Noninvasive Imaging of Intracellular Lipid Metabolism in Macrophages

Article pubs.acs.org/ac

Noninvasive Imaging of Intracellular Lipid Metabolism in Macrophages by Raman Microscopy in Combination with Stable Isotopic Labeling Christian Matthaü s,*,† Christoph Krafft,† Benjamin Dietzek,† Bernhard R. Brehm,‡ Stefan Lorkowski,∥,⊥ and Jürgen Popp†,§,⊥ †

Institute of Photonic Technology, Albert-Einstein-Straße, 9, 07745 Jena, Germany Katholische Klinik Koblenz, Innere Medizin/Kardiologie, Rudolf-Virchow-Str9, 56073 Koblenz, Germany § Institute for Physical Chemistry and Abbe Center of Photonics, Friedrich Schiller University Jena, Helmholtzweg 4, 07743 Jena, Germany ∥ Institute of Nutrition, Friedrich Schiller University Jena, Dornburger Straße 25, 07743 Jena, Germany ‡

ABSTRACT: Monocyte-derived macrophages play a key role in atherogenesis because their transformation into foam cells is responsible for deposition of lipids in plaques within arterial walls. The appearance of cytosolic lipid droplets is a hallmark of macrophage foam cell formation, and the molecular basics involved in this process are not well understood. Of particular interest is the intracellular fate of different individual lipid species, such as fatty acids or cholesterol. Here, we utilize Raman microscopy to image the metabolism of such lipids and to trace their subsequent storage patterns. The combination of microscopic information with Raman spectroscopy provides a powerful molecular imaging method, which allows visualization at the diffraction limit of the employed laser light and biochemical characterization through associated spectral information. In order to distinguish the molecules of interest from other naturally occurring lipids spectroscopically, deuterium labels were introduced. Intracellular distribution and metabolic changes were observed for serum albumin-complexed palmitic and oleic acid and cholesterol and quantitatively evaluated by monitoring the increase in CD scattering intensities at 0.5, 1, 3, 6, 24, 30, and 36 h. This approach may also allow for investigating the cellular trafficking of other molecules, such as nutrients, metabolites, and drugs.

M

or long-term lipid storage.5 Lipid droplets, for example in macrophages, are in a dynamic state of continuous and tightly regulated esterification and hydrolysis, whereby the hydrolysis of triglycerides is about 3−4 times faster than that of esterified cholesterol. Lipid accumulation can last up to several days.6 However, our knowledge about changes at the molecular level, such as triglycerides, cholesteryl esters, phospholipids, and other bioactive metabolites, is limited because these molecules cannot be visualized directly in microscopy. To visualize uptake and the intracellular fate of individual lipids, without the need of external fluorescent labels, it is possible to employ vibrational spectroscopic methods in combination with conventional microscopy. Molecular vibrations can be excited by either infrared radiation or Raman scattering using monochromatic laser excitation. The resolution depends on the wavelength of the employed laser light, the numerical aperture of the microscopic objective, and the sampling interval.7,8 Raman microscopy has been successfully applied for monitoring cellular composition and metabolic

acrophages are the phagocytic cells of the cellular immune system, which play a key role in many physiological and pathophysiological processes.1 As phagocytes, they engulf and digest cellular debris, pathogens, and other extraneous material. Within the arterial wall, macrophages are particularly responsible for scavenger receptor-mediated removal and subsequent recycling of cytotoxic modified lowdensity lipoproteins (LDL). Lipids obtained from ingested lipoproteins are exported to acceptor particles, such as highdensity lipoproteins (HDL), for transport back to the liver (i.e., reverse cholesterol transport).2 Fatally, scavenger receptormediated uptake is not subjected to negative feedback regulation so that macrophages take up lipids uncontrolledly and excessively.3 If their export capacity is overwhelmed, macrophages inevitably store lipids as triglycerides and cholesteryl esters in cytosolic dropletlike organelles which cause a foamy appearance of the cells.2 This transformation of macrophages into so-called foam cells is a hallmark of early atherosclerosis which is the most common cause of death in Western developed countries.4 In contrast to previous assumptions, recent studies indicate that cytosolic lipid droplets of eukaryotic cells are by no means static cellular structures which are used only for intermediate© 2012 American Chemical Society

