High-resolution imaging of dietary lipids in cells and tissues by NanoSIMS analysis

Abstract

Nanoscale secondary ion MS (NanoSIMS) imaging makes it possible to visualize stable isotope-labeled lipids in cells and tissues at 50 nm lateral resolution. Here we report the use of NanoSIMS imaging to visualize lipids in mouse cells and tissues. After administering stable isotope-labeled fatty acids to mice by gavage, NanoSIMS imaging allowed us to visualize neutral lipids in cytosolic lipid droplets in intestinal enterocytes, chylomicrons at the basolateral surface of enterocytes, and lipid droplets in cardiomyocytes and adipocytes. After an injection of stable isotope-enriched triglyceride-rich lipoproteins (TRLs), NanoSIMS imaging documented delivery of lipids to cytosolic lipid droplets in parenchymal cells. Using a combination of backscattered electron (BSE) and NanoSIMS imaging, it was possible to correlate the chemical data provided by NanoSIMS with high-resolution BSE images of cell morphology. This combined imaging approach allowed us to visualize stable isotope-enriched TRLs along the luminal face of heart capillaries and the lipids within heart capillary endothelial cells. We also observed examples of TRLs within the subendothelial spaces of heart capillaries. NanoSIMS imaging provided evidence of defective transport of lipids from the plasma LPs to adipocytes and cardiomyocytes in mice deficient in glycosylphosphatidylinositol-anchored HDL binding protein 1. Nanoscale secondary ion MS (NanoSIMS) imaging makes it possible to visualize stable isotope-labeled lipids in cells and tissues at 50 nm lateral resolution. Here we report the use of NanoSIMS imaging to visualize lipids in mouse cells and tissues. After administering stable isotope-labeled fatty acids to mice by gavage, NanoSIMS imaging allowed us to visualize neutral lipids in cytosolic lipid droplets in intestinal enterocytes, chylomicrons at the basolateral surface of enterocytes, and lipid droplets in cardiomyocytes and adipocytes. After an injection of stable isotope-enriched triglyceride-rich lipoproteins (TRLs), NanoSIMS imaging documented delivery of lipids to cytosolic lipid droplets in parenchymal cells. Using a combination of backscattered electron (BSE) and NanoSIMS imaging, it was possible to correlate the chemical data provided by NanoSIMS with high-resolution BSE images of cell morphology. This combined imaging approach allowed us to visualize stable isotope-enriched TRLs along the luminal face of heart capillaries and the lipids within heart capillary endothelial cells. We also observed examples of TRLs within the subendothelial spaces of heart capillaries. NanoSIMS imaging provided evidence of defective transport of lipids from the plasma LPs to adipocytes and cardiomyocytes in mice deficient in glycosylphosphatidylinositol-anchored HDL binding protein 1. Lipids are crucial for the formation of membranes, as fuel for oxidative metabolism, and for the generation of signaling molecules (1GunstoneF. D.HarwoodJ. L.DijkstraA. J.. 2012. The Lipid Handbook with CD-ROM. CRC Press, Boca Raton, FL.Google Scholar). The ability to visualize lipids within cells and tissues is important, but the experimental approaches for creating high-resolution images of lipids within cells and tissues are quite limited. The most commonly used technique for lipid imaging in tissues is fluorescence microscopy after the delivery of fluorescently labeled lipids (2Schultz C. Neef A.B. Gadella Jr, T.W. Goedhart J. Imaging lipids in living cells.Cold Spring Harb. Protoc. 2010; 2010: pdb.top83PubMed Google Scholar, 3Kuerschner L. Moessinger C. Thiele C. Imaging of lipid biosynthesis: how a neutral lipid enters lipid droplets.Traffic. 2008; 9: 338-352Crossref PubMed Scopus (303) Google Scholar). A drawback of that approach is that some lipids are not readily available as fluorescent compounds, and one must worry about how a fluorescent tag affects the biological properties of the lipid. To overcome these issues, label-free lipid imaging techniques have been developed, such as coherent anti-Stokes Raman scattering microscopy (4Evans C.L. Xie X.S. Coherent anti-Stokes Raman scattering microscopy: chemical imaging for biology and medicine.Annu. Rev. Anal. Chem. (Palo Alto Calif.). 2008; 1: 883-909Crossref PubMed Scopus (992) Google Scholar) and stimulated Raman scattering microscopy (5Freudiger C.W. Min W. Saar B.G. Lu S. Holtom G.R. He C. Tsai J.C. Kang J.X. Xie X.S. Label-free biomedical imaging with high sensitivity by stimulated Raman scattering microscopy.Science. 