers used (data not shown). No staining of endothelial cells of blood vessels was observed within the spinal cord.

Using our anti-plasmolipin antibody, we wanted to reveal whether or not plasmolipin expression is also restricted to myelinated structures in the brain, as seen in the PNS and spinal cord. We performed plasmolipin-specific immunocytochemistry to different coronal sections of the rat adult brain. As shown in Figs 3(c and d), our experiments resulted in significant staining of myelinated tracts. Coronal forebrain sections presented intense staining of, for example, the myelinated genu of the corpus callosum (GCC), myelinated tracts of the caudate putamen (Cpu), the anterior part of the anterior commissure (aC) and the lateral olfactory tract (LO) (Fig. 3c). Coronal sections of the hind brain showed significant staining of the ventral pyramidal tract (PY), tracts within the caudal part of the pontine reticular nucleus (prNc), the middle (mCP) or superior cerebellar peduncle (sCP) and the white matter (wm) of the cerebellum (Fig. 3d).

Besides the ubiquitous localization in myelin, we also observed distinct neuronal plasmolipin expression in coronal and sagital sections of adult rat brains. Plasmolipin-immunoreactivity of neuronal cell bodies was observed in the neocortex (Figs 4a and b) as well as in the granular and pyramidal cell layers of the hippocampal formation (Figs 4c–e). Comparison of the immunofluorescence labelling by confocal microscopy clearly identified these plasmolipin-positive cells as neurons due to their extensive co-immunostaining with NeuN-immunoreactivity (Figs 4a and c). Interestingly, plasmolipin-expressing neurons are enriched in cortical layer II (Figs 4a and b). However, as shown in Fig. 4, our experiments revealed that most but not all of the cortical or hippocampal neurons are plasmolipin positive, indicating a distinct but restricted expression of plasmolipin in neuronal subpopulations of adult rat brain. In contrast to our observations of neuronal plasmolipin expression in cortex or hippocampus, co-staining of plasmolipin with the neuronal markers NeuN (Fig. 4f) and MAP-2 (data not shown) revealed no distinct neuronal expression of plasmolipin in the cerebellar stratum granulosum. Figure 4(f) shows that cerebellar plasmolipin expression is restricted to the myelinated fibres of the marrow and stratum granulosum. Staining with a GFAP-specific antibody never showed co-localization of plasmolipin, either in spinal cord (data not shown) or in the brain (Fig. 4d).

image

Figure 4. Immunolocalization of plasmolipin [Pla (green); a–f], neuron-specific nuclear protein [NeuN (red); a, c, f], microtubule-associated protein 2 [MAP-2 (red); b, d], and glial fibrillary acidic protein [GFAP (red); e] in coronal (b, d, e) and sagital (a, c, f) paraffin sections of the adult rat brain by confocal laser scanning microscopy. Immunofluorescence showing the cellular localization of plasmolipin protein, not only in myelinated fiber tracts, e.g. of the alveus hippocampus (d), or cerebellum (f), but also in neuronal subpopulations of the neocortex (a, b) or of the hippocampus formation (c, d, e). Co-localization of the antigens appears yellow. alv, alveus hippocampus; CA1, CA4, CA1-, CA4-neuronal field of the Ammon's horn; dg, dentate gyrus; sg, stratum granulosum; sm, stratum moleculare; I, II, III, IV, layers of the neocortex. Scale bars: 100 μm in a–f; 50 μm in c (close up).

Download figure to PowerPoint

Epithelial expression of plasmolipin

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

As myelin is a very specialized laminar membraneous organelle with common characteristics of apical epithelial membranes and as an asymmetric plasmolipin distribution appeared to occur in nephron tubuli (Sapirstein et al. 1992b), we carefully investigated the cellular localization of plasmolipin in the epithelial tissues kidney and stomach. Immunostaining of paraffin sections of adult rat kidney revealed intense plasmolipin immunoreactivity in both kidney cortex and medulla (Figs 5a and b). Figure 5(a) shows the expected distinct staining of nephron tubuli within the cortex. An extensive labelling of the luminal side of the plasmamembrane was observed (Fig. 5a) as a striking feature of plasmolipin expression. Interestingly, neither the epithelial cells of the capsid glomeruli nor the glomeruli itself showed any plasmolipin staining (Fig. 5a). Within the medulla, not only the distal tubuli but also the collective ducts (Fig. 5b) showed a restricted luminal expression of plasmolipin. Furthermore, our experiments demonstrate that asymmetric expression of plasmolipin is not a unique characteristic of epithelial cells in rat kidney. Likewise, examination of plasmolipin distribution in the fundus area of rat stomach resulted in strong staining of the apical surface of the glandular stomach epithelium (Fig. 5c). A distinct labelling of the faveolae gastrica could be observed (Fig. 5c, close up). In addition, mucoidic epithelial cells of the pars pylorica glands also showed a very polarized staining on the luminal side of their plasma membranes (Fig. 6d). This apical expression of plasmolipin is observed already during post-natal development (data not shown) and might indicate an important role of plasmolipin in epithelial cell polarization.

