©Copyright 1992 by The Humana Press Inc. All rights of any nature whatsoever reserved. 1044-7393!92!1601-02-0033 $02.40
Primary Cultures of Astrocytes from Rat as a Model for Biotin Deficiency in Nervous Tissue PILAR RODRIGUEZ-POMBO, 1 LAWRENCE SWEETMAN, 2 AND MAGDALENA UGARTE* I 'Centro de Biologia Molecular, Facultad de Ciencias, Universidad Autonoma, 28049 Madrid, Spain; and 2 Biochemical Genetics Laboratory, Childrens Hospital of Los Angeles, Los Angeles, CA 90027 Received March 6, 1991; Accepted August 7, 1991
ABSTRACT The activities and biotin-dependence of the three mitochondrial biotin-dependent carboxylases: pyruvate carboxylase, propionyl CoA carboxylase, and f3-methylcrotonyl CoA carboxylase of primary culture of astrocytes have been examined. An increase of the three mitochondrial carboxylase activities was observed during cell growth, as was the case for developing rat brain. Mitochondrial carboxylase activities from 3-wk-old primary cultures of astrocytes were higher than those in the neonatal rat brain. When astrocytes were grown in a 10% serumenriched medium supplemented with avidin to bind biotin, the mitochondrial carboxylase activities were reduced to 15% of control value. Consistent with these results, after 3 wk in culture, the 3-hydroxyisovaleric acid concentration in the growth medium was tenfold higher than the controls. In this culture condition, cellular growth and the nonbiotin-dependent enzyme, glutamine synthetase, were not modified with respect to control. Primary cultures from newborn rat brain hemispheres are suggested as an experimental approach to the study of biotin deficiency in nervous tissue.
Index Entries: Biotin deficiency; primary culture of astrocytes; mitochondrial carboxylase activities; multiple carboxylase deficiency (MCD); 3-hydroxyisovaleric acid excretion. *A u th or to whom all correspondence and reprint requests should be addressed. Molecular and Chemical Neuropathology
33
Vol. 16, 1992
Rodriguez-Pombo, Sweetman, and Ugarte
34
Abbreviations: 3-HIVA: 3-Hydroxyisovaleric acid; GFAP: glial fibrillary acidic protein; GS: glutamine synthetase; MCD: multiple carboxylase deficiency; ,3MCC: i3-methylcrotonyl CoA carboxylase; PC: pyruvate carboxylase; PCC: propionyl CoA carboxylase.
INTRODUCTION The important role of biotin in human physiology has been highlighted by the recognition of two inborn errors of biotin metabolism in humans. The biotin-responsive multiple carboxylase deficiencies are inherited disorders of metabolism, in which a characteristic pattern of organic acids appears in the urine of patients as a consequence of a deficiency of the biotin-dependent carboxylases. There are four biotin-dependent enzymes in mammals: propionyl CoA carboxylase (PCC: EC 6.4.1.3), pyruvate carboxylase (PC: EC 6.4.1.1.), /3-methylcrotonyl CoA carboxylase ((3MCC: EC 6.4.1.4), and acetyl CoA carboxylase (ACC: EC 6.4.1.2). All but acetyl CoA carboxylase are located in the mitochondrial subcellular compartment. The most prominent features of biotin deficiency in humans are those affecting the skin and the nervous system. However, the obvious difficulty to carry out biochemical studies in CNS from patients have led to the use of experimental models (Sander et al., 1982; Schrijver et al., 1979). Because the brain contains different metabolic compartments, which at least partly correspond to different cell types (Balazs et al., 1973), we found it worthwhile to study the biotin-dependent enzymes in selected types of cells in culture to establish an easier model for studying biotin metabolism in nervous tissue than using biotin-deficient rats. The development of techniques to grow cerebral tissue in primary cultures (Booher and Sensenbrenner, 1972) has facilitated biochemical investigations of specific cerebral cell types. Primary cultures of astrocytes have been particularly well characterized (Hansson et al., 1980, 1982) and described as potential tools in neurochemical research (Schousboe, 1980). The large content of glial fibrillary acidic protein (GFAP) and the high activity of glutamine synthetase activity (GS, EC. 6.3.1.2) have been used as markers of their astrocytic nature (Bock et al., 1977; Hertz et al., 1985; Norenberg and Martinez-Hernandez, 1979). Immunohistochemical (Shank et al., 1985) and enzymatic (Yu et al., 1983) studies have shown an astrocytic localization of pyruvate carboxylase. Therefore, we used primary cultures of astrocytes as a model to study biotin metabolism in the nervous system. The aim of the present investigation was to study the mitochondrial carboxylases in astrocytes and to define a model of biotin deficiency in these cells.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Biotin-Restricted Astrocytes for MCD
35
METHODS Cell Culture Primary astroglial cell cultures were prepared by aseptically isolating cerebral hemispheres of newborn Wistar rats according to a modification of a previously described method (Booher and Sensenbrenner, 1972). The tissue was freed of meninges and capillary vessels. Mechanical dissociation was achieved by sequential passage through two needles of decreasing diameter (0.9 and 0.7 mm) in the presence of 20% serum-enriched medium (modified Eagle's Minimum Essential Medium containing 2 mM glutamine, 200 000 U/L penicillin, 100 mg/L streptomycin, and 20% (v/v) heat-inactivated fetal calf serum). The resulting cell suspension was seeded into six-well trays (Costar), at a density of 2.5 x 10 5 cells/cm 2 . The culture dishes were incubated at 37°C with a humidified atmosphere of 5% CO2 in air. The medium was changed on d 2 and subsequently every 3-4 d. The serum content was reduced to 10% at the end of the first week. The cells were harvested on d 21 by rinsing twice with sterile PBS and incubation with a trypsin-EDTA solution. At this time the cells were positive for GFAP as was demonstrated by fluorescence microscopy, and were therefore judged to be astrocytes (Bock et al., 1977).