2008; 322: 1857-1861Crossref PubMed Scopus (1671) Google Scholar), but the lateral resolution and sensitivity of these approaches are limited (6Folick A. Min W. Wang M.C. Label-free imaging of lipid dynamics using coherent anti-Stokes Raman scattering (CARS) and stimulated Raman scattering (SRS) microscopy.Curr. Opin. Genet. Dev. 2011; 21: 585-590Crossref PubMed Scopus (68) Google Scholar). More recently, imaging MS has been utilized to visualize lipids (7Gode D. Volmer D.A. Lipid imaging by mass spectrometry - a review.Analyst. 2013; 138: 1289-1315Crossref PubMed Scopus (167) Google Scholar, 8Goto-Inoue N. Hayasaka T. Zaima N. Setou M. Imaging mass spectrometry for lipidomics.Biochim. Biophys. Acta. 2011; 1811: 961-969Crossref PubMed Scopus (136) Google Scholar). New ionization methods, primary ion beams, and newer instrument designs have made imaging MS more powerful. MALDI techniques, TOF secondary ion MS (SIMS), and magnetic sector SIMS represent complementary techniques for lipid imaging (9Kraft M.L. Klitzing H.A. Imaging lipids with secondary ion mass spectrometry.Biochim. Biophys. Acta. 2014; 1841: 1108-1119Crossref PubMed Scopus (88) Google Scholar). In that order, these techniques have increasing spatial resolution but decreasing molecular specificity. MALDI techniques are able to identify and quantify lipids with a resolution of a few micrometers; TOF-SIMS retains the ability to identify specific lipids and can achieve 0.5–1 µm lateral resolution. By measuring the elemental composition of lipids, a nanoscale SIMS (NanoSIMS) instrument yields images of lipids with up to 50 nm lateral resolution. We describe the use of NanoSIMS imaging to create high-resolution images of lipids in cells and tissues. With NanoSIMS imaging, a primary ion beam (Cs+ or O−) is used to bombard a “surface of interest” (e.g., a section of mouse tissue), releasing secondary ions that can be detected by a magnetic mass analyzer (Fig. 1) (10Boxer S.G. Kraft M.L. Weber P.K. Advances in imaging secondary ion mass spectrometry for biological samples.Annu. Rev. Biophys. 2009; 38: 53-74Crossref PubMed Scopus (250) Google Scholar). The NanoSIMS instrument detects up to seven distinct masses in parallel. Due to a special lens design and the characteristics of the mass analyzer, these instruments have an unusual combination of high lateral resolution, high mass resolution, and high sensitivity (11Van TendelooG.Van DyckD.PennycookS. J., editors. 2012. Handbook of Nanoscopy. Vol. 1. Wiley-VCH, Weinheim, Germany.Google Scholar). As with other MS approaches, the NanoSIMS has the ability to detect all isotopes of each element within the sample. This characteristic makes it possible to image and quantify stable isotope-labeled lipids in tissues at high lateral resolution. An advantage of using stable isotope-labeled lipids for imaging is that the stable isotopes have little or no effect on the biochemical properties of the lipids. Here we describe NanoSIMS imaging for visualizing stable isotope-labeled lipids in cells and tissues (Fig. 2). After feeding mice either 13C- or 2H-labeled fatty acids, we harvested tissues and imaged tissue sections with a NanoSIMS 50 instrument. We also visualized lipids in mice after an intravenous injection of triglyceride-rich LPs (TRLs) enriched in 13C or 2H lipids. In many of our studies, we performed backscattered electron (BSE) imaging and NanoSIMS imaging on the same tissue sections. BSE images provide high-resolution morphological information, making it possible to correlate the chemical information provided by NanoSIMS imaging to subcellular features of individual cells. NanoSIMS imaging was effective in documenting defective lipid transport in the setting of glycosylphosphatidylinositol-anchored HDL binding protein not the capillary and the of TRLs is mice mice have been S. high protein a in the of PubMed Scopus Google Scholar). mice a and in a with a a mice have plasma and S. high protein a in the of PubMed Scopus Google Scholar, L. Wang in 2010; PubMed Scopus Google Scholar, C. A. plasma in protein 2010; PubMed Scopus Google Scholar, A. C. A. is for the of triglyceride-rich lipoproteins in 2014; PubMed Scopus Google Scholar). by of labeled fatty acids or labeled was in and lipids to either or We of fatty acids or of 2H-labeled for each In other we mice by a with TRLs