image

Figure 5. Immunostaining showing cellular localization of plasmolipin protein in cross sections of adult rat kidney (a, b) and stomach (c, d). Plasmolipin protein is predominantly localized in the apical membrane of several rat epithelia. Extensive plasmolipin staining is observed on both luminal sides of tubular epithelial cells of the kidney cortex (a) and collective ducts within the kidney marrow (b). In rat stomach as well the glandular epithelium (c) as mucoidic cells of the pars pylorus showed polarized plasmolipin distribution (d). CG, capsid glomeruli; G, glomerulus; arrow heads, luminal side of the plasma membrane; arrows, faveolae gastricae. Scale bars: 50 μm in a–d, 20 μm in a (close up), c (close up).

Download figure to PowerPoint

image

Figure 6. Western blot analyses of TX-100 extracted plasmolipin protein from purified (a) rat brain and (b) sciatic nerve myelin. Purified myelin was floated in sucrose step gradients after TX-100 extraction at 4°C, 37°C and pre-treated with β-cyclodextrin (β-CD). One-millilitre fractions were collected from the top [5% (w/v), sucrose, fraction 1–5] to the bottom [40% (w/v), sucrose, fraction 10–12]. Rafts containing insoluble fractions were found at the 5%/30% sucrose interphase (fraction 5–7). Part (b) was reproduced with modifications from Hasse et al. (2002); with permission.

Download figure to PowerPoint

Plasmolipin in myelin is associated with lipid rafts

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

The observation that apical and basolateral membrane compartments differ characteristically in their lipid and protein composition led to the lipid raft hypothesis, predicting that certain proteins of the apical membrane are sequestered in the trans-Golgi-network into glycosphingolipid and cholesterol enriched microdomains prior to their transport to the apical membrane (Simons and van Meer 1988; Rothberg et al. 1990; Simons and Ikonen 1997; Ledesma et al. 1998). As myelin biogenesis and maintenance requires specific polarized sorting and transport of myelin components, we proved whether plasmolipin protein of central myelin is also associated with lipid rafts. Due to their physical properties, rafts are resistant to detergent at low temperature. We carried out extraction of purified myelin from adult rats using the non-ionic detergent TX-100. Lipid rafts were retained as insoluble lipid rich microdomains at 4°C, floating on sucrose gradients at a low density (Simons and Ikonen 1997). Aliquots of the resulting gradient fractions were analyzed by western blots and revealed that plasmolipin is highly enriched in the detergent insoluble fractions of CNS myelin (Fig. 6a). At 4°C, only minor portions of plasmolipin could be detected in the corresponding soluble fractions 11–12 (Fig. 6a). However, central plasmolipin is completely solubilized upon TX-100 extraction at 37°C (Fig. 6a), a fundamental criterion for lipid rafts-associated proteins (Brown and Rose 1992). Furthermore, treatment of purified myelin with β-cyclodextrin (β-CD), a carbohydrate molecule that depletes membrane cholesterol prior to TX-100 extraction at 4°C led to the almost complete solubilization of plasmolipin from central myelin (Fig. 6a). In this respect, plasmolipin clearly differs from MBP, a protein also localized in the compact myelin sheath of the CNS and PNS. In contrast to plasmolipin, MBP prepared from CNS myelin was solubilized in TX-100 at 4°C (data not shown).

Plasmolipin of peripheral myelin is phosphorylated

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

Rafts have been described as membrane domains with specialized functions in signal transduction. Therefore, we were interested to determine whether plasmolipin of peripheral nerves and, in particular, plasmolipin associated with lipid rafts (Fig. 6b) is phosphorylated. To detect putative phosphorylated plasmolipin, protein immuno-precipitation experiments were carried out. Protein fractions from adult rat sciatic nerves and purified myelin were initially incubated in the presence of pervanadate to inhibit protein phosphatase activity. Then, these fractions were immunoprecipitated with a suitable anti-phophoserine antibody, separated by SDS–PAGE and immunoblotted with Pla-4 antibody. Significant amounts of phosphorylated plasmolipin protein could be detected in both sciatic nerve and purified myelin (Fig. 7a). As a control, immunoprecipitation using an irrelevant anti-rMDC15 antibody did not precipitate plasmolipin from adult purified myelin (Fig. 7a). Interestingly, the peripheral 4-TM-myelin protein PMP22, which also contains a putative serine phosphorylation site (Ser57; Spreyer et al. 1991), could not be obtained in our precipitates of nerve and myelin (Fig. 7a). Further anti-phophoserine immunoprecipitation and succeeding immunoblotting experiments were performed from sucrose gradient fractions of purified myelin extracted with TX-100 at 4°C (Fig. 6b). Figure 7(b) shows significant amounts of phosphorylated plasmolipin protein in the insoluble lipid raft fraction 6 of young (P15) and adult (ad) peripheral myelin membranes. In contrast, no phosphorylated plasmolipin could neither be detected in the soluble fractions 12, nor when material of fraction 6 was immunoprecipitated with the unrelated anti-rMDC15 antibody (Fig. 7b).