Biotin Deficient Status -
Cells were grown for 7 d under routine culturing condition. The cultures were then divided into two groups and cultured for an additional 14 d. Group I (control) cultures were maintained on 10% serum-enriched medium. Group II (biotin-restricted) cultures were divided into two categories: (a) 10% serum-enriched medium with 50 U/L avidin; (b) 10% dialyzed serum-enriched medium (modified Eagle's Minimum Essential Medium containing 2 mM glutamine, 200 000 U/L penicillin, 100 mg/L streptomycin, and 10% (v/v) heat-inactivated dialyzed fetal calf serum from GIBCO). Growth curves were obtained by removing cells from dishes at various stages of culture, staining with a trypan blue solution, and counting the suspension in a hemocytometer.
Flow Cytometry Analysis Staining of cells with anti-GFAP and immunofluorescence analysis was done by a modification of the method described in Jackson et al. (1984). Cells seeded in 35-mm Petri dishes were harvested by trypsinization and resuspended in PBS at a final concentration of 4x 10 6 cells/mL. The cell suspension was fixed in 70% (v/v) methanol for 10 min at - 20°C and subsequently incubated in 20% fetal calf serum in PBS, 1% (w/v)
Molecular and Chemical Neuropathology
Vol. 16, 1992
36
Rodriguez-Pornbo, Sweetman, and Ugarte
BSA, 0.1% (w; v) azide, 30 min, mouse antihuman GFAP (Amersham Bucks, UK) (1/200), 30 min; and fluorescein-conjugated sheep antimouse Ig (Amersham Bucks, UK) (1/50), 30 min. Following incubation, cells were washed three times with cold PBS, 1% (w/v) BSA, 0.1% (w/v) azide, resuspended in PBS, 0.1% (w/v) azide, and analyzed using an EPICS profile cytofluorograph. The background level of fluorescence was established by incubation of cells with the fluorescein-conjugated sheep antimouse Ig alone. For the determination of specific percent positive cells, gates were set to exclude the background level.
Enzymatic Determinations Cellular extracts: The cellular pellet was resuspended in 20 mM TrisHCI buffer, pH 7.8, or 50 mM imidazole buffer, pH 7.2, and disrupted by freezing-thawing three times in liquid N2. This crude homogenate was used as the source of carboxylase enzymes. Tissue homogenates: Rat brain hemispheres were homogenized 1/10 (w/v) in 20 mM Tris-HC1 buffer, pH 7.8, with a glass-glass homogenizer and disrupted by freezing-thawing as described above. Carboxylase activities were assayed by measuring the enzyme-dependent incorporation of 14 C]bicarbonate into acid-nonvolatile products (Suormala et al., 1985). Glutamine synthetase activity was determined spectrophotometrically (Berl, 1966). Protein was determined by the method of Lowry et al. (1951). [
3-Hydroxyisovaleric Acid Determination by Stable Isotope Dilution QC-MS Samples of incubation media from group I and group Ila cells grown as described above were obtained 7 d after a change of medium. Aliquots of 8 mL were frozen at -20°C, lyophilized, and analyzed according to Jakobs et al. (1984), modified by the use of methyl esters and a DB-wax (0.53 mm id x 15 m) capillary column, 1.5-µm film (JEW Scientific) . Lyophilized supernatants of the corresponding cellular pellet disrupted by freezing-thawing were assayed in the same way.
RESULTS
Characterization of Cell Type To quantitate the cells that expressed GFAP, we stained cells with anti-GFAP immunofluorescently and then analyzed the cells by flow cytometry. As shown in Fig. 1, the stain with antisera against GFAP was homogeneous and positive for more than 90% of cells in culture. The negative control was well displaced from positive.
Molecular and Chemical Neuropathology
Vol. 16, 1992
37
Biotin -Restricted Astrocytes for MCD
m 200 UI
200
z J J W LI
100
100
IQ J W X n
0
W
10 0 10 1102 FLUORESCENCE INTENSITY
103
Fig. 1. Flow cytometry analysis of immunofluorescence antiGFAP stained cells. Primary culture of astrocytes maintained for 21 d under routine culturing conditions were fixed with methanol, stained with a-IGG-fluorescein (A) (negative control) or anti-GFAP-a-IgG-fluorescein (B) and analyzed by flow cytometry at 510 nm. The panels show relative number of cells vs relative log fluorescence on a three-decade scale.
Developmental Changes in Mitochondrial Carboxylases We have studied the developmental profile of mitochondrial carboxylase activities from astrocytes as a function of time in culture, compared with the postnatal changes of these activities in vivo. In rat brain (Fig. 2), mitochondrial carboxylase activities at 12 h after birth were about 1.6-3fold higher than those of 6-d-old animals. Between 6 and 30 d of age, the mitochondrial carboxylase activities increased about 2.3-3.7-fold and reached the adult value by 30 d. In cultured astrocytes (Fig. 3), mitochondrial carboxylase activities increased twofold between 7 and 21 d. The activities at 21 d in culture were about 2-3-fold higher than those of 12-h-old animals and between 1.5-2-fold higher than 30-d-old animals.
Biotin-Deficient Status Cellular Growth Under Various Biotin -Restricted Culture Conditions Growth curves of control (Group I) and biotin-restricted cultures of astrocytes (Group II a and b) were compared (Fig. 4). After a short lag time following the initial plating, the viable cell number increased exponentially up to 10 d in the three culture conditions. At 15 d, both control (Group I) and 10% serum-enriched medium with 50 U/L avidin cultures (Group II a) reached stable and similar cell numbers but in 10% dialyzed serum-enriched medium culture (Group II b), the cell numbers decreased drastically.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Rodriguez -Pombo, Sweetman, and Ugarte
38 0 a 3
I
"
C
^^
E E 0
2
_A E
9
i
V
a
E N C LU
iu
LU
JU
HOUii
Age (days)
Fig. 2. Development of mitochondrial carboxylases activities in rat brain. Brain hemispheres from rats of different ages were homogenized and assayed for PC (s), PCC (A), and (MCC (W) activities. Results are the mean value ± SD of at least three determinations. C ai 0 I0 rn E
4
c_ 93 E 0 E
C
T
2
I
^I
Li
M
a E N C
w 0
lU
15
LU
Days in culture Fig. 3. Development of mitochondrial carboxylases in primary cultures of astrocytes. Primary culture of astrocytes grown under routine culturing conditions were harvested from dishes at each time-point and assayed for PC (•), PCC (A), and (3MCC (s). Results are the mean value ± SD of at least nine different cultures in triplicate.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Biotin-Restricted Astrocytes for MCD
E E Ln
m 0
U LU
5
r0
' 4.5
C, 0
0 10
15
Time (days)
Fig. 4. Growth curves of primary culture of astrocytes in different culture conditions. Primary culture of astrocytes maintained for 7 d under routine culturing conditions were transferred at 0 time to 10% serum-enriched medium (Group I, •), 10% serum-enriched medium with 50 U/L avidin (Group IIa, A), or 10% dialyzed serum-enriched medium (Group Ilb, n ). The viable cell numbers were determined by staining of cells with a trypan blue solution and counting in a hemocytometer. Results correspond to a typical experiment.
