Anti-inflammatory activity of low molecular weight polysialic...

来自 : 发布时间：2024-09-19

Anti-inflammatory activity of low molecular weight polysialic acid on human macrophages AbstractOligosialic and polysialic acid (oligoSia and polySia) of the glycocalyx of neural and immune cells are linear chains, in which the sialic acid monomers are 伪2.8-glycosidically linked. Sialic acid-binding immunoglobulin-like lectin-11 (SIGLEC-11) is a primate-lineage specific receptor of human tissue macrophages and microglia that binds to 伪2.8-linked oligoSia. Here, we show that soluble low molecular weight polySia with an average degree of polymerization 20 (avDP20) interacts with SIGLEC-11 and acts anti-inflammatory on human THP1 macrophages involving the SIGLEC-11 receptor. Soluble polySia avDP20 inhibited the lipopolysaccharide (LPS)-induced gene transcription and protein expression of tumor necrosis factor-伪 (Tumor Necrosis Factor Superfamily Member 2, TNFSF2). In addition, polySia avDP20 neutralized the LPS-triggered increase in macrophage phagocytosis, but did not affect basal phagocytosis or endocytosis. Moreover, polySia avDP20 prevented the oxidative burst of human macrophages triggered by neural debris or fibrillary amyloid-尾1鈥?2. In a human macrophage-neuron co-culture system, polySia avDP20 also reduced loss of neurites triggered by fibrillary amyloid-尾1鈥?2. Thus, treatment with polySia avDP20 might be a new anti-inflammatory therapeutic strategy that also prevents the oxidative burst of macrophages. IntroductionSIGLEC11 is a primate-lineage specific gene with a human-lineage specific expression profile that belongs to the human CD33-related SIGLEC receptor family1. Expression of SIGLEC-11 was detected in tissue macrophages such as Kupffer cells of the liver and microglia of the brain1,2. SIGLEC-11 recognizes 伪2.8-linked sialic acid (Sia), preferentially consisting of three monomers1,2, making it likely that oligosialic acid chains (oligoSia) represent ligands for SIGLEC-11. Artificial engagement of flag-tagged human SIGLEC-11 transduced into mouse microglia through crosslinking with flag-specific antibodies dampened the lipopolysaccharide (LPS)-induced pro-inflammatory gene transcription in mouse microglia3. Furthermore, the crosslinking of the flag-tagged SIGLEC-11 led to a decrease in phagocytosis of apoptotic cellular material3. In a co-culture system SIGLEC11 gene transduced murine microglia showed protection of neurons against LPS-induced neurotoxicity3. Although the neuroprotective effect was abrogated when cultured neurons were treated with sialidases to deplete Sia from the glycocalyx3; the question whether oligoSia and/or possibly polysialic acid (polySia) can induce functional effects via SIGLEC-11 receptors remained unsolved.PolySia in mammals is attached to glycoproteins of neural and immune cells such as neural cell adhesion molecule (NCAM; CD56), CD36 or neuropilin-2 (for review see4). The length of polymer chains measured as degree of polymerisation (DP) was shown to vary between DP10 to approximately DP2004,5. The mechanisms that regulate the addition of polySia to the carrier proteins as well as the mechanisms that control the chain length of added polymers are not fully understood. It has been demonstrated that polySia expression is tightly linked with the level of expressed polysialyltransferases6 and membrane associated neuraminidases5. Some neuroinvasive bacteria like Escherichia coli K1 can also produce a capsular polysaccharide, which in some substrains is chemically identical to polySia found in the human host7,8,9. By means of molecular mimicry the polySia coat supports escape of bacteria from immune recognition.In this study, we analyzed the effect of soluble low molecular weight polySia on human SIGLEC-11 expressing THP1 cell derived macrophages. Low molecular weight polySia with an average DP20 exhibited an anti-inflammatory effect that was inhibited by a knock-down of the SIGLEC-11 receptor. Furthermore, low molecular weight polySia inhibited inflammatory phagocytosis without affecting basal phagocytosis and endocytosis. In addition, low molecular weight polySia completely prevented the phagocytosis associated oxidative burst of human macrophages. Moreover, in a human macrophage-neuron co-culture system, polySia avDP20 inhibited the radical mediated neurotoxicity triggered by fibrillary amyloid-尾1鈥?2.ResultsLow molecular weight polySia interacts with SIGLEC-11OligoSia in form of 伪2.8-linked Sia has been shown to bind to SIGLEC-11 receptors that are expressed on human tissue macrophages and microglia1,2, but the interaction of SIGLEC-11 with soluble polySia has not been analyzed. Therefore, we analyzed the binding of a soluble low molecular weight polySia to SIGLEC-11 by enzyme-linked immunosorbent assay (ELISA). To obtain low molecular weight polySia, a bacterial derived polySia (50鈥?0鈥塳Da) that was chemical identical to human polySia was fragmented by self-hydrolysis induced by mild heating10. Then, the obtained polySia fragments were separated by strong anion-exchange high-performance liquid chromatography (SAX-HPLC). A major sample representing a polySia fraction with a molecular weight between 4 and 8鈥塳Da (here named avDP20) was isolated and selected as low molecular weight polySia for further functional experiments. The molecular size and purity of the fractioned polySia were confirmed by analytic high-performance liquid chromatography (HPLC) and polyacrylamide gel electrophoresis (Fig. 1a,b). The separation of polySia by SAX-HPLC allowed to separate different lengths of polySia (avDP18, avDP20 and avDP22), all representing compositions of polySia with defined range of molecular weights (Fig. 1a,b). To evaluate the binding between the selected polySia avDP20 and SIGLEC-11, a recombinant human SIGLEC-11 Fc-fusion (rhSIGLEC-11/Fc) protein was added to a protein-A coated plate. PolySia avDP20 was conjugated with biotin (Fig. 1c). Afterwards, different concentrations of biotinylated polySia avDP20 were added to the plate. Biotinylated dextran as a linear polysaccharide with a similar molecular weight (~5鈥塳Da) was used as control. PolySia avDP20 bound to the rhSIGLEC-11/Fc fusion protein in a concentration dependent manner, while no binding of dextran was observed (Fig. 1d).Figure 1(a) HPLC analysis of oligoSia with a degree of polymerization of 6 (oligoSia DP6) and polySia with average degree of polymerization 18, 20 and 22 (polySia avDP18, avDP20 and avDP22) compared to hydrolyzed polySia. DMB-labeled oligoSia DP6, polySia avDP18, avDP20 and avDP22 were eluted by a NaCl buffer gradient. The HPLC retention times were monitored. (b) Polyacrylamide gel electrophoresis of polySia avDP18, avDP20 and polySia avDP22 visualized by stains-all solution. PolySia avDP20 showed a molecular weight of ~6鈥塳Da. Dextran sulfate with molecular weights of ~3鈥塳Da and ~7鈥塳Da served as molecular weight controls. Representative image out of at least three independent experiments is shown. (c) Polyacrylamide gel electrophoresis of biotinylated polySia avDP20 (Biot. av DP20) visualized by stains-all solution. Majority of biotinylated polySia avDP20 is still intact after biotinylation. Representative image out of at least three independent experiments is shown. (d) ELISA to demonstrate the interaction between polySia with avDP20 and SIGLEC-11. SIGLEC-11 Fc-fusion protein was coated to an ELISA plate. Biotinylated polySia avDP20 (0.01, 0.05, 0.25, 1.25 and 6.25鈥壩糶/ml) or biotinylated dextran (0.01, 0.05, 0.25, 1.25 and 6.25鈥壩糶/ml) were added and detected by HRP conjugated streptavidin, followed by a TMB reaction. PolySia avDP20 bound to SIGLEC-11 Fc-fusion protein, while there was negligible binding of biotinylated dextran. Data are presented as relative binding (values of OD 450 measurements) mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni.Full size imageThus, a direct binding of soluble polySia with avDP20 to SIGLEC-11 was detectable.Low molecular weight polySia acts anti-inflammatory on human macrophages via the SIGLEC-11 receptorData from mouse microglia ectopically expressing the human SIGLEC11 gene co-cultured with neurons suggested that sialic acids (Sias) of the neuronal glycocalyx might have anti-inflammatory activity3. To test whether polySia avDP20 act anti-inflammatory on macrophages, human macrophages differentiated from THP1 monocytes were used. The human macrophages showed protein expression of SIGLEC-11 as determined by flow cytometry with a newly generated monoclonal antibody directed against SIGLEC-11 (Fig. 2a). The specificity of the antibody was confirmed on HEK-T cells transfected with a human SIGLEC11 overexpression plasmid. The SIGLEC-11 antibody (clone 3EH) bound to the SIGLEC11 transfected cells; while SIGLEC16 plasmid, as the closest homolog of SIGLEC11 and empty control plasmid transfected cells showed no substantial binding (Fig. 2a).Figure 2PolySia acts anti-inflammatory on human macrophages via SIGLEC-11 receptors.