Figure 1.
S100B transcription only occurs in barrier organs.
Reverse-transcriptase PCR results for S100B (top row) and β-actin (bottom row) mRNA in various rat tissues. S100B mRNA was found in brain tissue and testis; kidney tubule cells also expressed measurable levels of S100B mRNA. No S100B mRNA was found in organs where S100B protein levels were previously demonstrated by us and others (see [24]). These results summarize outcomes from at least four repetitions. Changing the number of PCR cycles in the protocol from 40 to 35 (see Methods) did not result in any significant changes.
Table 1.
Summary of experimental results testing for the presence of S100B at the protein, mRNA levels or after injection of labeled S100B.
Figure 2.
Endothelial cells take up circulating but not CNS-derived S100B.
(A) Lack of significant endogenous immunoreactivity for S100B in BBB endothelial cells. The arrows point to faintly stained capillaries in the hippocampus (CA1 region). This staining was accounted for by glial end feet positive for S100B. Note the endogenous S100B immunostaining of astrocytes located within the pyramidal cell layer. (B) Uptake of circulating (exogenous) S100B by endothelial cells after injection of green S100B is observed in most systemic vessels. The example shows the appearance of an arteriole and capillaries in a lymph node, indicated by arrowheads. (C1) and (C2) show the uptake of exogenous, circulating S100B by endothelial cells of the retinal barrier. Note the uptake indicated by arrows and the various retinal layers. (D1-3) Co-localization of CD31 and exogenous S100B indicates uptake by endothelial cells. Endothelial cells take up S100B after systemic injection of labeled protein. The results shown here use an immunohistochemical validation by an endothelium-specific marker to corroborate the results in Figure 1. In fact, the cells demonstrating S100B uptake (green) were also positive for the endothelial marker CD31. See also Figure S3 A for CNS BBB endothelial cells. G = ganglion cell layer, IN = inner nuclear layer, ON = outer nuclear layer, RC = rods and cones, C = choroid, S = sclera. AF 488-tagged S100B was injected to achieve a serum concentration of 0.12 ng/mL to mimic blood-brain barrier disruption (ref. 13).
Figure 3.
Circulating S100B fails to invade barrier organs; however S100B gene transcription and protein synthesis occurs in both brain and testis.
Injection (exogenous) strategies demonstrate privilege of barrier organs to transendothelial diffusion of S100B while immunodetection of endogenous S100B demonstrates brain- and testis specific S100B protein by astrocytes and Sertoli cells. mRNA detection in the same barrier organs confirms S100B expression. (A1) shows the lack of fluorescent signal (Alexa Fluor (AF) 488, in green) in brain regions where endogenous S100B was readily detected (A2). The section used for immunohistochemistry contained the CA2 sector of the hippocampus. Note that, as expected, S100B was present in glial cells but not in neurons; neuronal cell bodies in hippocampal CA2 region are seen as unstained ghosts. Testicular tissue yielded similar results albeit in testis the barrier is established by Sertoli and not endothelial cells (B1 and B2). Note that intravascular S100B was restricted to the stroma of the seminiferous tubules in the testis where endogenous S100B was not present. S100B+ cells in the stroma (arrowheads) are CD4+ dendritic cells (insert in B2). DAPI (blue) was added as a nuclear stain (B1). AlexaFlour 488-tagged S100B was injected to mimic blood-brain barrier disruption (e.g., ref. 13). The labeled protein was allowed to circulate 3 hrs. prior to tissue harvesting. The mRNA bands shown reflect levels of expression by brain and testis. Sem. T. = seminiferous tubule.
Figure 4.
Circulating S100B is taken up by peripheral immune cells.
AlexaFlour 488-tagged S100B protein accumulated in Langerhans cells of the skin (A1, 2), thymic dendritic cells (B1, 2), and in dendritic cells in lymph nodes (a para-aortic lymph node is shown in C1, 2). Note the lack of uptake by non-immune surrounding tissue. Also note the typical appearance of dendritic cells in nodal tissues. Injection consisted of AlexaFlour 488-tagged S100B at a concentration of 0.12 ng/mL introduced via tail vein and allowed to circulate 2–3 hrs. prior to tissue harvesting; DAPI (blue) was added as a nuclear stain. Note the different magnifications among the panels, with scale bar = 200 µm in B1, 100 µm in A2 and 50 µm in all other panels. Rats were injected with a combination of AlexaFlour 594-tagged S100A1 (red) and AlexaFlour 488-tagged S100B (green). Skin (D) and splenic (E) tissue revealed that Langerhans cells (arrows in D) and splenic cells within the germinal center (GC), mantel zone (MnZ) and marginal zone (MaZ) demonstrate uptake of both S100A1 and S100B; however, the extent and intensity of S100B uptake was much greater as also evident in the merged figures (D3, E3). F shows the preferential segregation of S100B at the membrane of dendritic cells in skin (F1 and F2) and thymus (F3); greater detail of a Langerhans cell from D2 (box) is shown in F1 in order to highlight membrane staining. Injection consisted of AlexaFlour 488-tagged S100B at a concentration of 0.10–0.12 ng/mL introduced via tail vein and allowed to circulate 2–3 hrs. prior to tissue harvesting; DAPI (blue) was added as a nuclear stain. GC = germinal center, MnZ = mantle zone, MaZ = marginal zone.