Kinetics of Decrease of the Mitochondrial Carboxylase Activities in a Biotin-Restricted Medium. Figure 5 illustrates the decrease of PC, PCC, and ,3MCC activities for primary cultures of astrocytes growing in 10% serum-enriched medium with 50 U/L avidin. The decrease has been normalized as percentage of control activities at different times of cultures. The rates of disappearance were similar for the three activities. The apparent exponential constant estimated by the analysis of the data using a curve-fitting program was 0.44/ d. During the experiment, GS activity, a nonbiotin-dependent enzyme, was not changed.
3- Hydroxyisovaleric Acid Excretion Figure 6 shows a typical selected ion-monitoring chromatogram of 8 mL of growth media from astrocytes grown during 3 wk in control or avidin-supplemented medium both with 50 nmol of D5,63-HIVA as internal standard. The concentration of 3-HIVA in these media were determined to be 2.8 µM in the control and 29.3 p.M in the medium with avidin. The concentration of 3-HIVA from control or avidin growth media of three primary culture of astrocytes have been summarized in Table 1.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Rodriguez -Pombo, Sweetman, and Ugarte
40 a,CV 0
c
100
0 ti
0
75
`. 50 A 25 E C
w n 1z
is
Days cultured in biotin-restricted medium
Fig. 5. Kinetic of the decreasing in the mitochondrial carboxilase activities from astrocytes. Primary culture of astrocytes maintained for 7 d under routine culturing conditions were transferred at 0 time to 10% serum-enriched medium with 50 U/L avidin. Cell were harvested from dishes at each time-point and assayed for PC (s), PCC (A), ,MCC (s), and GS (0). Results are the mean value ± SD of triplicate determinations of at least three different cultures. (*) 9 cultures. The 100% value was obtained by measuring PC, PCC, ,3MCC, and GS from astrocytes maintained under control culture conditions at the same timepoints. The mean concentration of 3-HIVA in the biotin-deficient media were ten times higher than the control mean. However, the 3-HIVA concentrations in cellular extracts were similar for the two culture conditions. The values determined were 1.01 p.M for control cells and 0.94 µM for biotindeficient cells.
DISCUSSION After 3 wk, our primary cultures of cells from newborn rat brain consisted mainly of astrocytes as confirmed by the large portion of the cells in such cultures containing GFAP. Similar results have been reported before (Hansson et al., 1984). In 3-wk-old cultures of astrocytes, PC, PCC, and ,3MCC activities were twofold higher than those measured in the neonatal brain or in 1-wk-old cultures, indicating that the enzymes develop even after the
Molecular and Chemical Neuropathology
Vol. 16, 1992
41
Biotin-Restricted Astrocytes for MCD
rn 10 0
100
r•^
50
_T VI C
a
a }
N \
0
0
•0 100
100
w
50
m 50 E
0
3
2.5
2
2
3
2.5
Time (minutes)
Time (minutes)
Fig. 6. Selected ion-monitoring chromatogram of 3-HIVA from control or avidin-supplemented growth media of primary culture of astrocytes. Ammonia chemical ionization selected ion-monitoring chromatograms (M + H) of methyl Do 3-HIVA (upper panel) and methyl D5,6 3-HIVA (lower panel) from 8 mL of growth media from group I cell (A) and group IIa (B), both with 50 nmol of D5,63HIVA as internal stable isotope standard.
Table 1
3-HIVA Concentration in Control or Avidin-Supplemented Growth Media from Primary Culture of Astrocytes" [3-HIVA] (1M)
Sample Control medium
2.8 2.3
1 2
2.6±0.4 (mean±SD)
Avidin-suppl. medium
1 2
42.5 29.3
3
28.4
33.4±7.9 (mean+SD) aExperimental conditions are described in Fig. 6.
Molecular and Chemical Neuropathology
Vol. 16, 1992
42
Rodriguez-Pombo, Sweetman, and Ugarte
cells have been removed from their natural environment. Besides, as previously observed for a number of other enzymes (Kimelberg et al., 1978; Schousboe et al., 1980), the close correspondence between development of mitochondrial carboxylase activities in vivo and in primary cultured astrocytes suggests that astrocytic proliferation may account for the mitochondrial carboxylases occurrence in vivo. The establishment of a model of biotin deficiency in astrocytes requires culture conditions that allow the development of cells in culture while reducing the mitochondrial carboxylase activities. A drastic decrease in the number of cells was observed when primary cultures of astrocytes were grown in dialyzed FCS compared with the growth similar to the control observed in the presence of avidin. It is possible that the effect of dialyzed FCS reflects a deficiency in another growth factor or oligoelement other than biotin, because of the normal rate of cellular growth observed for astrocytes cultured in avidin-supplemented medium. Moreover, when cells were grown in the presence of avidin for 7 d, mitochondrial carboxylase activities were reduced to 15% of the control value, without any change in the GS activity. These results led us to select the use of avidin as the means to reduce the free biotin levels in the culture medium. Two relevant factors could be implicated in the reduction of the activities, dilution owing to astrocytic proliferation in a biotin-restricted medium, and degradation rate of the holocarboxylases. However, the similar rate of disappearance for the three mitochondrial carboxylases, together with previous observations about a different rate of turnover for the biotin enzymes in human cell lines (Chandler et al., 1985), indicate that the major effect was dilution caused by cell division. A direct inactivation of the carboxylases by the binding of avidin to the covalently bound biotin is unlikely because the three carboxylases are mitochondrial and avidin probably cannot enter the mitochondria. Finally, consistent with the block in mitochondrial carboxylase activities, a very significant increase in the 3-HIVA concentration was observed in growth medium from biotin-deficient astrocytes with respect to control. The deficiency of I3MCC leads to accumulation of 3-methylcrotonyl CoA, which is hydrated to 3-hydroxyisovaleric acid by crotonase. A diffusion of high concentrations of 3-HIVA from the intracellular space to the medium could be the cause of the nondifferent values for the intracellular levels of 3-HIVA in control or deficient cells. In conclusion, growth of primary culture of astrocytes in a defined biotin-restricted condition seem to be a good model to resemble the metabolic state associated to MCD. In contrast with the results obtained in brain of biotin-deficient rats (Sander et al., 1982; Suchy and Wolf, 1986), mitochondrial carboxylase activities were reduced to background levels. Moreover, the most characteristic metabolite in MCD, 3-HIVA, was very significantly elevated in biotin-deficient astrocytes compared to normal.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Biotin-Restricted Astrocytes for MCD
43
Studies in progress using this model of deficiency may contribute to provide evidence to the idea (Shank and Campbell, 1984; Shank et al., 1985) that, astrocytes supply one or more citric acid cycle intermediates to synaptic terminals, thereby serving an anaplerotic function necessitated by the synthesis and release of amino acid neurotransmitters.