that enriched in 13C- or 2H-labeled lipids. To stable isotope-enriched 13C fatty acids to mice by gavage, a The mice of was The plasma was of in a and at for at in a The TRLs from the of the and of in a and at for The TRLs in a of with of stable isotope-enriched TRLs after or the mice and tissue sections for BSE and NanoSIMS We a that a mouse S. C. A. S.G. in that binding to the endothelial cell 2011; PubMed Scopus Google Scholar, A. S.G. within the of are crucial for the binding of Chem. 2009; PubMed Scopus Google Scholar, S. J. A. the of the glycosylphosphatidylinositol-anchored high protein in binding Chem. 2011; PubMed Scopus Google Scholar, S. A. of and endothelial Lipid PubMed Scopus Google Scholar). We in and with and was on a of was from a of of cell in studies, was and After or the mice and with of by of electron microscopy at a of sections for NanoSIMS In some studies, was with for NanoSIMS with in A. C. A. is for the of triglyceride-rich lipoproteins in 2014; PubMed Scopus Google Scholar, S. A. of and endothelial Lipid PubMed Scopus Google Scholar). The in with in for in and with at in the The with increasing of and for by with for with increasing of for for for for at and at in a nm with a on an The sections on a and to We used chemical it is and has been used and for in NanoSIMS in mice M.L. C. imaging mass spectrometry cell and PubMed Scopus Google Scholar, J.C. Wang M. imaging mass spectrometry protein in PubMed Scopus Google Scholar). tissues for imaging in and in for at in and in in The in and in in for at in and at for and in of and and for in in and for at of nm on an with a The sections on with and and in of for at and for at nm on and imaged with a electron A electron that can be used to create high-resolution images of a tissue are that are from the surface of a to of primary with the of within the tissue The of backscattered by is from the BSE is able to on the of in the is used to biological for it lipids and for imaging and New Scholar). in tissue sections more of the tissue tissue sections with are for BSE The resolution of is limited by the of the can be by the of the primary electron beam T. W. of biological and using ion beam and BSE J. Scholar). A primary electron beam is used to achieve high-resolution BSE In our studies, tissue sections imaged by of and the sections to an electron for BSE BSE images with a beam with a and a lens and the tissue SIMS imaging is on MS of secondary from a surface of a of (e.g., a tissue with primary Due to the primary ion beam and the a of (e.g., and or are from the surface of the sample. of these are In most SIMS secondary ions can be detected special are to achieve of neutral A. J. A. S. J. of secondary neutral mass secondary ion mass spectrometry in of in and 2011; PubMed Scopus Google Scholar). of SIMS instruments have on increasing and ionization to achieve increasing the of the mass and spatial resolution SIMS of in and imaging and A. 2013; Scopus Google Scholar, by Scholar). The NanoSIMS 50 instrument is a high-resolution SIMS instrument to spatial resolution the high sensitivity to detect The NanoSIMS the same ion as other SIMS but the primary ion beam the from a The NanoSIMS instrument a design of ion that the same lens the primary ion beam and secondary ions from the sample. This design has a a and secondary ion are of primary ion for the and the of and secondary NanoSIMS instruments can achieve spatial of 50 nm for a primary ion beam and nm for a primary ion beam 2012. Imaging In Handbook of D. and S. J. editors. Weinheim, Scholar). An design of the NanoSIMS is the mass analyzer in a with a section and This mass analyzer makes it possible for the NanoSIMS instrument to achieve high ion and high mass resolution of the secondary combined with high lateral resolution (11Van TendelooG.Van DyckD.PennycookS. J., editors. 