SEARCH

SEARCH BY CITATION

Keywords:

  • albumin;
  • growth-associated protein-43;
  • metabolism;
  • neurons;
  • oleic acid;
  • protein kinase C

Abstract

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

Unlike in the adult brain, the newborn brain specifically takes up serum albumin during the postnatal period, coinciding with the stage of maximal brain development. Here we report that albumin stimulates oleic acid synthesis by astrocytes from the main metabolic substrates available during brain development. Oleic acid released by astrocytes is used by neurons for the synthesis of phospholipids and is specifically incorporated into growth cones. Oleic acid promotes axonal growth, neuronal clustering, and expression of the axonal growth-associated protein-43, GAP-43; all these observations indicating neuronal differentiation. The effect of oleic acid on GAP-43 synthesis is brought about by the activation of protein kinase C, since it was prevented by inhibitors of this kinase, such as H-7, polymyxin or sphingosine. The expression of GAP-43 was significantly increased in neurons co-cultured with astrocytes by the presence of albumin indicating that neuronal differentiation takes place in the presence of oleic acid synthesized and released by astrocytes in situ. In conclusion, during brain development the presence of albumin could play an important role by triggering the synthesis and release of oleic acid by astrocytes, which induces neuronal differentiation.

Abbreviations used
BSA

bovine serum albumin

DIV

days in vitro

DMEM

Dulbecco's modified Eagle's medium

FCS

fetal calf serum

GAP-43

growth-associated protein-43

H-7

1-(5-isoquinolinyl sulfonyl)-2-methyl piperazine

NGF

nerve growth factor

PVDF

polyvinylidene difluoride.

The presence of high concentrations of albumin in the brain and the CSF of newborn mammals, unlike the situation in adults, has been known for two decades (Trojan and Uriel 1979; Saunders and Mollgard 1981) but the physiological relevance of this phenomenon has been elusive. The presence of albumin in the brain may be due to the existence of a mechanism through which albumin is transferred from the blood to the brain and CSF that is active only in immature rats (Habgood et al. 1992; Ohsugi et al. 1992). In addition, the in situ synthesis of albumin can also contribute to the increased albumin levels found in the newborn brain (Dziegielewska et al. 1984). This protein has been found in neurons and glial cells during development (Mollgard et al. 1988) but not in the adult brain. Since the presence of albumin in the brain is developmentally regulated, it has been suggested that this protein could play an important role in neural cell differentiation (Saunders and Mollgard 1981).

Albumin has been shown to have important functions in brain cells, such as the regulation of astrocyte proliferation or the control of intracellular calcium levels (Nadal et al. 1995; Nuñez and García-Sancho 1996). It has also been reported that albumin regulates brain cell metabolism (Vicario and Medina 1992; Tildon 1993; Vicario et al. 1993; Tabernero et al. 1999). Thus, we have shown in primary cultured astrocytes that albumin strongly increases the flux of glucose and lactate through the pyruvate dehydrogenase-catalyzed reaction (Tabernero et al. 1999). This effect was dose-dependent and specific to albumin and was not mimicked by other proteins such as γ-globulin or compounds of similar molecular weight such as dextran. In addition, the increase in pyruvate dehydrogenase activity promoted by albumin was counteracted by fatty acids, suggesting that albumin increases pyruvate dehydrogenase activity by sequestrating free fatty acids and/or their CoA-derivatives (Tabernero et al. 1999).

Although glucose is the main substrate for the brain, lactate and ketone bodies also play an important role during development. Thus, immediately after delivery, plasma lactate concentrations increase sharply, reaching concentrations as high as 9 mm (Persson and Tunell 1971; Cuezva et al. 1980; Juanes et al. 1986). Under these circumstances, lactate becomes an important metabolic substrate for the brain (Arizmendi and Medina 1983; Fernández and Medina 1986; Vicario et al. 1991; Vicario and Medina 1992; for a review, see Medina et al. 1992). In addition, glucose from glycogen is released by astrocytes as lactate (Dringen et al. 1993), which is presumably used by neurons as a source of energy and carbon skeletons (Tabernero et al. 1993; Vicario et al. 1993). Additionally, lactate plays a role in metabolite trafficking between neurons and astrocytes (Magistretti et al. 1993; Poitry-Yamate et al. 1995). Once milk fat has provided precursors for the synthesis of ketone bodies, 3-hydroxybutyrate and acetoacetate are readily used as metabolic substrates for the brain throughout the suckling period (Robinson and Williamson 1980).

The growth-associated protein, GAP-43, is a marker of neuronal differentiation. This protein is located in growing axons where it is bound to the membrane, possibly by two acylated cysteine residues at positions 3 and 4 of the N-terminus (Skene and Virág 1989). The carboxyl terminal part of the protein contains an F motif that binds to F-actin, thus building the internal structure of axons (for a review, see: Oestreicher et al. 1997). GAP-43 expression is restricted to the nervous system during development and regeneration (for a review, see: Benowitz and Routtenberg 1997) and its transcriptional regulation is brought about by transcription factors belonging to the basis-helix-loop family (Kinney et al. 1996) while post-transcriptional regulation is brought about by nerve growth factor (NGF) which, by activating protein kinase C, confers stability to GAP-43 mRNA (Perrone-Bizzozero et al. 1993).

In this work, we show that albumin stimulates the synthesis and release of oleic acid from astrocytes. Oleic acid reaches neurons, where it behaves as a neurotrophic factor inducing neuronal differentiation.

