0306-4522/91$3.00+ 0.00

N.zuroseienceVol. 45, No. 1, pp. 37-45, 1991 in Great Britain

Pergamon Press plc 0 1991 IBRO

Printed

IMMUNOCYTOCHEMICAL LOCALIZATION OF N-ACETYL-ASPARTATE WITH MONOCLONAL A~IBODI~S M. L. SIMMONS,*C. G. FRONDOZA~ and J. T. COYLE*~ *Departments of Neuroscience, Psychiatry, Pediatrics and Pharmacology, and TDepartment

of Immunology and Infectious Diseases, The Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD 21205, U.S.A.

Abs~ret-N-A~tyl-aspartate

is found in high concentrations in all areas of the brain, but is undetectable

in non-neuronal tissue. In order to characterize the cellular localization of N-acetyl-aspartate in brain, highly specific monoclonal antibodies against N-acetyl-aspartate were produced by fusing spleen lymphocytes obtained from mice immunized with N-acetyl-aspartate conjugated to ~yro~ob~in by ~b~i~de with P3/x63-Ag8.653 mouse myeloma cells. Clones were selected which secrete I@h(k) antibodies highly specific for conjugated ~-a~tyl-as~te. Only 36% loss-r~tivity with conjugated N-acetyl-aspartate-giutamate was observed at high antibody concentrations, whereas no cross-reactivity ( c I %) was observed with conjugated N-acetyl-glutamate or aspartate. Preincubation of the antibodies with 0.5 mg/ml conjugated N-acetyl-aspartate blocked immunoreactivity more than WY&,while preincubation with conjugated N-acetyl-aspartate-glutamate and free N-acetyl-aspartate had no effect. Immunocytochemical staining has shown that N-acetyl-aspartate-like immunoreactivity is localized in neurons, which are widely distributed throughout the brain. The immunoreactive neurons exhibited intense staining of the perikarya, proximal dendrites and axons. No consistent pattern of distribution of immunoreactivity was observed with regard to primary neurotransmitter characteristics of stained neurons although neurons with long projections or extensive arbors, such as pyramidal cells in cortex, locus coeruleus, motor neurons and Purkinje cells, stained much more intensively than local circuit neurons,

N-Acetyl-L-aspartate (BAA) is the second most abundant of the free and conjugated amino acids in the brain, after glutamate, with concentrations in the millimolar range.‘**“r2’ In contrast, NAA levels in non-nervous tissues are very 10w.‘~,~ NAA is synthesized from L-aspartate and acetyl-coA by L-aspartate N-acetyl transferase, an enzyme localized in the brain.29 NAA is also produced by the enzymatic cleavage of N-a~tyl-L-asp~te-Lglutamate (NAAG), a putative neurotransmitter/ neuromodulator.23*26 Recent clinical studies have implicated NAA in some forms of neurodegenerative disease. Several studies report significant elevation of NAA in the urine and plasma of children with progressive brain degeneration.1’~‘4*‘s Alterations in NAA levels have

$To whom correspondence should be addressed at: McLean Ho&al. 115 Mill W. Belmont. MA 02718-9106. U.S.A. A&e&i&: Ag8, P3jx63Ag8.653 mouse myelima cell line; ALS, amyotrophic lateral sclerosis; BSA, bovine serum albumin; EDAC, I-ethyl-3-(3-dimethylaminopropyl)carbodiimide; FCS, fetal calf serum; -LI, -like immunoreactivity; MAb, monoclonal antibody; NAA, N-acetyl+aspartate; NAAG, N-acetyl-L-aspartyl+ glutamate; N-Ac-Asn, N-acctyl-asparagine; N-AGG~u, N-acetyl-glutamate; -IR, immunoreactivity; NGS, normal goat serum; PBS, phosphate-buffered saline; PBSFTN, PBS wntaining 5% FCS, 0.2% Triton X-100 and 0.02% sodium azide; RIA, ~dio~rn~o~y; TBS, Tris-btiered saline; TTBS, TBS containing 0.05% Tween 20. 37