2012. Handbook of Nanoscopy. Vol. 1. Wiley-VCH, Weinheim, Germany.Google Scholar). properties it possible to image stable isotopes in tissue sections at a resolution to identify subcellular features of cells. In a primary beam was used to bombard the and secondary ions and and secondary ion are by ionization in is by the of the the surface to of secondary ion can achieve a after a of primary ion An of was to achieve of and for imaging of lipids in tissue sections. The of 13C is the of 2H is biological have been labeled with 13C- or 2H-labeled lipids, the NanoSIMS is of measuring the of 13C or 2H the to the of these mice are 13C- or 2H-labeled fatty acids, it is possible to create a high-resolution image of the of 13C- and 2H-labeled lipids in cells and tissues. In a images are on the or the in the The is to in cell morphology. In our studies, on of from mice and of from of from mice and of from of heart from mice and of heart from on heart from of and NanoSIMS images are or To a NanoSIMS it is to the primary ion beam and image The was used to the of the primary ion A primary was used to images of tissue sections A primary or was used to achieve spatial resolution images of capillaries. The in was used to NanoSIMS The and images with a with of can be in lipid droplets and the capillary are to a and in of can be with the and by BSE and NanoSIMS imaging are complementary techniques for imaging biological and can be performed on the same tissue of the imaging on tissue sections nm are for BSE and NanoSIMS After BSE the tissue sections with nm and to the NanoSIMS instrument for The of from the BSE image can be with an within the NanoSIMS instrument. BSE and NanoSIMS imaging makes it possible to correlate the chemical information from the NanoSIMS with the provided by BSE After administering 13C fatty acids to mice by for we performed NanoSIMS imaging on of NanoSIMS of The was high in cytosolic lipid droplets in the of and in chylomicrons that been along the basolateral surface of In the same NanoSIMS imaging 13C in the cytosolic lipid droplets of (Fig. and heart (Fig. In these studies, the 13C of cytosolic lipid droplets in was the of The 13C of lipid droplets in the heart was the 13C in heart from lipid was NanoSIMS images after an intravenous injection of the 13C of cytosolic lipid droplets of cardiomyocytes in these was imaging of 13C lipids in and and images of from a mouse that been 13C fatty acids by for and images of the heart from the same and images of heart from a mouse that was with 13C In of 13C to lipid µm µm µm µm µm image To NanoSIMS imaging can detect in plasma metabolism, we and mice 13C fatty acids by for and heart and sections for NanoSIMS The more 13C in lipid droplets of mice in lipid droplets from mice (Fig. The in lipid droplets of was in (Fig. The was in heart (Fig. The in lipid droplets was in mice in mice (Fig. The not the of the in heart the of lipid droplets in cardiomyocytes are in mice the of the capillary of mice was in a of 13C TRLs in the plasma of mice to defective of the (Fig. 13C is in in a in tissue sections of mice that have not been lipids has a in a of secondary ions we and mice a available fatty by for and performed NanoSIMS imaging on heart We observed a in heart (Fig. the 2H in lipid droplets was (Fig. The was in cardiomyocytes from mice (Fig. As we observed a high in the heart capillary of the mouse (Fig. defective of 2H in the tissue and of PubMed Scopus Google Scholar) that of lipids in tissues (e.g., tissue and of mice is by lipid in the To the of that we the of and mice that been fatty acids by for (Fig. In the lipid droplets and 13C was In lipid droplets and the 13C of lipid droplets was The 13C in other of the from lipid was in in To the of stable isotope-enriched lipids, we performed NanoSIMS and BSE imaging on tissue sections from mice that been an intravenous injection of 13C- or TRLs and or By BSE imaging, we observed on the luminal surface of capillaries (Fig. NanoSIMS imaging on the same section that these by TRLs enriched in 13C (Fig. or 2H (Fig. In we detected 2H in a cytosolic lipid of a that lipids can from the plasma to parenchymal cells within we NanoSIMS imaging was able to detect endothelial cell protein that and on the luminal surface of capillary endothelial cells. We and mice with a After the mice with and sections of heart for After capillaries by BSE imaging, NanoSIMS was performed on the same sections. We observed of along the luminal face of heart capillaries in binding of the to (Fig. As no was detected in heart capillaries of mice the of on endothelial cells and the of (Fig. we image heart capillaries of mice by we TRLs along the luminal face of endothelial cells (Fig. By we able to high-resolution images of TRLs in capillary endothelial cells (Fig. By we have observed subendothelial TRLs in the capillaries of a mouse heart after injection of TRLs (Fig. To these we performed BSE and NanoSIMS imaging on heart capillaries. In we BSE and NanoSIMS images of a heart capillary from a mouse that been 13C fatty acids by A of 13C was observed