Materials and methods

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

Materials

Dulbecco's modified Eagle's medium (DMEM), DMEM + Ham's F12, insulin, transferrin, progesterone, putrescine, selenium, T3, T4, dexamethasone, penicillin, streptomycin, poly-l-lysine, cytosine arabinoside, bovine serum albumin (fatty acid-free), oleic acid, anti-mouse Ig FITC, monoclonal antibodies against GAP-43, protease inhibitors, polyacrylamide, anti-mouse Ig biotin, avidin–peroxidase conjugate and the AEC substrate staining kit were purchased from Sigma-Aldrich Chemical Co. (Madrid, Spain). Fetal calf serum (FCS), DNase and trypsin were obtained from Boehringer Mannheim (Barcelona, Spain). d-[6–14C]glucose, L-[U-14C]lactate, [1(2)-14C]acetate, [1–14C]oleic acid and polyvinylidene difluoride (PVDF) membrane were purchased from New England Nuclear (Boston, USA) and D-[3–14C]3-hydroxybutyrate was from American Radiolabeled Chemicals Inc. (St Louis, USA). Trizol Reagent, agarose and BioPrime DNA labelling kit were from Gibco BRL, Life Technologies (Barcelona, Spain). Nitrocellulose sheets and Bio-Spin 30 chromatography columns were from Bio-Rad (Madrid, Spain). The cDNA encoding GAP-43 was kindly donated by Drs P. Skene and R. Mirsky (Curtis et al. 1992; Skene and Virág 1989). Films were Kodak Tmax 400 Asa. Autoradiographic emulsion K2 was from Ilford (Cheshire, UK). Other chemicals were purchased from Sigma-Aldrich Chemical Co. (Madrid, Spain) or Merck (Barcelona, Spain).

Animals

Albino Wistar rats, fed ad libitum on a stock laboratory diet (49.8% carbohydrates, 23.5% protein, 3.7% fat, 5.5% (w/v) minerals and added vitamins and amino acids), were used for the experiments. Rats were maintained on a 12-h light–dark cycle. Females with a mean weight of 250 g were caged with males overnight and conception was considered to occur at 01:00 h; this was verified the following morning by the presence of spermatozoa in the vaginal smears. For preparing neurons in primary culture, fetuses at 17.5 days of gestation were delivered by rapid hysterectomy after cervical dislocation of the mother. Postnatal day 1 newborn rats or 7-day-old rats were used to prepare astrocytes or oligodendrocytes in culture, respectively.

Cell culture

Neuron and astrocyte cultures were prepared and cultured in DMEM + 10% FCS as previously reported (Tabernero et al. 1993). When the effect of BSA + oleic acid was tested in the culture, neurons were grown in defined serum-free medium (1 : 1 mixture of DMEM and Ham's F12 supplemented with 5 µg/mL insulin, 100 µg/mL transferrin, 1 mm pyruvate, 50 U/mL penicillin and 37.5 U/mL streptomycin). Co-culture of neurons and astrocytes was performed by plating dissociated embryonic (17.5 days) brain cells on 14 DIV astrocytes. Oligodendrocytes were prepared from optic nerves as previously described (Raff et al. 1983) and were cultured in defined medium (1 : 1 mixture of DMEM and Ham's F12 supplemented with 5 µg/mL insulin, 100 µg/mL transferrin, 60 ng/mL progesterone, 16 µg/mL putrescine, 160 ng/mL selenium, 400 ng/mL T4, 10 ng/mL T3, 0.035% (w/v) BSA, 38 ng/mL dexamethasone, 50 U/mL penicillin and 37.5 U/mL streptomycin and 0.5% (FCS). Cells were plated onto poly-l-lysine (10 µg/mL)-coated Petri dishes at a density of 1.5 × 105 cells/cm2 for neurons or 1.0 × 105 cells/cm2 for astrocytes or oligodendrocytes.

Cell incubation

Cell incubations were carried out as previously described (Tabernero et al. 1993). Fourteen days in vitro (DIV) astrocytes, 3 DIV neurons or 5 DIV oligodendrocytes grown on Petri dishes (25 cm2) were incubated at 37°C for 1 h with 1.5 mL of Elliot buffer (Elliot 1969) (11 mm sodium phosphate, 122 mm NaCl, 4.8 mm KCl, 0.4 mm KH2PO4, 1.2 mm MgSO4 and 1.3 mm CaCl2; pH 7.4) containing cold and radiolabelled substrates in the absence or presence of BSA, as indicated.

Analysis of fatty acids and phospholipids by HPLC

Lipids were extracted from the incubation medium (fatty acids) or from cells (phospholipids) with a mixture of chloroform/methanol (2 : 1, v/v) as described by Folch et al. (1957). Fatty acids were analyzed as 4-bromophenacyl ester derivatives by a gradient-elution HPLC method as described by Elliott and Parkin (1989). Phospholipid species were separated by an isocratic HPLC method as described (Bolaños and Medina 1993; Tabernero et al. 1993). Fractions (0.5 mL) of the eluate were collected and the radioactivity measured by liquid scintillation counting.

Allocation of radioactive oleic acid incorporated into neurons

Neurons were cultured in defined medium in the presence of 2% (w/v) BSA plus 100 µm oleic acid in Petri dishes. After 3 days, 10 µCi of [1–14C]oleic acid was added per dish. After 1 h or 12 h, the dishes were washed several times with phosphate-buffered saline plus 2% (w/v) BSA and processed for autoradiography, as described (Levi et al. 1983). Briefly, neurons were fixed with 3.5% glutaraldehyde and covered with a thin layer of an autoradiographic emulsion. After 5 days at 4°C, the cells were developed and photomicrographs were taken.

Immunocytochemistry

Cells were fixed in 4% formaldehyde for 20 min and permeabilized with methanol for 10 min at − 20°C. Mouse monoclonal GAP-43 antibodies (1 : 1000) were applied for 1 h, followed by anti-mouse Ig FITC (1 : 100) for 1 h. Phase-contrast or fluorescence photographs were taken.