(a) Flow cytometry analysis of HEK-T-cells transfected with either control plasmid, SIGLEC11 plasmid or SIGLEC16 plasmid. GFP-positive cells were gated and stained with a SIGLEC-11 (clone 3 EH) antibody. SIGLEC-11 expression was detected on the SIGLEC11 transfected cells. Control antibody: irrelevant isotype antibody. Representative histograms of n鈥?鈥? independent experiments. (b) Gene transcription of SIGLEC7, SIGLEC10 and SIGLEC11 in human macrophages. SIGLEC11 shows the highest transcript levels. GAPDH: internal control. H2O: control without cDNA. Representative image of at least n鈥?鈥? independent experiments. (c) TNFSF2 transcription of macrophages treated with LPS (1鈥壩糶/ml) alone or with LPS plus polySia avDP20 (1.5鈥壩糓), oligoSia with degree of polymerization 6 (SiaDP6; 1.5鈥壩糓) or sialic acid (Sia; 1.5鈥壩糓). Only polySia avDP20 reduced the LPS induced gene transcription of TNFSF2. Mean鈥?/鈭掆€?SEM of n鈥?鈥? independent experiments. ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni. (d) Gene transcripts for SIGLEC11 were analyzed by qRT-PCR. Knock-down of SIGLEC11 by short hairpin lentiviral vectors (ShSiglec-11) in macrophages reduced the gene transcription of SIGLEC11. ShControl: scrambled short hairpin vector. Mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. ***p鈥?lt;鈥?.001; t-test. (e) SIGLEC-11 expression on macrophages was analyzed by flow cytometry. Knock-down of SIGLEC11 by short hairpin vectors (ShSiglec-11) reduced the expression of SIGLEC-11. ShControl: scrambled short hairpin vector. (f) Effects of SIGLEC11 knock-down on TNFSF2 gene transcription. Macrophages were treated with LPS (1鈥壩糶/ml) alone or with LPS plus polySia avDP20 (1.5鈥壩糓). PolySia avDP20 reduced the TNFSF2 gene transcription that were transduced with the control vector (ShControl), but failed to reduce the TNFSF2 gene transcription levels after knock-down of SIGLEC11 (ShSiglec-11). Mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. **p鈥?lt;鈥?.01, ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni. (g) Effects of SIGLEC11 knock-down on TNFSF2 protein secretion. Macrophages were treated 24鈥塰ours with LPS (1鈥壩糶/ml) alone or with LPS plus polySia avDP20 (1.5鈥壩糓). PolySia avDP20 reduced the TNFSF2 protein secretion that were transduced with the control vector (ShControl), but failed to reduce the TNFSF2 protein secretion levels after knock-down of SIGLEC11 (ShSiglec-11). Mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. **p鈥?lt;鈥?.01; ANOVA followed by Bonferroni.Full size imageNext, we analyzed whether polySia avDP20 had anti-inflammatory effects on the gene transcription of the pro-inflammatory cytokine tumor necrosis factor-伪 (Tumor Necrosis Factor Superfamily Member 2, TNFSF2) that was induced in the macrophages by treatment with 1鈥壩糶/ml LPS. First, gene transcription of SIGLEC7, SIGLEC10 and SIGLEC11 was analyzed in human macrophages, obtained from THP1 monocytes, before and after treatment with LPS. Data show that human macrophages clearly transcribed SIGLEC11, while SIGLEC10 was barely detected and SIGLEC7 was inducible by LPS (Fig. 2b). Afterwards the effect of polySia avDP20 on gene transcription of TNFSF2 was investigated. Results showed that polySia avDP20 prevented the LPS-induced gene transcription of TNFSF2 at the concentration of 1.5鈥壩糓, while Sia as monomer or oligoSia with a DP of 6 had no effect (Fig. 2c). In detail, 1.5鈥壩糓 polySia avDP20 antagonized the LPS-triggered TNFSF2 gene transcription from 2.34鈥?/鈭掆€?.05 to 1.38鈥?/鈭掆€?.16 (Fig. 2c). Then, we asked whether the anti-inflammatory effect of polySia avDP20 on human macrophages was mediated via SIGLEC-11. In order to modify the expression of SIGLEC-11 on macrophages, we transduced cells with a lentiviral knock-down construct targeting SIGLEC11. Reduced gene transcription of SIGLEC11 in the human macrophage was confirmed by qRT-PCR after lentiviral knock-down with a SIGLEC11 targeting vector compared to a control vector (Fig. 2d). In detail, lentiviral knock-down of SIGLEC11 reduced its relative gene transcript from 1鈥?/鈭掆€?.04 to 0.36鈥?/鈭掆€?.05 (Fig. 2d). Flow cytometry analysis also confirmed reduced cell surface expression of SIGLEC-11 on the human macrophages after lentiviral knock-down of SIGLEC11 (Fig. 2e). Then, we analyzed whether the anti-inflammatory effect of polySia avDP20 on TNFSF2 gene transcription was mediated via SIGLEC-11 receptors of human macrophages. Therefore, we stimulated the transduced macrophage cells (that received either SIGLEC11 knock-down vector or control vector) with LPS (1鈥壩糶/ml) and treated them with 1.5鈥壩糓 polySia avDP20 for 3 or 24鈥塰ours to assess transcription and protein level (Fig. 2f,g). Knock-down of SIGLEC11 neutralized the anti-inflammatory effect of polySia avDP20 on the LPS-induced cytokine gene transcription of TNFSF2 (Fig. 2f) and protein secretion of TNFSF2 (Fig. 2g). In detail, LPS-induced relative TNFSF2 gene transcription was unchanged from 2.49鈥?/鈭掆€?.09 to 2.68鈥?/鈭掆€?.19 after lentiviral knock-down of SIGLEC11, while it was reduced from 2.75鈥?/鈭掆€?.11 to 1.85鈥?/鈭掆€?.2 after transduction with the control plasmid (Fig. 2f). Likewise, LPS-induced TNFSF2 protein secretion was unchanged from 278鈥?/鈭掆€?.3 to 266鈥?/鈭掆€?1.2鈥塸g/ml after lentiviral knock-down of SIGLEC11, while it was reduced from 268鈥?/鈭掆€?.8 to 153鈥?/鈭掆€?8.9 after transduction with the control plasmid (Fig. 2g). Thus, polySia avDP20 acted anti-inflammatory on the LPS-induced TNFSF2 gene transcription and protein secretion of human macrophages via the SIGLEC-11 receptor.Low molecular weight polySia acts anti-inflammatory on macrophages at nanomolar concentrationsNext, we tested a wider concentration range of polySia avDP20 on human macrophages. Therefore, we also analyzed the effect of polySia avDP20 on the metabolic activity of macrophages. Even relative high concentrations of polySia avDP20 up to 150鈥壩糓 did not significantly interfere with cell viability as determined by the MTT-assay (Fig. 3a). The half maximum toxic concentration in the 24鈥塰ours MTT-assay was determined as TC50m鈥?鈥?48.78鈥壩糓. Next, we determined the half maximum effective concentration of polySia avDP20 leading to a 50% reduction of the TNFSF2 gene transcripts induced in the human macrophages by 1鈥壩糶/ml LPS. This half effective concentration was determined as EC50m鈥?鈥?.14鈥壩糓 (Fig. 3b). Thus, the therapeutic index of polySia avDP20 on human macrophages, calculated as the ratio of the compound鈥檚 50% toxic concentration to the compound鈥檚 50% effective concentration, was TI(m/m)鈥?鈥塗C50m/EC50m鈥?鈥?062.71.Figure 3PolySia acts at nanomolar concentrations and prevents LPS-stimulated production of TNFSF2.(a) Metabolic activity of human macrophages was determined by the MTT-assay. Metabolic activity was reduced after addition of 500鈥壩糓 and 1500鈥壩糓 polySia avDP20 compared to the untreated control (UT), indicating reduced cell viability. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. *p鈥?lt;鈥?.05, **p鈥?lt;鈥?.01; ANOVA followed by Bonferroni. (b) Gene transcripts for TNFSF2 of human macrophages after treatment with LPS (1鈥壩糶/ml) and different concentrations of polySia avDP20 for 3鈥塰ours. At a concentration of 0.05鈥壩糓 or higher polySia avDP20 reduced the LPS-induced gene transcription of TNFSF2. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni. (c) Gene transcription of TNFSF2 after treatment with two different LPS concentrations (1鈥壩糶/ml and 50鈥壩糶/ml) and polySia avDP20 (1.5鈥壩糓) for 3鈥塰ours. PolySia avDP20 also reduced the gene transcription of TNFSF2 induced by 50鈥壩糶/ml LPS. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. *p鈥?lt;鈥?.05, ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni. (d) TNFSF2 protein secretion after treatment with two different LPS concentrations (1鈥壩糶/ml and 50鈥壩糶/ml) and polySia avDP20 (1.5鈥壩糓) for 24鈥塰ours. PolySia avDP20 reduced the protein secretion of TNFSF2 induced by 50鈥壩糶/ml LPS. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. *p鈥?lt;鈥?.05, ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni.Full size imageNext, we analyzed whether polySia avDP20 could also antagonize the TNFSF2 gene transcription induced by higher concentration of LPS. Therefore, human macrophages were treated for 3鈥塰ours with 50鈥壩糶/ml LPS plus 1.5鈥壩糓 polySia avDP20 (Fig. 3c). Treatment with 50鈥壩糶/ml LPS increased the gene transcription of TNFSF2 from 1鈥?/鈭掆€?.08 to 7.3鈥?/鈭掆€?.8, while addition of 1.5鈥壩糓 polySia avDP20 antagonized this increase to only 2.59鈥?/鈭掆€?.1 (p鈥?鈥?.001; Fig. 3c).Then, we asked whether polySia avDP20 also interferes with the LPS-stimulated TNFSF2 protein secretion of human macrophages. Macrophages were treated for 24鈥塰ours with LPS鈥?/鈭掆€?.5鈥壩糓 polySia avDP20; and then TNFSF2 protein secretion by macrophages was determined by ELISA (Fig. 3d). LPS (50鈥壩糶/ml) led to a production of 978.86鈥?/鈭掆€?6.98鈥塸g/ml TNFSF2 protein in the supernatant. However addition of polySia avDP20 (1.5鈥壩糓) reduced the secreted TNFSF2 protein to 741.06鈥?/鈭掆€?2.01鈥塸g/ml (p鈥?鈥?.001; Fig. 3d).Thus, polySia avDP20 showed a relative high therapeutic index on cultured human cells and acted anti-inflammatory on human macrophages at the lowest concentration of 0.05鈥壩糓 without clear signs of toxicity up to 150鈥壩糓 as determined by the MTT-assay.Low molecular weight polySia inhibits LPS-triggered phagocytosis of microbeads, but does not affect endocytosis of nanobeadsSince inhibitory Siglecs could increase endocytosis and reduce phagocytosis11, we analyzed the effect of polySia avDP20 on basal and inflammation triggered uptake of microbeads and nanobeads. Phagocytic uptake of microbeads was relatively increased from 100鈥?/鈭掆€?1.4% to 144鈥?/鈭掆€?.8% after treatment with 1鈥壩糶/ml LPS (Fig. 4a). Treatment with 1.5鈥壩糓 polySia avDP20 prevented this LPS-dependent increase in phagocytosis to 103鈥?/鈭掆€?.7% without interference with basal phagocytosis of macrophages (Fig. 4a). To compare the effect of different lengths of sialic acids, phagocytic uptake of microbeads was investigated after treatment of cells with 1.5鈥壩糓 of Sia, oligoSia DP6 and polySia avDP20. Treatment with 1.5鈥壩糓 polySia avDP20 reverted the LPS-triggered increase in phagocytosis from 157鈥?/鈭掆€?% to 106鈥?/鈭掆€?% without affecting basal macrophage phagocytosis (Fig. 4b). Then, we analyzed whether SIGLEC-11 is involved in the inhibitory effect of polySia avDP20 on the LPS-stimulated microbead phagocytosis. Lentiviral knock-down of SIGLEC11 in the macrophages abolished the inhibitory effect of polySia avDP20 on phagocytosis compared to control plasmid transduced cells, which showed a reduction in microbeads phagocytosis from 171鈥?/鈭掆€?% to 124鈥?/鈭掆€?% (Fig. 4c). Next, we analyzed the effect of polySia avDP20 on endocytosis. Basal endocytosis of nanobeads was unchanged after treatment with polySia avDP20. Interestingly, also the LPS-triggered increase in phagocytosis of nanobeads from 100鈥?/鈭掆€?.1% to 212鈥?/鈭掆€?4.3% was not significantly modulated by polySia avDP20 (Fig. 4d). Cytochalasin D and methyl-beta-cyclodextrin, which are inhibitors of phagocytosis and endocytosis, respectively, were used as controls (Fig. 4a鈥揹).Figure 4PolySia does not affect basal phagocytosis and homeostatic endocytosis.(a) Phagocytosis of microbeads as determined by flow cytometry. PolySia avDP20 did not change phagocytosis of microbeads. However, the LPS-induced increase in microbeads phagocytosis was antagonized by 0.5鈥壩糓 and 1.5鈥壩糓 polySia avDP20. The phagocytosis inhibitor cytochalasin D was used as control. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. **p鈥?lt;鈥?.01, ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni. (b) Phagocytosis of microbeads as determined by flow cytometry. Sia, oligoSia DP6 and polySia avDP20 did not change phagocytosis of microbeads. However, LPS-induced increase in microbead phagocytosis was antagonized by 1.5鈥壩糓 polySia avDP20. The phagocytosis inhibitor cytochalasin D was used as control. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. *p鈥?lt;鈥?.05; **p鈥?lt;鈥?.01; ANOVA followed by Bonferroni. (c) Phagocytosis of microbeads by transduced THP1-derived macrophages as determined by flow cytometry. In control plasmid (Sh Control) transduced cells, polySia avDP20 was able to antagonize the effect of LPS on microbead phagocytosis. However, in SIGLEC11 knock-down plasmid (Sh SIGLEC-11) transduced macrophages, polySia avDP20 had no effect on LPS-induced microbead phagocytosis. The phagocytosis inhibitor cytochalasin D was used as control. Data are presented as mean鈥?/鈭掆€塖EM of at least n鈥?鈥? independent experiments. *鈥?lt;鈥?.05; ANOVA followed by Bonferroni. (d) Endocytosis of nanobeads as determined by flow cytometry. PolySia avDP20 did not change the homeostatic basal nanobeads endocytosis. Furthermore, the inflammatory LPS-induced increase of nanobead endocytosis was not antagonized by polySia avDP20. The endocytosis inhibitor methyl-beta-cyclodextrin was used as control. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni.Full size imageThus, polySia avDP20 does not affect basal phagocytosis and homeostatic endocytosis, but prevents the LPS-triggered increase in the inflammation associated phagocytosis.Low molecular weight polySia prevents the phagocytosis associated respiratory burstSiglecs have been shown to block the oxidative burst associated with phagocytosis of neural debris in a murine culture system12 and in vivo in Siglec-E deficient mice13. To study the effect of polySia avDP20 on phagocytosis of neural debris and the associated oxidative burst, we incubated human macrophages with fluorescently labeled neural debris. First, we determined the uptake of neural debris into the cells by confocal microscopy and 3D-reconstruction. We observed ingestion of neural debris into macrophages (Fig. 5a). After treatment with polySia avDP20 the number of macrophages showing uptake of neural debris was slightly decreased (Fig. 5b). In detail, 1.5鈥壩糓 polySia avDP20 reduced the relative uptake of neural debris by macrophages from 1鈥?/鈭掆€?.06 to 0.7鈥?/鈭掆€?.07 (p鈥?鈥?.007; Fig. 5b).Figure 5PolySia avDP20 reduced macrophage phagocytosis of neural debris and prevented the associated oxidative burst.(a) Phagocytosis of neural debris by human macrophages. Neural debris was labelled with a red fluorescent dye and added for 1.5鈥塰our to macrophages. Cells were fixed and counter labelled with an antibody directed against CD11b. Confocal 3D-reconstruction of a macrophage cell that has ingested neural debris is shown. Nuclei are labelled by DAPI. Scale bar: 50鈥壩糾. (b) Quantification of the uptake of neural debris into macrophages. Cells were treated by polySia avDP20. PolySia avDP20 reduced the uptake of neural debris at a concentration of 1.5鈥壩糓. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. **p鈥?lt;鈥?.01; ANOVA followed by Bonferroni. (c) Release of superoxide as detected by DHE. Human macrophages both challenged with neural debris and pre-treated with different concentrations of polySia avDP20. Superoxide release triggered by neural debris was inhibited by 1.5鈥壩糓 of polySia avDP20. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. **p鈥?lt;鈥?.01; ANOVA followed by Bonferroni. (d) Scavenging of the superoxide release by superoxide dismutase SOD1 and Trolox. Human macrophages were challenged with debris and superoxide release was determined by DHE. The release of superoxide was scavenged by addition of SOD1 or Trolox. The radical scavenger Trolox and the SOD1 confirmed that DHE detected extracellular production of superoxide. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. ***p鈥?lt;鈥?.001; ANOVA followed by Bonferroni.Full size imageTo determine the phagocytosis associated oxidative burst, we used the superoxide-sensitive fluorescent dye dihydroethidium (DHE). Treatment of macrophages with neural debris stimulated the relative superoxide production from 1鈥?/鈭掆€?.09 to 1.6鈥?/鈭掆€?.1 (p鈥?鈥?.004), while 1.5鈥壩糓 polySia avDP20 completely prevented the debris induced stimulation of superoxide (0.9鈥?/鈭掆€?.09; p鈥?