Received: May 22, 2012 Accepted: September 7, 2012 Published: September 7, 2012 8549

dx.doi.org/10.1021/ac3012347 | Anal. Chem. 2012, 84, 8549−8556

Analytical Chemistry

Article

events within cells.9−18 In order to distinguish molecules of interest from their cellular environment, it is feasible to introduce stable isotopic labels.19−21 The first studies using deuterated phospholipids to study biomembranes were reported as early as 1978,22 and long-term intracellular buildup of cholesterol has been investigated in macrophages exposed to polymeric microspheres using Raman microscopy.23 Here, we introduce the results of uptake and storage dynamics of two common fatty acids and cholesterol in a well-established cell model as a first attempt to visualize macrophage foam cell formation directly at the level of individual lipid species. We analyzed uptake dynamics and distribution of deuterated derivatives of saturated and nonsaturated fatty acids in human cells, palmitic acid and oleic acid, as well as deuterated cholesterol. Our studies show that Raman microscopy in combination with stable deuterium labeling of compounds is a feasible tool for investigating the subcellular distribution of compounds and metabolites in in vitro cell models.

Raman Image Acquisition. For acquiring Raman images of triglyceride-laden cells, mature THP-1 macrophages were incubated for up to 36 h with either 400 μM d31-palmitic acid (Sigma-Aldrich) or 400 μM d33-oleic acid (Sigma-Aldrich) bound to BSA (molar ratio 1:4). For acquiring Raman images of cholesterol-laden cells, mature THP-1 macrophages were incubated for 24, 48, and 72 h with 50 μM cholesterol2,2,3,4,4,6-d6 (Dr. Ehrenstorfer, Augsburg, Germany) mixed with lipid-free BSA (molar ratio 1:4) and in the presence of 10% FBS similar to the protocol of Jepson et al.26 Cells were then rinsed with PBS and fixed in 4% paraformaldehyde in PBS at room temperature. Cells were grown onto calcium fluoride slides (Crystal, Berlin, Germany) and examined as follows. Raman spectra were acquired using a Confocal Raman Microscope Model CRM 300 (WITec, Ulm, Germany). Excitation at 488 nm (about 5 mW at the sample) is provided by an air-cooled Ar+ ion laser (Lasos Lasertechnik, Jena, Germany). A Nikon Fluor water immersion objective (60×/ 1.00 numerical aperture, working distance = 2.0 mm) was used in the studies reported here. The sample is located on a piezo electrically driven microscope scanning stage. Spectra are collected at a 0.5 μm grid with a dwell time of 0.25 s. Image Analysis and Data Processing. For extracting spectral information, several factor methods such as principle component analysis (PCA) or vector component analysis (VCA) have shown high potential for the evaluation of Raman data sets and been described in detail.27,28 All images presented here were reconstructed using a VCA based on the N-FINDR algorithm, described by Winter et al.29,30 Prior to this dimension reduction of the data set, all spectra were cleared from cosmic rays. The CD/CH ratios, for the uptake quantification for each cell, were evaluated by dividing the integrated Raman scattering intensities of the CD (2050− 2275 cm−1) and CH (2800−3020 cm−1) stretches for every pixel of the image, using image processing features of the WITec software (WITec). Graph plots were evaluated using Origin (OriginLab, Northampton, MA).