Western blotting

Proteins (90 µg) were extracted from the cells and applied to a 10% SDS–polyacrylamide gel under reducing conditions and transferred to a nitrocellulose sheet, as described (Morgan et al. 1991). The nitrocellulose sheet was then exposed for 12 h to monoclonal GAP-43 antibody (1 : 1000). Anti-mouse Ig biotin (1 : 1000) was applied for 4 h, followed by avidin–peroxidase conjugate (1 : 2000) for 1 h. The AEC substrate staining kit was applied for 10 min, prepared according to the manufacturer's instructions.

Northern blotting

Total RNA was extracted with Trizol reagent followed by RNA precipitation with isopropanol and purification with 75% ethanol according the manufacturer's instructions. The RNA (20 µg) was then run on an agarose/formaldehyde gel and transferred to a PVDF membrane for GAP-43 mRNA detection. The membrane was then hybridized with biotin-labelled cDNA at 65°C for 18 h. The cDNA encoding GAP-43 was excised from plasmid pGEM3 using EcoRI and was labelled with biotin-14-dCTP using the BioPrime DNA labelling kit. Unincorporated deoxyribonucleoside triphosphates were removed by chromatography through a Bio-Spin 30 chromatography column. For GAP-43 mRNA detection, the PVDF membrane was rinsed, washed and blocked according to the manufacturer's instructions. mRNA detection was carried out by exposing the membrane to avidin–peroxidase conjugate (1 : 2000) for 1 h and the AEC substrate staining kit for 10 min, following the manufacturer's instructions.

In situ hybridization

The biotin-labelled GAP-43 cDNA probe was used to detect GAP-43 mRNA in neuronal cultures, as described (Morgan et al. 1994). Cells were hybridized with 0.3 µg/mL biotin cDNA at 65°C for 18 h and washed several times in decreasing concentrations of standard saline citrate buffer. Biotin-labelled GAP-43 cDNA probe was detected as described for northern blotting.

Statistical analyses

Statistical analyses were carried out using Student's t-test.

Results

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

Astrocytes synthesize and release oleic acid in the presence of albumin

We have previously reported that lactate is used by astrocytes as a source of energy and carbon skeletons for lipid synthesis and that this process is enhanced in the presence of bovine serum albumin (BSA) (Tabernero et al. 1993; Vicario et al. 1993; Tabernero et al. 1999). In this work, we found that a large amount of the radioactivity incorporated into the lipophilic fraction under these circumstances was released into the incubation medium. In a preliminary analysis, TLC findings suggested that the radioactivity was incorporated into free fatty acids, which presumably remained soluble in the incubation medium because of their affinity to BSA (Spector and Fletcher 1978). The chloroform/methanol extract of the incubation medium was then derivatized and chromatographed by HPLC and eluate fractions were collected for radioactivity measurements. On the basis of the resolution times (Figs 1a and b), internal standard additions, and the UV/VIS spectra (results not shown) of the radioactive fractions (Fig. 1c), the fatty acid was identified as oleic acid. Similar experiments were carried out with neurons in culture but no detectable amount of oleic acid was found in the incubation medium (data not shown), suggesting that the release of oleic acid is an astrocyte-specific process.

image

Figure 1.  Albumin induces the synthesis of oleic acid in astrocytes from different substrates. (a) Fatty acid standards were analyzed by HPLC as described in Methods; retention times in parentheses. (b) Astrocytes were incubated in Elliot buffer containing 10 mm lactate, 100–150 dpm/nmol [U-14C]lactate and 2% (w/v) BSA for 1 h. Washed chloroform/methanol extracts of the incubation medium were derivatized and analyzed by HPLC to determine the presence of fatty acids. Retention times are shown in parentheses. (c) Radioactivity recovered in the eluate of these experiments. (d) Astrocytes were incubated as described above in the presence of increasing concentrations of BSA. The oleic acid synthesized was quantified as nmol of lactate transformed into oleic acid and the values shown are means ± SEM (n = 4–7). (e) Astrocytes were incubated in the presence of 5 mm glucose and 200–300 dpm/nmol [6-14C]glucose or 10 mm lactate and 100–150 dpm/nmol [U-14C]lactate or 5 mm 3-hydroxybutyrate (3-HB) and 200–300 dpm/nmol [1-14C]3-HB or 5 mm acetate and 200–300 dpm/nmol [1(2)-14C]acetate, in the absence or presence of 2% (w/v) BSA. The oleic acid synthesized in the medium was expressed as nmol of substrate transformed into oleic acid and values are means ± SEM (n = 4–7). ***p < 0.001 as compared with the absence of BSA.

Download figure to PowerPoint

BSA stimulated oleic acid synthesis from lactate in a dose-dependent way (Fig. 1d), the saturation effect being reached at about 2% (w/v) BSA. The effect of BSA on oleic acid synthesis was also observed when glucose, 3-hydroxybutyrate or acetate were used as substrates (Fig. 1e). It should be mentioned that lactate, glucose and 3-hydroxybutyrate are the main metabolic fuels for brain development (for a review, see: Medina et al. 1992). Likewise, the substrate concentrations used in these experiments were the maximum plasma concentrations found in vivo (for a review, see: Medina et al. 1992).

