Published online before print June 12, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||


* Institute of Human Genetics, University of Regensburg, Regensburg, Germany; and
Center of Excellence for Fluorescent Bioanalytics, Regensburg, Germany
2 Correspondence: Institute of Human Genetics, University of Regensburg, Franz-Josef-Strauss-Allee 11, 93053 Regensburg, Germany. E-mail: thomas.langmann{at}klinik.uni-regensburg.de
|
|
|---|
-stimulated primary microglia using real-time quantitative RT-PCR validation. Importantly, a unique transcriptional phenotype with anti-inflammatory and IFN-
counter-regulatory properties partially related to alternatively activated macrophages was identified. Using functional cell assays, we found that CSPG-DS-stimulated microglia possess increased phagocytic capacity but lack direct cytotoxic effects such as secretion of NO. Furthermore, conditioned media from CSPG-DS-treated microglia did not diminish the viability or cause apoptosis of cultured photoreceptor cells and partially rescued these cells from IFN-
-induced apoptosis. Taken together, our data provide a unique transcript dataset and important in vitro findings about the functional properties of CSPG-DS-activated microglia. These might be starting points to explore the in vivo role of CSPG-DS as a bioactive microglia regulator and its potential, therapeutic application in immune-related, neurodegenerative disorders.
Key Words: neurodegeneration microglia activation neuroprotection proteoglycans
Microglia cells exert various neurotrophic and regulatory functions in the CNS and the retina, sustaining the immune privilege of the nervous system [1 , 2 ]. However, chronic and exaggerated microglia activation contributes to tissue damage and neuronal apoptosis in neurodegenerative diseases including retinal dystrophies [3 , 4 ]. Hence, bioactive molecules attenuating microglia overactivation could be envisioned as neuroprotective therapy options. Recent evidence suggests that a disaccharide degradation product of the matrix component chondroitin sulfate proteoglycan-disaccharide (CSPG-DS) inhibits retinal degeneration by partially affecting microglia activation [5 6 7 ].
Here, we have comprehensively characterized the influence of CSPG-DS on immortalized BV-2 microglia cells and primary murine microglia by using genome-wide expression profiling and studied their effects on cultured photoreceptor cells with several functional analyses. BV-2 microglia-like cells were cultured as described elsewhere [8
] and were stimulated with 50 µg/ml CSPG-DS (Sigma, Steinheim, Germany) for 24 h. IFN-
(20 ng/ml, Sigma) was used as a proinflammatory control to induce a classically activated phenotype of BV-2 cells [9
]. In the first set of experiments, BV-2 cells were stimulated with CSPG-DS or control vehicle (PBS) for 24 h. Total RNA was extracted using the RNeasy Protect Mini kit (Qiagen, Hilden, Germany) and quality-controlled with the RNA 6000 Nano LabChip (Agilent Technologies, Santa Clara, CA, USA). DNA microarray analysis was carried out with Affymetrix 430 2.0 mouse genome arrays according to the Affymetrix standard protocol. Duplicate microarrays were performed with RNA from independent cultures of unstimulated and CSPG-DS-treated cells. Data analysis was performed as described previously using the ChipInspector software (Genomatix Software GmbH, Munich, Germany) [10
, 11
], and minimal information about a microarray experiment criteria was met [12
]. The microarray dataset of this study is publicly available (http://www.ncbi.nlm.nih.gov/geo/). Applying a significance analysis of microarrays at a false discovery rate of 5%, a minimum log2 ratio of 1.5 (fold change of 2.9), and a threshold level of 3 significant, single probes, 29 significantly up-regulated and 25 down-regulated genes were identified (Supplemental Table 1). For a functional annotation of these CSPG-DS-regulated transcripts, Database for Annotation, Visualization, and Integrated Discovery [10
] and the GenMap annotator and pathway profiler (Gladstone Institutes, San Francisco, CA, USA) were used. Of interest, clearly separated biological pathways were represented by the induced and repressed genes, respectively (Fig. 1A
and 1B
). The major functional categories in the CSPG-DS-induced group (Fig. 1A)
were phagocytosis, detoxification, migration, immune regulation, and antigen presentation. A thorough comparison with gene expression profiles reported in genomic databases and in the literature showed a large overlap with the transcriptional phenotype of alternatively activated or the M2 type of tissue macrophages [11