■

EXPERIMENTAL SECTION Cell Culture. THP-1 monocytes (ATCC, Manassas, VA) were grown in a RPMI 1640 medium (PAA, Coelbe, Germany) supplemented with 10% fetal bovine serum and 0.1 mg/mL penicillin/streptomycin/L-glutamine (PAA) as previously described.24 Cells were cultured at 37 °C in a humidified 5% CO2 atmosphere. Monocytes were differentiated into macrophages using 100 ng/mL phorbol-12-myristate-13-acetate (PMA, Fisher Scientific, Schwerte, Germany) and 50 μM βmercaptoethanol (Roth, Karlsruhe, Germany) in a supplemented RPMI 1640 medium according to described protocols.25 After 96 h, mature macrophages were incubated as indicated in the figures and processed as described below. Quantification of Triglyceride Accumulation. Mature THP-1 macrophages were incubated for up to 36 h with either 400 μM palmitic acid (Sigma-Aldrich) or 400 μM oleic acid (Sigma-Aldrich) bound to lipid-free bovine serum albumin (BSA; molar ratio 1:4). Cells were then washed twice with PBS and harvested in PBS containing 5% Triton X-100. Cell suspension was sonicated three times on ice. An aliquot of the cell suspension was subjected to a BCA Protein Assay (Thermo Fisher Scientific, Bonn, Germany) for cellular protein quantification. For dissolving triglycerides, remaining samples were incubated at 80 °C for 10 min, mixed thoroughly for 5 min at 80 °C, and then chilled to room temperature. This incubation process was repeated twice. Cell lysates were centrifuged for 5 min at 16000g; 50 μL of the cell suspension was analyzed with a Triglyceride GPO-PAP kit (Roche Diagnostics, Mannheim, Germany). Absorption was measured at 544 nm using a Fluostar Optima (BMG Labtech, Offenburg, Germany). Quantification of Cholesterol Accumulation. Mature THP-1 macrophages were incubated for 24, 48, and 72 h with 50 μM cholesterol (Sigma-Aldrich) incubated with lipid-free BSA (molar ratio 1:4), in the presence of 10% FBS following a modification of the protocol described by Jepson et al.26 For quantification of the cellular total cholesterol content, a CHOD-PAP Kit from Roche Diagnostics was used according to the manufacturer’s instructions. The total cholesterol was normalized to intracellular protein content as measured by a BCA Protein Assay (Thermo Fisher Scientific). Absorption was measured at 544 nm using a Fluostar Optima (BMG Labtech) instrument.

■

RESULTS AND DISCUSSION Raman Spectra of Pure Deuterium-Labeled Fatty Acids and Cholesterol. A Raman spectrum of crystalline d31-palmitic acid is shown in Figure 1A. The spectrum is dominated by the signals of CD stretching vibrations between 2000 and 2300 cm−1. The strongest bands are observed for symmetric CD2 stretches at 2097 cm−1. The peak at 2194 cm−1 can be assigned to symmetric chain-end CD3 stretches. The bands in the lower-wavenumber region, likely due to CD2 deformations, are far less pronounced. The Raman spectrum of crystalline d33-oleic acid is depicted in Figure 1B. Band positions and band shapes of symmetric and antisymmetric CD stretching modes are very similar to those of d31-palmitic acid. Unsaturated CD stretches are typically less pronounced and appear at 2242 cm−1. In addition to CD bands, the scattering due to CC stretching can be observed at 1622 cm−1. A Raman spectrum of crystalline cholesterol, deuterated at positions 2, 3, and 4 of the A ring, as well as at position 6 of the B ring, is shown in Figure 1C. Stretches of CD bonds are observed between 2000 and 2275 cm−1. Moieties of CD2 at positions 2 and 4 are chemically very similar, and their symmetric stretching modes are centered at 2081 and 2122 cm−1, respectively. The band at 2240 cm−1 can be assigned to the CD vibration next to the double bond in ring B. The 8550



Article

FINDR algorithm is shown in red and apparently colocalizes with cytosolic lipid droplets. The endmember spectrum associated with regions of the image shown in red is plotted in Figure 3A and represents the

Figure 1. Representative Raman spectra of pure (A) d31-palmitic acid, (B) d33-oleic acid, and (C) d6-2,2,3,4,4,6-cholesterol. Most pronounced scattering intensities originate from CD2 and CD3 stretching vibrations between 2000 and 2300 cm−1. Specific CC bands are observed at 1622 cm−1 (for the deuterated fatty acids) and at 1653 cm−1 (for the deuterated cholesterol).