also been reported in adult patients with the neurodegenerative disease amyotrophic lateral sclerosis (ALS), where NAA is increased two- to three-fold in the cerebrospinal fluid, and decreased 40% in spinal cord of patients diagnosed with ALS.24 While the function of NAA remains unknown, several roles have been postulated. One suggestion is that NAA donates its a&y1 group for lipid synthesis, particularly during the developments period corresponding to myelination of fibers.4~z5Another suggestion is that NAA may serve as the N-terminal for non-ribosomal synthesis of NAA-eontaining neuropeptides, including NAAG.= McIntosh and Coope?6 have proposed that NAA merely serves as an organic anion in neurons, similar to isothianate in the squid axon. NAA is not thought to function as a neurotransmitter, as it is neither excitatory nor inhibitory to spinal motoneurons.3 Although NAA is a metabolite of NAAG, several lines of evidence suggest that it has functions independent of NAAG. For instance, NAA and NAAG have different regional distributions in the brain. NAAG levels vary lo-fold and exhibit a rostrocaudal gradient. In contrast, NAA levels in the CNS are more uniform, ranging from 50 nmol/mg protein in spinal cord to 73 mnol/mg protein in hypothalamus.‘* NAA and NAAG also differ in their phylogenetic distributions,” subcellular localizations,2’ and developmental profiles.‘3*‘8 While studies have suggested that NAA is localized there have been no rein neuronal cytoplasm, 12+19

M. L.

38

SIMMONS

ports, to our knowledge, of the immunocytochemical localization of NAA in specific cell groups. In the present study, a monoclonal antibody specific for NAA was produced to directly identify brain cells containing NAA, and thereby further study its possible roles in the nervous sytem.

EXPERIMENTAL

PROCEDURES

Antigens

NAA, aspartate, and other amino acids were obtained from Sigma (St Louis, MO). NAAG was obtained from Bachem (Torrance, CA). These small molecules were conjugated to large carrier proteins using I-ethyl-3-(3-dimethylamino-propyl)carbodiimide (EDAC; Sigma) as previously described.’ NAA was conjugated to thyroglobulin (Sigma) for immunization; NAA, NAAG, aspartate, and other amino acids were conjugated to bovine serum albumin (BSA; Sigma) for use in the radioimmunoassay (RIA) described below. Conjugation efficiency was determined by calculating the fraction of tritiated amino acid (New England Nuclear, Boston, MA) associated with the protein after precipitation with 5% trichloroacetic acid. Immunization

of mice

Female Balb/c mice (Harlan, Walkersville, MD) were immunized subcutaneously in several sites with 2OOpg NAA-thyroglobulin emulsified in 1: 1 Freund’s complete adjuvant (Difco, Detroit, MI). They were boosted two months later with identical injections. One month thereafter they were boosted subcutaneously with 5Opg NAAthyroglobulin emulsified in 1: 1 Freund’s incomplete adjuvant (Difco). After three days, they were bled from the retro-orbital plexus, and serum was assayed for antibody titer using the RIA described below. Three days before fusion, mice, which were positive for antibody secretion, were boosted intraperitoneally with 50pm NAA-thyroglobulin in phosphate-buffered saline (PBS). Monoclonal

antibody production

Cells of the mouse myeloma line P3/x63-Ag8,653 (Ag8) were grown in RPM1 1640 (Gibco, Grand Island, NY) containing 10% fetal calf serum (FCS; Flow Laboratories, McLean, VA), 20pg/ml I-azaguanine (Sigma), and 50 rg/ml gentamycin sulfate (Schering, Kenilworth, NJ). After the mouse was killed by cervical dislocation, the spleen was aseptically removed and reduced to a single cell suspension in RPM1 1640 by gently scraping it with wire mesh. Spleen cells were incubated in a 75-mm flask (Costar, Cambridge, MA) containing RPM1 1640 with 10% FCS for at least 1 h, in order to allow macrophages to adhere to the flask. Spleen lymphocytes and Ag8 cells were decanted from the flasks, washed several times in RPM1 1640, and counted using a hemocytometer. Fusion of 10s lymphocytes with lo* Ag8 cells was done using polyethylene glycol(4000 mol.wt; Sigma) and dimethyl suifoxide (Baker, Phillipsburg, NJ).8 Cells were resuspended in hybridoma medium containing 30% conditioned Ag8 medium (conditioned hybridoma medium and 13.6/cg/ml aminopterin, and 7.6 pg/ml thymidine (Sigma). Cells were plated at a density of 4 x 10’ cells/well in 96-well flat-bottom plates (Costar). Hybridoma colonies appeared nine to 21 da& after.the fusion in at least 35% of the wells. Selected hybridomas were cloned bv limiting dilution in hybridoma medium containing 50% conditioned hybridoma medium without 13.6 pg/ml hypoxanthine, 0.18 pg/ml aminopterin, and 7.6 pg/ml thymidine.

~1 al.