] and little accordance with classically activated microglia [12
]. Several transcripts in the up-regulated cluster including Gas6, selenoprotein P, C-type lectin 4n (Clec4n), and Clec7a were previously linked to trophic, anti-inflammatory, and tissue-remodeling functions of embryonic phagocytes [13
] and phagocytic microglia [14
]. Moreover, this gene set contains several markers for migrating phagocytic microglia with documented neuroprotective functions, including the detoxification enzymes carbonyl reductase 2 [15
] and GST m1 [16
]; the migration regulators Cxcr4 [17
], heparanase [18
], and Cd39/ENTPDase 1 [19
]; and the immune attenuator TNFL superfamiliy 12/TNF-like weak inducer of apoptosis [20
]. Of note, we also identified a strong and inducible expression of the costimulator Cd28, which is normally present only on T cells and neutrophils [21
]. In the group of down-regulated genes, cell-cyle and nucleotide metabolism genes were significantly over-represented, together with the proinflammatory genes Ifi44, Irf7, and Cxcl10 [12
], which were all highly repressed by CSPG-DS. These expression patterns clearly point toward a phagocytic and mainly anti-inflammatory activation profile of CSPG-DS-stimulated microglia.
![]() View larger version (38K): [in a new window] |
Figure 1. Gene-expression profiles of CSPG-DS and IFN- -stimulated microglia. (A and B) Pie chart with over-represented functional categories of the 29 significantly up-regulated (A) and 25 down-regulated (B) transcripts in 24 h, 50 µg/ml CSPG-DS-treated BV-2 microglia-like cells. (C–J) Real-time quantitative RT-PCR (qRT-PCR) analysis of unstimulated, 50 µg/ml CSPG-DS, 20 ng/ml IFN- , and CSPG-DS + IFN- -cotreated BV-2 and primary microglia cells (MG). The data shown are mean fold changes ± SD of three independent biological replicates for BV-2 cells and two independent experiments for primary microglia. Histocompatibility 2 class II antigen A β 1 (H2Ab1), Fc R IIIa (FcgrIIIa), Cbr2, growth arrest-specific 6 (Gas6), Cd28, IFN-induced protein 44 (Ifi44), IFN-regulatory factor 7 (Irf7), and Cxcl10 transcript abundance was normalized to a set of three reference genes (Atp5b, Gusb, and Hprt1), which were not significantly different in all analyzed cDNA sources.
|

comparative threshold method for relative quantitation using vehicle-stimulated cells as calibrators. Normalization to three stable reference genes was performed as described earlier [22
]. The differential expression of all 29 up-regulated and the top 10 down-regulated genes in BV-2 cells could be replicated by real-time qRT-PCR (Supplementary Table 1).
As several biological differences between immortalized BV-2 cells and primary microglia have been described [23
], we next determined whether the transcriptional effects of CSPG-DS are also present in cultured primary microglia. Furthermore, we were interested in evaluating a potential counter-regulatory role of CSPG-DS on IFN-
-mediated gene expression and performed costimulation experiments. We selected two genes, H2Ab1 and FcgrIIIa, which showed particularly high microarray expression changes in the group of phagocytosis-related transcripts. CSPG-DS treatment of BV-2 and primary microglia cells for 24 h strongly enhanced mRNA levels of both genes to a similar extent as IFN-
, whereas coadministration of CSPG-DS with IFN-
did not show an additive effect (Fig. 1C
and 1D)
. Interestingly, a trend to higher responses was identified in primary microglia compared with BV-2 cells in most gene expression experiments. In both cell types, treatment with CSPG-DS also strongly increased Gas6, Cbr2, and Cd28 transcripts (Fig. 1E
1F
1G)
, whereas mRNA levels of the proinflammatory mediators Ifi44, Irf7, and Cxcl10 were potently repressed (Fig. 1H
1I
1J)
. Of note, for all of these six CSPG-DS-regulated genes, IFN-
had the opposite effect. In combined treatment experiments, CSPG-DS was able to abolish the inhibitory activity of IFN-
on Cbr2, Gas6, and Cd28 transcription (Fig. 1E
1F
1G)
and at least dampened the induction of Ifi44 with a minor effect on Irf7 and Cxcl10 transcription (Fig. 1H
1I
1J)
. These data reveal that CSPG-DS and IFN-
-stimulated cells represent completely polarized phagocyte phenotypes with different genetic programs. It is currently an open question whether CSPG-DS acts via a surface receptor or directly penetrates into the cells to initiate an anti-inflammatory signaling cascade. A potential mechanism could be the interference with NF-
B activation, which has been demonstrated recently for T cells stimulated with a heparan sulfate disaccharide [24
].