鈥?.003; Fig. 5c). As a control we scavenged the superoxide by addition of superoxide dismutase-1 (SOD1) or by addition of Trolox, a water-soluble analog of vitamin E, into the medium. Both scavengers neutralized the superoxide that was induced by neural debris (Fig. 5d). The inhibitory effect of 1.5鈥壩糓 polySia avDP20 on superoxide was as strong as the scavengers Trolox or SOD1 (Fig. 5d).We also determined the effect of polySia avDP20 on the uptake and superoxide production triggered by the Alzheimer麓s disease associated fibrillary amyloid-尾1鈥?2. We incubated human macrophages with biotinylated fibrillary amyloid-尾1鈥?2 and visualized the uptake into the cells by confocal microscopy and 3D-reconstruction (Fig. 6a). PolySia avDP20 slightly reduced the uptake of amyloid-尾1鈥?2 into macrophages. In detail, 1.5鈥壩糓 polySia avDP20 reduced relative uptake of fibrillary amyloid-尾1鈥?2 from 1鈥?/鈭掆€?.08 to 0.61鈥?/鈭掆€?.06 (p鈥?鈥?.003; Fig. 6b). Furthermore, we analyzed the oxidative burst triggered by amyloid-尾1鈥?2. Again, production of superoxide release was determined via the fluorescence dye DHE. Treatment of macrophages with amyloid-尾1鈥?2 stimulated the relative superoxide production from 1鈥?/鈭掆€?.05 to 1.4鈥?/鈭掆€?.1 (p鈥?鈥?.004), while 1.5鈥壩糓 polySia avDP20 prevented the amyloid-尾1鈥?2 induced stimulation of the superoxide production (1.08鈥?/鈭掆€?.03; p鈥?鈥?.03; Fig. 6c). SOD1 and Trolox scavenged the amyloid-尾1鈥?2 triggered superoxide molecules (Fig. 6d). Again, the inhibitory effect of polySia avDP20 on fibrillary amyloid-尾1鈥?2 was as strong as the scavengers.Figure 6PolySia reduced macrophages phagocytosis of fibrillary amyloid-尾1鈥?2 and prevented the associated oxidative burst.(a) Phagocytosis of fibrillary biotin-conjugated amyloid-尾1鈥?2 by human macrophages. Cells were fixed and stained with fluorescent labelled streptavidin and an antibody directed against CD11b. Nuclei were labelled by DAPI. Confocal 3D-reconstruction of a macrophage cell that showed ingested amyloid-尾1鈥?2 material inside the cell. Scale bar: 50鈥壩糾. (b) Quantification of the uptake of amyloid-尾1鈥?2 into macrophages. Cells were treated for 1鈥塰our by polySia avDP20. PolySia avDP20 prevented the uptake of amyloid-尾1鈥?2 at a concentration of 1.5鈥壩糓. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. **p鈥?lt;鈥?.01, ANOVA followed by Bonferroni. (c) Release of superoxide as detected by DHE of the human macrophages challenged with fibrillary amyloid-尾1鈥?2 and treated with different concentrations of polySia avDP20. Superoxide release triggered by amyloid-尾1鈥?2 was inhibited by 0.5鈥壩糓 and 1.5鈥壩糓 polySia avDP20. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. *鈥?lt;鈥?.05, **p鈥?lt;鈥?.01, ANOVA followed by Bonferroni. (d) Scavenging of the superoxide release by superoxide dismutase SOD1 and Trolox. Human macrophages were challenged with fibrillary amyloid-尾1鈥?2 and superoxide release was determined by DHE. The radical scavenger Trolox and the SOD1 confirmed that DHE detected extracellular production of superoxide. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. *鈥?lt;鈥?.05, **p鈥?lt;鈥?.01; ANOVA followed by Bonferroni.Full size imageIn summary, polySia avDP20 reduced the phagocytosis of neural debris and fibrillary amyloid-尾1鈥?2 and completely inhibited the associated superoxide release.Neuroprotective activity of low molecular weight polySiaNext, we tested polySia avDP20 in a co-culture of human macrophage and neurons. THP1 derived macrophage cells were added to the neurons and neurite length was determined after 48鈥塰ours of co-culture (Fig. 7a). Addition of macrophages to neurons slightly decreased the relative neurite length (Fig. 7b). Addition of fibrillary amyloid-尾1鈥?2 to macrophage-neuron co-culture further reduced the relative neurite length (Fig. 7b). In detail, relative neurite length was reduced from 1鈥?/鈭掆€?.08 to 0.59鈥?/鈭掆€?.03 after addition of macrophages and amyloid-尾1鈥?2. At the applied concentration of 1鈥壩糓 amyloid-尾1鈥?2 alone had no neurite reducing effect (Fig. 7b). Next, we added polySia avDP20 (1.5鈥壩糓) to the macrophage-neuron co-culture. PolySia avDP20 did not interfere with the general neurite-reducing effect of the macrophages, but it completely inhibited the neurite-reducing effect of fibrillary amyloid-尾1鈥?2 (Fig. 7c). In detail, addition of fibrillary amyloid-尾1鈥?2 reduced the relative neurite length from 0.91鈥?/鈭掆€?.08 to 0.59鈥?/鈭掆€?.03 (p鈥?鈥?.034), while treatment with polySia avDP20 antagonized this neurotoxic effect (0.9鈥?/鈭掆€?.05; p鈥?鈥?.036; Fig. 7c). Trolox was added to the co-culture as a control to scavenge the radical-mediated neurotoxic effect. Trolox completely antagonized all neurite reducing effects of macrophages.Figure 7Neuroprotective effect of polySia with average degree of polymerization 20.(a) Co-culture of human macrophages and neurons. Cells were double-immunostained with antibodies directed against the macrophage marker CD11b and the neuronal marker neurofilament. Nuclei were labelled by DAPI. Co-cultures were treated with fibrillary amyloid-尾1鈥?2 and polySia avDP20 plus fibrillary amyloid-尾1鈥?2. Loss of neurites observed in amyloid-尾1鈥?2 treated co-cultures was prevented by addition of polySia avDP20. Scale bar: 50鈥壩糾. (b) Relative neurite length of human neurons was determined with or without addition of THP1 macrophages and/or fibrillary amyloid-尾1鈥?2 after 48鈥塰ours. Addition of amyloid-尾1鈥?2 alone did not affect the relative neurite length. Treatment of macrophage-neuron co-culture with amyloid-尾1鈥?2 further reduced the relative neurite length. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. *p鈥?lt;鈥?.05; ANOVA followed by Bonferroni. (c) Relative neurite length of human neurons in macrophage-neuron co-culture. The co-culture was incubated with fibrillary amyloid-尾1鈥?2 and/or polySia avDP20. The amyloid-尾1鈥?2 induced reduction in the relative neurite length was antagonized by polySia avDP20. Data are presented as mean鈥?/鈭掆€塖EM of n鈥?鈥? independent experiments. *p鈥?lt;鈥?.05; ANOVA followed by Bonferroni.Full size imageThus, polySia avDP20 prevented the negative effects mediated by fibrillary amyloid-尾1鈥?2 and radicals on the neurite length of human neurons.DiscussionHere we show that low molecular weight polySia (avDP20) acted anti-inflammatory on human THP1 macrophages by inhibiting the LPS-induced gene transcription and protein secretion of TNFSF2 and by preventing the oxidative burst associated with phagocytosis of debris or fibrillary amyloid-尾1鈥?2. Furthermore, we demonstrate that the anti-inflammatory effect of polySia avDP20 on human macrophages is mediated via SIGLEC-11 receptors. PolySia avDP20 interacting with SIGLEC-11 as determined by ELISA and lentiviral knock-down of SIGLEC11 neutralized the anti-inflammatory effect of polySia avDP20 on TNFSF2 gene transcription and protein secretion. Thus, it is likely that the anti-inflammatory effect of polySia avDP20 is mediated via SIGLEC-11, although we cannot exclude that polySia also might act on other SIGLEC receptors of monocytes/macrophages recognizing 伪2.8-linked Sia (e.g. SIGLEC-5 or SIGLEC-7). SIGLEC-11 has a immunoreceptor tyrosine-based inhibitory motif (ITIM) in the cytosolic tail2, which can counteract the immunoreceptor tyrosine-based activation motif (ITAM) signalling of DAP12/TYROBP14. Upon target recognition by TYROBP-associated phagocytic receptors, the tyrosine of the ITAM will become phosphorylated and activates a series of downstream events, which trigger phagocytosis and NADPH-oxidase mediated release of superoxide14,15. This activatory signaling is counter-regulated by the ITIM containing SIGLEC receptors, which upon ligand binding could inhibit downstream signaling by de-phosphorylation of ITAM tyrosine residues11,14.Expression of SIGLEC-11 was detected in human tissue on specialized tissue macrophages including microglia of the brain, Kupffer cells of the liver, lamina propria macrophages of the intestine and perifollicular cells in the spleen2. Furthermore, expression of SIGLEC-11 was immunostained in macrophages invading inflammatory tissue of the stomach, the tonsil or appendix2. However, it is still unclear how SIGLEC-11 is regulated in human macrophages and how far the THP1-derived macrophage model used in this study can reflect a defined polarization status of macrophages. In cultured murine microglial cells, that ectopically were expressing human SIGLEC-11, the LPS induced gene transcription of the pro-inflammatory mediators, interleukin-1尾 (IL-1尾) and nitric oxide synthase-2 (NOS2) were reduced after artificial stimulation of SIGLEC-11 with cross-linking antibodies3. Furthermore, SIGLEC-11 expression on murine microglia had a neuroprotective effect, which was diminished by sialidase (EndoN) treatment of neurons that led to the removal of 伪2.8-linked Sias from