Calcium and phosphate levels in blood are tightly regulated

Transcription

1 Integrative Physiology Clearance of Fetuin-A Containing Calciprotein Particles Is Mediated by Scavenger Receptor-A Marietta Herrmann,* Cora Schäfer,* Alexander Heiss, Steffen Gräber, Anne Kinkeldey, Andrea Büscher, Martin M.N. Schmitt, Jörg Bornemann, Falk Nimmerjahn, Martin Herrmann, Laura Helming, Siamon Gordon, Willi Jahnen-Dechent Rationale: Fetuin-A is a liver-derived plasma protein involved in the regulation of calcified matrix metabolism. Biochemical studies showed that fetuin-a is essential for the formation of protein-mineral complexes, called calciprotein particles (CPPs). CPPs must be cleared from circulation to prevent local deposition and pathological calcification. Objective: We studied CPP clearance in mice and in cell culture to identify the tissues, cells, and receptors involved in the clearance. Methods and Results: In mice, fetuin-a containing CPPs were rapidly cleared by the reticuloendothelial system, namely Kupffer cells of the liver and marginal zone macrophages of the spleen. Macrophages from scavenger receptor-ai/ii (SR-A)-deficient mice cleared CPPs less efficiently than macrophages from wild-type mice, suggesting that SR-AI/II is involved in CPP binding and endocytosis. Accordingly, we found reduced clearance of CPPs in SR-A/MARCO deficient mice. Conclusions: We could demonstrate that fetuin-a containing CPPs facilitate the clearance of mineral debris by macrophages via SR-A. Since the same receptor also contributes to the uptake of modified low-density lipoprotein particles in atherosclerosis, defective endocytosis of both types of particle may impinge on lipid as well as mineral debris clearance in calcifying atherosclerosis. (Circ Res. 2012;111: ) Key Words: fetuin-a mineral homeostasis calcification scavenger receptor atherosclerosis Calcium and phosphate levels in blood are tightly regulated through the concerted action of calciotropic hormones and phosphatonins, which regulate uptake in the gut, deposition in the skeleton, and excretion in the kidney. 1 Nevertheless, calcification, the deposition of mineral precipitates in the vasculature and in soft tissues, is a common pathological event, especially in dialysis patients. 2 Even under physiological conditions, blood is considered a metastable aqueous calcium-phosphate system sustaining mineral precipitation once crystals are nucleated. William Neuman aptly stated that we all suffer Lot s wife s problem, the imminent danger of turning into a pillar of salt. 3 Thus, a mechanism is required to safeguard against the disposal of mineral nuclei from circulation to prevent pathological calcification. In This Issue, see p 505 Various in vitro studies have shown that physiological salt solutions containing proteins spontaneously form proteinmineral complexes, which are initially soluble but precipitate with time. A series of studies showed that cell culture media, containing fetal calf serum or human serum, sustain the spontaneous formation of nano-sized mineralo-protein complexes even in absence of any cells and further that the mineralo-protein complexes slowly grow, aggregate, and precipitate. 4 6 It was pointed out that the formation of protein-mineral particles occurs by default, representing common physiological remnants linked to normal calcium homeostasis. The formation of mineral particles is found in many organisms at different evolutionary levels. 6,7 Spon- Original received November 23, 2011; revision received June 27, 2012; accepted July 2, In June 2012, the average time from submission to first decision for all original research papers submitted to Circulation Research was days. From Helmholtz Institute for Biomedical Engineering, Biointerface Group (M.H., C.S., A.H., S.G., A.K., A.B., W.J.-D.), Institute for Molecular Cardiovascular Research (M.M.N.S.), and Department of Pathology, Electron Microscopic Facility (J.B.), RWTH Aachen University, Germany; the Department of Biology, Institute of Genetics, Friedrich-Alexander University of Erlangen-Nuremberg, Germany (F.N.); the Department for Internal Medicine 3, Institute for Clinical Immunology, Friedrich-Alexander University of Erlangen-Nürnberg, Germany (M.H.); Institute for Medical Microbiology, Immunology, and Hygiene, Technische Universität München, Germany (L.H.); and Sir William Dunn School of Pathology, University of Oxford, United Kingdom (S.G.). *These authors share first authorship. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Willi Jahnen-Dechent, PhD, Helmholtz Institute for Biomedical Engineering, Biointerface Group, RWTH Aachen University, Pauwelsstrasse 30, Aachen, Germany. willi.jahnen@rwth-aachen.de 2012 American Heart Association, Inc. Circulation Research is available at DOI: /CIRCRESAHA

2 576 Circulation Research August 17, 2012 Non-standard Abbreviations and Acronyms acldl acetylated low-density lipoprotein AsF asialofetuin ASGP-R asialoglycoprotein receptor BSA bovine serum albumin BMM bone marrow derived macrophage CPP calciprotein particle CR-3 complement receptor 3 F-A fetuin-a KO knockout MARCO macrophage receptor with collagenous structure MMM marginal zone metallophilic macrophage MZM marginal zone macrophage Poly I polyinosinic acid RAGE receptor for advanced glycation end products RES reticuloendothelial system SR scavenger receptor t 1/2 half life TLR toll-like receptor TRIF toll/il-1 receptor (TIR) domain-containing adaptor WT wild-type taneously forming soluble protein-mineral particles are variously described as mineralo-protein complexes, 6 nanons, 8 fetuin-a/albumin-mineral complexes, 4 fetuin-mineral complexes, 9 or calciprotein particles The latter term, calciprotein particles (CPPs), was chosen by analogy to the lipoprotein particle, which carries otherwise insoluble lipids in the circulation. We and others have shown that the plasma protein fetuin-a plays a critical role in the formation and stabilization of such protein-mineral complexes. 15 Fetuin-A circulates in the blood and acts as a systemic inhibitor of pathological mineralization by sequestering calcium-phosphate nuclei to form soluble CPPs. Fetuin-A is a hepatic plasma protein belonging to the subgroup 3 of the cystatin superfamily, which also includes fetuin-b, histidine-rich glycoprotein, and kininogen. 16 Fetuin- A deficient mice show severe soft tissue calcification Pathological calcification in dialysis patients is associated with low fetuin-a serum levels Fetuin-mineral particles were found in rats treated either with bisphosphonate or adenine. 24,25 We identified ascites-derived CPPs in a patient with calcifying peritonitis, 12 suggesting that in blood, CPPs may only accumulate to detectable amounts under extreme circumstances. The recent detection of CPPs in serum from dialysis patients supports this view and furthermore shows that technical reasons may prevent the detection of CPPs in patients. 