ACKNOWLEDGMENTS M. U. acknowledges the economic support of the Fundacion Juan March during her sabbatical in San Diego, CA. This work has been supported by a grant of Fondo de Investigaciones Sanitarias.
REFERENCES Balazs R., Patel A. J., and Richter D. (1973) Metabolic compartments in the brain: Their properties and relation to morphological structures, in Metabolic Compartmentation in the Brain (Balazs R. and Cremer J. E., eds.), pp. 167-184, Macmillan, London. Berl S. (1966) Glutamine synthetase determination of its distribution in brain during development. Biochemistry 5, 916-922. Bock E., Moller M., Nissen C., and Sensenbrenner M. (1977) Glial fibriliary acidic protein in primary astroglial cell cultures derived from newborn rat brain. FEBS Lett. 83, 207-211. Booher J. and Sensenbrenner M. (1972) Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology 2, 97-105. Chandler C. S. and Ballard F. J. (1985) Distribution and degradation of biotincontaining carboxylases in human cell lines. Biochem. J. 232, 385-393. Hansson E., Sellstrbm A., Persson L. I., and Rbnnback L. (1980) Brain primary culture. A characterization. Brain Res. 188, 233-246. Hansson E., Ronnback, L., Lowenthal A., Noppe M., Alling C., Karlsson B., and Sellstrom A. (1982) Brain primary culture—a characterization (part II). Brain Res. 231, 173-183. Hansson E., Ronnback L., Persson L. I., Lowenthal A., Noppe M., Alling C., and Karlsson B. (1984) Cellular composition of primary cultures from cerebral cortex, striatum, hippocampus, brainstem and cerebellum. Brain Res. 300, 9-18. Hertz L., Juurlink B. H. J., and Szuchet S. (1985) Cell cultures, in Handbook of Neurochemistry, vol. 8 (Lajtha A., ed.), pp. 603-661, Plenum, New York. Jackson C. W., Brown L. K., Somerville B. C., Lyles S. A., and Look A. T. (1984) Two-color flow cytometric measurement of DNA distributions of rat megakaryocytes in unfixed, unfractionated narrow cell suspensions. Blood 63, 768-778.
Molecular and Chemical Neuropathology
Vol. 16, 1992
44
Rodriguez-Pombo, Sweetman, and Ugarte
Jakobs, C., Sweetman L., Nyhan W. L., and Packman S. (1984) Stable isotope dilution analysis of 3-hydroxyisovaleric acid in amniotic fluid: Contribution to the prenatal diagnosis of inherited disorders of leucine catabolism. J. Inher. Metab. Dis. 7, 15 -20. Kimelberg H. K., Narumi S., and Bourke R. S. (1978) Enzymatic and morphological properties of primary rat brain astrocyte culture, and enzyme development in vivo. Brain Res. 153, 55 -57. Lowry 0. H., Rosebrough N. J., Farr A. L., and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. Norenberg M. D. and Martinez-Hernandez A. (1979) Fine structural localization of glutamine synthetase in astrocytes of rat brain. Brain Res. 161, 303-310. Sander J. E., Packman S., and Townsend J. (1982) Brain pyruvate carboxylase and the pathophysiology of biotin-dependent diseases. Neurology 32, 878-880. Schousboe A. (1980) Primary cultures of astrocytes from mammalian brain as a tool in neurochemical research. Cell. Mol. Biol. 26, 505-513. Schousboe A., Nissen C., Bock E., Sapirstein V. S., Juurlink R. H. J., and Hertz L. (1980) Biochemical development of rodent astrocytes in primary culture, in Tissue Culture in Neurobiology (Giacobini E., Vernadakis A., and Shahar A., eds.), pp. 397-409, Raven, New York. Schrijver J., Dias Th., and Hommes F. A. (1979) Some biochemical observations on biotin deficiency in the rat as a model for human pyruvate carboxylase deficiency. Nutr. Metab. 23, 179-191. Shank R. P. and Campbell G. L. (1984) a-Ketoglutarate and malate uptake and metabolism by synaptosome: Further evidence for an astrocyte-to-neuron metabolic shuttle. J. Neurochem. 42, 1153-1161. Shank R. P., Bennett G. S., Freytag S. 0., and Campbell G. L. (1985) Pyruvate carboxylase an astrocyte-specific enzyme implicated in the replenishment of amino acid neurotransmitter pools. Brain Res. 329, 364-367. Suchy S. F. and Wolf B. (1986) Effect of biotin deficiency and supplementation on lipid metabolism in rats: Cholesterol and lipoproteins. Am. J. Clin. Nutr. 43, 831-838. Suormala T., Wick H., Bonjour J. P., and Baumgartner E. R. (1985) Rapid differential diagnosis of carboxylase deficiencies and evaluation for biotin-responsiveness in a single blood sample. Clin. Chim. Acta 145, 151-162. Yu A. C. H., Drejer J., Hertz L., and Schousboe A. (1983) Pyruvate carboxylase activity in primary cultures of astrocytes and neurons. J. Neurochem. 41, 1484-1487.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Primary Cultures of Astrocytes from Rat as a Model for Biotin Deficiency in Nervous Tissue PILAR RODRIGUEZ-POMBO, 1 LAWRENCE SWEETMAN, 2 AND MAGDALENA UGARTE* I 'Centro de Biologia Molecular, Facultad de Ciencias, Universidad Autonoma, 28049 Madrid, Spain; and 2 Biochemical Genetics Laboratory, Childrens Hospital of Los Angeles, Los Angeles, CA 90027 Received March 6, 1991; Accepted August 7, 1991
ABSTRACT The activities and biotin-dependence of the three mitochondrial biotin-dependent carboxylases: pyruvate carboxylase, propionyl CoA carboxylase, and f3-methylcrotonyl CoA carboxylase of primary culture of astrocytes have been examined. An increase of the three mitochondrial carboxylase activities was observed during cell growth, as was the case for developing rat brain. Mitochondrial carboxylase activities from 3-wk-old primary cultures of astrocytes were higher than those in the neonatal rat brain. When astrocytes were grown in a 10% serumenriched medium supplemented with avidin to bind biotin, the mitochondrial carboxylase activities were reduced to 15% of control value. Consistent with these results, after 3 wk in culture, the 3-hydroxyisovaleric acid concentration in the growth medium was tenfold higher than the controls. In this culture condition, cellular growth and the nonbiotin-dependent enzyme, glutamine synthetase, were not modified with respect to control. Primary cultures from newborn rat brain hemispheres are suggested as an experimental approach to the study of biotin deficiency in nervous tissue.