To assess whether the particular gene expression profile measured in CSPG-DS-stimulated BV-2 cells and primary microglia also translates into phagocytic, anti-inflammatory, and neuroprotective properties, functional assays were performed. We first used a phagocytosis assay to test the capability of CSPG-DS to induce particle ingestion. Blue Latex beads (1 µm) from polystyrene (Sigma) were added to the wells at a concentration of 1 µl beads/ml, and cells were washed with PBS after 8 h incubation. Representative micrographs (Fig. 2A
2B
2C
2D
2E
2F
) showed that bead engulfment was only moderate in unstimulated BV-2 cells and primary microglia, respectively (Fig. 2A
and 2D)
but was strongly enhanced by CSPG-DS in both cell types (Fig. 2B
and 2E)
. IFN-
also increased phagocytosis but less potently than CSPG-DS (Fig. 2C
and 2F)
. For quantitation, the phagocytosis potential expressed as percent-positive cells was monitored by counting the optical bead density in up to 100 cells from six individual wells using light microscopy. This analysis confirmed our initial findings and further revealed that primary microglia showed a significantly higher phagocytic response after CSPG treatment than BV-2 cells (Fig. 2G
and 2H)
. Our data clearly indicate a prominent bioactive effect of CSPG-DS on the debris-clearing function of microglia. This may be mediated by stimulating prophagocytic transcriptional mechanisms, such as an increased expression of phagocytic receptors (Fig. 1C
and 1D)
, or an enhancement of particle uptake by direct opsonization.
![]() View larger version (36K): [in a new window] |
Figure 2. Effects of CSPG-DS, IFN- , and CSPG-DS + IFN- costimulation on BV-2 and primary microglia phagocytosis (A–H) and NO secretion (I and J). BV-2 cells (A–C) and primary microglia (D–F) treated with vehicle (A and D), 50 µg/ml CSPG-DS (B and E), or 20 ng/ml IFN- (C and F) were incubated with Latex blue beads for 8 h and analyzed for uptake by phase-contrast microscopy. Representative micrographs are shown for each condition. (G and H), The optical bead density was counted for up to 100 cells in six fields of view. Three independent experiments for BV-2 cells and two independent experiments for primary microglia were performed, and the mean ± SD of each phagocytosis index given as percent-positive cells is displayed. CSPG-DS and IFN- significantly increased phagocytosis of both cell types compared with controls (P<0.05; Students t-test). In addition, CSPG-DS was more potent than IFN- in increasing the phagocytic activity of primary microglia. (I and J) NO release of CSPG-DS, IFN- , and CSPG-DS + IFN- -stimulated BV-2 cells and primary microglia. IFN- induced a strong and significant (P<0.05, Students t-test) secretion of NO in the cell culture medium, whereas CSPG-DS does not influence NO concentrations. Coincubation with CSPG-DS significantly reduced IFN- -induced NO secretion in primary microglia (P<0.05; Students t-test). Co, Control.