Table 2.
Comparison of S100B and S100A1 sequences to underscore similarities (normal character) and differences (in bold).
Figure 5.
Splenic S100B-positive cells change in both number and morphology following pilocarpine-induced seizure.
S100B positive cells can be viewed within splenic follicles both in an unmanipulated animal (A), as well as following simulation of BBBD via IV injection of Alexa Fluor 488-labeled S100B (B) and BBBD from pilocarpine administration (C). The pattern of staining is preserved in all 3 conditions, however, staining is clearly augmented in BBBD simulation and to a much greater degree in actual induced BBBD. S100B-labeled lymphocytes and dendritic cells can be observed in all regions of the splenic follicle. Note that the morphology of labeled cells also appears to change with induction of BBBD in (C), where cells appeared to have a more dendritic and interconnected staining pattern than in other conditions. The identity of these cells was verified via the immune cell markers, CD4 and CD8 immunostaining (D1–D2); S100B+ and CD4+/CD8+ double positive cells are again found throughout the follicle, with emphasized staining in the marginal zone vs other areas. Injection consisted of AlexaFlour 488-tagged S100B at a concentration of 0.10–0.12 ng/mL introduced via tail vein and allowed to circulate 2–3 hrs prior to tissue harvesting. For induced BBBD, the animal was treated with pilocarpine and spleen was removed prior to onset of status epilepticus. Sections in A and C were treated with mouse anti-S100B antibody (Ab) and donkey anti-mouse 2° Ab conjugated to FITC (Jackson). In D1–D2, sections were treated with rabbit anti-S100B Ab and donkey rabbit 2° Ab conjugated to Texas Red (Jackson) and rat anti-CD4 or CD8 antibody and mouse anti-rat 2° Ab conjugated to FITC (Jackson). DAPI was added as a nuclear stain. In E–F, sections shows rats injected with S100B tracer in control (E1–E3) and pilocarpine administrated rats (F1–F3). Co-localization of S100B+CD86 (F1) and high individual staining of CD86+ (F2) in pilocarpine compared to controls E1 and E2 indicates that activated immune cells capture S100B. The dendritic cell nature of these cells was further demonstrated by their CD86+ staining (Figure S3). Scale bar in A = 100 µm for all images. GC = germinal center, MnZ = mantle zone, MaZ = marginal zone, WP = white pulp, RP = red pulp.
Figure 6.
BBBD precedes clinical seizures and result in the formation of S100B autoantibodies.
(A) Data were obtained by performing serum analysis in patients affected by seizures who underwent monitoring in the Epilepsy unit. Blood was drawn as soon as changes consistent with seizure development were observed. These consisted of a combination of behavioral and EEG changes. False positive samples (blood drawn, no subsequent seizure within 15 minutes) were discarded. Pre-ictal samples refer to blood drawn before seizures (at least 4 hours from preceding episode). Note the sharp increase in S100B at time of ictal event. The asterisks indicate p<0.05 (ANOVA). (B). Serum samples were taken from patients undergoing osmotic disruption of the BBB [49], [50] to improve chemotherapic efficiency. This procedure causes repeated BBBD and S100B elevations at each cycle of treatment [85]. The positive linear relationship between number of treatments with iatrogenic disruption of the BBB and % increase in serum S100B autoantibodies is shown. Auto-antibody titers show a sharp increase around treatment/month 4; after this time levels remained constant. The asterisks indicate p<0.05 (ANOVA). Statistical significance was achieved only after 5 cycles of BBBD. The data remained statistically different from pre-BBBD throughout procedure 7. These data represent the average of values for three patients undergoing treatment for primary CNS lymphoma, with induced osmotic breaching of the BBB; monthly treatments were performed for a total of 9 months. All three patients completed 6 treatments; two of the three completed 7 and one completed 9.
Figure 7.
Schematic representation of the events that follow BBBD.
See text for details.