the cell surface3. Thus, our current data with an anti-inflammatory action of polySia avDP20 on human SIGLEC-11 expressing macrophages are in line with these recent in vitro findings.The effect of polySia avDP20 could be explained by multimerization of several SIGLEC-11 receptors. Normally, interaction between a single protein such as a SIGLEC receptor and a carbohydrate such as Sia is of relative low affinity, often insufficient for induction of a biological signal16,17. Therefore, it is necessary that several protein receptors pair together in a way that the carbohydrate ligands are recognized by a cluster of receptors14. This multivalency seems to be essential for protein-carbohydrate interaction and response in a cellular system16,17. Thus, two or more multimerized SIGLEC-11 receptors might recognize one stretch of polySia that is then leading to a biological effect. Short oligoSia would fail to induce such a multimerization of SIGLEC-11 receptors. Indeed, our data show that oligoSia DP6 did not antagonize the LPS-induced increase of TNFSF2 gene transcripts in the human macrophages.PolySia avDP20 also modulated the inflammation associated endocytosis/ phagocytosis. Microbeads are normally ingested by phagocytic receptors, while nanobeads are mainly taken up by receptor-mediated endocytosis. Interestingly, polySia avDP20 antagonized the LPS-triggered increase in phagocytosis without affecting the basal homeostatic phagocytosis and receptor-mediated endocytosis. While beads allow a distinction between basal and inflammation-triggered phagocytosis, debris and fibrillary amyloid-尾1鈥?2 always trigger an inflammatory response in macrophages. Thus, the slightly reduced uptake of debris and fibrillary amyloid-尾1鈥?2 by treatment with polySia avDP20 could be explained by antagonization of the inflammation-triggered phagocytosis. The finding that polySia avDP20 binding to SIGLEC-11 interferes with inflammatory phagocytosis is in agreement with our recent analysis on cultured mouse microglia, where we showed that ectopic expression of SIGLEC-11 in mouse microglia slightly reduced the uptake of debris3. Phagocytosis can be a defense mechanism to remove invading pathogens. In contrast, receptor-mediated endocytosis enables homeostatic uptake of macromolecules into the cells that will be re-utilized. Thus, the discrete effects of polySia avDP20 on inflammatory phagocytosis and receptor-mediated endocytosis as well as basal phagocytosis suggest a role of SIGLEC-11 in recognition of endogenous structures and anti-inflammatory signaling.PolySia avDP20 completely prevented the superoxide release triggered by neural debris or fibrillary amyloid-尾1鈥?2. Recently, mouse Siglec-E has been shown to inhibit the production of superoxide radicals of phagocytes in vitro and in vivo12,13. Thus, human SIGLEC-11 acts similar to mouse Siglec-E suggesting that the ITIM-signaling receptors negatively interfere with the oxidative burst. While polySia avDP20 completely prevented the oxidative burst, phagocytosis was only partially reduced suggesting that polySia avDP20 still allows clearance of endogenous material without oxidative burst.PolySia avDP20 acted neuroprotective in a fibrillary amyloid-尾1鈥?2 challenged human macrophage-neuron co-culture system. Amyloid-尾1鈥?2 increased the neurite diminishing activity of human macrophages. This is in line with the literature, where it was shown that even relative low concentrations of amyloid-尾 can reduce neuronal numbers in the presence of microglia18. The neurotoxic effect of fibrillary amyloid-尾1鈥?2 was neutralized in our human co-culture system by polySia avDP20. Experiments with the radical scavenger Trolox indicate that the neurite diminishing effect of fibrillary amyloid-尾1鈥?2 triggered by macrophages was mediated by reactive oxygen species (ROS). Thus, the protective effect of polySia avDP20 in this co-culture system could be explained by prevention of the macrophages oxidative burst.Our data indicate that polySia avDP20 might be used as a novel therapy for neuroinflammatory diseases by preventing the phagocytes superoxide production and still allowing basal phagocytosis and endocytosis of endogenous structures. PolySia has been used before in therapy approaches, mainly by improving the pharmacokinetics of recombinant therapeutic proteins and antibodies19,20,21. Therefore, therapeutic proteins were decorated with polySia by posttranslational modification with terminal Sias by enzymatic or chemical means19,20,21. The polysialylated conjugates had greater stability, lower immunogenicity and an increased blood circulation time due to the chemico-physical properties of these hydrophilic long chain molecules.Our data now demonstrate that low molecular weight polySia (avDP20) can interact with SIGLEC-11 and can act anti-inflammatory on SIGLEC-11 expressing human macrophages. Thus, polySia avDP20 is a novel interesting candidate for inflammatory diseases involving increased production of superoxide by macrophages.MethodsFragmentation and separation of polySiaPurified 伪2.8-linked polysialic acid (~50鈥?0鈥塳Da8,9; University of Hanover and Carbosynth Limited, UK) was used for fragmentation and anion exchange high-performance liquid chromatography (HPLC) separation and purification. Then strong anion exchange high-performance liquid chromatography (SAX-HPLC) was applied for its separation and purification. First, samples were heated at 65鈥壜癈 for 90鈥塵inutes to induced spontaneous hydrolysis. Then, fragmented polySia was subjected to a 53鈥塵l strong anion-exchange column (Hi Load 26/10 Q, GE-Healthcare) and separated via a HPLC system coupled to a photometric UV detector at 205/280鈥塶m (Amersham Bioscience/GE Healthcare), utilizing a 2鈥塎 NH4HCO3 buffer as solvent with a flow-rate of 4鈥塵l/minute (Table 1). The flow-through was collected in 90 tubes with a respective volume of 8鈥塵l. Each three consecutive tubes were pooled to get 30 fractions. Fractions 3鈥? were collected and analyzed by gel electrophoresis. PolySia with average degree of polymerization (here labeled as polySia avDP20) was selected for further experiments. To get rid of buffer residues the samples were lyophilized and solved in water (Ampuwa, Fresenius).Table 1 Elution conditions of the HPLC for separation of the polySia fractions.Full size tableDetermination of the polySia concentrationDetermination of the concentration of polySia avDP20 was performed with a thiobarbituric acid based method22. Therefore, polySia was pretreated with 1鈥塎 H2SO4 at 80鈥壜癈 for 1鈥塰our in order to hydrolyze the polymer into single n-acetylneuraminic acid (Sia monomers). A standard containing concentrations form 0鈥?0鈥壩糶 Sia monomers (Nacalai Tesque ING, Japan) was prepared. The test samples and the standard were treated with 25鈥壩糽 of 25鈥塵M periodic acid in 0.125鈥塎 H2SO4 and incubated at 37鈥壜癈 for 30鈥塵inutes. Afterwards, 20鈥壩糽 of 2% sodium arsenite solution (in 0.5鈥塏 HCl) was added to each sample to reduce the excess of periodate. After 2鈥塵inutes at room temperature, 200鈥壩糽 of 2-thiobarbituric acid (0.1鈥塎, pH 9) was given to the samples. Subsequently, a heating step (7.5鈥塵inutes at 99鈥壜癈) was performed to induce the formation of a red colored complex. The solution was cooled on ice for 5鈥塵inutes, afterwards shaken with 500鈥壩糽/sample acid butanol (butan-1-ol plus 5% of 12 N HCl) and then centrifuged to separate the two phases. The intensity of the colorful upper phase was measured via spectrometer at 549鈥塶m. Quantification was performed based on the Sia monomers standard.Gel electrophoresis of polySiaPolySia avDP20 was analyzed by 18% Tris-Glycine polyacrylamide gel (Life Technologies GmbH; 2.5鈥塰ours electrophoresis at 125鈥塚). Sulphated dextrans of different sizes were used as marker (TdB Consultancy). Subsequently, the gel was stained via a protocol from Goldberg and Warner for at least 2鈥塰ours at room temperature with Stains-All (Sigma) solution (30鈥塵M Tris, 25% isopropanol, 7.5% formamide and 0.025% (w/v) at a pH of 8.8). To clear the background, the gel was washed with deionized water including 25% isopropanol afterwards.Analytic HPLC analysis of oligoSia DP6 and polySia with avDP20Oligosialic acid with a degree of polymerization of 6 (oligoSia DP6) was obtained from Nacalai Tesque, Japan. The labeling of oligoSia DP6, polySia avDP18, avDP20, avDP22 and hydrolyzed colominic acid with 1,2-diamino-4,5-methylenedioxybenzenewas (DMB) was carried out over at least 48鈥塰ours