Trafficking of PLP/DM20 and cAMP signaling in immortalized jimpy oligodendrocytes

Infection Procedure

Both 158N and 158JP oligodendrocytes were cultured on coverslips in 24-microwell plates at 3,000 cells/well. After 4 days, the cells were used for transfection. Transfection was performed by the calcium phosphate procedure as described above with 0.1–0.2 μg DNA/coverslip with one of four purified plasmids (nPLP-EGFP, nDM20-EGFP, jpPLP-EGFP, or jpDM20-EGFP). After 4-h incubation, cultures were washed twice with DMEM containing 5% calf serum and then incubated for 24, 48, and 72 h in the same medium. Cells were then fixed in 4% paraformaldehyde (PFA) for 15 min at room temperature (RT) and then stored at 4°C in phosphate buffer saline (PBS).

Generation of Vectors and Expression Constructs

PLP-EGFP and DM20-EGFP fusion constructs were made by initially cloning full-length PLP and DM20 cDNA (kindly provided by Dr. A. Fannon, Mount Sinai, NY) into the BamHI site of the expression vector pCMV, and then inserting the EGFP gene in frame with PLP or DM20 at the 3′ end. The constructs were isolated by transformation of LE392 competent cells, followed by Kanimycin selection, and confirmed by gel electrophoresis and restriction analysis. The sequence of the inserts and PLP-EGFP/DM20-EGFP fusion points were verified by DNA sequencing. The performance of the construct expressions was tested by transfection into NT2 cells and then by detection of the fluorescent PLP/DM20 products. The monoclonal antibody O10 (kindly provided by Dr. K. Nave) combined with rhodamine-conjugated secondary antibody was used to attest the correct insertion of PLP/DM20 by performing live staining. Both EGFP fluorescence and rhodamine O10 immunofluorescence confirmed that PLP-EGFP and DM20-EGFP were synthesized and transported to the cell surface. Jimpy PLP-EGFP and jimpy DM20-EGFP were constructed using the same protocols as their wild-type counterparts. Since jimpy PLP and DM20 cDNAs (kindly provided by Dr. A. Fannon, Mount Sinai, NY) differ from wild type at the 3′ end, a different set of primers were used for PCR amplification. The size and orientation of the constructs were tested after transformation into LE392 cells.

Immunodetection of Myelin Antigens and Other Cellular Components

Cells grown on 14 mm glass coverslips in 24-well plates were fixed in 4% paraformaldehyde (PFA) in PBS for 30 min. Cultures were then saturated by incubation for 1 h with 10% normal goat serum or in 3% blocking solution in PBS (Roche, France). After several washes in PBS, cells were then incubated with one of the following antibodies; rabbit anti-PLP directed to the sequence 117–129 peptide (Trifilieff et al., 1986) or to PLP C-terminal (Benjamins et al., 1989). Appropriate fluorescein Alexa 488 and Alexa 546-labeled secondary antibodies were used (Molecular Probes, The Netherlands). All incubations were performed at RT. Surface antigen detection was performed by incubation of the primary antibody with live cells for 30 min at 37°C followed by fixation with 4% PFA for 5 min. PLP-EGFP and DM20-EGFP expression in transfected cells was visualized with fluorescence microscopy either directly or after incubation with a rabbit polyclonal antibody to EGFP (Clontec, France) and then with Alexa-conjugated antirabbit IgG (Molecular Probes). Coverslips were mounted in Aquapolymount (Polysciences, Germany) for light microscopic observation (Leica DM and Olympus BX 60 microscopes). Transfected cells were also examined by confocal microscopy using the ACAS laser cytometer (Meridian Instruments).

Dibutyryl cAMP Treatment

After plating immortalized jimpy and normal cells at 3,000 cells/coverslip in 24-well plates for 4 days, transfected and nontransfected cultures were treated with N⁶,2′-O-dibutyryladenosine 3′,5′-cyclic monophosphate (dbcAMP; Sigma) at 1 mM for 72 h. Treated cultures were then rinsed and fixed with 4% PFA for 20 min for immunocytochemical staining.