26 Thus, CPP-like protein-mineral complexes in normal subjects are either present in nondetectable amounts or cleared extremely fast from the circulation like rigid particles 27,28 and dead cell remnants. 29 However, the dynamics and mechanistic basis of CPP clearance from the circulation has not been addressed thus far. In the present study, we show for the first time that CPPs are effectively cleared from the circulation in mice and report their organ distribution after clearance as well as the involvement of scavenger receptor-a in the process. Our findings suggest that pathological mineralization in high-risk situations may be slowed down by stimulation of reticuloendothelial system (RES) clearance. Methods An expanded method section is available in the Online Data Supplement. Protein Purification Bovine fetuin-a, asialofetuin, and albumin were purified by gel filtration. Proteins were labeled with Alexa488 or Alexa546 carboxylic acid, succinimidyl ester using a labeling kit. CPPs were prepared as described before. 10 Animals Mice were injected intravenously with monomeric fetuin-a, CPP, asialofetuin type II, or bovine serum albumin. Mice were euthanized at different time points after injection to establish the serum clearance and organ distribution of different preparations. Serum samples taken immediately after injection were assigned 100% injected dose. Preparation of Serum and Tissue Samples Tissues extracts and serum samples were separated using SDS- PAGE, the fluorescence signal was scanned, and images were quantified by densitometry. Macrophage detection on cryosections was performed using MOMA-1, MARCO, MAdCAM-1, CD68, and F4/80 primary antibodies. Second antibody was Alexa546-labeled goat anti-rat antibody, and DAPI nuclear staining was applied. Fluorescence signals were quantified by histomorphometry. Cell Culture Experiments Bone marrow derived macrophages (BMMs) and RAW cells were used to study binding and uptake. Binding of fluorescent acldl, CPPs, and fetuin-a monomer was assayed using detached BMMs incubated in reaction tubes for 60 minutes at 4 C. Cellassociated fluorescence was measured by flow cytometry. For uptake measurements, adherent cells were incubated for 10 minutes with fluorescent fetuin-a monomer or CPPs preincubated (or not) in 30% serum. For inhibition studies, cells were treated with inhibitors or antibodies for 30 and 60 minutes, respectively. Multiphoton Microscopy Carotid arteries were dissected from C57BL/6 ApoE/Fetuin-A / mice. Carotids were perfused with fluorescent CPPs for 60 minutes and examined using laser-assisted multiphoton microscopy. Results Clearance of Protein-Mineral Complexes From Circulation Is Fast and Efficient To study clearance of mineral debris from circulation, we synthesized CPPs comprising fluorescence-labeled fetuin-a, calcium, and phosphate All protein and particle preparations were purified before use to remove aggregates and contaminants influencing the clearance (Online Figure I). We injected CPPs intravenously into mice and measured clearance from the blood (Figure 1). Blood was collected at various time points and clearance calculated from densitometry of the fluorescence signal of serum samples separated by SDS-PAGE. For comparison, we injected labeled monomeric fetuin-a (F-A), asialofetuin (AsF), and bovine serum albumin (BSA). It is known that AsF is rapidly cleared by the

3 Herrmann et al Clearance of Calciprotein Particles minutes), but AsF monomer was cleared even faster than both fetuin-a monomer and CPPs (t 1/2 43 minutes). This finding corroborated the presence of a high-affinity ASGP-R, strongly discriminating between fully glycosylated fetuin-a and asialofetuin (t 1/2 149 minutes versus 43 minutes). In summary, the faster clearance of CPPs versus fetuin-a monomer suggested that clearance was greatly enhanced by mineral complex formation of fetuin-a. Figure 1. In vivo clearance of fluorescence-labeled protein. Fluorescence-labeled proteins (160 g) were injected intravenously into mice, and blood was drawn at indicated time points to determine the amount of remaining circulating protein. Shown are the fluorescence signals as percentage of the value measured one minute after injection. BSA had the longest half-life (t 1/2 ) of all proteins tested, and in descending order: fetuin-a monomer CPPs asialofetuin. a, Statistically significant different from fetuin-a monomer at the same time point, P 0.01; b, statistically significant different from CPP, P n 4. asialoglycoprotein receptor (ASGP-R), which binds terminal galactose residues on desialylated plasma proteins and is predominantly expressed on the sinusoidal surface of hepatocytes. Figure 1 shows that the clearance of CPPs was roughly 3 times faster than the clearance of fetuin-a monomer with serum half-lives t 1/2 of 45 minutes and 149 minutes, respectively. CPP clearance was also faster than BSA clearance (t 1/2 CPPs Are Taken Up by the Reticuloendothelial System and Degraded Rapidly We reasoned that the clearance of CPPs from blood should be mirrored by their organ accumulation. Figure 2 illustrates CPP clearance and accumulation in the liver, a major organ of the RES. The fluorescence micrographs of liver sections demonstrate localization of fluorescent CPPs to a specific cell population (Figure 2A through 2E). Antibody staining showed that CPPs were taken up by F4/80- and CD68- positive liver macrophages, Kupffer cells (Figure 2F through 2G), whereas F4/80-negative liver sinusoidal endothelial cells did not accumulate CPPs. Transmission electron microscopy showed that CPPs formed electron-dense, dark, small, needlelike crystals densely packed in vesicles in Kupffer cells (Online Figure II). Taking fluorescence of the Kupffer cells as a proxy of cellular clearance, uptake of fluorescencelabeled CPPs peaked at 2 minutes and decreased thereafter (Figure 2A through 2E). Densitometry of fluorescence micrographs showed that 55% of the fluorescent material present 2 minutes after injection had disappeared 8 minutes later, suggesting rapid clearance and degradation of CPPs in Kupffer cells (Figure 2H). The half-life of CPPs in Kupffer cells was 3 minutes. Thus, the clearance rate for CPPs was considerably faster than that determined from serum disap- Figure 2. Localization and degradation of CPPs in the liver. A through E, The fluorescence of CPPs in liver sections diminished rapidly from 2 to 180 minutes, suggesting maximum clearance within 10 minutes and degradation thereafter. F and G, Immunostaining with macrophage specific antibodies CD68 and F4/80 (red) showed colocalization (yellow) with CPPs (green) inside macrophages. H, Decrease in fluorescence of liver sections measured by histomorphometry as shown in A through E. Values shown are mean SEM. Size of bars in A through G, 50 m. n 4.

