Platelet-derived hyaluronidase 2 cleaves hyaluronan into fragments that trigger monocyte-mediated production of proinflammatory cytokines

Abstract

Hyaluronan (HA) occurs in the body as a large, hydrating, space-filling, carbohydrate polymer in the extracellular matrix; it has both anti-angiogenic and immunosuppressive properties. Cleavage of HA results in the generation of variably sized fragments that stimulate multiple angiogenic and inflammatory responses in a size-specific manner. In this study, we report that platelets, as well as their megakaryocyte precursors, are unusual among somatic cells in that they contain only hyaluronidase 2 (HYAL2) but not HYAL1. Platelet HYAL2 is sufficient to cleave HA into fragments that are specific for inflammatory and angiogenic signaling; this process occurs in the absence of HYAL1, which is necessary in all other tissues to perform further HA degradation. Platelets can bind to HA, some of which derives from the stressed microvessel endothelial cell surface. Platelet-derived HYAL2 cleaves HA into fragments that stimulate mononuclear leukocytes in the immediate microenvironment to produce proinflammatory cytokines, including interleukin-6 and interleukin-8. Platelets, thus, are not only involved in hemostasis, the earliest step in wound healing, but are also important in the signaling of subsequent inflammatory and angiogenic steps. We hypothesize that aberrations in these sequential steps can promote chronic inflammation, as found in inflammatory bowel disease. The platelet may thus provide an interface between acute and chronic inflammation, wound healing, and their subsequent fibrotic responses.

Figure 1

HYAL1 and HYAL2 are differentially expressed during colitis, with HYAL2 being specifically associated within blood vessels. Serial sections of distal colon from a mouse that had developed a dextran sodium sulfate-induced inflammation were stained with antisera raised against the hyaluronidase HYAL1 and HYAL2 (green) and with DAPI stain for nuclei (blue). A: HYAL1 appears throughout the inflamed tissue and is most prominently associated with leukocytes. HYAL2-specific staining (B and higher magnification, D) is most prominent on small anuclear structures within focused areas of the mucosa, which when viewed at higher magnification (D) have the appearance of blood vessels (arrows, B). D: The red color indicates autofluorescent molecules in the tissue including the bright elastic lamina defining the blood vessels, and some red blood cells in the lumen in the magnified image. C: H&E-stained adjacent (not serial) section provided for tissue orientation. The area noted (dotted line) is characteristic of the area captured in the fluorescent images (serial sections A and B). As we have previously shown, features of colitis include: a highly damaged epithelial layer (E) surrounding the intestinal lumen (L); a pronounced leukocyte infiltrate (I) present above the mucosal muscle (MM) layer, and in the expanded submucosa (S); more prominent blood vessels (bv); and an expanded external muscle layer (EM).

Figure 2

Murine platelets and megakaryocytes contain exclusively the HA-degrading enzyme HYAL2. A: An in vivo clot from inflamed mouse colon tissue is stained for the platelet marker CD42b (green) and HYAL2 (red) with specific antisera and fluorescently labeled secondary antibodies. Nuclei are labeled blue with DAPI and observed by confocal microscopy. The co-localization of HYAL2 and CD42b appears yellow. Lack of blue DAPI staining indicates high purity of the platelet preparation. B: Mouse bone marrow aspirate demonstrating that megakaryocytes express HYAL2 (red). Identification of megakaryocytes in this preparation is based on the characteristic large size (∼50 μm) and fragmented DAPI-stained nuclear material (blue). They contain the Hyal 2 enzyme in abundance before fragmenting into platelets.

Figure 3

Human platelets contain exclusively the HA-degrading enzyme HYAL2. Paraffin sections of an in vitro clot generated by thrombin treatment of human platelet-rich plasma stained with antisera raised against HYAL1 and HYAL2 (red). Platelets specifically stained for HYAL2 with no evidence for HYAL1 staining. Platelet identity was confirmed with CD42b co-staining (yellow).