Northern Blots of ICER and GAPDH

Total RNAs were isolated from confluent cultures in 100 mm individual petri dishes using Trizol (Invitrogen, Life Technologies, France). cDNA synthesis from the RNAs was performed using the Riboclone cDNA Synthesis System AMV RT kit (Promega). Total RNA (4 μg) in the presence of 2.5 μg oligo dT₁₅ primers (Promega) in 15 μl H₂O was heated at 70°C for 5 min. First-strand buffer and RNAsine ribonuclease inhibitor (40 units, Promega) were then added and incubated at 42°C for 5 min. Avian myeloblastosis virus (AMV) reverse transcriptase (30 U; Promega) was used at 42°C for 1 h. cDNAs were then subjected to PCR with specific primers to ICER (sense, 5′-AACATGGCTGTAACTGGAGA; antisense, 5′-TTACTCTGCTTTATGGCAAT) and GAPDH (sense, 5′-TCCACTCACGGCAAATTCAACG; antisense 5′-CAGGCGGCACGTCAGATCCACG). Amplification was performed in 10 mM Tris-HCl buffer, pH 9, containing 50 mM KCl, 1.5 mM MgCl₂, 0.2 mM dNTPs, 0.1% Triton X-100, 0.2 mg/ml BSA, and 2 U Taq DNA polymerase (Appligene, France) for 35 cycles in thermocycler Progene (Techne, U.K.) at 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min, and then, at the end 10 min, at 72°C for ICER. GAPDH was amplified with 35 cycles, 94°C for 1 min, 69°C for 1 min, and 72°C for 1 min. The probes were purified using GFX columns (AmershamPharmacia Biotech, France) and then labeled with dCTP-P³² using Random Primed DNA kit (Roche). The labeled probes were then purified on Sephadex G25 gel (AmershamPharmacia Biotech). After electrophoresis on agarose gel, the RNAs (10 μg/sample) were transferred to hybridization membrane Hybond-N (AmershamPharmacia Biotech). Hybridization with ICER and GAPDH probes were performed at 42°C for 12 h. For autoradiography, the hybridized nylon membrane was exposed to Kodak film at −80°C for 4 days and then developed.

Electrophysiological Recordings

For electrophysiological experiments, cells were cultured as described above and plated in 35 mm noncoated dishes at low cell density (2,000 cells per dish). Recordings were performed on 2- or 5-day-old cells using solutions of the following composition (in mM): NaCl 132, KCl 2, HEPES 10, MgCl₂ 2, CaCl₂ 0.5, D-glucose 10 for the bath and KCl 125, NaCl 5, EGTA 5.5, CaCl₂ 1, MgCl₂ 2, HEPES 10 for the pipette medium. The pH was adjusted to 7.4 with NaOH and 7.2 with KOH for the bath and the pipette medium, respectively. Resting membrane potential was measured in the whole cell mode of the patch-clamp technique (Hamill et al., 1981) using an Axopatch 200B amplifier (Axon Instruments, CA) and the clampex routine of the pClamp software package (Axon Instruments). After gigaseal establishment with pipettes of 3 to 4 MΩ resistance, the breakthrough into whole cell was achieved in current-clamp mode, permitting resting potential determination before dialysis of the cell with the pipette medium. Records were filtered at 1 kHz and acquired (2 kHz sampling frequency) with a LabRack interface card (Scientific Instruments, OH).

O10 Antibody Immunoreactivity With PLP/DM20 in Nontransfected Oligodendrocyte Lines

Our previous study (Feutz et al., 2001), using antibody that detects only the PLP isoform, showed that PLP was distributed similarly in cell bodies and processes in both normal and jimpy oligodendrocyte lines and it is associated with plasma membrane. The monoclonal antibody O10, which detects a conformationally sensitive PLP/DM20 epitope, also showed positive immunoreactivity on paraformaldehyde-fixed cells from both normal and jimpy cell lines (Fig. 1A and C). This observation further confirms that both normal and jimpy PLP and DM20 are translated and transported to membrane. However, when normal and jimpy live cells were incubated with O10 antibody to detect the surface epitope, only normal cells were immunofluorescent (Fig. 1B), while the jimpy cells did not exhibit any significant specific immunofluorescence (Fig. 1D), indicating PLP/DM20 is not properly inserted into the plasma membrane.

Distribution of Normal and Jimpy Fusion Protein PLP-EGFP and DM20-EGFP in Transfected Normal and Jimpy Cells

Normal 158N cells transfected with the vectors nPLP-EGFP (Fig. 2A) or with nDM20-EGFP (Fig. 2B) for 48 h expressed the fusion fluorescent protein in the cell body and its fine processes. Similarly, jimpy 158JP cells transfected with the same vectors express both normal PLP (Fig. 2C) and DM20 (Fig. 2D) in cell perikarya and processes. Both normal and jimpy transfected cells expressed high level of normal transfectant PLP or DM20 as judged by the immunofluoresence detection with anti-GFP antibody. When jimpy PLP or jimpy DM20 expression vectors were introduced into the normal cells, jimpy PLP (Fig. 3A) and jimpy DM20 (Fig. 3B) were detected in cell bodies as well as in distal cell processes. Also, high levels of jimpy PLP (fig. 3C) and jimpy DM20 (Fig. 3D) were detected in jimpy cells in both cell body and its extensions.