4 578 Circulation Research August 17, 2012 pearance (t 1/2 45 minutes) shown above. We attribute this to the fact that CPP preparations also contain monomeric fetuin-a 14 greatly distorting the apparent serum clearance rate but not the rate of accumulation of CPPs in Kupffer cells. Fetuin-A monomer was cleared by unspecified cells with slow kinetics typical for long circulating plasma proteins like albumin (not shown). Of note, quantification of the more or less homogeneous fluorescence-signal of both monomer types was disturbed by tissue auto-fluorescence of the liver. In summary, no cell type specific accumulation was observed for fetuin-a monomers, whereas a rapid macrophage specific uptake was observed for CPPs. Fetuin-A Monomer and CPP Show Different Organ Distribution To quantify the organ distribution of injected CPPs and monomeric fetuin-a, we analyzed perfused organs of mice. We measured the amount of intact (high molecular weight) and degraded (low molecular weight) fetuin-a in cellular extracts by SDS-PAGE and quantitative fluorescence imaging (Online Figure III). The background fluorescence was arbitrarily set to the average mean fluorescence signal observed in extracts from those organs that scored negative by immunofluorescence (Online Figure IV). We attribute any residual signal in these organs to insufficient blood removal during organ perfusion. Fetuin-A monomer accumulated to 1.5-fold background level in liver, kidney, and bone marrow (Online Figure IVA), suggesting that these organs may be involved in the clearance, excretion, and deposition of monomeric fetuin-a. Online Figure IVB shows that the distribution of CPPs was very different from the monomer. Strong fluorescence signals were detected in liver extracts immediately after injection of CPPs (see also Figure 2A through 2E and 2H) and decreased continuously. The highest amount was present at 30 minutes after injection and attained 8-fold background level. At this time, the CPP liver content corresponded to 31% of the total injected dose of CPP. A decrease over time was seen for both the intact as well as the degraded portion of CPP associated fetuin-a. A comparable pattern was seen in the spleen revealing up to 5-fold increase over background, yet a lower content than the liver (range, 5- to 2.5-fold in spleen and 8- to 4.6-fold in liver). All other tissues showed comparatively constant fluorescence values and thus did not contribute to clearance and excretion of CPPs. Fluorescent protein content was also consistently low in myocardium, axillary lymph nodes, pancreas, gonads, and skeletal muscle (not shown). AsF showed a specific accumulation in liver, confirming published data. 31,32 The second highest signal was detected in kidney showing mostly degraded asialofetuin (not shown). This suggests elimination of AsF from circulation in the liver and subsequent excretion through the kidney. In summary, the analysis of organ extracts showed that CPPs were predominantly cleared by liver and spleen. In both organs, the labeled CPPs were also degraded as evident from the initial rapid increase and the following steady decline in signal intensity in liver and spleen extracts compared with other tissues. Figure 3. Localization of CPPs in the spleen. Spleens were harvested 30 minutes after CPP injection. Spleen sections show Alexa488-labeled CPPs (green) and cell type specific surface markers (red). A and B, CPPs show partial colocalization (yellow) with the MOMA-1 antigen of the marginal metallophilic macrophages. C and D, No colocalization with MAdCAM-1 as a marker for sinus lining cells; C and D, almost complete colocalization with MARCO-positive marginal zone macrophages. WP indicates white pulp; MZ, marginal zone; and RP, red pulp. Scale bar in A, C, and E, 200 m; in B, D, and F, 50 m. Marginal Zone Macrophages of the Spleen Clear Protein-Mineral Particles Besides the specific uptake of CPPs in the liver, the quantification of organ extracts suggested the spleen as the second major organ involved in clearing CPPs from the circulation. We analyzed the clearance of CPPs in spleen by fluorescence microscopy and used cell type specific markers to determine the cell type responsible for CPP clearance. The spleen harbors several types of macrophages, localized in the red pulp or the marginal zone. The white pulp is mostly free of macrophages but contains dendritic cells. Marginal metallophilic macrophages (MMM) were stained with MOMA-1/ sialoadhesin antibody (Figure 3A and 3B) and marginal zone macrophages (MZM) were stained with antibody against MARCO (macrophage receptor with collagenous structure) (Figure 3E and 3F). 33 MMM and MZM macrophages in the marginal zone are separated by a layer of sinus lining cells, which stain positive for the cell adhesion molecule MAdCAM-1 (Figure 3C and 3D). MZM are located exclusively in the outer side of the sinus lining cell layer of the

5 Herrmann et al Clearance of Calciprotein Particles 579 marginal zone (pointing to the red pulp), the MMM are localized mainly on the inner side of the marginal sinus pointing to the white pulp, and a fraction of weakly MOMA-1 positive cells is also found on the outer side. Figure 3 shows that CPP accumulated in the marginal zone, but were absent in white and red pulp. The sinus lining cell marker MAdCAM-1 (Figure 3C and 3D) separated a few CPP lo / MOMA pos macrophages from CPP hi /MARCO pos macrophages (Figure 3E and 3F), suggesting that most CPPs were taken up by MARCO pos MZM, whereas the contribution of MOMA-1 pos MMM was clearly lower. In Vitro Clearance of CPPs by Macrophages Liver and spleen are both part of the RES or mononuclear phagocytic system. The RES comprises phagocytes that clear endogenous and particulate matter including colloidal particles and bacteria. To study the macrophage clearance of CPPs we performed binding and endocytosis assays using the macrophage-like murine cell line RAW and primary BMMs. Figure 4A and 4B show the results of a typical experiment testing endocytosis of fetuin-a monomer and CPPs at 37 C. Uptake was measured by