at 4鈥壜癈 as previously described by Inoue et al.23. Before the run the sample was set to a pH of 8 to improve binding of the Sia to the column. HPLC retention time was monitored via a Mono Q 5/50 column operated with a flow speed of 0.5鈥塵l/minute (GE Healthcare) in a linear gradient from 0鈥?00鈥塵M NaCl in order to elute and separate the single fractions of polySia avDP18, avDP20 and avDP22.Determination of the interaction between SIGLEC-11 and polySia avDP20 by ELISAThe polySia avDP20 was coupled with a biotin molecule at the terminus of the polySia avDP20 chain. Therefore, polySia with avDP20 was oxidized to an aldehyde by sodium metaperiodate. Afterwards, the hydrazide coupled biotin (Thermo Scientific) was bound at room temperature to the aldehyde group to form a hydrazone bond. Purification was carried out with a desalting column (HiTrap Desalting Column GE Healthcare). A 96-well plate was coated with 10鈥壩糶/ml pierce recombinant protein-A (Thermo Scientific) and incubated overnight at 4鈥壜癈. All following steps were carried out at room temperature. The plate was washed 3 times with PBS+ 0.05% tween20 and blocked with 3% BSA for 1鈥塰our. Blocking solution was removed and 5鈥壩糶/ml of recombinant human SIGLEC-11/Fc chimera (R D system) was added to the wells and incubated for 2鈥塰ours. Next, the plate was washed 3 times with PBS+ 0.05% tween20 and blocked with 3% BSA for 1鈥塰our. Afterwards, blocking solution was removed and different concentrations of biotinylated-polySia avDP20 (0.01, 0.05, 0.25, 1.25, 6.25鈥壩糶/ml) or biotinylated dextran (5鈥塳Da, NANOCS) serving as control (0.01, 0.05, 0.25, 1.25, 6.25鈥壩糶/ml diluted in PBS) were added and incubated for 2鈥塰ours. The plate was washed 3 times with PBS鈥?鈥?.05% tween20 and incubated with streptavidin-HRP (0.5鈥壩糽/ml, BD Bioscences, Pharmingen) for 1鈥塰our, followed by 3 times washing with PBS+ 0.05% tween20. At the end 100鈥壩糽 TMB (Sigma) were added for 15鈥塵inutes and the reaction was stopped by 100鈥壩糽 1N HCL. The signal was measured at 450鈥塶m by an ELISA plate reader (PerkinElmer).Flow cytometry analysisA novel SIGLEC-11 specific mouse monoclonal antibody (clone 3 EH) binding up to a dilution of 1:128鈥塊 to the SIGLEC-11 specific peptide ISISHDNTSALE was generated by Abmart. To confirm the binding specificity of the antibody, HEK293-T cells were transfected with plenti-EF1伪-SIGLEC11- or -SIGLEC16-IRES-GFP over-expression plasmid as well as a plenti-EF1伪-IRES-GFP control plasmid. Transfection procedure was carried out using Lipofectamine 2000 (Invitrogen). In brief, 2.5鈥壝椻€?06 cells were plated on 10鈥塩m cell culture dishes. Regular cell culture medium was changed directly before the procedure to Opti-MEM medium with 10% FCS. The plasmids (5鈥壩糶) and Lipofectamine 2000 (90鈥壩糽) were combined according to the manufacturer鈥檚 instructions and added for 24鈥塰ours to the cells. Afterwards, cells were detached mechanically and washed with PBS. Cells were stained for protein expression by the SIGLEC-11 specific antibody (1:500) followed by Cy-5 fluorescence-labelled secondary antibody (Jackson Laboratories Inc, USA, 1:200). Control samples were incubated with an isotype control and secondary fluorescence labelled antibodies. Analysis was performed with a flow cytometer (BD, Calibur) and the FlowJo 8.7 Software (Tree Star Inc.).THP1 cell cultureThe human monocytic cell line THP1 (kindly provided by V. Hornung, University of Bonn) derived from an acute monocytic leukemia patient was used to obtain macrophages. THP1 monocytes were cultured in medium containing RPMI (Gibco) supplemented with 10% defined fetal bovine serum (Gibco), 1% Penicillin-Streptomycin (Gibco), 1% L-glutamine (Gibco) and 1% sodium pyruvate (Gibco). Macrophage cultures were differentiated repeatedly from frozen stocks of THP1 cells. At least one week before experiment, THP1 monocytes were moved to THP1 differentiation medium containing RPMI with 1% Penicillin-Streptomycin, 1% L-glutamine, 1% sodium pyruvate plus 1% N2 supplement (Gibco) and 1% chicken serum (Gibco). To differentiate THP1 cells into a macrophage phenotype, the cells were treated in THP1 differentiation medium with 0.5鈥壩糶/ml phorbol-12-myristate-13-acetate (PMA, Sigma) for 3鈥塰ours. Afterwards, cells were washed 2 times with 37鈥壜癈 warm medium and were kept in PMA-free THP1 differentiation medium for at least 24鈥塰ours. Then a serum free medium (THP1 differentiation medium without chicken serum) was applied to the macrophages.Phagocytosis of microbeads and endocytosis of nanobeadsTo investigate phagocytosis activity, macrophages were plated/differentiated on 6-well-plates in a concentration of 2.5鈥壝椻€?05 cells per well. Cells were stimulated for 24鈥塰ours with 1鈥壩糶/ml LPS (InvivoGen, E.coli 0111:B4 strain) prior to the addition of micro- or nanobeads. PolySia was added to the micro- or nanobeads and incubated at 4鈥壜癈. After 1鈥塰our, beads (microbeads from Polyscience Inc., PE labeled, 1鈥壩糾 bead diameter; nanobeads from Life technologies, yellow-green labeled, 0.04鈥壩糾 bead diameter) were added in a concentration of 1鈥壩糽 beads/ml medium (microbeads) or 0.1鈥壩糽 beads/ml medium (nanobeads) for 1鈥塰our (microbeads) or 5鈥塵inutes (nanobeads) to the cells. All steps were performed under regular cell culture conditions (37鈥壜癈, 5% CO2). For analysis, media was removed and cells were treated for 1鈥?鈥塵inutes with trypsin (Gibco) in order to get rid of beads sticking to the surface of macrophage cells. Afterwards cells were washed 3-times with PBS before being detached mechanically. In addition, nanobeads treated cells were fixed for 15鈥塵inutes with 2% paraformaldehyde (PFA) solution before flow cytometry analysis to stop endocytosis. Flow cytometry analysis was performed with a FACS Calibur (BD Biosciences).Determination of cell viability by the MTT-assayCell viability was determined by the 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Millipore). Cells were plated on a 96 well plate (2.5鈥壝椻€?04) treated for 24鈥塰ours with different concentrations of polySia avDP20. To generate physiologic conditions a pH of 7 was adjusted for polySia avDP20 at higher concentrations (15鈥壩糓 or higher) using sodium hydroxide and 50鈥壩糓 Tris. Additionally, the colour of phenol red in the culture medium was closely monitored to confirm absence of major pH changes. After 20鈥塰ours of stimulation, 10鈥壩糽 of MTT-reagent was added to each well and cells were cultured for another 4鈥塰ours. Then, cells were lysed with isopropanol containing 0.04 N HCL to measure the dye formazan. The light absorbance of the purple formazan dye was determined by a spectrophotometer at a wavelength of 570鈥塶m with a reference wavelength of 630鈥塶m (Perkin Elmer, Envision Multiplate Reader). Macrophage cell numbers were determined via a Neubauer cell chamber. The data from the MTT-assay of the macrophages were normalized to the cell number. Measured absorbance values were divided through the number of cells per well to determine the turnover per single cell. All values were compared to unstimulated control cells to receive the relative changes of cell proliferation and viability.Gene transcript analysis of SIGLEC7, SIGLEC10, SIGLEC11 and TNFSF2Total RNA was collected from cells via the RNeasy kit system (Quiagen). Reverse transcription of the RNA was performed using Super Script III reverse transcriptase (Life Technologies) and hexamer random primers (Roche). SIGLEC7 (for 5鈥?GCTCAGCCTCTCAGGGTAAC-3鈥? rev 5鈥?GGGTCGTGAACCCTCAAACA -3鈥?, SIGLEC10 (for 5鈥?GGACTCAGACAGAAACCCCG-3鈥? rev 5鈥?TGTGGCTTTCTGATTCCGCT-3鈥? and SIGLEC11 (for 5鈥?CACTGGAAGCTGGAGCATGG-3鈥? rev 5鈥?ATTCATGCTGGTGACCCTGG-3鈥? PCR was performed at Tm of 60鈥壜癈 for 35 cycles with GAPDH as loading control (for 5鈥?CTGCACCACCAACTGCTTAG-3鈥? rev 5鈥?TTCAGCTCAGGGATGACCTT-3鈥?. Quantitative qRT-PCR with specific oligonucleotides was performed with SYBR Green PCR Master Mix (Qiagen) using the ABI 5700 Sequence Detection System (PerkinElmer). The qRT-PCR was running for 40 cycles with a Tm of 60鈥壜癈 using the same primer as for the PCR plus TNFSF2 (for 5鈥?GACAAG CCTGTAGCCCATGT-3鈥?rev 5鈥?AGGAC CTGGGAGTAGATGAGG-3鈥?. The 未未CT method with GAPDH as internal standard was performed for qRT- PCR quantification.Plasmid construction, viral particle production and lentiviral transductionPlasmids for lentiviral knock-down of SIGLEC11 (shRNASig11: TRCN0000062842, Open Biosystems) were obtained from a knock-down library in a human pLKO.1 lentiviral shRNA target gene set backbone (Open Biosystems). A