Confocal microscopy using the Meridian laser cytometer confirmed the distribution of jimpy PLP (Fig. 4) and DM20 (not shown) in jimpy cells observed with whole cell fluorescence microscopy. Higher concentration (red color) of PLP was observed in the cell body surrounding the nucleus and also at the emergence of cell processes. Lower levels were present in the distal cell extensions (Fig. 4). Jimpy DM20 was also observed in the cell processes but the highest concentration was found in the cell perikaryon (not shown).

The normal cell line (Fig. 5A and C) transfected with normal PLP showed a bright immunofluorescence using O10 antibody on live cells (Fig. 5A) but when transfected with jimpy PLP they did not show any specific fluorescence (Fig. 5C). The background fluorescence observed in normal cells was due mainly to the presence of endogenous PLP in normal cells. The normal PLP can be detected on the cell surface of jimpy cells (Fig. 5B). In contrast, jimpy PLP was not detected on live jimpy cells with O10 antibody (Fig. 5D). Normal and jimpy DM20 showed similar behavior in both transfected cell lines (not shown). Ten to 20 cultures on coverslips obtained from three to five transfection experiments were usually used to generate the present data throughout the article. Each coverslip showed 20 to 30 fluorescent transfected cells.

Electrophysiological Recording

Because resting membrane potential was described as an index of cell maturation and plasma membrane composition in oligodendroglial progenitors (Knutson et al., 1997), we investigated whether failure to insert jimpy PLP in proper orientation into plasma membrane of 158JP cells influences this resting membrane potential. Recording pipette potential was recorded before and after breakthrough into the whole cell. After membrane breakdown, the potential signal showed a sharp negative shift, which generally peaked at a minimum before a slow evolution to more depolarized steady-state values. The peak value was considered as a good estimation of the resting potential (RP) before cell dialysis. Oligodendrocytes, used here as a control, showed as expected (Knutson et al., 1997) a statistically significant (t-test, P < 0.05) negative shift in RP according to the age of the cells in culture. The RP obtained at 2 and 5 days after plating were −31.3 ± 8.9 and −48.1 ± 4.5 mV, respectively (Fig. 6). The 158N cells that are, like the oligodendrocytes, sensitive to cAMP showed the same negative shift amplitude in RP values. The absolute values, however, were more depolarized (−10.1 ± 1.7 and −28.1 ± 2.6 mV for 2- and 5-day-old cells, respectively; P < 0.0005). In contrast, the 158JP cells did not exhibit any RP evolution during the 5 days of culture. A mean value of −14.2 ± 2.8 mV was obtained at 5 days of cell culture, which is not significantly different from the −12.5 ± 1.9 mV measured at 2 days.

Expression of Transfectant PLP and DM20 and Differentiation of Oligodendrocyte Lines Induced by dbcAMP

Dramatic morphologic transformation of 158N cells from flat to process bearing cells after 72-h treatment with dbcAMP was clearly observed when examined under a phase contrast microscope or when stained for PLP (Fig. 7A and B). The cell bodies became round and smaller, had many emerging cell processes, and were more brightly stained with PLP, while jimpy cells remained insensitive to similar treatment (Fig. 7C and D).

After 72 h, all the fluorescent cells from the normal cell line 158N transfected with the EGFP alone showed transformation in their morphology after treatment with dbcAMP (Fig. 8). Interestingly, most of the normal cells transfected with nPLP-EGFP or nDM20-EGFP plasmids and consequently overexpressed PLP or DM20 lost their potential of differentiation (Fig. 8). Only 20% to 25% of the fluorescent cells exhibited moderate transformation. In contrast, after similar treatment, jimpy 158JP cells transfected with the control plasmid EGFP remained insensitive to dbcAMP (Fig. 8), as did jimpy cells transfected with nPLP-EGFP or nDM20-EGFP (Fig. 8). The expression of normal fusion protein PLP-EGFP or DM20-EGFP in jimpy cells did not restore the sensitivity to dbcAMP treatment.