evaluating cellassociated fluorescence using flow cytometry. Despite similar amounts of labeled fetuin-a present in both monomer and CPP preparations offered to the macrophages, significantly more CPPs than fetuin-a monomer were endocytosed. The cell-associated fluorescence signal of fetuin-a monomer ranged between 2% to 6% of the signal detected for CPPs (P for all time points). These results are in full agreement with the in vivo clearance (Figures 1 through 3 and Online Figure IV) showing consistently more efficient uptake of CPPs than monomeric fetuin-a. To investigate CPP binding we incubated BMMs with fluorescent CPPs at 4 C. Figure 4C illustrates binding of fluorescent CPPs to BMMs, which decreased on competition with unlabeled CPPs suggesting a specific and saturable binding mechanism. The initial cell-associated fluorescence signal was reduced from 100 to 90, 76, and 60 (P 0.05), when a similar amount, 2-fold excess or 5-fold excess of unlabeled CPPs were added, respectively. To determine the uptake mechanism for CPPs we treated macrophages with known inhibitors of endocytosis pathways. Treating BMMs with cytochalasin D, an inhibitor of actin polymerization, reduced the uptake of CPPs. The cell-associated fluorescence decreased from 103 to 44 (P 0.001, Figure 4D) indicating an active actin-mediated uptake mechanism. Preincubation with chlorpromazine, an inhibitor of clathrin-mediated endocytosis, reduced the cellassociated fluorescence from 103 to 57 (P 0.001). In contrast, genistein and 5-(N,N-dimethyl)amiloride, inhibitors of caveolae-mediated endocytosis and macropinocytosis, respectively, showed minor if any inhibition of CPP endocytosis. Collectively, these findings suggested an active, clathrinmediated route of uptake involving the cytoskeleton and a saturable clearance receptor. This prompted us to test the participation of common uptake mechanisms of particulate matter and pathogens. Inhibition with Ly hydrochloride, an inhibitor of phosphoinositol-3-kinase, which is involved in signal transduction of Fc -receptor mediated endocytosis, had no effect (Figure 4E). Similarly, inhibition of the Figure 4. Macrophage clearance of CPPs. A, Adherent RAW cells were incubated for 10 to 180 minutes with fluorescent fetuin-a (F-A) monomer or CPPs. Cell-associated fluorescence was measured by flow cytometry. Filled gray curves depict untreated cells. B, Endocytosis of CPPs was up to 50-fold higher than uptake of fetuin-a monomer. C, Binding of fluorescent CPPs was studied in BMM at 4 C. Increasing amounts of unlabeled CPPs reduced the binding of labeled CPPs. D through F, BMMs (D) or RAW cells (E and F) were incubated with fluorescent CPPs or serum-coated CPPs for 10 minutes. Cells were treated with inhibitors for 30 minutes before addition of CPPs. Cytochalasin D, chlorpromazine, and poly-i reduced CPP uptake. Serum absorption to CPPs diminished the overall uptake of CPPs (P 0.01) yet improved specificity of uptake in that it caused a greater percentage of inhibition of uptake by poly-i. *P 0.05, **P 0.01, ***P mannose receptor with mannan and of the receptor for advanced glycation end products (RAGE) with glycated serum albumin had minor if any effect on CPP endocytosis (Online Figure V). In contrast, inhibition of scavenger receptors (SRs) using polyinosinic acid (poly-i) reduced cellassociated fluorescence from 172 to 124 (P 0.01) and 100 (P 0.001) when macrophages were preincubated with 2 ng/ L or10ng/ L poly-i, respectively (Figure 4D), indicating involvement of SR in CPP clearance. To mimic in vivo

6 580 Circulation Research August 17, 2012 uptake, we preincubated CPPs with serum before addition to cells. Uptake of serum-coated CPPs was reduced in comparison to uncoated CPPs (Figure 4D through 4F). Like uncoated CPPs, uptake of serum-coated CPPs was inhibited by cytochalasin D. The cell-associated fluorescence decreased from 53 to 20 (P 0.001). A slight decrease of cell-associated fluorescence was likewise observed with chlorpromazine (54 versus 43), whereas CPP uptake was unaffected by inhibition of caveolae-mediated endocytosis or macropinocytosis. Inhibition of CPP uptake with poly-i was more pronounced in serum-coated CPPs than uncoated CPPs. Cell-associated fluorescence decreased from 126 to 100 and 40 with 2 ng/ L and 10 ng/ L (P 0.001) poly-i, respectively. In summary, in vitro clearance studies confirmed that CPPs were readily taken up by macrophages through a SR/clathrindependent pathway, whereas monomeric fetuin-a was endocytosed barely. Scavenger Receptor-A Is Involved in CPP Clearance Our endocytosis studies pointed to a major role of SRs in the uptake of CPPs. The family of SRs is characterized by their involvement in scavenging modified forms of LDL. 34 We focused on SR-AI/II and the class B scavenger receptor CD36 but initially also interrogated other macrophage receptors that might also be involved in CPP clearance. We studied CPP uptake in the presence of blocking antibodies directed against a panel of innate immune receptors alone and in combination, namely complement receptor 3 (CR3), CD16/CD32, toll-like receptors TLR2 and TLR4, CD36, and sialoadhesin. We found that these receptors contributed little if any to CPP endocytosis (Online Figure VI). In addition, we tested the involvement of macrophage receptors, which might be involved in CPP uptake in BMMs derived from the respective knock-out mice. We included Fc -receptor, involved in the clearance of immunoglobulins, as well as the opsonic serum proteins C-reactive protein and serum amyloid P. 35,36 We studied annexin 5 and 6, which have been reported to mediate endocytosis of fetuin-a, 37,38 apoptotic cells, 39 and enhanced lipoprotein particle endocytosis, 40 respectively. The role of TLR-mediated signaling was further studied in TRIFdeficient mice. 41 We also tested galectin1 and galectin3, 2 lectin receptors putatively interacting with fetuin-a glycosyl chains. We found that whereas CD36, annexins, galectins, TRIF, and Fc -receptor were not essential for CPP binding and/or uptake (Online Figure V), SR-AI/II