Oleic acid synthesized by astrocytes is incorporated into neuronal phospholipids

In order to investigate whether the oleic acid released by astrocytes was used by neurons or not, we took advantage of the fact that acetate is a poor substrate for neurons (results not shown; see also: Sonnewald et al. 1993) but a good precursor of oleic acid in astrocytes (Fig. 1e). Accordingly, astrocytes were incubated for 1 h in Elliot buffer containing 5 mm acetate + 200–300 dpm/nmol of [1(2)-14C]acetate and 2% (w/v) BSA. This astrocyte medium was then used to incubate neurons for 1 h. The incubation medium before exposure to astrocytes was used to incubate neurons as a control. The incorporation of radioactivity into neuronal phospholipids was followed by HPLC of chloroform/methanol lipid extracts, after which the eluate fractions were collected for liquid scintillation counting. As expected, the radioactivity from acetate was only poorly incorporated into neuronal phospholipids when the incubation medium was not preincubated with astrocytes (Fig. 2a). However, when the incubation medium had been preincubated with astrocytes, the putative oleic acid synthesized from acetate was actively incorporated into neuronal phosphatidylcholine and phosphatidylethanolamine (Fig. 2a). No incorporation into sphingomyelin was detectable (data not shown). The incorporation into total phospholipids was eight-fold higher with the preincubated medium in comparison with the nonpreincubated one.

image

Figure 2. Oleic acid synthesized by astrocytes is incorporated into neuronal phospholipids. (a) Astrocytes were incubated for 1 h in Elliot buffer containing 5 mm acetate + 200–300 dpm/nmol of [1(2)-14C]acetate and 2% (w/v) BSA and this incubation medium, was then used to incubate neurons for 1 h (preincubated; ▪). The incubation medium before exposure to astrocytes was used to incubate neurons for 1 h as a control (non-preincubated; □). Chloroform/methanol extracts from neurons were analyzed by HPLC and the eluate followed for radioactivity. ***p < 0.001 as compared with nonpreincubated medium. (b) Neurons, astrocytes or oligodendrocytes were incubated in Elliot buffer containing 5 mm glucose, 100 µm oleic acid and 9000–12000 dpm/nmol [1-14C]oleic acid in the presence of 2% (w/v) BSA and the incorporation into phospholipids (PL) was followed as described above. ***p < 0.001; **p < 0.01 as compared with neurons. (c) Neurons were incubated in Elliot buffer containing 5 mm glucose, 100 µm oleic acid and 9000–12000 dpm/nmol [1–14C]oleic acid in the presence of 2% (w/v) BSA (▪) or 0.01% (v/v) DMSO (□). **p < 0.01; *p < 0.05 as compared with in the presence of BSA. (d) Neurons were incubated in Elliot buffer containing 5 mm glucose, 100 µm oleic acid and 9000–12000 dpm/nmol [1-14C]oleic acid and increasing concentrations of BSA. Results are expressed in dpm or in nmol of oleic acid incorporated into phosphatidylethanolamine (PE), phosphatidylcholine (PC) or total phospholipids and are means ± SEM (n = 4).

Download figure to PowerPoint

To evaluate oleic acid incorporation into neuronal lipids, neurons were exposed to authentic 100 µm oleic acid plus 9000–12000 dpm/nmol [1-14C]oleic acid and the radioactivity incorporated into phospholipids was analyzed by HPLC. Cultured astrocytes and oligodendrocytes were used as controls. Oleic acid was incorporated into neuronal phosphatidylethanolamine and phosphatidylcholine, while no incorporation into the sphingomyelin fraction was detectable (Fig. 2b). The rates of oleic acid incorporation into phospholipid species were much lower in oligodendrocytes and astrocytes than in neurons, except those observed for sphingomyelin synthesis, which were only detectable in oligodendrocytes. The rate of phospholipid synthesis from oleic acid in neurons was about seven-fold and three-fold higher than the rates found in astrocytes and oligodendrocytes, respectively (Fig. 2b).

The incorporation of oleic acid into neuronal phosphatidylethanolamine and phosphatidylcholine depends on the concentration of BSA present in the incubation medium, a maximum being reached at about 1.5% (w/v) of BSA (Fig. 2d). To determine whether the effect of BSA was due to the ability of BSA to act as a vehicle for oleic acid, neurons were exposed to oleic acid diluted in DMSO. Under these circumstances, the incorporation of oleic acid into neuronal phospholipids was very low (Fig. 2c).

In an attempt to localize the newly synthesized phospholipids, neurons were thoroughly washed after 1 h or 12 h of incubation with radiolabelled oleic acid and, after exposure to an autoradiographic emulsion, photomicrographs were taken. Neurons but not astrocytes or oligodendrocytes were stained (Fig. 3a), suggesting that oleic acid is mainly incorporated into neurons. Radiolabelled oleic acid was concentrated at the sites of newly emerging neurites, while cell bodies were poorly stained (Figs 3b and c).

image

Figure 3.  Radiolabelled oleic acid is found at the base of growing neurites. Neurons cultured in defined medium in the presence of 2% (w/v) BSA and 100 µm oleic acid for 3 days were exposed for 1 or 12 h to 10 µCi/mL of [1-14C]oleic acid. After washing with 2% (w/v) BSA, cells were fixed and exposed to an autoradiographic emulsion, developed and photographed as described in methods. After 1 h, radioactivity was seen to be preferentially located in the neuronal plasma membrane at the base of growing neurites (b). This effect was much more evident after 12 h of exposure to radioactivity, when the clusters of neurons were surrounded by radioactivity (c). Note that neither astrocytes nor oligodendrocytes [labelled in (a) as a and o, respectively] were stained. Bar = 70 µm.