The activation of ICER gene by dbcAMP in the normal immortalized line was detected by northern blots. After agarose gel electrophoresis of the RNAs (10 μg/sample) from dbcAMP-treated and nontreated 158N and 158 JP cell, the Hybond-N membrane was hybridized with the ICER and GAPDH probes. ICER mRNA level was increased substantially by 2 h after dbcAMP treatment and then completely downregulated after 12 h (Fig. 9A). In contrast, jimpy immortalized cell line did not show any significant transient increase in ICER transcriptional level following the same treatment with dbcAMP (Fig. 9B). The housekeeping gene GAPDH mRNA levels were detected as a control experiment (Fig. 9C and D).

Distribution of Endogenous Normal and Jimpy PLP

We previously reported that both normal and jimpy PLP were detected at low levels with the antibody to the sequence 117–129AA of PLP in the immortalized normal and jimpy cell lines (Feutz et al., 2001). This antibody recognizes both normal and jimpy PLP but not DM20 because this sequence is deleted from DM20 and is before the jimpy mutation in intron 4 (Trifilieff et al., 1986). The 010 monoclonal antibody, which recognizes an extracellular domain of PLP related to the large loop of normal PLP/DM20 (Jung et al., 1996), showed that the live jimpy immortalized cells did not react with this antibody. In contrast, live normal cells and both normal and jimpy permeabilized cells were positive for O10. The present observation confirms the previous conclusion regarding the abnormal insertion of jimpy PLP/DM20 into the plasma membrane (Jung et al., 1996; Thomson et al., 1997; Feutz et al., 2001). This abnormal insertion of jimpy PLP/DM20 into the plasma membrane may underlie the abnormal cell signaling in jimpy. Neither bFGF nor cAMP induced stimulation of cell proliferation and differentiation respectively (Fetuz et al., 1995, 2001). The defective membrane resting potential reported in this study and other defects such as elevated levels of intracellular Ca²⁺ (Knapp et al., 1999) and progesterone (Le Goascogne et al., 2000) may be also related to the abnormal insertion of jimpy PLP/DM20 into plasma membrane of oligodendrocytes.

Expression of Transfectant PLP/DM20

When normal PLP and DM20 were expressed after transfection in both immortalized normal or jimpy oligodendrocytes, the two proteins showed roughly similar distributions in the cell bodies and processes. However, more intense fluorescence for PLP/DM20 was observed in the cell body around the cell nucleus where the protein is processed through the endoplasmic reticulum (Schwob et al., 1985; Roussel et al., 1987; Lees, 1998). It is interesting to note that jimpy immortalized cells were able to translocate normal PLP and DM20 to the cell processes distant from the cell bodies. This observation suggests that the protein transportation machinery in jimpy immortalized oligodendrocytes allowed synthesis and translocation of these two proteins. Our observation is in agreement with previous observations in vivo that showed the presence of MBP in the scattered myelin sheaths in jimpy brain (Duncan et al., 1989) and uniform distribution of CA II in jimpy oligodendrocytes (Ghandour and Skoff, 1988).

The expression of jimpy PLP and jimpy DM20 after transfection of 158N cells or 158JP cells was detected in the whole cell body and in cell processes. Previous observations showed that jimpy PLP is transported from the cell body to the cell processes in vivo and in vitro. In jimpy brain, jimpy PLP was detected in some myelin sheaths present in jimpy brain using the same PLP-specific antibody used in the present study (Roussel et al., 1987) and also in jimpy oligodendrocyte processes in culture (Feutz et al., 1995). Our observations differ from several studies that report jimpy and other mutated PLP and DM20 proteins are abnormally transported and completely sequestered within cells, which may cause cell toxicity (Gow et al., 1994, 1998; Jung et al., 1996; Thomson et al., 1997). These studies were done in cell lines like Cos7 that are unrelated to oligodendrocytes and that do not normally process PLP. They showed that after cell transfection, abnormal accumulation of jimpy PLP gene products occurred in the perinuclear area and particularly in the endoplasmic reticulum. We cannot exclude the possibility of partial accumulation of jimpy PLP/DM20 in oligodendrocyte perikarya but, more importantly, detectable amounts are transported to the plasma membrane. This association with the plasma membrane most likely affects maturation of the plasma membrane as shown with electrophysiological properties. The capacity of both immortalized normal and jimpy oligodendrocytes to transport the jimpy PLP and DM20 may be explained by the oligodendrocytic origin of the cell lines. Oligodendrocytes are quite unique, compared to the other cell types, characterized by very efficient myelin protein transportation (Kim and Pfeiffer, 1999; Simons et al., 2000; Krämer et al., 2001). Even though the oligodendrocytes are immortalized with the large T-antigen, transportation of myelin proteins in immortalized oligodendrocytes has more significance to the biology of oligodendrocytes than cell lines not derived from oligodendrocytes. The cellular distribution of the fusion proteins confirms the transport of jimpy PLP/DM20 towered the plasma membrane while the live staining for O10 antibody of transfected normal and jimpy cells confirmed the abnormal insertion of this protein (Fig. 10).