Index Entries: Biotin deficiency; primary culture of astrocytes; mitochondrial carboxylase activities; multiple carboxylase deficiency (MCD); 3-hydroxyisovaleric acid excretion. *A u th or to whom all correspondence and reprint requests should be addressed. Molecular and Chemical Neuropathology
33
Vol. 16, 1992
Rodriguez-Pombo, Sweetman, and Ugarte
34
Abbreviations: 3-HIVA: 3-Hydroxyisovaleric acid; GFAP: glial fibrillary acidic protein; GS: glutamine synthetase; MCD: multiple carboxylase deficiency; ,3MCC: i3-methylcrotonyl CoA carboxylase; PC: pyruvate carboxylase; PCC: propionyl CoA carboxylase.
INTRODUCTION The important role of biotin in human physiology has been highlighted by the recognition of two inborn errors of biotin metabolism in humans. The biotin-responsive multiple carboxylase deficiencies are inherited disorders of metabolism, in which a characteristic pattern of organic acids appears in the urine of patients as a consequence of a deficiency of the biotin-dependent carboxylases. There are four biotin-dependent enzymes in mammals: propionyl CoA carboxylase (PCC: EC 6.4.1.3), pyruvate carboxylase (PC: EC 6.4.1.1.), /3-methylcrotonyl CoA carboxylase ((3MCC: EC 6.4.1.4), and acetyl CoA carboxylase (ACC: EC 6.4.1.2). All but acetyl CoA carboxylase are located in the mitochondrial subcellular compartment. The most prominent features of biotin deficiency in humans are those affecting the skin and the nervous system. However, the obvious difficulty to carry out biochemical studies in CNS from patients have led to the use of experimental models (Sander et al., 1982; Schrijver et al., 1979). Because the brain contains different metabolic compartments, which at least partly correspond to different cell types (Balazs et al., 1973), we found it worthwhile to study the biotin-dependent enzymes in selected types of cells in culture to establish an easier model for studying biotin metabolism in nervous tissue than using biotin-deficient rats. The development of techniques to grow cerebral tissue in primary cultures (Booher and Sensenbrenner, 1972) has facilitated biochemical investigations of specific cerebral cell types. Primary cultures of astrocytes have been particularly well characterized (Hansson et al., 1980, 1982) and described as potential tools in neurochemical research (Schousboe, 1980). The large content of glial fibrillary acidic protein (GFAP) and the high activity of glutamine synthetase activity (GS, EC. 6.3.1.2) have been used as markers of their astrocytic nature (Bock et al., 1977; Hertz et al., 1985; Norenberg and Martinez-Hernandez, 1979). Immunohistochemical (Shank et al., 1985) and enzymatic (Yu et al., 1983) studies have shown an astrocytic localization of pyruvate carboxylase. Therefore, we used primary cultures of astrocytes as a model to study biotin metabolism in the nervous system. The aim of the present investigation was to study the mitochondrial carboxylases in astrocytes and to define a model of biotin deficiency in these cells.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Biotin-Restricted Astrocytes for MCD
35
METHODS Cell Culture Primary astroglial cell cultures were prepared by aseptically isolating cerebral hemispheres of newborn Wistar rats according to a modification of a previously described method (Booher and Sensenbrenner, 1972). The tissue was freed of meninges and capillary vessels. Mechanical dissociation was achieved by sequential passage through two needles of decreasing diameter (0.9 and 0.7 mm) in the presence of 20% serum-enriched medium (modified Eagle's Minimum Essential Medium containing 2 mM glutamine, 200 000 U/L penicillin, 100 mg/L streptomycin, and 20% (v/v) heat-inactivated fetal calf serum). The resulting cell suspension was seeded into six-well trays (Costar), at a density of 2.5 x 10 5 cells/cm 2 . The culture dishes were incubated at 37°C with a humidified atmosphere of 5% CO2 in air. The medium was changed on d 2 and subsequently every 3-4 d. The serum content was reduced to 10% at the end of the first week. The cells were harvested on d 21 by rinsing twice with sterile PBS and incubation with a trypsin-EDTA solution. At this time the cells were positive for GFAP as was demonstrated by fluorescence microscopy, and were therefore judged to be astrocytes (Bock et al., 1977).
Biotin Deficient Status -
Cells were grown for 7 d under routine culturing condition. The cultures were then divided into two groups and cultured for an additional 14 d. Group I (control) cultures were maintained on 10% serum-enriched medium. Group II (biotin-restricted) cultures were divided into two categories: (a) 10% serum-enriched medium with 50 U/L avidin; (b) 10% dialyzed serum-enriched medium (modified Eagle's Minimum Essential Medium containing 2 mM glutamine, 200 000 U/L penicillin, 100 mg/L streptomycin, and 10% (v/v) heat-inactivated dialyzed fetal calf serum from GIBCO). Growth curves were obtained by removing cells from dishes at various stages of culture, staining with a trypan blue solution, and counting the suspension in a hemocytometer.