Download figure to PowerPoint

Oleic acid synthesized by astrocytes promotes neuronal differentiation and GAP-43 expression

Because oleic acid was specifically incorporated into the sites of neuronal sprouting, the effect of oleic acid on axonal growth was measured by culturing neurons in defined medium in the presence of BSA plus oleic acid. BSA or BSA plus palmitic acid were used as controls. Axon growth was also studied by following the expression of GAP-43, a protein marker of axonal differentiation (Skene 1989). Accordingly, neurons were fixed and labelled with antibodies to GAP-43. In the presence of oleic acid, the cell bodies of neurons aggregated together, projecting thin, well-defined axons, characterized by GAP-43-immunoreactivity, which reached neighboring neuronal groups. Axon length was doubled in the presence of oleic acid (Fig. 4e). Phase and fluorescence photomicrographs showed that the increase in axon length caused by the presence of oleic acid was accompanied by an enhancement of GAP-43 expression (Figs 4a–d). In addition, western blot analyses (Fig. 4f) revealed that the concentration of immunoreactive GAP-43 was significantly increased in neurons exposed to oleic acid (Fig. 4f). Moreover, the observed increase in GAP-43 protein was due to an enhancement of the synthesis of the protein because GAP-43 mRNA was also increased in these circumstances (Fig. 5). Thus, northern blot analyses revealed a significant increase in the GAP-43 mRNA concentration in the presence of oleic acid (Fig. 5a). Furthermore, in situ hybridization analysis showed that in the presence of oleic acid GAP-43 mRNA levels increased in most of the neurons (Fig. 5b). Since oleic acid activates protein kinase C (Khan et al. 1992) and since the stability of GAP-43 mRNA depends on protein kinase C activity (Perrone-Bizzozero et al. 1993), we investigated the effects of several inhibitors of protein kinase C on the enhancement of GAP-43 mRNA levels caused by the presence of oleic acid. The presence of inhibitors of protein kinase C, such as H-7, polymyxin or sphingosine (Perrone-Bizzozero et al. 1993), prevented the increase in GAP-43 mRNA due to the presence of oleic acid (Figs 5b and c).

image

Figure 4. Oleic acid induces neuronal differentiation characterized by the enhancement of GAP-43 expression. Neurons were cultured for 3 days in defined medium in the absence or presence of 2% (w/v) BSA, 2% (w/v) BSA + 100 µm oleic acid, or 2% (w/v) BSA + 100 µm palmitic acid. (a–d) Neurons were fixed and processed for immunocytochemistry for GAP-43 as described in Materials and methods. Phase-contrast (left panels) and fluorescence photographs (right panels) were taken of the same field. Note the enhancement of immunofluorescence and differentiation observed in neurons cultured in the presence of BSA + oleic acid. (e) Axon length was quantified in digitized photographs by measuring the length of the GAP-43 immunoreactive neurites with the NIH Image software. The results are means ± SEM (n = 40). ***p < 0.001 as compared with BSA + oleic acid. (f) Western blot analysis for GAP-43. Lanes are labelled according to the culture additions. Newborn brain tissue was used as a positive control. Bar = 85 µm.

Download figure to PowerPoint

image

Figure 5.  Oleic acid increases neuronal GAP-43 mRNA by a protein kinase C-dependent mechanism. Neurons were cultured for 2 days in defined medium in the absence or in the presence of 2% (w/v) BSA, 2% (w/v) BSA + 100 µm oleic acid or 2% (w/v) BSA + 100 µm palmitic acid. For the experiments with the protein kinase C inhibitors, cells were pretreated for 30 min with 50 µm H-7, 250 U/mL polymyxin B or 2 µm sphingosine before the addition of BSA or BSA + oleic acid and the inhibitors were maintained up to the end of the experiment. (a) Northern blot analysis of GAP-43 mRNA. Lanes are labelled according to the culture additions. RNA from newborn brain was used as a positive control. RNA loadings are shown in the lower panel, visualized using ethidium bromide. (b) The percentages of GAP-43 mRNA-positive cells assessed by in situ hybridization are expressed as means ± SEM (n = 3); ***p < 0.001 as compared with BSA. (c) Northern blot analysis of GAP-43 mRNA in the presence of protein kinase C inhibitors. Lanes are labelled according to the culture additions. RNA loadings are shown in the lower panel, visualized using ethidium bromide.

Download figure to PowerPoint

To investigate whether the neuronal differentiation promoted by oleic acid would take place in the presence of oleic acid synthesized and released by astrocytes in situ, neurons were co-cultured with astrocytes and GAP-43 synthesis was followed. Since our results had shown that astrocytes synthesize and release oleic acid only in the presence of BSA (Fig. 1), we expected that the presence of BSA in the co-culture would enhance neuronal differentiation. Accordingly, co-culture of astrocytes and neurons was performed in the absence or presence of BSA and immunocytochemistry, western and northern blot analyses of GAP-43 were carried out (Fig. 6). The expression of GAP-43 was significantly increased in the presence of BSA as compared with the co-culture in the absence of BSA (Figs 6c–f). However, the level of GAP-43 was not significantly modified by the presence of BSA in neurons or astrocytes cultured apart (Figs 6a and b), indicating that the effect of BSA only takes place when neurons and astrocytes are co-cultured.

image

Figure 6. Albumin induces GAP-43 expression in the co-culture of neurons and astrocytes. Fourteen DIV astrocytes, neurons or astrocyte–neuron co-cultures were grown for two additional days in defined medium in the absence or presence of 2% (w/v) BSA. (a and b) GAP-43 immunocytochemistry of astrocytes or neurons grown in defined medium in the presence of 2% (w/v) BSA. (c and d) GAP-43 immunocytochemistry of astrocyte–neuron co-culture grown in defined medium in the absence or in the presence of 2% (w/v) BSA. (e) Western blot analysis. Lanes are labelled according to the source of the protein, and culture additions. Newborn brain was used as a positive control. (f) Northern blot analysis. GAP-43 mRNA was visualized using a biotin-labelled probe, as described in Methods. Lanes are labelled according to the source of the mRNA, and culture additions. Bar = 70 µm.