was critically involved. Figure 5A shows that SR-AI/II deficient macrophages showed strongly reduced uptake of fluorescence-labeled CPPs compared with wild-type macrophages. Depending on the CPP concentration, the cell-associated fluorescence in SR-AI/II deficient macrophages ranged from 57% to 73% of wild-type (P 0.01). This difference was even more pronounced when CPPs were preincubated with serum (Figure 5B). Here the fluorescence signal was reduced to 44% to 49% (P 0.001). Remarkably, CPP uptake in SR-AI/II deficient BMM was as low as fetuin-a monomer uptake (Figure 5B). We also investigated CPP binding to macrophages at 4 C. In agreement with our results from endocytosis assays, SR-AI/II Figure 5. SR-A is involved in binding and uptake of CPPs. BMMs of wild-type (WT) and SR-A deficient mice (SR-A KO) were studied. A and B, Uptake assay: Adherent cells were incubated at 37 C for 10 minutes with CPPs (A) with fetuin-a monomer or with serum-coated CPPs (B). Uptake of CPPs and serum-treated CPPs was concentration-dependent and saturable. SR-A KO macrophage uptake of CPPs was reduced. Uptake of serum pretreated fetuin-a monomer was low and did not show any difference between WT and KO. C, Binding assay: Cells were seeded and incubated on ice for 45 minutes with CPPs or fetuin-a monomer. The binding of CPPs was concentration-dependent and saturable. Macrophages from SR- A deficient mice showed reduced binding compared with wildtype. Binding of fetuin-a monomer to macrophages of all cell sources was minimal. a, P 0.05; b, P 0.01; c, P compared with wild-type. deficient BMMs showed diminished CPP binding; the fluorescence signal was reduced to 63% to 78% compared with wild-type BMMs (P 0.05 for 0.05 mg/ml and mg/ml CPPs). This difference in binding was less pronounced than the difference in uptake. This may be explained by the general tendency of CPPs to adhere to the cell membrane, as CPPs also bound to various cell lines including nonprofessional phagocytes, HeLa, Cos1, and Cos7 cells (not shown). We confirmed the involvement of SRs in CPP binding and endocytosis by studying CPP clearance in SR-A/MARCO deficient mice. These mice have combined deficiency in the putative CPP receptor SR-AI/II, detected in cell-based assays (Figure 5), and MARCO, a class A SR expressed on CPP pos MZM in the spleen (Figure 3). Figure 6A illustrates reduced clearance of CPPs from blood in SR-A/MARCO deficient mice (8% clearance 10 minutes after injection) compared

7 Herrmann et al Clearance of Calciprotein Particles 581 Figure 7. Competitive binding of CPPs and acldl. BMMs from wild-type mice were incubated on ice for 45 minutes with CPPs containing fetuin-a in varying in the presence or absence of 10 g/ml acldl. AcLDL decreased CPP binding at all concentrations of fetuin-a (P 0.01). Figure 6. CPP clearance in SR-A/MARCO deficient mice. Fluorescent CPPs were injected intravenously into SR-A/ MARCO deficient (KO) and wild-type (WT) mice (n 3). Mice were exsanguinated 10 minutes after administration of CPPs. A, Blood clearance of CPPs was determined as percentage decrease in fluorescence signal compared with the value 1 minute after injection. B, Tissue protein extracts were analyzed by SDS-PAGE and fluorescence scanning densitometry. The values are measured against 10 ng labeled fetuin-a internal standard. C, Fluorescence micrographs of liver and spleen sections showing Alexa488-labeled CPPs (green) and MAdCAM-1 (red) as a marker for sinus lining cells. Note that CPPs demarcated the outer marginal zone. Cell nuclei were stained with DAPI (blue). Scale bars, 200 m (liver); 100 m (spleen). CPP clearance was lower in KO mice, resulting in decreased organ accumulation and thus decreased fluorescence signal of CPPs. **P with wild-type (WT) mice (13%). CPP accumulation in liver (1.42 versus 1.18 ng/100 mg organ weight in WT and KO mice, respectively) and spleen (1.04 versus 0.45 ng/100 mg organ weight in WT and KO mice, respectively, P 0.01) were likewise reduced (Figure 6B). Fluorescence microscopy confirmed reduced CPP clearance in liver and spleen of SR-A/MARCO deficient mice (Figure 6C). CPPs in Atherosclerosis To support the role of SR-AI/II in CPP binding, we conducted a competitive binding assay and incubated wild-type BMM with CPP in presence or absence of acetylated LDL (acldl), a known ligand of SR-AI/II. Figure 7 shows that the binding of CPPs to macrophages was decreased in the presence of acldl. The cell-associated fluorescence was diminished to 43% to 71%, dependent on CPP concentrations studied (P 0.001). The competitive binding of CPPs and acldl prompted us to test the fate of CPPs in atherosclerosis. We used mice deficient in fetuin-a and apolipoprotein E (ApoE) as a model of calcified atherosclerosis. 19 High-resolution microscopy of live explanted carotid arteries of these mice revealed that CPPs accumulated in the plaque area (Figure 8A and 8B). Immunofluorescence showed that CPPs colocalized with CD68-positive macrophages (Figure 8C and 8D). In summary, our results suggest that CPPs are bound and taken up by macrophages using SR-AI/II mediated endocytosis. Marginal zone macrophages and Kupffer cells participate in splenic and hepatic uptake, respectively. This endocytic uptake mechanism discriminates against monomeric fetuin-a, which circulates much longer in vivo and distributes differently in the body. Figure 8. CPPs in atherosclerosis. A and B, Ex vivo multiphoton microscopy of an atherosclerotic carotid artery perfused with Alexa546-labeled CPPs (red). The blue signal derives from autofluorescence of collagen fibers. A, Plaque-free area showing CPP accumulation in the vessel lumen and bound to the endothelium. B, Strong cellular CPP accumulation was detected in the atherosclerotic plaque within the bifurcation zone of the carotid artery. C and D, Immunofluorescence staining of carotid arteries. Macrophages were stained by CD68 (green), DAPI was used as nuclear counterstain (light blue). CPPs (red) accumulate within the vessel lumen (C) and the plaque (D) but not in the artery wall. Scale bars, 100 m.