image

Figure 7. Detection of phosphorylated plasmolipin protein after immunoprecipitation. (a) Adult sciatic nerve proteins and purified peripheral myelin were immunoprecipitated using anti-phosphoserine antibody. Western blots demonstrated the presence of phosphorylated plasmolipin protein (Pla) in anti-phosphoserine precipitates of both sciatic nerve (SN) and purified peripheral myelin (My). In contrast, no phosphorylated plasmolipin was detected in purified myelin precipitated with an unrelated polyclonal anti-rMDC15 control antibody (–). (b) Phosphorylated plasmolipin was detected in the lipid rafts containing TX-100 insoluble fraction 6 (F6) of young (P15: post-natal day 15) and adult (ad) purified myelin after immunoprecipitation with the anti-phosphoserine antibody. No plasmolipin was detected in immunoprecipitated TX-100 soluble fractions 12 (F12). –, Precipitation of fraction 6 (P15) with the unrelated antibody anti-rMDC15 led to a complete loss of plasmolipin detection; +, positive control containing either (a) an aliquot of original purified adult myelin, or (b) an aliquot of original fraction 6 (P15).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References

We generated a peptide-specific antibody (Pla-4), directed against the C-terminal portion of rat plasmolipin. Western blots carried out with this polyclonal antibody revealed recognition of a distinct 20 kDa band in protein fractions prepared from rat sciatic nerve, brain and kidney. While the antibody specifically detected both endogenously and recombinantly expressed plasmolipin (Fig. 1b), pre-adsorption of the immunogenic C-terminal peptide resulted in the lack of immunostaining proving specificity of our antibody.

Plasmolipin immunoreactivity could be detected in myelinated structures of adult rat sciatic nerve, spinal cord and brain (Fig. 3). This is in line with previous in situ-hybridization experiments (Gillen et al. 1996), which clearly confined plasmolipin transcripts to oligodendrocytes and myelinating Schwann cells of the CNS and PNS, respectively. Further, the post-natal developmental expression profiles of plasmolipin protein in rat brain (Sapirstein et al. 1992b) and sciatic nerve (Fig. 2a) resemble the time course of myelination, suggesting a close correlation of plasmolipin function to myelin biogenesis or maintenance. However, peripheral nerve injury revealed an interesting feature that distinguishes plasmolipin protein from other established peripheral myelin proteins (Fig. 2b). After a traumatic lesion of peripheral nerve axons, degenerate and myelinforming Schwann cells dedifferentiate and display strong down-regulation of myelin gene expression (for review see Bosse et al. 2001). Accordingly, we observed a strong and rapid decrease of plasmolipin protein in the distal stumps of both crushed (data not shown) and transected sciatic nerves (Fig. 2b). However, in contrast to the established myelin protein P0 (Fig. 2b), re-induction of plasmolipin expression not only occurred during nerve regeneration, when re-myelination is induced by re-established axon-Schwann cell contacts. Plasmolipin protein re-expression was also observed in the distal part of a transected and ligated sciatic nerve, where axon regeneration and re-myelination was lacking, suggesting an additional role different from myelin biogenesis and maintenance.

The neuronal expression of plasmolipin observed in the neocortex and hippocampus of adult rat brain (Figs 4a and c) indicates a function of this protein different from myelin. Confocal co-localization of plasmolipin and NeuN in neurons of the neocortex and hippocampus (Figs 4a and c) clearly documented neuronal plasmolipin expression. Interestingly, most neurons of the neocortex and hippocampal formation (Fig. 4a–e), but no neurons of the cerebellar stratum granulosum (Fig. 4f) or spinal cord (data not shown), are plasmolipin positive. Thus the present study comprises the first in vivo data of a distinct but restricted neuronal expression of plasmolipin in the rat adult brain.