Download figure to PowerPoint

Discussion

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

Although glucose is the main metabolic substrate for the brain, ketone bodies and lactate may also play an important role as a source of energy and carbon skeletons for the developing brain. In both humans and rats, brain development mainly takes place in the postnatal period. Under these circumstances, the ketone bodies synthesized by the newborn liver from milk fat are used by the brain to fulfil most of the energy and carbon requirements of the tissue (Robinson and Williamson 1980). However, immediately after birth, when glucose is scarce and ketone bodies are not yet available, lactate is used by the brain not only as a source of energy but also as a precursor for lipid synthesis (for a review, see: Medina et al. 1992). We have recently shown that albumin, a serum protein that is specifically taken up by the brain during development (Habgood et al. 1992; Ohsugi et al. 1992; Saunders and Mollgard 1981; Trojan and Uriel 1979), actively enters astrocytes and significantly enhances glucose and lactate utilization (Tabernero et al. 1999), suggesting that albumin may regulate substrate utilization by astrocytes. In this paper, we show that BSA increases oleic acid synthesis from lactate, glucose, acetate and 3-hydroxybutyrate (Fig. 1e) in a dose-dependent way (Fig. 1d). Therefore, our results suggest that during development astrocytes use metabolic substrates to synthesize oleic acid which is released to the extracellular medium. This prompted us to investigate the destination of oleic acid.

Although the oleic acid released by astrocytes is also used  by oligodendrocytes to synthesize sphingomyelin, our results suggest that the main destination of the oleic acid is neurons, where it is preferentially incorporated into phosphatidylcholine and phosphatidylethanolamine (Fig. 2b). Since the incorporation of oleic acid into neuronal structures depends on the presence of albumin (Figs 2c and d), it is feasible to propose that albumin could be used as a vehicle for the transport of oleic acid between astrocytes and neurons. If so, albumin would stimulate oleic acid synthesis in the astrocyte and would then transport it to neurons and oligodendrocytes.

Under our experimental conditions, oleic acid was the only fatty acid synthesized by astrocytes (Figs 1b and c), suggesting that this phenomenon has a specific purpose. In fact, the single double-bond of oleic acid is enough to increase sharply the fluidity of biological membranes (Alberts et al. 1994). Because the membrane fluidity is very critical for neurons, the incorporation of oleic acid-borne phospholipids into a discrete area of the membrane could substantially change membrane properties. In agreement with this, oleic acid was preferentially incorporated into neurite bases (Fig. 3), suggesting that an increased fluidity is required at the sites of newly emerging axons and/or dendrites. This would facilitate the sprouting of the membrane during neurite growth together with an enhanced flexibility for axon orientation.

GAP-43 is conspicuously present during brain development but its content decreases sharply in adult life, when the presence of GAP-43 is constrained to high-plasticity neuronal regions or certain exclusive synapses during long-term potentiation (Skene and Virág 1989). Consequently, GAP-43 may play an important role in neuronal differentiation. In support of this, our results show that the presence of oleic acid significantly increases the synthesis of GAP-43, which is distributed along axonal structures (Fig. 4). Indeed, the presence of oleic acid led to the aggregation of neurons in the typical gray/white-matter fashion observed in vivo. In addition, the presence of oleic acid elongated axons, which contacted other neurons, thus mimicking the neuronal networks observed in the CNS. This phenomenon was accompanied by an increase in the synthesis of GAP-43, which may play an important role in axonal build-up. It is therefore reasonable to propose that oleic acid promotes neuronal differentiation in culture.

Although albumin is unable to cross the blood brain barrier during adult life, there is convincing evidence supporting the notion that albumin enters the CNS during the postnatal development of nervous tissue (Saunders and Mollgard 1981). This phenomenon coincides with the increase in blood and brain albumin concentrations that occurs during the postnatal period (Dziegielewska et al. 1981). Because the greater part of brain development, particularly neuronal differentiation, occurs postnatally in nonprecocious species (Bayer 1995) such as humans and rats, it can be suggested that albumin would play an important role in brain development. Consistent with this, our results show that the presence of albumin alone is able to sustain neuronal differentiation in co-culture of astrocytes and neurons (Fig. 6). This suggests that the oleic acid synthesized and released by astrocytes in the presence of albumin causes neuronal differentiation by promoting axonal growth and neuronal clustering.

It should be mentioned that oleic acid and NGF may share a common mechanism in the regulation of GAP-43 synthesis. In this context, oleic acid and NGF have been shown to activate protein kinase C (Khan et al. 1992; Perrone-Bizzozero et al. 1993), which enhances GAP-43 expression by conferring stability to GAP-43 mRNA (Perrone-Bizzozero et al. 1993). In agreement with this, our results suggest that the effect of oleic acid on GAP-43 synthesis is brought about by the activation of protein kinase C. Thus, the presence of inhibitors of protein kinase C such as H-7, sphingosine or polymyxin (Perrone-Bizzozero et al. 1993) prevented the enhancement of GAP-43 mRNA caused by oleic acid (Fig. 5). Therefore, it may be suggested that albumin promotes oleic acid synthesis by astrocytes, thus collaborating with NGF in neuronal differentiation in a coordinated fashion that persist as long as the passage of albumin to the brain is allowed. Such NGF–oleic acid collaboration would later decay coinciding with the plunge in oleic acid synthesis caused by the cessation of the passage of albumin to the brain. In addition, it has been reported that albumin enters the adult brain under hypoxic conditions (Plateel et al. 1997) and after insult to the blood–brain barrier. This tempts us to speculate that albumin may also play a role in the neural repair response that follows such insults.

It should further be mentioned that Down syndrome patients show lower blood albumin concentrations than their healthy counterparts (Nelson 1961), a fact that is accompanied by decreased levels of monounsaturated fatty acids in brain phospholipids (Shah 1979). These biochemical alterations may be associated with the decrease in neuronal interconnection frequency observed in such patients (Elul et al. 1975). Whether the deficit in blood albumin during brain development may be responsible for the decrease in oleic acid concentrations in the brain, resulting in the plunge in intelligence quotient (IQ) observed in these patients after birth (Morgan 1979), remains to be elucidated.