pLenti 6.2/V5_DEST Gateway Vector (Life technologies) without or scrambled target gene served as control vector. For lentiviral virus production, HEK293-FT cells (Invitrogen) were transfected with the targeting and packaging plasmids (Addgene). For transfection, calcium chloride (CaCl2) was used. Medium was changed 16鈥塰ours post-transfection. Supernatant was collected 48鈥塰ours after transfection. For concentration, supernatant containing the viral particles was incubated for 1.5鈥塰ours with 50% Polyethylene glycol (Roth GmbH), 4鈥塎 NaCl and 1xPBS at 4鈥壜癈. Subsequently the solution was centrifuged with 4500鈥塯 at 4鈥壜癈 for at least 30鈥塵inutes. The viral particle containing pellet was resuspended in 1鈥塵l of medium and added to the target cells. After 48鈥塰ours of incubation transduced cells were selected by adding 1鈥壩糶/ml puromycin. The knock-down efficiency was determined by qRT-PCR and flow cytometry.TNFSF2 protein detection by ELISAFor the detection of TNFSF2 release by THP1 cells a Quantikine ELISA kit (R D Systems) was used. Macrophages were stimulated with 1 or 50鈥壩糶/ml LPS and 1.5鈥壩糓 polySia avDP20 for 24鈥塰ours. Supernatant was harvested and used according to the manufacturer鈥檚 instructions. Optical density of the ELISA samples was determined by a spectrophotometer at a wavelength of 450鈥塶m with a reference wavelength of 560鈥塶m (Perkin Elmer, Envision Multiplate Reader).Generation of human neurons from iPS cell linesHuman iPS cell line (Foreskin-1, WiCell24) was used for generation of neurons out of primitive neural stem cells (pNSCs). Therefore, iPS cells were cultured on feeder cells to form small colonies. Next, medium was changed to neural stem cell medium (DMEM/F12: Neurobasal; GIBCO) in the presence of leukaemia inhibiting factor (LIF; Millipore, 10鈥塶g/ml) and three small molecules CHIR99021 (inhibitor of GSK-3尾, Axon Medchem, 4鈥壩糓) and SB431542 (inhibitor of TGF-尾 and activin receptors; Axon Medchem, 3鈥壩糓) and Compound E (inhibitor of 纬-secretase; Axon Medchem, 0.1鈥壩糓) for 10 days. To induce differentiation towards neurons, pNSCs were dissociated by accutase (PAA) and cultured on poly-L-ornithine (Sigma, 0.15鈥塵g/ml) plus laminin (Sigma, 1鈥壩糶/ml) coated cell culture dishes in neural stem cell medium (DMEM/F12:Neurobasal, plus LIF, CHIR99021 and SB431542) till cell attached and formed small colonies. Then, medium was changed to neuronal differentiation medium (DMEM/F12, plus N2 and B27 supplements, GIBCO) plus ascorbic acid (TOCRIS, 0.2鈥塵M) and cyclic adenosine monophosphate (Sigma, 300鈥塶g/ml) in presence of brain derived neurotrophic factor (BDNF; Prospect, 10鈥塶g/ml) and glial cell line-derived neurotrophic factor (GDNF; Prospect, 10鈥塶g/ml) for 2 weeks. Medium containing the neurotrophic factors was changed every second day.Phagocytosis of debrispNSCs from an established induced pluripotent (iPS) stem cell line (WiCell, foreskin-1)24 as described above were incubated with 40鈥塶M okadaic acid (Sigma) for 24鈥塰ours. Then, cellular debris was collected and centrifuged. Debris was washed with PBS and stained with 鈥淒il-Derivatives for Long-Term Cellular Labeling鈥?Molecular Probes (Invitrogen) according to the supplier麓s manual. Macrophages were pre-incubated for 1鈥塰our with different concentrations of polySia avDP20 (0.15, 0.5 and 1.5鈥壩糓) followed by 1.5鈥塰our incubation with 5鈥壩糶/鈥壩糽 pre-stained cell debris. Cells were fixed for 15鈥塵inutes with 4%PFA, blocked and permeabilized for 60鈥塵inutes with a solution that contained bovine serum albumin (10% BSA) and normal goat serum (5% nGS) and 0.1% TritonX-100. Then cells were stained with a CD11b antibody (Becton Dickinson) followed by a secondary Alexa488-conjugated antibody (Molecular probes). For analysis, images were randomly scanned and 3D-reconstructions were obtained by a confocal laser scanning microscope (Fluoview 1000, Olympus). To determine the ratio of cells having ingested fluorescently labeled material, five images per experiment group were obtained and all cells on the images were quantified using Image J software (NIH).Phagocytosis of fibrillary amyloid-尾1鈥?2To obtain fibrillary amyloid-尾1鈥?2 forms, biotinylated amyloid-尾1鈥?2 (Bachem) was incubated for 72鈥塰ours at 37鈥壜癈 as previously described25. Macrophages were pre-incubated for 1鈥塰our with different concentrations of polySia avDP20 (0.15, 0.5 and 1.5鈥壩糓), followed by 1.5鈥塰our incubation with 2鈥壩糓 fibrillary biotinylated amyloid-尾1鈥?2. Cells were fixed for 15鈥塵inutes with 4% PFA, blocked and permeabilized for 60鈥塵inutes with a solution that contained 10% BSA and 5% nGS and 0.1% TritonX-100. Afterwards, cells were incubated with a CD11b antibody (Becton Dickinson) followed by a secondary Alexa 488-conjugated antibody directed against rat IgG (Molecular probes) and streptavidin-Cy3 (Jackson ImmunoResearch). For analysis, images were randomly obtained with a confocal laser scanning microscope and the fluorescent labeled amyloid-尾1鈥?2 was visualized inside macrophages by 3D-reconstruction (Fluoview 1000, Olympus). To determine the ratio of cells having ingested fluorescent marked amyloid-尾1鈥?2, five images per experimental group were collected and all cells of the images were analyzed using Image J software (NIH).Detection of superoxide by dihydroethidiumTo measure the relative production of superoxide by macrophages, cells were differentiated in 4-chamber culture dishes. After 48鈥塰ours cells were treated with either 10鈥壩糓 fibrillary amyloid-尾1鈥?2 or 5鈥壩糶/渭l debris for 15鈥塵inutes with or without 1鈥塰our polySia avDP20 pre-incubation. To test the antioxidant effect of SOD1 or Trolox as positive controls, macrophage cells were pre-incubated 1鈥塰our either with 60鈥塙/ml SOD1 (Serva) or 40鈥壩糓 Trolox (Cayman), then fibrillary amyloid-尾1鈥?2 or debris added to them. Afterwards, cells were washed 2-times with Krebs-HEPES-buffer and afterwards incubated for 15鈥塵inutes with 30鈥壩糓 DHE solution (diluted in Krebs-HEPES-buffer). Finally cells were washed 2-times with Krebs-HEPES-buffer and fixed for 15鈥塵inutes with 0.25% glutaraldehyde and 4% PFA. In total, three images were randomly collected per experimental group by confocal laser scanning microscopy (Fluoview 1000, Olympus). All cells of the collected images were analyzed by Image J software (NIH).Macrophage-neuron co-culture and analysis by immunocytochemistryTHP1 macrophage cells (ratio of 1:5 of macrophage:neurons) and 1鈥壩糓 fibrillary amyloid-尾1鈥?2 were added to the iPS cell-derived neurons (see above) with or without polySia avDP20 for 48鈥塰ours in differentiation medium without growth factors. To use Trolox as a positive control, 40鈥壩糓 Trolox were added at the same time with or without amyloid-尾 or polySia avDP20. Cells were fixed for 15鈥塵inutes with 4% PFA, blocked and permeabilized for 60鈥塵inutes with a solution that contained 10% BSA and 5% nGS and 0.1% TritonX-100. Next, immunostained with monoclonal rat anti-CD11b (BD pharmingen) and neurofilament (Sigma) antibodies overnight at 4鈥壜癈 followed by secondary Cy3-conjugated goat antibody directed against rat IgG (Dianova) Alexa488-conjugated antibody directed against rabbit IgG (Molecular Probes) for 2鈥塰ours at room temperature. five images were randomly collected from each experimental setup by confocal laser scanning microscopy (Fluoview 1000, Olympus) and total lengths of neuronal branches from neurofilament stained neurites was determined by the NIH ImageJ/NeuronJ software.Statistical analysisNormal distribution of data was checked by Kolmogorov-Smirnov test. Levene鈥檚 test was used to check equal distribution of data. Data are presented as mean鈥?