8 582 Circulation Research August 17, 2012 Discussion Soluble colloidal protein-mineral complexes are now considered byproducts of general mineral homeostasis. They are variously called CPPs by analogy to lipoprotein particles 15 or mineralo-protein complexes. 6 To prevent local deposition of such granules, especially in conditions with excess amounts of calcium and phosphate in the body, a clearance mechanism is required mediating CPP recycling or disposal by dissolution, degradation, or storage. We studied the clearance and the disposal of in vitro generated fetuin-a containing CPPs in mice. Several studies have investigated the clearance of fetuin-a monomer, and especially of its desialylated form, asialofetuin. Endocytosis of fetuin-a monomer was found to be mediated by annexins expressed on the cell surface. 37,38 Asialofetuin, like many asiologlycoproteins, is rapidly cleared from the circulation by the ASGP-R expressed by hepatocytes. 32,42,43 To study the role of fetuin-a in clearance of CPPs, we compared the clearance of fetuin-a monomer, its desialylated form asialofetuin, serum albumin, and fetuin-a containing CPPs in vivo. Asialofetuin was cleared from the circulation within minutes (t 1/2 43 minutes), confirming the highly efficient ASGP-R mediated endocytosis of this ligand (Figure 1), and we detected accumulation of asialofetuin monomer in liver tissue (not shown), the major site of ASGP-R expression. This finding confirmed the wellestablished activity of the ASGP-R 32 and thus validated our experimental approach. Using CPPs, we observed a fast clearance (t 1/2 4 minutes) mainly by macrophages of the RES represented by hepatic Kupffer cells and splenic MZM macrophages (Figures 2 and 4). These cells are known to be involved in the clearance of aged blood cells 44 and particulate matter. 45 In contrast, liver sinusoid endothelial cells and hepatocytes, the prime sites of ASGP-R expression did neither accumulate fetuin-a monomer nor CPPs (Figure 2). Thus, the particulate form of fetuin-a bound to CPPs invoked a clearance mechanism that overcame the clearance of fetuin-a monomer and the well-established ASGP-R mediated endocytosis of asialofetuin. Using a cell culture assay, blocking reagents, macrophages from knockout (KO) animals, and live SR-A/MARCO mice, we could show that macrophage uptake of CPPs is receptormediated and that SR-AI/II is essential for CPP endocytosis. Similarly, it was previously shown that the uptake of polystyrene nanospheres, 50 nm in size, coated with fetuin-a into cultured rat Kupffer cells, could be partially inhibited by poly-i as well as by anti SR-A antibodies, and a similar result was observed in liver perfusion studies in rats. 46 In addition to SR-AI/II, we propose that the specific architecture of the RES constituting the mechanical microfilter and the specific flow restrictions imposed by sinusoidal barriers in spleen marginal zone versus red pulp 33 directs CPP clearance to Kupffer cells and MZM macrophages, since SR-A expression alone is insufficient for CPP clearance. For instance, SR-AI/II positive liver endothelial cells 47 and red pulp macrophages 48,49 failed to take up CPPs (Figures 2 and 3). Similarly, particulate preparations of formaldehydetreated albumin were taken up by Kupffer cells via scavenger receptors but not by endothelial cells. 50 In the current study, cell-based assays were focused on SR-AI/II and CD36, whereas MARCO could not be examined in detail, because MARCO is not expressed in BMMs. Reduced CPP clearance in SR-A/MARCO double-deficient mice, however, suggests that in addition to SR-A MARCO might also be involved in CPP clearance. Ongoing studies will clarify this point. Regarding the ligand responsible for SR-AI/II binding, our data suggest that additional serum proteins other than fetuin-a might modulate CPP binding and clearance by macrophages, whereas fetuin-a is mainly responsible for the stabilization of CPPs. It is currently unknown which serum proteins contribute to the clearance of CPPs and fetuin-a stabilized mineral debris. We anticipate a situation similar to the clearance synapse identified for the clearance of apoptotic cells containing many yet circumscriptive factors. 29 The identification of albumin as a major contributor in mineralo-protein complexes 6,8,9,12 suggests that albumin may contribute to CPP clearance. Indeed, we found that preincubation of CPPs with albumin decreased CPP endocytosis similar to serum pretreatment (not shown), which might be attributed to the dysopsonic activity of albumin. 51 Apolipoprotein A1, a known SR-AI/II ligand, 52 which was previously identified in mineralo-protein complexes, 6 may likewise contribute to CPP uptake. In conclusion, our data show that SR-AI/II is responsible for the major part of endocytosis of CPPs. The finding that SR-AI/II, a well-known scavenger receptor for LDL particles 34 and lipid debris, likewise participates in the clearance of CPPs and mineral debris may help explain why atherosclerotic plaques frequently calcify, especially in patients with perturbed mineral homeostasis. Several studies in humans show that fetuin-a colocalizes with atherosclerotic plaques in calcifying atherosclerosis In mice, the combined deficiencies for fetuin-a and apolipoprotein E caused an exacerbated phenotype of calcifying atherosclerosis 19 that may involve CPP uptake by plaque macrophages, further supporting a combined role for the clearance of lipids and mineral debris in the pathogenesis of atherosclerosis. CPP clearance by macrophages may drive these cells into the calcifying myeloic cell type that was recently described in several publications. 56,57 Sources of Funding This work was funded by Deutsche Forschungsgemeinschaft priority program Principles of Biomineralization and Research Training Groups 1035 Biointerface and 1508 EuCAR. None. Disclosures References 1. Quarles LD. Endocrine functions of bone in mineral metabolism regulation. J Clin Invest. 2008;118: Block GA, Raggi P, Bellasi A, Kooienga L, Spiegel DM. Mortality effect of coronary calcification and phosphate binder choice in incident hemodialysis patients. Kidney Int. 2007;71: Neuman WF, Urist MR. Bone material and calcification mechanisms. Fundam Clin Bone Physiol. 1980: Wu C-Y, Martel J, Young D, Young JD. Fetuin-A/albumin-mineral complexes resembling serum calcium granules and putative nanobacteria:

9 Herrmann et al Clearance of Calciprotein Particles 583 demonstration of a dual inhibition-seeding concept. PLoS One. 2009;4: e Young J, Martel J, Young L, Wu C, Young A, Young D. Putative nanobacteria represent physiological remnants and culture by-products of normal calcium homeostasis. PLoS One. 2009;4:e Young JD, Martel J, Young D, Young A, Hung C-M, Young L, Chao Y-J, Young J, Wu C-Y. Characterization of granulations of calcium and apatite in serum as pleomorphic mineralo-protein complexes and as precursors of putative nanobacteria. PLoS One. 2009;4:e Ryall R. The future of stone research: rummagings in the attic, Randall s plaque, nanobacteria, and lessons from phylogeny. Urol Res. 2008;36: Raoult D, Drancourt M, Azza S, Nappez C, Guieu R, Rolain J-M, Fourquet P, Campagna B, La Scola B, Mege J-L, Mansuelle P, Lechevalier E, Berland Y, Gorvel J-P, Renesto P. Nanobacteria are mineralo fetuin complexes. PLoS Pathog. 2008;4:e Price PA, Nguyen TMT, Williamson MK. Biochemical characterization of the serum fetuin-mineral complex. J Biol Chem. 2003;278: Heiss A, DuChesne A, Denecke B, Grötzinger J, Yamamoto K, Renné T, Jahnen-Dechent W. Structural basis of calcification inhibition by alpha 2-hs glycoprotein/fetuin-a: formation of colloidal calciprotein particles. J Biol Chem. 2003;278: Heiss A, Jahnen-Dechent W, Endo H, Schwahn D. Structural dynamics of a colloidal protein-mineral complex bestowing on calcium phosphate a high solubility in biological fluids. Biointerphases. 2007;2: Heiss A, Eckert T, Aretz A, Richtering W, Van Dorp W, Schafer C, Jahnen-Dechent W. Hierarchical role of fetuin-a and acidic serum proteins in the formation and stabilization of calcium phosphate particles. J Biol Chem. 2008;283: Heiss A, Pipich V, Jahnen-Dechent W, Schwahn D. Fetuin-A is a mineral carrier protein: small angle neutron scattering provides new insight on fetuin-a controlled calcification inhibition. Biophys J. 2010;99: Wald J, Wiese S, Eckert T, Jahnen-Dechent W, Richtering W, Heiss A. Formation and stability kinetics of calcium phosphate fetuin-a colloidal particles probed by time-resolved dynamic light scattering. Soft Matter. 2011;7: Jahnen-Dechent W, Heiss A, Schäfer C, Ketteler M. Fetuin-A regulation of calcified matrix metabolism. Circ Res. 2011;108: Lee C, Bongcam-Rudloff E, Sollner C, Jahnen-Dechent W, Claesson-Welsh L. Type 3 cystatins; fetuins, kininogen and histidine-rich glycoprotein. Front Biosci. 2009;14: Schäfer C, Heiss A, Schwarz A, Westenfeld R, Ketteler M, Floege J, Müller-Esterl W, Schinke T, Jahnen-Dechent W. The serum protein alpha 2-heremans-schmid glycoprotein/fetuin-a is a systemically acting inhibitor of ectopic calcification. J Clin Invest. 2003;112: Westenfeld R, Schäfer C, Smeets R, Brandenburg VM, Floege J, Ketteler M, Jahnen-Dechent W. Fetuin-A (ahsg) prevents extraosseous calcification induced by uraemia and phosphate challenge in mice. Nephrol Dial Transplant. 2007;22: Westenfeld R, Schäfer C, Krüger T, Haarmann C, Schurgers LJ, Reutelingsperger C, Ivanovski O, Drueke T, Massy ZA, Ketteler M, Floege J, Jahnen-Dechent W. Fetuin-A protects against atherosclerotic calcification in CKD. J Am Soc Nephrol. 2009;20: Ketteler M, Bongartz P, Westenfeld R, Wildberger J, Mahnken A, Bohm R, Metzger T, Wanner C, Jahnen-Dechent W, Floege J. Association of low fetuin-a (ahsg) concentrations in serum with cardiovascular mortality in patients on dialysis: a cross-sectional study. Lancet. 2003;361: Stenvinkel P, Wang K, Qureshi A, Axelsson J, Pecoits-Filho R, Gao P, Barany P, Lindholm B, Jogestrand T, Heimbürger O, Holmes C, Schalling M, Nordfors L. Low fetuin-a levels are associated with cardiovascular death: impact of variations in the gene encoding fetuin. Kidney Int. 2005;67: Wang A, Woo J, Lam C, Wang M, Chan I, Gao P, Lui S-F, Li P, Sanderson J. Associations of serum fetuin-a with malnutrition, inflammation, atherosclerosis and valvular calcification syndrome and outcome in peritoneal dialysis patients. Nephrol Dial Transpl. 2005;20: Moe S, Reslerova M, Ketteler M, O Neill K, Duan D, Koczman J, Westenfeld R, Jahnen-Dechent W, Chen N. Role of calcification inhibitors in the pathogenesis of vascular calcification in chronic kidney disease (CKD). Kidney Int. 2005;67: Price PA, Thomas GR, Pardini AW, Figueira WF, Caputo JM, Williamson MK. Discovery of a high molecular weight complex of calcium, phosphate, fetuin, and matrix gamma-carboxyglutamic acid protein in the serum of etidronate-treated rats. J Biol Chem. 2002;277: Matsui I, Hamano T, Mikami S, Fujii N, Takabatake Y, Nagasawa Y, Kawada N, Ito T, Rakugi H, Imai E, Isaka Y. Fully phosphorylated fetuin-a forms a mineral complex in the serum of rats with adenineinduced renal failure. Kidney Int. 2009;75: Smith ER, Ford ML, Tomlinson LA, Rajkumar C, McMahon LP, Holt SG. Phosphorylated fetuin-a-containing calciprotein particles are associated with aortic stiffness and a procalcific milieu in patients with pre-dialysis CKD. Nephrol Dial Transplant Jenkin C, Rowley D. The role of opsonins in the clearance of living and inert particles by cells of the reticuloendothelial system. J Exp Med. 1961;114: Beduneau A, Ma Z, Grotepas C, Kabanov A, Rabinow B, Gong N, Mosley R, Dou H, Boska M, Gendelman H. Facilitated monocyte-macrophage uptake and tissue distribution of superparmagnetic iron-oxide nanoparticles. PLoS One. 2009;4:e Elliott M, Ravichandran K. Clearance of apoptotic cells: implications in health and disease. J Cell Biol. 2010;189: Tolleshaug H. Binding and internalization of asialo-glycoproteins by isolated rat hepatocytes. Int J Biochem. 1981;13: Drevon CA, Tolleshaug H, Carlander B, Berg T. Metabolism of asialoglycoproteins in cultured rat hepatocytes: evidence for receptor mediated uptake and degradation which is not feed-back regulated. Int J Biochem. 1983;15: Tozawa R, Ishibashi S, Osuga J, et al. Asialoglycoprotein receptor deficiency in mice lacking the major receptor subunit: its obligate requirement for the stable expression of oligomeric receptor. J Biol Chem. 2001;276: Mebius RE, Kraal G. Structure and function of the spleen. Nat Rev Immunol. 