Acknowledgements

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

This work was supported by FIS, DGES and the Junta de Castilla y León, Spain. E. Lavado was a recipient of a fellowships from the University of Salamanca and A. Velasco is a recipient of a BEFI fellowship from the Fondo de Investigaciones Sanitarias. We thank Professors R. Mirsky and P. Skene for GAP-43 plasmid. We also thank Professor R. Mirsky for helping with the discussion. We are grateful for the technical assistance of T. del Rey and thank N. Skinner for help in writing the manuscript.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Alberts B., Bray D., Lewis J., Raff M., Roberts K., Watson J. D. (1994) Molecular Biology of the Cell. Garland Publishing Inc, New York.
  • Arizmendi C. & Medina J. M. (1983) Lactate as an oxidizable substrate for rat brain in vivo during perinatal period. Biochem. J. 214, 633635.
  • Bayer S. A. (1995) Neurogenesis and neuronal migration, in The Rat Nervous System (Paxinos, G., ed.), pp. 10411076. Academic Press, Sydney.
  • Benowitz L. I. & Routtenberg A. (1997) GAP-43: an intrinsic determinant of neuronal development and plasticity. TINS 20, 8491.
  • Bolaños J. P. & Medina J. M. (1993) Lipogenesis from lactate in fetal rat brain during late gestation. Pediatr. Res. 33, 6671.
  • Cuezva J. M., Moreno F. J., Medina J. M., Mayor F. (1980) Prematurity in the rat. I. Fuels and gluconeogenic enzymes. Biol. Neonate 37, 8895.
  • Curtis R., Stewart H. J. S., Hall S. M., Wilkin G. P., Mirsky R., Jessen K. R. (1992) GAP-43 is expressed by nonmyelin-forming Schwann cells of the peripheral nervous system. J. Cell Biol. 116, 14551464.
  • Dringen R., Gebhardt R., Hamprecht B. (1993) Glycogen in astrocytes, possible function as lactate supply for neighboring cells. Brain Res. 62, 208214.
  • Dziegielewska K. M., Evans C. A. N., Lai P. C. W., Lorscheider F. L., Malinowska D. H., Mollgard K., Saunders N. R. (1981) Proteins in cerebrospinal fluid and plasma of fetal rats during development. Dev. Biol. 83, 193200.
  • Dziegielewska K. M., Evans C. A. N., New H., Reynolds M. L., Saunders N. R. (1984) Synthesis of plasma proteins by rat fetal brain and choroid plexus. Int. J. Dev. Neurosci. 2, 215222.
  • Elliott J. M., B. H., Parkin K. L. (1989) An improved liquid chromatographic method for the quantitative determination of the free fatty acids in milk products. J. Dairy Sci. 72, 24782482.
  • Elliot K. A. C. (1969) The use of brain slices, in Handbook of Neurochemistry, Vol. 2 (Lajtha, A., ed.), pp. 103114. Plenum Press, New York.
  • Elul R., Hanley J., Simmons J. I. (1975) Non-Gaussian behavior of the EEG in Down's syndrome suggests decreased neuronal connections. Acta Neurol. Scand. 51, 2128.
  • Fernández E. & Medina J. M. (1986) Lactate utilization by the neonatal rat brain in vitro. Competition with glucose and 3-hydroxybutyrate. Biochem. J. 234, 489492.
  • Folch J., Lees M., Sloane-Stanley G. H. (1957) A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226, 497509.
  • Habgood M. D., Sedgwick J. E. C., Dziegielewska K. M., Saunders N. R. (1992) A developmentally regulated blood-cerebrospinal fluid transfer mechanism for albumin in immature rats. J. Physiol. (Lond.) 456, 181192.
  • Juanes M. C., Arizmendi C., Medina J. M. (1986) Attenuation of postnatal hypoxia in the premature newborn rat by maternal treatment with dexamethasone: its relationship with lung phospholipid content. Biol. Neonate 50, 337344.
  • Khan W. A., Blobe G. C., Hannun Y. A. (1992) Activation of protein kinase C by oleic acid. J. Biol. Chem. 267, 36053612.
  • Kinney W. R., McNamara R. K., Valcourt E., Routtenberg A. (1996) Prolonged alteration in E-box binding after a single systemic kainate injection: potencial relation to F1/GAP-43 gene expression. Mol. Brain Res. 38, 2536.
  • Levi G., Wilkin G. P., Ciotti M. T., Johnstone S. (1983) Enrichment of differentiated, stellate astrocytes in cerebellar interneuron cultures as studied by GFAP immunofluorescence and autoradiographic uptake patterns with [3H]-aspartate and [3H]-GABA. Brain Res. 10, 227241.
  • Magistretti P., Sorg O. YuN., Martin J., Pellerin L. (1993) Neurotransmitters regulate energy metabolism in astrocytes: implications for the metabolic trafficking between neural cells. Dev. Neurosci. 15, 306312.
  • Medina J. M., Vicario C., Juanes M., Fernández E. (1992) Biochemical adaptations to early extrauterine life, in Perinatal Biochemistry (Herrera, E. and Knopp, R., eds), pp. 233258. CRC Press, Boca Raton, FL.
  • Mollgard K., Dziegielewska K. M., Saunders N. R., Zakut H., Soreq H. (1988) Synthesis and localization of plasma proteins in the developing human brain. Dev. Biol. 128, 207221.
  • Morgan S. (1979) Development and distribution of intellectual and adaptative skills in Down syndrome children: implications for early intervention. Ment. Retard. 17, 247249.