Flow Cytometry Analysis Staining of cells with anti-GFAP and immunofluorescence analysis was done by a modification of the method described in Jackson et al. (1984). Cells seeded in 35-mm Petri dishes were harvested by trypsinization and resuspended in PBS at a final concentration of 4x 10 6 cells/mL. The cell suspension was fixed in 70% (v/v) methanol for 10 min at - 20°C and subsequently incubated in 20% fetal calf serum in PBS, 1% (w/v)
Molecular and Chemical Neuropathology
Vol. 16, 1992
36
Rodriguez-Pornbo, Sweetman, and Ugarte
BSA, 0.1% (w; v) azide, 30 min, mouse antihuman GFAP (Amersham Bucks, UK) (1/200), 30 min; and fluorescein-conjugated sheep antimouse Ig (Amersham Bucks, UK) (1/50), 30 min. Following incubation, cells were washed three times with cold PBS, 1% (w/v) BSA, 0.1% (w/v) azide, resuspended in PBS, 0.1% (w/v) azide, and analyzed using an EPICS profile cytofluorograph. The background level of fluorescence was established by incubation of cells with the fluorescein-conjugated sheep antimouse Ig alone. For the determination of specific percent positive cells, gates were set to exclude the background level.
Enzymatic Determinations Cellular extracts: The cellular pellet was resuspended in 20 mM TrisHCI buffer, pH 7.8, or 50 mM imidazole buffer, pH 7.2, and disrupted by freezing-thawing three times in liquid N2. This crude homogenate was used as the source of carboxylase enzymes. Tissue homogenates: Rat brain hemispheres were homogenized 1/10 (w/v) in 20 mM Tris-HC1 buffer, pH 7.8, with a glass-glass homogenizer and disrupted by freezing-thawing as described above. Carboxylase activities were assayed by measuring the enzyme-dependent incorporation of 14 C]bicarbonate into acid-nonvolatile products (Suormala et al., 1985). Glutamine synthetase activity was determined spectrophotometrically (Berl, 1966). Protein was determined by the method of Lowry et al. (1951). [
3-Hydroxyisovaleric Acid Determination by Stable Isotope Dilution QC-MS Samples of incubation media from group I and group Ila cells grown as described above were obtained 7 d after a change of medium. Aliquots of 8 mL were frozen at -20°C, lyophilized, and analyzed according to Jakobs et al. (1984), modified by the use of methyl esters and a DB-wax (0.53 mm id x 15 m) capillary column, 1.5-µm film (JEW Scientific) . Lyophilized supernatants of the corresponding cellular pellet disrupted by freezing-thawing were assayed in the same way.
RESULTS
Characterization of Cell Type To quantitate the cells that expressed GFAP, we stained cells with anti-GFAP immunofluorescently and then analyzed the cells by flow cytometry. As shown in Fig. 1, the stain with antisera against GFAP was homogeneous and positive for more than 90% of cells in culture. The negative control was well displaced from positive.
Molecular and Chemical Neuropathology
Vol. 16, 1992
37
Biotin -Restricted Astrocytes for MCD
m 200 UI
200
z J J W LI
100
100
IQ J W X n
0
W
10 0 10 1102 FLUORESCENCE INTENSITY
103
Fig. 1. Flow cytometry analysis of immunofluorescence antiGFAP stained cells. Primary culture of astrocytes maintained for 21 d under routine culturing conditions were fixed with methanol, stained with a-IGG-fluorescein (A) (negative control) or anti-GFAP-a-IgG-fluorescein (B) and analyzed by flow cytometry at 510 nm. The panels show relative number of cells vs relative log fluorescence on a three-decade scale.
Developmental Changes in Mitochondrial Carboxylases We have studied the developmental profile of mitochondrial carboxylase activities from astrocytes as a function of time in culture, compared with the postnatal changes of these activities in vivo. In rat brain (Fig. 2), mitochondrial carboxylase activities at 12 h after birth were about 1.6-3fold higher than those of 6-d-old animals. Between 6 and 30 d of age, the mitochondrial carboxylase activities increased about 2.3-3.7-fold and reached the adult value by 30 d. In cultured astrocytes (Fig. 3), mitochondrial carboxylase activities increased twofold between 7 and 21 d. The activities at 21 d in culture were about 2-3-fold higher than those of 12-h-old animals and between 1.5-2-fold higher than 30-d-old animals.
Biotin-Deficient Status Cellular Growth Under Various Biotin -Restricted Culture Conditions Growth curves of control (Group I) and biotin-restricted cultures of astrocytes (Group II a and b) were compared (Fig. 4). After a short lag time following the initial plating, the viable cell number increased exponentially up to 10 d in the three culture conditions. At 15 d, both control (Group I) and 10% serum-enriched medium with 50 U/L avidin cultures (Group II a) reached stable and similar cell numbers but in 10% dialyzed serum-enriched medium culture (Group II b), the cell numbers decreased drastically.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Rodriguez -Pombo, Sweetman, and Ugarte
38 0 a 3
I
"
C
^^
E E 0
2
_A E
9
i
V
a
E N C LU
iu
LU
JU
HOUii
Age (days)
Fig. 2. Development of mitochondrial carboxylases activities in rat brain. Brain hemispheres from rats of different ages were homogenized and assayed for PC (s), PCC (A), and (MCC (W) activities. Results are the mean value ± SD of at least three determinations. C ai 0 I0 rn E
4
c_ 93 E 0 E
C
T
2
I
^I
Li
M
a E N C
w 0
lU
15
LU
Days in culture Fig. 3. Development of mitochondrial carboxylases in primary cultures of astrocytes. Primary culture of astrocytes grown under routine culturing conditions were harvested from dishes at each time-point and assayed for PC (•), PCC (A), and (3MCC (s). Results are the mean value ± SD of at least nine different cultures in triplicate.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Biotin-Restricted Astrocytes for MCD
E E Ln
m 0
U LU
5
r0
' 4.5
C, 0
0 10
15
Time (days)
Fig. 4. Growth curves of primary culture of astrocytes in different culture conditions. Primary culture of astrocytes maintained for 7 d under routine culturing conditions were transferred at 0 time to 10% serum-enriched medium (Group I, •), 10% serum-enriched medium with 50 U/L avidin (Group IIa, A), or 10% dialyzed serum-enriched medium (Group Ilb, n ). The viable cell numbers were determined by staining of cells with a trypan blue solution and counting in a hemocytometer. Results correspond to a typical experiment.