/鈭掆€塖EM of at least 3 independent experiments. Data were analyzed by student t-test for experiments with two groups only and by one-way ANOVA followed by Bonferroni test (when equal variances were assumed) using SPSS20 software.Use of human cell linesAll experiments were performed in accordance with relevant guidelines and regulations. All experiments involving established and commercially available human cell lines have been performed in accordance with the ethical guidelines of the University of Bonn and the national legislations.Additional InformationHow to cite this article: Shahraz, A. et al. Anti-inflammatory activity of low molecular weight polysialic acid on human macrophages. Sci. Rep. 5, 16800; doi: 10.1038/srep16800 (2015). ReferencesHayakawa, T. et al. A human-specific gene in microglia. Science 309, 1693 (2005).CAS聽 PubMed聽Google Scholar聽 Angata, T. et al. Cloning and characterization of human Siglec-11. A recently evolved signaling that can interact with SHP-1 and SHP-2 and is expressed by tissue macrophages, including brain microglia. J Biol Chem 277, 24466鈥?4474 (2002).CAS聽 Article聽Google Scholar聽 Wang, Y. Neumann, H. Alleviation of neurotoxicity by microglial human Siglec-11. J Neurosci 30, 3482鈥?488 (2010).CAS聽 Article聽Google Scholar聽 Hildebrandt, H. Dityatev, A. Polysialic Acid in Brain Development and Synaptic Plasticity. Top Curr Chem 366, 55鈥?6 (2013).Article聽Google Scholar聽 Sato, C. Kitajima, K. Disialic, oligosialic and polysialic acids: distribution, functions and related disease. J Biochem 154, 115鈥?36 (2013).CAS聽 Article聽Google Scholar聽 Oltmann-Norden, I. et al. Impact of the polysialyltransferases ST8SiaII and ST8SiaIV on polysialic acid synthesis during postnatal mouse brain development. J Biol Chem 283, 1463鈥?471, 10.1074/jbc.M708463200 (2008).CAS聽 Article聽 PubMed聽Google Scholar聽 Troy, F. A. 2nd. Polysialylation: from bacteria to brains. Glycobiology 2, 5鈥?3 (1992).CAS聽 Article聽Google Scholar聽 Rode, B. et al. Large-scale production and homogenous purification of long chain polysialic acids from E. coli K1. J Biotechnol 135, 202鈥?09 (2008).CAS聽 Article聽Google Scholar聽 Bice, I. et al. Downstream processing of high chain length polysialic acid using membrane adsorbers and clay minerals for application in tissue engineering. Engineering in Life Sciences 13, 140鈥?48, 10.1002/elsc.201200041 (2013).CAS聽 Article聽Google Scholar聽 Manzi, A. E., Higa, H. H., Diaz, S. Varki, A. Intramolecular self-cleavage of polysialic acid. J Biol Chem 269, 23617鈥?3624 (1994).CAS聽 PubMed聽Google Scholar聽 Linnartz-Gerlach, B., Kopatz, J. Neumann, H. Siglec functions of microglia. Glycobiology 24, 794鈥?99, 10.1093/glycob/cwu044 (2014).CAS聽 Article聽 PubMed聽Google Scholar聽 Claude, J., Linnartz-Gerlach, B., Kudin, A. P., Kunz, W. S. Neumann, H. Microglial CD33-Related Siglec-E Inhibits Neurotoxicity by Preventing the Phagocytosis-Associated Oxidative Burst. J Neurosci 33, 18270鈥?8276 (2013).CAS聽 Article聽Google Scholar聽 Schwarz, F. et al. Siglec receptors impact mammalian lifespan by modulating oxidative stress. eLife 4, 10.7554/eLife.06184 (2015).Linnartz, B. Neumann, H. Microglial activatory (immunoreceptor tyrosine-based activation motif)- and inhibitory (immunoreceptor tyrosine-based inhibition motif)-signaling receptors for recognition of the neuronal glycocalyx. Glia 61, 37鈥?6 (2013).Article聽Google Scholar聽 Linnartz, B., Wang, Y. Neumann, H. Microglial immunoreceptor tyrosine-based activation and inhibition motif signaling in neuroinflammation. Int J Alzheimers Dis Volume 2010, 10.4061/2010/587463 (2010).Cohen, M. Varki, A. Modulation of glycan recognition by clustered saccharide patches. International review of cell and molecular biology 308, 75鈥?25, 10.1016/B978-0-12-800097-7.00003-8 (2014).CAS聽 Article聽 PubMed聽Google Scholar聽 Dam, T. K. Brewer, C. F. Lectins as pattern recognition molecules: the effects of epitope density in innate immunity. Glycobiology 20, 270鈥?79 (2010).CAS聽 Article聽Google Scholar聽 Neniskyte, U., Neher, J. J. Brown, G. C. Neuronal death induced by nanomolar amyloid beta is mediated by primary phagocytosis of neurons by microglia. J Biol Chem 286, 39904鈥?9913, 10.1074/jbc.M111.267583 (2011).CAS聽 Article聽 PubMed聽 PubMed Central聽Google Scholar聽 Lindhout, T. et al. Site-specific enzymatic polysialylation of therapeutic proteins using bacterial enzymes. Proc Natl Acad Sci USA 108, 7397鈥?402, 10.1073/pnas.1019266108 (2011).ADS聽 Article聽 PubMed聽Google Scholar聽 Gregoriadis, G., Jain, S., Papaioannou, I. Laing, P. Improving the therapeutic efficacy of peptides and proteins: a role for polysialic acids. Int J Pharm 300, 125鈥?30 (2005).CAS聽 Article聽Google Scholar聽 Constantinou, A., Epenetos, A. A., Hreczuk-Hirst, D., Jain, S. Deonarain, M. P. Modulation of antibody pharmacokinetics by chemical polysialylation. Bioconjug Chem 19, 643鈥?50, 10.1021/bc700319r (2008).CAS聽 Article聽 PubMed聽Google Scholar聽 Aminoff, D. Methods for the quantitative estimation of N-acetylneuraminic acid and their application to hydrolysates of sialomucoids. Biochem J 81, 384鈥?92 (1961).CAS聽 Article聽Google Scholar聽 Inoue, S., Lin, S. L., Lee, Y. C. Inoue, Y. An ultrasensitive chemical method for polysialic acid analysis. Glycobiology 11, 759鈥?67 (2001).CAS聽 Article聽Google Scholar聽 Yu, J. et al. Induced pluripotent stem cell lines derived from human somatic cells. Science 318, 1917鈥?920 (2007).CAS聽 ADS聽 Article聽Google Scholar聽 Liu, Y. et al. LPS receptor (CD14): a receptor for phagocytosis of Alzheimer鈥檚 amyloid peptide. Brain 128, 1778鈥?789 (2005).Article聽Google Scholar聽 Download referencesAcknowledgementsThis project was supported by the DFG (FOR1336, KFO177, SFB704) and the Hertie foundation. H.N is member of the DFG-funded excellence cluster 鈥業mmunoSensation鈥?(EXC 1023). This project was kindly supported by a grant from Bayer Pharma AG (鈥榝rom targets to novel drugs鈥?2014-08-1139). We very much thank Rita Gerady-Schahn and Ajit Varki for helpful comments and open discussions. We thank Veit Hornung for kindly provide us THP monocytes. We thank WiCell for providing us iPS (Foreskin)-1 cells25. We thank Jessica Reinartz and Rita Jietou for excellent technical support.Author informationAffiliationsInstitute of Reconstructive Neurobiology, University Hospital Bonn, University of Bonn, Bonn, 53127, GermanyAnahita Shahraz,聽Jens Kopatz,聽Rene Mathy,聽Shoba Kapoor,聽Vlad Sch眉tza聽 聽Harald NeumannInstitute of Biochemistry and Molecular Biology, University Hospital Bonn, University Bonn, Nussallee 11, Bonn, 53115, GermanyJoachim Kappler,聽Dominic Winter聽 聽Volkmar GieselmannInstitute of Technical Chemistry, Leibnitz University Hannover, Callinstra脽e 5, Hannover, 30167, GermanyThomas ScheperAuthorsAnahita ShahrazView author publicationsYou can also search for this author in PubMed聽Google ScholarJens KopatzView author publicationsYou can also search for this author in PubMed聽Google ScholarRene MathyView author publicationsYou can also search for this author in PubMed聽Google ScholarJoachim KapplerView author publicationsYou can also search for this author in PubMed聽Google ScholarDominic WinterView author publicationsYou can also search for this author in PubMed聽Google ScholarShoba KapoorView author publicationsYou can also search for this author in PubMed聽Google ScholarVlad Sch眉tzaView author publicationsYou can also search for this author in PubMed聽Google ScholarThomas ScheperView author publicationsYou can also search for this author in PubMed聽Google ScholarVolkmar GieselmannView author publicationsYou can also search for this author in PubMed聽Google ScholarHarald NeumannView author publicationsYou can also search for this author in PubMed聽Google ScholarContributionsA.S. and J.Ko. performed experiments, analyzed data and drafted part of the manuscript; J.Ko., V.S., J.Ka., D.W., T.S. and V.G. generated and analyzed the polysialic acid fractions; A.S., R.M. and S.K. analyzed the interaction between polySia avDP20 and SIGLEC11; HN initiated and planned the study and wrote the manuscript.Ethics declarations Competing interests A.S. (PhD student of the University of Bonn), J.K. (Postdoctoral fellow of the University of Bonn) and H.N. (Professor at the University of Bonn) are named inventors on a patent application related to the use of polysialic acid for neurodegenerative diseases filed by the Universities of Bonn and Cologne. Rights and permissions This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article鈥檚 Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Reprints and PermissionsAbout this articleCite this articleShahraz, A., Kopatz, J., Mathy, R. et al. Anti-inflammatory activity of low molecular weight polysialic acid on human macrophages. Sci Rep 5, 16800 (2015). https://doi.org/10.1038/srep16800Download citationReceived: 10 May 2015Accepted: 20 October 2015Published: 19 November 2015DOI: https://doi.org/10.1038/srep16800 Divya G. Shastry, Flaviyan Jerome Irudayanathan, Asher Williams, Mattheos Koffas, Robert J. Linhardt, Shikha Nangia Pankaj Karande Scientific Reports (2020) Sameera Iqbal, Lindsay M. Parker, Arun V. Everest-Dass, Edward S. X. Moh, Nima Sayyadi, Mark R. Hutchinson Nicolle H. Packer Molecular Neurobiology (2020) CommentsBy submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate. Sign up for the Nature Briefing newsletter 鈥?what matters in science, free to your inbox daily.