|
potently increased NO secretion in both cultures (Fig. 2I
and 2J)
. Interestingly, combined treatement with IFN-
and CSPG-DS resulted in significantly lower NO levels (Fig. 2I
and 2J)
. These data indicate that microglia activation by CSPG-DS is unlikely to cause direct cytotoxic effects but moreover, might even diminish cytotoxicity in a proinflammatory microenvironment. To test this hypothesis, a culture system of 661W photoreceptor cells with conditioned medium (CM) from BV-2 cells or microglia was established. 661W cells were incubated for 24 h in their own medium or with 10% culture supernatants from unstimulated, CSPG-DS, IFN-
, or IFN-
+ CSPG-DS-treated cells. 661W photoreceptor cell morphology was assessed by phase-contrast microscopy, cell viability was studied using the Cytotox96 nonradioactive cytotoxicity assay measuring lactate dehydrogenase (LDH) release (Promega), and apoptotic cell death was determined with the Caspase 3/7 glow assay (Promega). 661W cells in their own medium grew confluent after 24 h (Fig. 3A
), and the presence of CM from CSPG-DS-treated BV-2 cells did not affect growth (Fig. 3B)
. In contrast, coincubation with IFN-
-stimulated BV-2 supernatant strongly inhibited 661W proliferation with prominent cell-free areas present in the culture (Fig. 3C)
. When adding CM from CSPG-DS and IFN-
-stimulated BV-2 cells simultaneously, nearly normal cell characteristics were retained (Fig. 3D)
. Similar morphological changes of 661W cells were observed when cultured in the presence of primary microglia CM (data not shown). Along with these findings, a significantly decreased cell viability of 661W cells associated with a high level induction of caspase 3/7 activity was noticed with CM from IFN-
-treated BV-2 cells and primary microglia (Fig. 2E
2F
2G
2H)
, which is in agreement with earlier findings [25
]. In contrast, microglia supernatants from CSPG-DS stimulations alone neither influenced 661W viability (Fig. 2E
and 2F)
nor initiated caspase-related apoptotic signaling (Fig. 2G
and 2H)
. Remarkably, combined treatment of IFN-
conditioned microglia with CSPG-DS fully abrogated cell death of 661W cells as measured by LDH release and significantly reduced caspase 3/7 activation (Fig. 2E
2F
2G
2H)
. Direct incubation of photoreceptor cells with IFN-
or CSPG-DS had no effects (data not shown).
![]() View larger version (117K): [in a new window] |
Figure 3. Effects of CSPG-DS, IFN- , and CSPG-DS + IFN- -treated BV-2 cells and primary microglia on photoreceptor cell morphology, cell viability, and apoptosis. (A–D) Morphological changes of 661W cultures treated with CM from unstimulated (A), CSPG-DS (B), IFN- (C), or IFN- + CSPG-DS (D)-treated BV-2 cells. Control (A) and CSPG-DS (B) cultures grew confluent, whereas CM from IFN- -stimulated BV-2 cells inhibited full 661W cell growth (C). Coincubation with CSPG-DS (D) largely prevented formation of cell-free spaces and restored confluency. The viability of 661W cells was significantly (P<0.05; Students t-test) reduced to 44% and 39% after 24 h treatment with CM from IFN- -treated BV-2 cells (E) and primary microglia (F), respectively. In contrast, CSPG-DS CM per se did not affect 661W cell viability but rather, normalized cell viability when coadministered with IFN- CM (E and F). (G and H) Caspase activity expressed as relative light units (RLU) after incubation with different CM. IFN- CM elicited significant caspase 3/7 activation (P<0.05; Students t-test) in BV-2 (G) and primary microglia (H), whereas CSPG-DS had no effect above the vehicle-stimulated control. The presence of CSPG-DS in the CM significantly suppressed caspase 3/7 activation in IFN- conditioned 661 W cells (P<0.05; Students t-test). The data shown are mean values of three independent experiments ± SD for BV-2 cells and two independent experiments for primary microglia.
|
Received February 26, 2008; revised April 22, 2008; accepted May 15, 2008.
|
|
|---|
B activation, cell migration, and modulation of intracellular signaling J. Leukoc. Biol. 75,1139-1146This article has been cited by other articles:
![]() |
T. Langmann, C. C. L. Lai, K. Weigelt, B. M. Tam, R. Warneke-Wittstock, O. L. Moritz, and B. H. F. Weber CRX controls retinal expression of the X-linked juvenile retinoschisis (RS1) gene Nucleic Acids Res., November 1, 2008; 36(20): 6523 - 6534. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||