Kinetics of Decrease of the Mitochondrial Carboxylase Activities in a Biotin-Restricted Medium. Figure 5 illustrates the decrease of PC, PCC, and ,3MCC activities for primary cultures of astrocytes growing in 10% serum-enriched medium with 50 U/L avidin. The decrease has been normalized as percentage of control activities at different times of cultures. The rates of disappearance were similar for the three activities. The apparent exponential constant estimated by the analysis of the data using a curve-fitting program was 0.44/ d. During the experiment, GS activity, a nonbiotin-dependent enzyme, was not changed.
3- Hydroxyisovaleric Acid Excretion Figure 6 shows a typical selected ion-monitoring chromatogram of 8 mL of growth media from astrocytes grown during 3 wk in control or avidin-supplemented medium both with 50 nmol of D5,63-HIVA as internal standard. The concentration of 3-HIVA in these media were determined to be 2.8 µM in the control and 29.3 p.M in the medium with avidin. The concentration of 3-HIVA from control or avidin growth media of three primary culture of astrocytes have been summarized in Table 1.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Rodriguez -Pombo, Sweetman, and Ugarte
40 a,CV 0
c
100
0 ti
0
75
`. 50 A 25 E C
w n 1z
is
Days cultured in biotin-restricted medium
Fig. 5. Kinetic of the decreasing in the mitochondrial carboxilase activities from astrocytes. Primary culture of astrocytes maintained for 7 d under routine culturing conditions were transferred at 0 time to 10% serum-enriched medium with 50 U/L avidin. Cell were harvested from dishes at each time-point and assayed for PC (s), PCC (A), ,MCC (s), and GS (0). Results are the mean value ± SD of triplicate determinations of at least three different cultures. (*) 9 cultures. The 100% value was obtained by measuring PC, PCC, ,3MCC, and GS from astrocytes maintained under control culture conditions at the same timepoints. The mean concentration of 3-HIVA in the biotin-deficient media were ten times higher than the control mean. However, the 3-HIVA concentrations in cellular extracts were similar for the two culture conditions. The values determined were 1.01 p.M for control cells and 0.94 µM for biotindeficient cells.
DISCUSSION After 3 wk, our primary cultures of cells from newborn rat brain consisted mainly of astrocytes as confirmed by the large portion of the cells in such cultures containing GFAP. Similar results have been reported before (Hansson et al., 1984). In 3-wk-old cultures of astrocytes, PC, PCC, and ,3MCC activities were twofold higher than those measured in the neonatal brain or in 1-wk-old cultures, indicating that the enzymes develop even after the
Molecular and Chemical Neuropathology
Vol. 16, 1992
41
Biotin-Restricted Astrocytes for MCD
rn 10 0
100
r•^
50
_T VI C
a
a }
N \
0
0
•0 100
100
w
50
m 50 E
0
3
2.5
2
2
3
2.5
Time (minutes)
Time (minutes)
Fig. 6. Selected ion-monitoring chromatogram of 3-HIVA from control or avidin-supplemented growth media of primary culture of astrocytes. Ammonia chemical ionization selected ion-monitoring chromatograms (M + H) of methyl Do 3-HIVA (upper panel) and methyl D5,6 3-HIVA (lower panel) from 8 mL of growth media from group I cell (A) and group IIa (B), both with 50 nmol of D5,63HIVA as internal stable isotope standard.
Table 1
3-HIVA Concentration in Control or Avidin-Supplemented Growth Media from Primary Culture of Astrocytes" [3-HIVA] (1M)
Sample Control medium
2.8 2.3
1 2
2.6±0.4 (mean±SD)
Avidin-suppl. medium
1 2
42.5 29.3
3
28.4
33.4±7.9 (mean+SD) aExperimental conditions are described in Fig. 6.
Molecular and Chemical Neuropathology
Vol. 16, 1992
42
Rodriguez-Pombo, Sweetman, and Ugarte
cells have been removed from their natural environment. Besides, as previously observed for a number of other enzymes (Kimelberg et al., 1978; Schousboe et al., 1980), the close correspondence between development of mitochondrial carboxylase activities in vivo and in primary cultured astrocytes suggests that astrocytic proliferation may account for the mitochondrial carboxylases occurrence in vivo. The establishment of a model of biotin deficiency in astrocytes requires culture conditions that allow the development of cells in culture while reducing the mitochondrial carboxylase activities. A drastic decrease in the number of cells was observed when primary cultures of astrocytes were grown in dialyzed FCS compared with the growth similar to the control observed in the presence of avidin. It is possible that the effect of dialyzed FCS reflects a deficiency in another growth factor or oligoelement other than biotin, because of the normal rate of cellular growth observed for astrocytes cultured in avidin-supplemented medium. Moreover, when cells were grown in the presence of avidin for 7 d, mitochondrial carboxylase activities were reduced to 15% of the control value, without any change in the GS activity. These results led us to select the use of avidin as the means to reduce the free biotin levels in the culture medium. Two relevant factors could be implicated in the reduction of the activities, dilution owing to astrocytic proliferation in a biotin-restricted medium, and degradation rate of the holocarboxylases. However, the similar rate of disappearance for the three mitochondrial carboxylases, together with previous observations about a different rate of turnover for the biotin enzymes in human cell lines (Chandler et al., 1985), indicate that the major effect was dilution caused by cell division. A direct inactivation of the carboxylases by the binding of avidin to the covalently bound biotin is unlikely because the three carboxylases are mitochondrial and avidin probably cannot enter the mitochondria. Finally, consistent with the block in mitochondrial carboxylase activities, a very significant increase in the 3-HIVA concentration was observed in growth medium from biotin-deficient astrocytes with respect to control. The deficiency of I3MCC leads to accumulation of 3-methylcrotonyl CoA, which is hydrated to 3-hydroxyisovaleric acid by crotonase. A diffusion of high concentrations of 3-HIVA from the intracellular space to the medium could be the cause of the nondifferent values for the intracellular levels of 3-HIVA in control or deficient cells. In conclusion, growth of primary culture of astrocytes in a defined biotin-restricted condition seem to be a good model to resemble the metabolic state associated to MCD. In contrast with the results obtained in brain of biotin-deficient rats (Sander et al., 1982; Suchy and Wolf, 1986), mitochondrial carboxylase activities were reduced to background levels. Moreover, the most characteristic metabolite in MCD, 3-HIVA, was very significantly elevated in biotin-deficient astrocytes compared to normal.
