Induction of serine racemase expression and D-serine release from microglia by amyloid beta-peptide

Abstract

BACKGROUND: Roles for excitotoxicity and inflammation in Alzheimer's disease have been hypothesized. Proinflammatory stimuli, including amyloid beta-peptide (Abeta), elicit a release of glutamate from microglia. We tested the possibility that a coagonist at the NMDA class of glutamate receptors, D-serine, could respond similarly. METHODS: Cultured microglial cells were exposed to Abeta. The culture medium was assayed for levels of D-serine by HPLC and for effects on calcium and survival on primary cultures of rat hippocampal neurons. Microglial cell lysates were examined for the levels of mRNA and protein for serine racemase, the enzyme that forms D-serine from L-serine. The racemase mRNA was also assayed in Alzheimer hippocampus and age-matched controls. A microglial cell line was transfected with a luciferase reporter construct driven by the putative regulatory region of human serine racemase. RESULTS: Conditioned medium from Abeta-treated microglia contained elevated levels of D-serine. Bioassays of hippocampal neurons with the microglia-conditioned medium indicated that Abeta elevated a NMDA receptor agonist that was sensitive to an antagonist of the D-serine/glycine site (5,7-dicholorokynurenic acid; DCKA) and to enzymatic degradation of D-amino acids by D-amino acid oxidase (DAAOx). In the microglia, Abeta elevated steady-state levels of dimeric serine racemase, the apparent active form of the enzyme. Promoter-reporter and mRNA analyses suggest that serine racemase is transcriptionally induced by Abeta. Finally, the levels of serine racemase mRNA were elevated in Alzheimer's disease hippocampus, relative to age-matched controls. CONCLUSIONS: These data suggest that Abeta could contribute to neurodegeneration through stimulating microglia to release cooperative excitatory amino acids, including D-serine.

Figure 1

D-serine levels in microglial culture medium measured by HPLC. A. Chromatographic separation of amino acid standards. 1: L-Asp, R_t = 22 min; 2: L-Glu, R_t = 24.7'; 3: L-Ser, R_t = 26.2'; 4: D-Ser, R_t = 27.8'; 5: L-Gln, R_t = 29.3'; 6: Gly, R_t = 30.9'; 7: L-Arg, R_t = 33.2'. B. Chromatographic separation of actual microglia-conditioned medium. C. Primary microglia were incubated 20 h with no addition (Con) or 15 μM Aβ_1–42. Tracings are shown for aliquots of media from duplicates of each treatment. D. D-serine values are represented as the mean ± SEM of triplicates (*p < 0.01), and results are representative of three separate experiments.

Figure 2

Expression of serine racemase mRNA in activated microglia. Semi-quantitative RT-PCR was performed to detect mRNA for serine racemase and GAPDH in microglial cultures incubated 20 h in the absence (Con) or presence of proinflammatory stimuli. A. Primary microglia treated with 15 μM Aβ_1–42. [Densitometric analysis of racemase/GAPDH: Con: 5.12 ± 0.64; Aβ: 9.78 ± 0.3 (p 0.005)] B. HAPI microglial cell line treated with 15 μM Aβ_1–42. C. N9 microglial cell line treated with 300 ng/mL LPS or 10 nM sAPPα₆₉₅.

Figure 3

Induction of serine racemase by Aβ. Serine racemase protein was detected by western blot analysis of lysates of primary microglia. A. Microglial proteins were probed with antibody that either had (+) or had not (-) been preabsorbed to recombinant serine racemase. The detection was intentionally overdeveloped to demonstrate nonspecific bands distinct from the monomer and unreducible dimer. B. Microglia were incubated in triplicate for 12 h either with (+) or without (-) 15 μM Aβ_1–42. Arrowhead designates monomer and arrow dimer. Results are representative of three experiments. Densitometry of the dimer in digitized images indicated a significant difference between treated and untreated samples [cntrl: 139.97 ± 54.92, Aβ: 418.52 ± 74.37 (arbitrary units); p < 0.02, unpaired Student's t-test].

Figure 4

Responsiveness of serine racemase promoter to Aβ. The human serine racemase upstream regulatory region was cloned into a firefly luciferase reporter construct. HAPI microglial cells were cotransfected with this construct and a vector encoding Renilla luciferase under control of a constitutive promoter. After transfection, the cells were treated in serum-free medium with 0.3% DMSO ("Control"), 15 μM Aβ_1–42 or 100 ng/mL LPS. Luciferase activity was measured after 24 h and is represented as firefly luciferase signal, relative to Renilla luciferase signal in the same well (mean of quadruplicates ± SEM; * p < 0.02; ** p < 0.001). Results are representative of three separate experiments.

Figure 5

Elevations in apparent D-serine detected by neuronal bioassay. Primary hippocampal neurons were monitored for [Ca²⁺]_i during the application of conditioned medium (CM) from microglia (for the period indicated by the lower bar). DCKA (100 μM) was added as indicated by the bar thus labeled. A. Microglia were cultured 20 h in the absence (evenly dashed line) or presence of Aβ_1–42 for 20 h. CM from Aβ-treated cultures was added to the neurons at either a 1:100 or 1:18 dilution. B. Microglia were cultured 20 h in the absence (dashed line) or presence (solid line) of 300 ng/mL LPS. Both samples were added to neurons at a dilution of 1:10. C. CM from Aβ-treated cultures was incubated with DAAOx or a control buffer, then applied to neurons at a 1:18 dilution. Similar results were obtained with conditioned medium from LPS-stimulated microglia.

Figure 6

Suppression of microglial neurotoxicity by DCKA and DAAOx. Primary microglia were treated for 24 h in the absence (Con) or presence of 15 μM Aβ_1–42 (Aβ). The conditioned medium from these cultures was then diluted four-fold into the medium of primary hippocampal neuron cultures; neuronal viability was measured 24 h later by LDH release. Some neuronal cultures received simultaneous application of 1 or 10 μM DCKA, and additional sets were exposed to microglia-conditioned medium that had been pre-treated with DAAOx. Values represent the mean ± SEM of triplicate determinations, and the results are representative of three experiments (*p < 0.01 versus "no drug, +Aβ"). Similar data were obtained using MTT reduction as an index of viability.

Figure 7

Analysis of serine racemase mRNA in Alzheimer's disease. Total RNA was isolated from hippocampus of AD or age-matched control (AMC) brains.A. Semi-quantitative RT-PCR was performed with primers for serine racemase and GAPDH. B. Densitometric analysis of PCR products is represented graphically. (*p < 0.02)