BSA conjugates @g in 50~1 PBS) were coated on IMMUNOLON plates (Dynatech Labs. Chantilly. VA). Unbound sites were blocked with PBS containing 5% FCS. 0.2% Triton X-100 and 0.02% sodium azide (PBS FTN). The antigen-coated cells were incubated with 50~1 of the supernatant fluid for 1 h and subsequently washed three times with PBS-FTN. Iodinated secondary antibody (100,000 c.p.m./50 ~1 of [‘?51]goat and anti-mouse immunoglobulin; New England Nuclear, Boston. MA) was added to each well for another I-h incubation. The wells were washed three times and the bound complex was solubilized with 2 N NaOH for 30 min. Bound radiolabeled antibody was quantitated using a Beckman gamma counter. The assay was used to select mice for fusion, to screen hybridomas for production of anti-NAA antibodies, and to assess the cross-reactivity of selected antibodies. Isotype

of monoclonal

antibodies

The heavy and light chains of the monoclonal antibody (MAb) were determined by an enzyme-linked immunosorbent assay.’ Conditioned culture medium containing MAb was blotted onto nitrocellulose. The samples were air-dried, and free sites were blocked with 1% BSA for I h. After washing with Tris-buffered saline (TBS) containing 0.05% Tween 20 (TTBS), the blots were incubated for 40 min with rabbit anti-mouse IgA, IgM, IgGI, IgGZa, IgGZb, IgG3 heavy chains and kappa and lambda light chains (ICN Biomedicals, Costa Mesa, CA). After washing with TTBS, positive samples were revealed using an avidin-biotinperoxidase kit (anti-rabbit Vectastain ABC kit; Vector. Burlingame, CA) with 4-chloro-I-naphthol (Biorad. Rockville, MD) as substrate. Preparation

oj’ascites

containing monoclonal

antibody

Female Balb/c mice were primed by intraperitoneal injection of 0.5 ml pristane (Aldrich, Milwaukee, WI). The next day they were injected intraperitoneally with 1.5 x lo6 cloned hybridoma cells in serum-free medium. As&tic fluid developed one to two weeks later. Ascitic fluid was aspirated from the peritoneal cavity, and cells were removed by centrifugation at 200 g for IO min. Ascitic fluid WSS stored at -70°C. Competitive

binding experiments

MAbs from spent culture medium or ascitic fluid were incubated with various concentrations of NAA-BSA, other BSA conjugates, free NAA, or PBS overnight at 4°C. Residual antibody activity was assayed using the RIA or the immunocytochemical method described below. Immunocytochemistry

Adult male Spragu+Dawley rats (Harlan) were anesthetized by intraperitoneal injection of lOOmg/kg sodium pentobarbital and perfused transcardially with 500 ml PBS, followed by 500 ml 4% EDAC in PBS. Brain and spinal cord were removed and postfixed in 4% EDAC for at least 48 h, then in 4% formaldehyde overnight. Tissue was cryoprotected in 30% sucrose or 20% glycerol overnight or longer. Coronal sections, 40 pm thick, were cut on a frozen-stage sliding microtome and processed free-floating. Sections were incubated at 4°C for 48 h in PBS containing MAb, 0.1% Triton X-100, and 1% normal goat serum (NGS; Vector). MAb from conditioned culture medium was used at dilutions of 1:5 and 1:10, MAb from ascitic fluid was used at dilutions from 1: 5000 to I : 20,000. Solutions containing either hybridoma medium, normal mouse serum, or PBS instead of MAb were used as negative controls. Sections were washed in PBS and stained with an Solid-phase radioimmunoassay avidin-biotin-peroxidase kit (anti-mouse Vectastain ABC Antibody activity was quantified by RIA using [LZ51]goat Elite kit; Vector). The substiate solution used was PBS with 0.5 mg/ml 3,3’-diaminobenzidine tetrahydrochloride anti-mouse immunoglobulin (New England Nuclear, Boston, MA) as previously described.9 Briefly, BSA or (Aldrich), 0.2 mg/ml nickel ammonium sulfate (Fisher

39

Monoclonal antibodies to N-acetyl-aspartate

Scientific, Fair Lawn NJ), 0.28 mg/ml cobalt chloride (Fisher), and 0.01% hydrogen peroxide, Sections were washed ~orou~y in TBS and mounted onto gelatin-coated skies. Slides were allowed to dry overnight, and were then dehydrated and coverslipped.

-

16116Gp.

-

2c760$. 2M16up.