  • Morgan L., Jessen K. R., Mirsky R. (1991) The effects of cyclic AMP on differentiation of cultured Schwann cells: progression from an early phenotype (04+) to a myelin phenotype (Po+, GFAP-, N-CAM+, NGF-receptor–) depends on growth inhibition. J. Cell Biol. 112, 457467.
  • Morgan L., Jessen K. R., Mirsky R. (1994) Negative regulation of the Po gene in Schwann cells: suppression of Po mRNA and protein induction in cultured Schwann cells by FGF2 and TGFβ1, TGFβ2 and TGFβ3. Development 120, 13991409.
  • Nadal A., Fuentes E., Pastor J., McNaughton P. A. (1995) Plasma albumin is a potent trigger of calcium signals and DNA synthesis in astrocytes. Proc. Natl. Acad. Sci. USA 92, 14261430.
  • Nelson T. L. (1961) Serum protein and lipoprotein fractions in mongolism. Am. J. Dis. Child. 102, 369374.
  • Nuñez L. & García-Sancho J. (1996) Two different constituents of plasma increase cytosolic calcium selectively in neurons or glia of primary rat cerebellar cultures. J. Physiol. 490, 577583.
  • Oestreicher B. A., De Graan P. N. E., Gispen H. W., Verhaagen J., Schrama L. H. (1997) B-50, the growth associated protein-43: modulation of cell morphology and communication in the nervous system. Prog. Neurobiol. 53, 627686.DOI: 10.1016/s0301-0082(97)00043-9
  • Ohsugi M., Sato H., Yamamura H. (1992) Transfer of I-125-albumin from blood to brain in newborn rats and the effect of hyperbilirubinemia on the transfer. Biol. Neonate 62, 4754.
  • Perrone-Bizzozero N. I., Cansino V. V., Kohn D. T. (1993) Post-transcriptional regulation of GAP-43 gene expression in PC12 cells through protein kinase C-dependent stabilization of the messenger RNA. J. Cell Biol. 120, 12631270.
  • Persson B. & Tunell R. (1971) Influence of environmental temperature and acidosis on lipid mobilization in the human infant during the first two hours after birth. Acta Paediatr. Scand. 60, 385398.
  • Plateel M., Teissier E., Cecchelli R. (1997) Hypoxia dramatically increases the nonspecific transport of blood-borne proteins to the brain. J. Neurochem. 68, 874877.
  • Poitry-Yamate C. L., Poitry S., Tsacopoulos M. (1995) Lactate released by Muller glial cells is metabolized by photoreceptors from mammalian retina. J. Neurosci. 15, 51795191.
  • Raff M. C., Miller R. H., Noble M. (1983) A glia progenitor cell that develops in vitro into an astrocyte or an oligodendrocyte depending on culture medium. Nature 303, 390396.
  • Robinson A. M. & Williamson D. H. (1980) Physiological roles of ketone bodies as substrates and signals in mammalian tissues. Physiol. Rev. 60, 143187.
  • Saunders N. R. & Mollgard K. (1981) The natural internal environment of the developing brain beyond the barrier. TINS 4, 5660.
  • Shah S. N. (1979) Fatty acid composition of lipids of human brain myelin and synaptosomes: changes in phenylketonuria and Down's syndrome. Int. J. Biochem. 10, 477482.
  • Skene J. H. P. (1989) Axonal growth-associated proteins. Ann. Rev. Neurosci. 12, 127156.
  • Skene J. H. P. & Virág I. (1989) Posttranslational membrane attachment and dynamic fatty acylation of a neuronal growth cone protein, GAP-43. J. Cell Biol. 108, 613624.
  • Sonnewald U., Westergaard N., Schousboe A., Svendsen J. S., Unsgard G., Petersen S. B. (1993) Direct demonstration by <C-13>NMR spectroscopy that glutamine from astrocytes is a precursor for GABA synthesis in neurons. Neurochem. Int. 22, 1929.
  • Spector A. A. & Fletcher J. E. (1978) Transport of fatty acid in the circulation, in Disturbances in Lipid and Lipoprotein Metabolism (Dietschy, J. M., Gotto, A. M. and Ontko, J. A., eds), pp. 229249. American Physiological Society, Bethesda, MD.
  • Tabernero A., Bolaños J. P., Medina J. M. (1993) Lipogenesis from lactate in rat neurons and astrocytes in primary culture. Biochem. J. 294, 635638.
  • Tabernero A., Medina A., Sánchez-Abarca L. I., Lavado E., Medina J. M. (1999) The effect of albumin on astrocyte energy metabolism is not brought about through the control of intercellular Ca2+ concentrations but by free-fatty acid sequestration. Glia 25, 19.DOI: 10.1002/(sici)1098-1136(19990101)25:1<1::aid-glia1>3.0.co;2-2
  • Tildon J. T. (1993) Differential effects of serum proteins on substrate oxidation by isolated synaptosomes and cultured brain astrocytes. Dev. Neurosci. 15, 226232.
  • Trojan J. & Uriel J. (1979) Localisation intracellulaire de l'alpha-fetoprotéine et de la sérumalbumine dans le système nerveux central du rat au cours du developpement foetal et postnatal. C. R. Acad. Sci. Paris 289, 11571160.
  • Vicario C. & Medina J. M. (1992) Metabolism of lactate in the rat brain during the early neonatal period. J. Neurochem. 58, 3240.
  • Vicario C., Arizmendi C., Malloch J. G., Clark J. B., Medina J. M. (1991) Lactate utilization by isolated cells from early neonatal rat brain. J. Neurochem. 57, 17001707.
  • Vicario C., Tabernero A., Medina J. M. (1993) Regulation of lactate metabolism by albumin in rat neurons and astrocytes from primary culture. Pediatr. Res. 34, 709715.