2005;5: Greaves DR, Gordon S. The macrophage scavenger receptor at 30 years of age: current knowledge and future challenges. J Lipid Res. 2009; 50(Suppl):S282 S Nimmerjahn F, Ravetch J. Fcgamma receptors as regulators of immune responses. Nat Rev Immunol. 2008;8: Bharadwaj D, Mold C, Markham E, Du Clos TW. Serum amyloid p component binds to Fc gamma receptors and opsonizes particles for phagocytosis. J Immunol. 2001;166: Kundranda M, Ray S, Saria M, Friedman D, Matrisian L, Lukyanov P, Ochieng J. Annexins expressed on the cell surface serve as receptors for adhesion to immobilized fetuin-a. Biochim Biophys Acta. 2004;1693: Chen NX, O Neill KD, Chen X, Duan D, Wang E, Sturek MS, Edwards JM, Moe SM. Fetuin-A uptake in bovine vascular smooth muscle cells is calcium dependent and mediated by annexins. Am J Physiol Renal Physiol. 2007;292:F599 F Frey B, Schildkopf P, Rödel F, Weiss E-M, Munoz LE, Herrmann M, Fietkau R, Gaipl US. Annexina5 renders dead tumor cells immunogenic: implications for multimodal cancer therapies. J Immunotoxicol. 2009;6: Grewal T, Heeren J, Mewawala D, Schnitgerhans T, Wendt D, Salomon G, Enrich C, Beisiegel U, Jäckle S. Annexin VI stimulates endocytosis and is involved in the trafficking of low density lipoprotein to the prelysosomal compartment. J Biol Chem. 2000;275: Yamamoto M, Sato S, Hemmi H, Hoshino K, Kaisho T, Sanjo H, Takeuchi O, Sugiyama M, Okabe M, Takeda K, Akira S. Role of adaptor trif in the myd88-independent toll-like receptor signaling pathway. Science. 2003;301: Spiess M. The asialoglycoprotein receptor: a model for endocytic transport receptors. Biochemistry. 1990;29: Tolleshaug H, Berg T. Uptake and degradation of asialo-fetuin by isolated rat hepatocytes. Acta Biol Med Ger. 1977;36: Henson P, Hume D. Apoptotic cell removal in development and tissue homeostasis. Trends Immunol. 2006;27: De Jong W, Hagens W, Krystek P, Burger M, Sips A, Geertsma R. Particle size-dependent organ distribution of gold nanoparticles after intravenous administration. Biomaterials. 2008;29: Nagayama S, Ogawara K-i, Minato K, Fukuoka Y, Takakura Y, Hashida M, Higaki K, Kimura T. Fetuin mediates hepatic uptake of negatively charged nanoparticles via scavenger receptor. Int J Pharmaceutics. 2007; 329:

10 584 Circulation Research August 17, Blomhoff R, Eskild W, Berg T. Endocytosis of formaldehyde-treated serum albumin via scavenger pathway in liver endothelial cells. Biochem J. 1984; 218: Hughes D, Fraser I, Gordon S. Murine macrophage scavenger receptor: in vivo expression and function as receptor for macrophage adhesion in lymphoid and non-lymphoid organs. Eur J Immunol. 1995;25: Taylor PR, Martinez-Pomares L, Stacey M, Lin H-H, Brown GD, Gordon S. Macrophage receptors and immune recognition. Annu Rev Immunol. 2005;23: Jansen RW, Molema G, Harms G, Kruijt JK, van Berkel TJ, Hardonk MJ, Meijer DK. Formaldehyde treated albumin contains monomeric and polymeric forms that are differently cleared by endothelial and Kupffer cells of the liver: evidence for scavenger receptor heterogeneity. Biochem Biophys Res Commun. 1991;180: Absolom DR. Opsonins and dysopsonins: an overview. Methods Enzymol. 1986;132: Neyen C, Plüddemann A, Roversi P, Thomas B, Cai L, Van Der Westhuyzen DR, Sim RB, Gordon S. Macrophage scavenger receptor a mediates adhesion to apolipoproteins A-I and E. Biochemistry. 2009;48: Keeley FW, Sitarz EE. Identification and quantitation of alpha 2-hs-glycoprotein in the mineralized matrix of calcified plaques of atherosclerotic human aorta. Atherosclerosis. 1985;55: Moe S, Neal C, O Neill K, Brown K, Westenfeld R, Jahnen-Dechent W, Ketteler M. Fetuin-A and matrix gla protein (mgp) are important inhibitors of vascular calcification in CKD. J Am Soc Nephrol. 2003;14:692A 692A. 55. Emoto M, Mori K, Lee E, Kawano N, Yamazaki Y, Tsuchikura S, Morioka T, Koyama H, Shoji T, Inaba M, Nishizawa Y. Fetuin-A and atherosclerotic calcified plaque in patients with type 2 diabetes mellitus. Metabolism. 2010;59: Naik V, Leaf EM, Hu JH, Yang H-Y, Nguyen NB, Giachelli CM, Speer MY. Sources of cells that contribute to atherosclerotic intimal calcification: an in vivo genetic fate mapping study. Cardiovasc Res. 2012;94: Fadini GP, Albiero M, Menegazzo L, et al. Widespread increase in myeloid calcifying cells contributes to ectopic vascular calcification in type 2 diabetes. Circ Res. 2011;108: Novelty and Significance What Is Known? Deficiency in the plasma protein fetuin-a is associated with pathological calcifications. Fetuin-A protects against calcification by forming soluble proteinmineral complexes called calciprotein particles (CPP). CPP carry insoluble mineral that may otherwise deposit in soft tissues. What New Information Does This Article Contribute? Blood clearance of synthetic CPP was studied in mice. Liver and spleen macrophages cleared injected CPP within minutes. Scavenger Receptor-A predominantly mediated the clearance of CPP. Increased plasma levels of CPP are associated with aortic stiffness and a procalcific milieu in patients with chronic kidney disease. Thus, reduced CPP clearance could contribute to increased calcification observed in these patients. We studied CPP clearance in mice to identify the tissues, cell types, and receptors involved in CPP clearance. CPP were predominantly cleared by the liver and the spleen. Kupffer cells in the liver and marginal zone macrophages in the spleen were the major cell types involved in CPP clearance. Scavenger Receptor A was the predominant cellular receptor mediating CPP clearance. This finding is particularly significant, because Scavenger Receptor A is also involved in the clearance of lipoprotein particles. The role of CPP in calcification disease may be similar to the role of low-density lipoprotein in atherosclerosis. Atherosclerotic plaque macrophages take up both CPP and low-density lipoprotein, and this could trigger local vascular calcification. Thus, both defective systemic clearance and increased local uptake could contribute to plaque calcification in patients with a combined procalcific and proatherogenic milieu. Additional studies are needed to identify potential risks and benefits of CPP clearance.