Defect in cAMP Signaling and Expression of Normal and Jimpy PLP/DM20

The lack of dbcAMP stimulation in jimpy oligodendrocytes was previously reported (Feutz et al., 2001). After dbcAMP treatment, the normal oligodendrocyte line enhanced several-fold the levels of oligodendrocyte markers. In contrast, the jimpy oligodendrocyte line remained insensitive to similar treatment. The increased levels in oligodendrocyte markers is also associated with morphologic changes reflected by many emerging thin cell processes and rounded cell bodies (Feutz et al., 2001). The morphologic transformation of immortalized 158N normal oligodendrocytes and the absence of such transformation in 158JP jimpy cells were confirmed by the present data.

Cellular effects of cAMP elevation are multiple. In oligodendrocytes, the elevation of cAMP level resulted in acceleration of cell maturation (McMorris, 1983; Raible and McMorris, 1989; Sato-Bigbee and DeVries, 1996). Since the cAMP binding protein (CREB), an activator of cAMP inducible genes, did not show significant difference between normal and jimpy cells after stimulation (not shown), the ICER gene was selected because it exerts a direct and efficient negative feedback in cAMP signaling pathway (Molina et al., 1993; Folco and Koren, 1997; Krueger et al., 1999). ICER gene transcription plays a key role in this pathway and it is stimulated by the cAMP (Molina et al., 1993) but not by Ca²⁺ (Krueger et al., 1999) or by the membrane depolarization (Folco and Koren, 1997). The transient elevation of the transcription level of ICER may function to attenuate the cAMP response in normal immortalized oligodendrocytes but it fails to function in jimpy cells. The role of PLP in this pathway needs to be elucidated. Since jimpy immortalized cells showed higher levels of intracellular Ca²⁺ (Knapp et al., 1999), this high level might attenuate the stimulatory effect of cAMP on ICER. Thus, jimpy PLP/DM20 may modify the function of Ca²⁺ pumps leading to abnormal function of cAMP stimulation pathway. Alternatively, PLP may act itself as a membrane channel (Helynck et al., 1983) and jimpy PLP may alter the membrane functions directly.

The expression of normal PLP or DM20 after transfection of immortalized 158JP oligodendrocytes did not allow immortalized jimpy oligodendrocytes to restore their deficiency in cAMP stimulation. The jimpy mutation is dominant as previously reported (Nadon et al., 1994; Schneider et al., 1995), but the dominant negative character of the jimpy mutation did not prevent the expression of normal PLP or DM20 in jimpy cells. The weakly fluorescent jimpy cells as well as the strongly fluorescent ones remained insensitive to dbcAMP. Thus, the jimpy cells remained deficient for dbcAMP stimulation despite the presence of low and high level of normal PLP/DM20. More interestingly, when a high level of normal PLP and DM20 was induced by transfection with the vectors nPLP-EGFP and nDM20-EGFP, the differentiation of normal cells by the dbcAMP was reduced consequently. The highest level of PLP-EGFP or DM20-EGFP was found in cells that completely lost the differentiation potential. This observation at cellular level is in agreement with previous observations on the negative effects of PLP overexpression in oligodendrocytes in mice (Readhead et al., 1994; Ikenaka and Kagawa, 1995; Kagawa et al., 1995; Bradl et al., 1999).

Finally, the electrophysiological recordings of the resting plasma membrane potential showed the absence of a developmentally regulated negative shift in the membrane potential in jimpy cells compared to the normal native or immortalized oligodendrocytes. This result suggests that the absence of sensitivity of 158JP cells to cAMP is associated with the failure of jimpy cells to mature, most likely due to altered structure of the plasma membrane proteins in jimpy cells. The abnormal insertion of jimpy PLP/DM20 into plasma membrane may be the primary cause of such alterations. Preliminary results on membrane currents displayed by 158JP cells indicate that these cell types exhibit different ionic current pattern reported in native oligodendrocytes (Attali et al., 1997) or 158N cells. Moreover, because the death of oligodendrocytes may involve misfolded proteins and interactions of extracellular components with plasma membrane proteins (Casaccia-Bonnefil, 2000), our study shows an abnormal plasma membrane in jimpy oligodendrocytes that may also trigger and/or contribute to the apoptotic cascade.