Figure 3. Endmember spectra associated with the image analysis shown in Figure 2. (A) Spectrum 1 (top) corresponds to the red regions showing the distribution of d31-palmitic acid. The spectrum is overlaid by spectral features of unsaturated lipids stored within the lipid droplets. (B) Spectrum 2 (blue) is the endmember spectrum that is associated with the cell body, reflecting the protein composition of the cell.

signals at 2197 cm−1 are likely due to asymmetric stretching combinations of CD2 groups. The remaining part of the Raman spectrum is very complex because of the molecular structure of the sterane backbone of cholesterol.31 Accumulation of Triglycerides in Fatty Acid-Laden Macrophages. Figure 2A shows a bright field image of a macrophage incubated with serum albumin-complexed d31palmitic acid at a concentration of 400 μM for 30 min. The phase contrast image visualizes lipid droplets, the nucleus, and edges of the cytoplasm. A Raman microscopic image of the cell is provided in Figure 2B. The image was generated by plotting the integrated CH scattering intensities between 2800 and 3020 cm−1, reflecting protein or lipid density within the cell body. The highest scattering originates from lipid bodies. The image depicted in Figure 2C is the result of the N-FINDR spectral decomposition algorithm, which decomposes the Raman spectra data set into two main spectral components, usually termed endmembers. The overall protein distribution obtained from the spectral information is plotted in light blue and reconstructs the cell body with its main features. The distribution of d31-palmitic acid as obtained from the N-

contributions from the CD stretching vibrations of the aliphatic fatty acid chain between 2000 and 2300 cm−1. The spectral positions of the CD2 and CD3 stretches appear unaltered, indicating conformational integrity compared to the spectrum of the crystalline compound. CD intensities are in superposition with strong CH stretching intensities between 2800 and 3100 cm−1. All observed Raman bands are typical for molecules of the lipid family, especially for fatty acids and esterified glycerides. Specific assignments can be found in the literature.31 The composition of the lipid droplets depends on the lipid content in the environment of the cells. The endmember spectrum associated with the light blue regions plotted in Figure 3B shows all spectral features typical for protein vibrations. The vibrational spectra of cellular proteins have been described in detail.32

Figure 2. (A) Bright field and (B, C) Raman images of a THP-1 macrophage incubated with serum albumin- complexed d31-palmitic acid at a concentration of 400 μM for 30 min. The Raman image in (B) was generated by plotting the integrated CH scattering intensities, which reflect the density of the cell’s composition. Image (C) was reconstructed employing the N-FINDR spectral decomposition algorithm and shows the distribution of d31-palmitic acid in red and the protein composition of the cell in blue. 8551



Article

Figure 4. Raman spectra obtained from lipid droplets in THP-1 macrophages at different times of incubation with serum albumin-complexed (A) d31-palmitic acid and (C) d33-oleic acid, reflecting a continuous increase in triglyceride storage. The spectra shown in the insets (B and D) are enlarged for spectral regions between 1350 and 1800 cm−1 for better visualization of less-pronounced spectral features.

Figure 5. THP-1 macrophages incubated with serum albumin-complexed (A) d31-palmitic and (B) d33-oleic acid for the times indicated. CD/CH intensity ratios, given in the lower right corners, reflect the amounts of deuterated fatty acid molecules normalized to the protein composition of the cells, as described in the Experimental Section. (A) d31-palmitic acid and (B) d33-oleic acid are shown in red, and the cellular protein compositions are provided in blue.