within a heart capillary endothelial cell (Fig. In we BSE and NanoSIMS images of a capillary from a mouse after an intravenous injection of 13C was observed a capillary endothelial cell and in TRLs within the capillary BSE and NanoSIMS images of mouse BSE and NanoSIMS images of a heart capillary from a mouse that been 13C fatty acids by In the NanoSIMS a of 13C was observed an endothelial cell BSE and NanoSIMS images of a heart capillary of a mouse after an intravenous injection of 13C In the NanoSIMS a of 13C was observed an endothelial cell 13C was also observed in TRLs in the capillary the image in NanoSIMS images A of was by the BSE and that was BSE images and NanoSIMS images of heart capillaries from a mouse after an injection of 13C lipids capillary endothelial cells to on the In other of 13C to cytosolic lipid droplets and a that along the luminal surface of a BSE and NanoSIMS images of a heart capillary from a mouse after an intravenous injection of 13C A 13C was observed in the subendothelial 13C was also detected in TRLs that along the luminal face of capillary endothelial cells µm µm image We the to heart capillaries of mice (Fig. mice have of on capillary endothelial cells S. A. of and endothelial Lipid PubMed Scopus Google Scholar), but from the we that our of in the of TRLs was In our S. A. of and endothelial Lipid PubMed Scopus Google Scholar), we that has little or no on endothelial cells or on or plasma S. A. of and endothelial Lipid PubMed Scopus Google Scholar). In we BSE images and NanoSIMS images of heart capillaries of mice after an intravenous injection of 13C 13C lipids observed capillary endothelial cells to in the BSE In we BSE and NanoSIMS images of a heart capillary from a mouse after an intravenous injection of 13C In these we observed a in the subendothelial and TRLs that along the luminal face of the capillary endothelial cells. In the we describe NanoSIMS imaging of stable isotope-enriched lipids in cells and tissues. We mice either 13C- or 2H-labeled fatty acids or mice with TRLs 13C or 2H lipids. We able to the and of stable isotope-labeled lipids in the the of lipids endothelial and the and the subcellular of lipids within adipocytes and NanoSIMS imaging with 13C- and 2H-labeled fatty acids, but the of stable to was with 2H-labeled lipids. that with 2H fatty acids are for visualizing of stable in tissues (e.g., of stable at is a drawback to imaging with the NanoSIMS 50 instrument that was used in our studies, the to create the images that are for tissue morphology. the NanoSIMS 50 is to and the secondary ion that can be detected is one can for the to images with BSE A newer NanoSIMS instrument is of and and secondary ions in a NanoSIMS imaging was for visualizing the binding of a to the luminal face of capillaries. In we that to on endothelial cells and from the within T. S.G. C. for glycosylphosphatidylinositol-anchored protein in mouse tissues by Chem. 2010; PubMed Scopus Google Scholar). a fluorescently labeled to the luminal surface of as by microscopy on tissue sections S. A. of and endothelial Lipid PubMed Scopus Google Scholar, C. is for the of 2010; PubMed Scopus Google Scholar). By the binding of the fluorescently labeled was but along the of the capillary In the NanoSIMS the was observed in along capillaries but was not detected along the of the This is to the of in heart capillaries by S. A. of and endothelial Lipid PubMed Scopus Google Scholar). We that of in the NanoSIMS images is to the that we secondary ions from an of the tissue section In studies, Weber P.K. Kraft M.L. for imaging in with lipids using high-resolution secondary ion mass Chem. PubMed Scopus Google Scholar) used NanoSIMS imaging to detect a to binding was documented by NanoSIMS by secondary As as we are our are the to that NanoSIMS imaging is to detect the binding of a labeled in tissue sections. approach is making labeled with and one not to worry about an effect of on We that be many to use NanoSIMS imaging to visualize binding in M. C. S. ion beam imaging of 2014; PubMed Scopus Google Scholar). In the of NanoSIMS imaging to be for the of or to cells. The ability of NanoSIMS imaging to detect in lipid was by our in and and adipocytes the is by on endothelial cells and to the capillary S. A. of and endothelial Lipid PubMed Scopus Google Scholar, C. is for the of 2010; PubMed Scopus Google Scholar, S.G. A. an endothelial cell for Lipid 2011; PubMed Scopus Google Scholar). In the