As a prerequisite for normal functions, plasma membranes of epithelial cells are asymmetric and subdivided into apical and basolateral domains (Rodriguez-Boulan and Nelson 1989; Rodriguez-Boulan and Powell 1992). Therefore, we examined whether asymmetric protein distribution is a common feature of epithelial plasmolipin expression. Indeed, our study reveals a polarized plasmolipin distribution in both tissues examined. In rat adult kidney, we observed a pronounced plasmolipin staining at the luminal side of plasma membranes of tubular cells (Fig. 5a) and collective ducts within the medulla (Fig. 5b). Interestingly, the epithelia of the capsula glomeruli and the glomerulus itself were clearly devoid of plasmolipin immunoreactivity (Fig. 5a), indicating that polarized distribution of plasmolipin is not a general feature of kidney epithelial cells. Apical plasmolipin expression, in the fundus region of rat stomach (Fig. 5c) and in the pars pylorica glands (Fig. 5d) further supported asymmetric plasmolipin targeting in epithelial cells.

Whether or not plasmolipin expression in neural and non-neural tissues reflects diverse biological functions remains to be investigated. But the widespread expression pattern in the PNS, CNS and epithelial tissues with their different morphologies and functions argues against a single myelin-specific function. The similarities in plasmolipin structure and expression with MAL/VIP17 (Alonso and Weissman 1987; Schaeren-Wiemers et al. 1995; Frank et al. 1998), a member of the same gene-family (Magyar et al. 1997; Perez et al. 1997), may suggest putative functions of plasmolipin as an intracellular sorting and/or transport component or as a structural protein in apical membranes.

As the glycosphingolipid-enriched composition of apical membranes resembles that of myelin (Shayman and Radin 1991; Pfeiffer et al. 1993; Frank et al. 1998) myelin-producing cells are considered to be polarized cells and several myelin proteins of the CNS and PNS were identified to be associated with lipid rafts (Kim et al. 1995; Krämer et al. 1997; Frank et al. 1998; Kim and Pfeiffer 1999; Simons et al. 2000; Hasse et al. 2002). The biochemical data presented in this study (Fig. 6a) clearly revealed that plasmolipin is also a distinct component of CNS myelin lipid rafts. However, the determinants, signals or binding motifs responsible for the differential targeting of myelin proteins are largely unknown. In addition, it is noteworthy that a direct correlation of the biochemical data with cellular compartements is difficult as the resulting patterns of raft-associated proteins depend on the choice of detergent, the experimental protocol, as well as on the tissue or cell source (Krämer et al. 1997; Kim and Pfeiffer 1999; Simons et al. 2000; Hasse et al. 2002).

It has been proposed that lipid rafts provide platforms, not only for protein trafficking, but in addition may serve as dynamic platforms for signal transduction initiation (Brown and Rose 1992; Simons and Ikonen 1997; Brown and London 1998; Deans et al. 1998; Montixi et al. 1998; Simons and Toomre 2000; Zajchowski and Robbins 2002). Whether phosphorylation of plasmolipin (Fig. 7) is a direct consequence of its association with a specific raft micro-environment remains to be investigated. But the fact, that higher amounts of phosphorylated plasmolipin were detected in lipid raft fractions of peripheral myelin of young rats (P15) as compared with adult animals (Fig. 7b), is in line with the previous observation, that the relative amount of phosporylated P0 protein also declines during nerve maturation (Iyer et al. 2000). Thus, the phosphorylation of both compact myelin proteins is most prominent during the period of peak myelin deposition. It has been shown recently that phosphorylation of P0 could modulate its adhesive properties (Xu et al. 2001). Therefore, we propose that the phosphorylation of plasmolipin could be important for the formation and compaction of myelin and thus should be more critical during early events of myelin assembly than for the maintenance of mature myelin.

In summary, this study provides a comprehensive cellular protein expression analysis of plasmolipin in the nervous system and epithelial tissues. Although the finding of a predominant myelin-related expression in both the PNS and CNS suggests a structural role of plasmolipin during myelin biogenesis and maintenance, the re-induction of plasmolipin in axotomized non-regenerating peripheral nerve fragments, the restricted neuronal expression, the distinct lipid raft association and the observed polarized distribution of plasmolipin in different epithelial tissues implicates a more generalized function of plasmolipin, e.g. as a structural component of apical membranes or as an intracellular protein trafficking component.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Animals and surgery
  5. Antibodies
  6. Immunohistochemistry
  7. Myelin preparation
  8. Detergent extraction of purified myelin
  9. Cholesterol extraction
  10. Sucrose gradient centrifugation
  11. Western blot analyses
  12. Immunoprecipitation and phosphorylation
  13. Results
  14. Generation and characterisation of plasmolipin-specific polyclonal antibody
  15. Plasmolipin expression during peripheral (re)myelination
  16. Localization of plasmolipin protein
  17. Epithelial expression of plasmolipin
  18. Plasmolipin in myelin is associated with lipid rafts
  19. Plasmolipin of peripheral myelin is phosphorylated
  20. Discussion
  21. Acknowledgements
  22. References