Figure 4

The HA substrate for HYAL2 enzyme is also present in platelets, as well as in their megakaryocyte precursors. A, left: A human platelet-rich plasma preparation stained with antiserum to HYAL2 (red) and biotin-labeled HA-binding protein (green) demonstrates that platelets contain both HA and HYAL2. Preparations were also stained with DAPI nuclear stain, and the lack of blue staining as well as the infrequent autofluorescent, larger sized red blood cells (RBCs) indicate high purity of the platelet preparation. Right: Nonspecific staining control (NS). B: High magnification confocal image of a cytospin preparation of platelets within which membrane staining of HYAL2 (red) occurs, whereas HA granules (green) are present internally. C: Staining of mouse bone marrow aspirates with biotin-labeled HA-binding protein (green) and DAPI nuclear stain (blue) demonstrates that platelet precursor megakaryocytes are rich in HA compared with other bone marrow cells (left). Right: NS, nonspecific staining control.

Figure 5

Human platelets contain HYAL2 mRNA and protein. A: Western blot analysis of protein extracts derived from two different samples of human platelets isolated as described in the Materials and Methods and reacted with HYAL2-specific antiserum. Each lane contains an extract of platelets derived from 0.3 ml of whole blood. B: RT-PCR analysis of mRNA isolated from freshly isolated platelets. Analysis of RNA was performed using primers to detect HYAL1, HYAL2, and β-actin expression as described in the Materials and Methods. Lanes were loaded with reaction products from 0.1 μg of RNA per lane.

Figure 6

Platelets bind to HA produced on TNF-α-activated intestinal endothelial cell surfaces. Intestinal endothelial cells (detected by their blue nuclei) produce HA (green) during treatment with TNF-α for 18 hours [A (magnified image) and C] compared with control (B). When freshly isolated platelets (CD42b, positive, red) are added to the cultures at 4°C for 1 hour, some adhere to HA [A (magnified image) and C] whereas others can be washed away. D: HA (green) is rapidly degraded on incubation with platelets at 37°C, whereas the endothelial monolayer remains intact.

Figure 7

Platelets cleave endothelial-produced HA into signaling-sized fragments. TNF-α stimulated, ³H-glucosamine-labeled intestinal endothelial cells were incubated without or with platelets for 1 hour at 37°C. Supernatants were collected, the proteins exhaustively digested, and the glycoseaminoglycans precipitated (see Materials and Methods) and analyzed. A: HPLC analysis of radiolabeled molecules shows the relative quantity and size distribution of ³H-glucosamine-labeled glycoseaminoglycans in the endothelial cell supernatants. B: Fluorophore-assisted carbohydrate gel electrophoresis (FACE) confirms the presence of HA in the pooled fractions from endothelial cells co-incubated with platelets, but not in the control endothelial fractions.

Figure 8

Platelet-cleaved HA fragments activate monocytes. Fragments of HA were collected from the culture fluid of TNF-α-stimulated endothelial cells that were incubated with platelets for 1 hour at 37°C as described in the Materials and Methods. Fragments were purified by exhaustive protein digestion and ethanol precipitation, and the amount of HA determined by ELISA. Preparations of HA fragments were split and one half fully degraded with hyaluronidase-SD (HYAL-SD). Elutriation-purified monocytes (10⁷, 99% pure) were cultured in 2 ml of RPMI 1640 medium containing 2% FBS and treated without fragments (control), with platelet-created HA fragments (50 μg/ml), or with HYAL-SD-digested platelet-created HA fragments for 48 hours. Supernatants were analyzed for cytokine content by immunoblot (A), and normalized densitometric values calculated (B). White bars, control; black bars, HA fragment treated; and striped bars, hyaluronidase-SD-degraded fragment-treated. Protein blot is representative of four experiments using different preparations of platelet-clipped HA.

Figure 9

Diagram of platelet HYAL2 generating signaling fragments of HA, and their interactions in the catabolic cascades associated with wound healing and inflammation. A diagram of interactions that occur between the signaling fragments of HA generated by the HYAL2 of platelets and their involvement in the earliest stages of wound healing and the subsequent inflammatory phase. Aberrations in this pathway can lead to chronicity of inflammation and possibly to autoimmune disorders.