setting of the in tissue and is within the from with TRLs and to and delivery of lipid to parenchymal cells S. high protein a in the of PubMed Scopus Google Scholar, S. A. of and endothelial Lipid PubMed Scopus Google Scholar, in the tissue and of PubMed Scopus Google Scholar, C. is for the of 2010; PubMed Scopus Google Scholar, S.G. A. an endothelial cell for Lipid 2011; PubMed Scopus Google Scholar). NanoSIMS provide of the in an of stable isotope-enriched TRLs in the capillary and delivery of lipids to parenchymal cells. NanoSIMS imaging of the was a in lipid droplets and in mice in the mouse of S. high protein a in the of PubMed Scopus Google Scholar, C. A. plasma in protein 2010; PubMed Scopus Google Scholar, of by the and Chem. PubMed Scopus Google Scholar). the that capillaries in the are that of TRLs not transport of by C. A. plasma in protein 2010; PubMed Scopus Google Scholar). the of TRLs in tissue and is as in the a in in the tissue and of PubMed Scopus Google Scholar). In an of in the tissue and of PubMed Scopus Google Scholar) of fatty acids in the tissue of mice but of fatty acids in the that the defective in tissues of mice is by a in and lipid in the in the tissue and of PubMed Scopus Google Scholar). NanoSIMS images a of 13C in that In our studies, we performed BSE and NanoSIMS imaging on the same tissue This approach made it possible to of to by NanoSIMS imaging and made it possible to correlate the chemical information from the NanoSIMS with the high-resolution morphological information from the BSE This approach was quite we able to that most of the 13C or 2H lipid in parenchymal cells was in neutral lipid 13C or 2H in other of the cell was In the 13C the 13C in the from lipid was the of the BSE and NanoSIMS imaging allowed us to identify 13C- or TRLs that along the luminal surface of capillary endothelial cells. the allowed us to identify of 13C within capillary endothelial cells. In most these of in the approach made it possible to identify a stable isotope-enriched within the subendothelial TRLs also by provide evidence of lipid from the plasma to endothelial cells and parenchymal cells. we are not that we are the of the lipid transport By and we have of TRLs along the luminal surface of capillaries. in to the of in we of lipids endothelial cells In the of stable to In our NanoSIMS imaging studies, we that we have visualized the and transport of TRLs by endothelial cells. We that TRLs to the on capillary endothelial cells A. C. A. is for the of triglyceride-rich lipoproteins in 2014; PubMed Scopus Google Scholar, S. A. of and endothelial Lipid PubMed Scopus Google Scholar). We also that and endothelial cells S. A. of and endothelial Lipid PubMed Scopus Google Scholar, T. S.G. C. for glycosylphosphatidylinositol-anchored protein in mouse tissues by Chem. 2010; PubMed Scopus Google Scholar). We that most examples of stable capillary endothelial cells are to TRLs a with We are not that we are visualizing the of the fatty of is that fatty acids endothelial cells in but that the is that the are detected in tissue sections. is that the fatty acids and that the stable within an of 13C and 2H within cells and tissues. To it be to NanoSIMS imaging in mice after is in that the in NanoSIMS imaging be With one be able to visualize the transport of fatty acids endothelial cells and parenchymal fatty acids and More of capillary in the of A for lipid transport from by lateral in cell PubMed Scopus Google Scholar) that the fatty of along the plasma of endothelial cells. that is the it to be possible to visualize the of fatty acids along the plasma by NanoSIMS the fatty acids for these of be and not a one to identify a NanoSIMS that with We examples of TRLs in the subendothelial spaces of heart capillaries. We that the subendothelial TRLs a endothelial cells with In studies, A. C. tissue 2011; PubMed Scopus Google Scholar) that triglyceride-rich capillaries in We for NanoSIMS imaging in lipid NanoSIMS imaging of transport from or of to cells of the be it be to visualize transport endothelial cells in that for and S.G. A. an endothelial cell for Lipid 2011; PubMed Scopus Google Scholar). it be to fatty and delivery to cells and cells within the are The for with NanoSIMS imaging, for and Kraft for with tissue backscattered electron electron microscopy glycosylphosphatidylinositol-anchored HDL binding protein nanoscale secondary ion MS secondary ion MS electron microscopy triglyceride-rich