-

2c?Asdmo

f,

RESULTS Generation

of its-N-acetyZ-~-~~rtate

~nocl~~aI

0

7

antibodies

Prior to the fusion, sera from five immunized mice were assayed for antibody activity against NAA-BSA. Cross-reactivity of the antisera with NAAG-BSA and c~~~i~de (BSA-EDAC) was also assessed. In each case, the antiserum showed greater reactivity with NAA-BSA than NAAG-BSA or BSA-EDAC (Fig. 1). Mouse no.3 was chosen for the fusion. Two weeks after the fusion, hybrid colonies had appeared in 167 of 480 (35%) wells plated. All of these colonies produced antibodies that reacted with NAA-BSA, but after several weeks many colonies stopped secreting antibodies. Three hybrids that continued secreting were cloned by limiting dilution. Cell growth was observed for clones of all three hybrids, but only clones of hybrid IGlO produced antibodies to NAA. lGl0 clones grew in 41 of 120 (30%) wells plated, and all of them produced MAbs to NAA. Three clones-lBl1, 2C7, and 2D8-were selected because they appeared as single colonies and exhibited vigorous growth and secretion. Clone 2C7 was arbitrarily chosen for production of ascitic fluid in mice. The heavy and light chains of the MAb were determined as described in Experimental Procedures. Antibodies from clones 1B 11, 2C7, and 2DS were all found to be IgG2a kappa. Dilution curves for MAbs from culture supernatant and ascitic fluid are shown in Fig. 2. Antibody activity in supematant remained high up to a dilution of 1: 100, and declined between dilutions of 1: 100 and 1: 5000. As expected, as&tic fluid had a much higher titer than culture supematant. Antibody activity remained relatively constant up to a dilution of

1

2

3

4

5

Moun no.

Fig. 1. Antibody activity in mouse sent. Sera from five mice ~muni~d with N~-~~o~ob~in were assayed for activity to NAA-BSA, NAAG-BSA and BSA-EDAC on the solid-phase RJA. Data are expressed as specific d.p.m., and each bar represents the mean of two samples.

Fig. 2. Antibody titer curves. Culture supernatants of clones lB11, 2C7 and 2D8 were assayed for anti-N&4 activity at

various dilutions up to 1: 1000. As&c guid was assayed at various dilutions from 1: 1000 to 1: l,UOO,OUO. Data are expressed as specific d.p.m., and each point represents the mean + S.E.M. of at least three samples.

1: 10,000, and was still detectable 1: l,OOO,OOO.

at a dilution

Characterization of anti-N-acetyl-t-aspartate clonal antibodies

of

mono-

MAbs from culture su~matants of the three clones and from as&tic fluid all showed very similar patterns of cross-reactivity. At high antibody concentrations (undiluted culture supematant; 1: 1000 dilution of ascitic fluid), up to 6% cross-reactivity with NAAG-BSA and 4% cross-reactivity with BSA-EDAC was observed. At lower antibody concentrations (1: 100 dilution of culture supematant); 1: 100,000 dilution of ascitic fluid), cross-reactivity was 1% or less, while the signal to NAA-BSA remained high. The cross-reactivities of the 2C7 MAb with several amino acids that are structurally related to NAA are shown in Table 1. In addition, the MAbs were determined to have no cross-reactivity with less structurally related amino acids, including N-acetylglutamine, N-acetyl-glycine, lysine, and GABA. These studies indicate that these MAbs are highly specific for NAA. The specificities of the MAbs were further tested by preabsorption experiments. Culture supematant or ascitic fluid was incubated overnight with O-500 pg/ml NAA-BSA, then assayed for remaining activity to NAA-BSA. MAb activity decreased as the concentration of NAA-BSA used to preincubate increased (Fig. 3A). Preincubation with 5OO~g~ml NAA-BSA blocked 93-96% of MAb activity. In contrast, preincubation with high concentrations (500,~g/ml) of NAAG-BSA and BSA-EDAC had little effect on MAb activity (Fig. 3B). The MAbs were also not blocked by preincubation with 5 mM free NAA, suggesting that the antibodies preferentially recognize the conjugated conformation of NAA.

Sections through the brainstem were utilized to characterize the specificity of immunoreactivity since

M. L.

40

SIMMONSrf rd.

Table I. Cross-reactivity of monoclonal antibody Coniugate NAA-BSA NAAG--BSA N-AC-Glu-BSA N-Ac-Sdn-BSA Asp-BSA Glu-BSA BSA-EDAC

2C7 supernatant Undiluted I:100

2C7 ascitic fluid 1:lOOO 1:100,000

IOO.00

100.00

100.00

3.82 < 1.00 < I .oo < 1.00 < 1.oo