Molecular and Chemical Neuropathology
Vol. 16, 1992
Biotin-Restricted Astrocytes for MCD
43
Studies in progress using this model of deficiency may contribute to provide evidence to the idea (Shank and Campbell, 1984; Shank et al., 1985) that, astrocytes supply one or more citric acid cycle intermediates to synaptic terminals, thereby serving an anaplerotic function necessitated by the synthesis and release of amino acid neurotransmitters.
ACKNOWLEDGMENTS M. U. acknowledges the economic support of the Fundacion Juan March during her sabbatical in San Diego, CA. This work has been supported by a grant of Fondo de Investigaciones Sanitarias.
REFERENCES Balazs R., Patel A. J., and Richter D. (1973) Metabolic compartments in the brain: Their properties and relation to morphological structures, in Metabolic Compartmentation in the Brain (Balazs R. and Cremer J. E., eds.), pp. 167-184, Macmillan, London. Berl S. (1966) Glutamine synthetase determination of its distribution in brain during development. Biochemistry 5, 916-922. Bock E., Moller M., Nissen C., and Sensenbrenner M. (1977) Glial fibriliary acidic protein in primary astroglial cell cultures derived from newborn rat brain. FEBS Lett. 83, 207-211. Booher J. and Sensenbrenner M. (1972) Growth and cultivation of dissociated neurons and glial cells from embryonic chick, rat and human brain in flask cultures. Neurobiology 2, 97-105. Chandler C. S. and Ballard F. J. (1985) Distribution and degradation of biotincontaining carboxylases in human cell lines. Biochem. J. 232, 385-393. Hansson E., Sellstrbm A., Persson L. I., and Rbnnback L. (1980) Brain primary culture. A characterization. Brain Res. 188, 233-246. Hansson E., Ronnback, L., Lowenthal A., Noppe M., Alling C., Karlsson B., and Sellstrom A. (1982) Brain primary culture—a characterization (part II). Brain Res. 231, 173-183. Hansson E., Ronnback L., Persson L. I., Lowenthal A., Noppe M., Alling C., and Karlsson B. (1984) Cellular composition of primary cultures from cerebral cortex, striatum, hippocampus, brainstem and cerebellum. Brain Res. 300, 9-18. Hertz L., Juurlink B. H. J., and Szuchet S. (1985) Cell cultures, in Handbook of Neurochemistry, vol. 8 (Lajtha A., ed.), pp. 603-661, Plenum, New York. Jackson C. W., Brown L. K., Somerville B. C., Lyles S. A., and Look A. T. (1984) Two-color flow cytometric measurement of DNA distributions of rat megakaryocytes in unfixed, unfractionated narrow cell suspensions. Blood 63, 768-778.
Molecular and Chemical Neuropathology
Vol. 16, 1992
44
Rodriguez-Pombo, Sweetman, and Ugarte
Jakobs, C., Sweetman L., Nyhan W. L., and Packman S. (1984) Stable isotope dilution analysis of 3-hydroxyisovaleric acid in amniotic fluid: Contribution to the prenatal diagnosis of inherited disorders of leucine catabolism. J. Inher. Metab. Dis. 7, 15 -20. Kimelberg H. K., Narumi S., and Bourke R. S. (1978) Enzymatic and morphological properties of primary rat brain astrocyte culture, and enzyme development in vivo. Brain Res. 153, 55 -57. Lowry 0. H., Rosebrough N. J., Farr A. L., and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. Norenberg M. D. and Martinez-Hernandez A. (1979) Fine structural localization of glutamine synthetase in astrocytes of rat brain. Brain Res. 161, 303-310. Sander J. E., Packman S., and Townsend J. (1982) Brain pyruvate carboxylase and the pathophysiology of biotin-dependent diseases. Neurology 32, 878-880. Schousboe A. (1980) Primary cultures of astrocytes from mammalian brain as a tool in neurochemical research. Cell. Mol. Biol. 26, 505-513. Schousboe A., Nissen C., Bock E., Sapirstein V. S., Juurlink R. H. J., and Hertz L. (1980) Biochemical development of rodent astrocytes in primary culture, in Tissue Culture in Neurobiology (Giacobini E., Vernadakis A., and Shahar A., eds.), pp. 397-409, Raven, New York. Schrijver J., Dias Th., and Hommes F. A. (1979) Some biochemical observations on biotin deficiency in the rat as a model for human pyruvate carboxylase deficiency. Nutr. Metab. 23, 179-191. Shank R. P. and Campbell G. L. (1984) a-Ketoglutarate and malate uptake and metabolism by synaptosome: Further evidence for an astrocyte-to-neuron metabolic shuttle. J. Neurochem. 42, 1153-1161. Shank R. P., Bennett G. S., Freytag S. 0., and Campbell G. L. (1985) Pyruvate carboxylase an astrocyte-specific enzyme implicated in the replenishment of amino acid neurotransmitter pools. Brain Res. 329, 364-367. Suchy S. F. and Wolf B. (1986) Effect of biotin deficiency and supplementation on lipid metabolism in rats: Cholesterol and lipoproteins. Am. J. Clin. Nutr. 43, 831-838. Suormala T., Wick H., Bonjour J. P., and Baumgartner E. R. (1985) Rapid differential diagnosis of carboxylase deficiencies and evaluation for biotin-responsiveness in a single blood sample. Clin. Chim. Acta 145, 151-162. Yu A. C. H., Drejer J., Hertz L., and Schousboe A. (1983) Pyruvate carboxylase activity in primary cultures of astrocytes and neurons. J. Neurochem. 41, 1484-1487.
Molecular and Chemical Neuropathology
Vol. 16, 1992