8552



Article

Dynamics of Fatty Acid Uptake and Storage. Figure 4A shows the endmember spectra of the lipid fraction in cells incubated with serum albumin-complexed d31-palmitic acid between 0.5 and 36 h. Spectra were normalized to CH stretching intensities for quantitative comparison. The steady increase of CD stretching intensities for the observed time period is easy to notice. The inset of Figure 4B shows an enlarged portion of the spectra showing the peak associated with CO stretching of the carboxy groups at 1738 cm−1, this time, normalized to CH deformation modes around 1440 cm−1. The increase in the intensity implies that the deuterated fatty acid molecules are stored mostly or completely in the form of triglycerides as expected. In order to compare the results for the uptake of the saturated palmitic acid with a nonsaturated fatty acid, the incubation conditions were repeated for d33-oleic acid, which exhibits a single CC group at position 9. Figure 4C shows the endmember spectra representative for macrophages incubated with d33-oleic acid between 0.5 and 36 h. The intensities of the CD stretching regions are again steadily increasing as well as the other spectral features associated with d33-oleic acid. Inset D of Figure 4 also shows an increase in the CC stretching intensities with no change in the position or appearance of the band, indicating that no chemical alteration of the double bond occurred. The intensity of the CO stretch of the carboxyl group at 1745 cm−1 is also increasing over time, showing that most of the fatty acid molecules are likely esterified to triglycerides for storage. Using endmember abundance plots of randomly selected individual cells incubated with 400 μM d31-palmitic acid or d33oleic acid, the de novo formation of lipid droplets is visualized in Figure 5. Figure 5 (panels A and B) shows the Raman images of macrophages incubated with d31-palmitic acid and d33-oleic acid for different time periods between 30 min and 36 h, as indicated. After 30 min, all observed cells showed noticeable inclusions in the form of cytosolic lipid dropletlike structures that contain the deuterated fatty acid molecules likely as triglyceride derivatives. After 30 h, most of the cytoplasm of the cells is completely occupied by lipid droplets. In order to describe the uptake dynamics quantitatively, we evaluated the scattering intensities of the CD stretching vibrations for the reconstructed images and compared them with the CH stretching intensities from the molecules within the cell body. Since the intensities of CD and CH bonds in the Raman cross sections are similar, ratios of the integrated areas of the CD intensities versus the CH intensities can be interpreted as approximated percentage values. This ratio for each individual cell is given in the lower right of the corresponding images and ranges from small values around 0.05, with little variation at early time points, to values larger than 1, after 1 day of incubation. The individual values for the respective time points for palmitic acid are plotted in Figure 6A. The curve shows an exponential dependence of the intracellular accumulation that apparently approaches a plateau after 36 h and can be fitted to

Figure 6. Uptake dynamics for serum albumin-complexed (A) d31palmitic and (B) d33-oleic acid as calculated from the cells shown in Figure 5. Intracellular storage of the fatty acids as triglycerides approaches a plateau after 36 h of continuous incubation. In comparison, enzymatic quantification of triglyceride content of THP-1 control macrophages and THP-1 macrophages incubated with 400 μM serum albumin-complexed (A) palmitic acid or (B) oleic acid for the time periods indicated are scaled on the right y-axis.

After about 13 h, CD scattering intensities of the d31-palmitic acid residues start to exceed CH scattering intensities of the molecules of which the cell consists. The quantification for oleic acid shown in the plot in Figure 6B results in a curve fit similar to that of d31-palmitic acid with C∞ = 2.160 ± 0.319 and τ = 18.652 ± 6.585 h. Thus, uptake curves for both fatty acids are very similar; considering the relatively large variation from cell to cell, the uptake curves are virtually the same, and most of the fatty acids taken up are rapidly transformed into triglycerides for storage. In order to explore the possibility of also monitoring efflux or intracellular degradation pathways, a set of time lapse experiments was performed in which the cells were exposed to a fresh medium after 1 h of incubation with either 400 μM d31-palmitic acid or d33-oleic acid complexed to serum albumin. For both fatty acids, no significant decrease or change in intracellular distribution patterns was observed after up to 2 days of culture under normal growth conditions, in the absence of serum or other extracellular acceptors for fatty acids (data not shown). To confirm that the incubation of macrophages with serum albumin-complexed fatty acids results in the storage of fatty acids as triglycerides, immediately after uptake enzymatic assays were performed to quantify total cellular triglyceride content. As shown in Figure 6, total cellular triglyceride content increased steadily with the time of incubation. These findings

C(t ) = C∞ − (C∞ − C0)e−t / τ

where C∞ is the maximum or plateau value, which can be reached depending on the incubation conditions; C0 is the initial CD/CH value; and τ is the time point at which half of the plateau value is reached. For the curve shown in Figure 6A, the following values were calculated: C∞ = 2.107 ± 0.072 and τ = 21.109 ± 2.044 h. The initial CD/CH ratio is of course zero. 8553



Article

Figure 7. (A) Raman spectra obtained from lipid droplets in THP-1 macrophages at different times of incubation with d6-2,2,3,4,4,6-cholesterol. Spectra demonstrate colocalization of cholesterol with other lipid species, such as triglycerides, likely of serum origin. Distribution of d6-cholesterol is shown in yellow and (B) cellular protein composition is given in blue.

Interestingly, intensities of CD stretching modes apparently do not further increase after 24 h of incubation. In comparison with all other Raman bands, relative intensities after 48 and 72 h are very similar, indicating a level of saturation with d6cholesterol within the lipid droplets. However, the size and amount of the droplets are still increasing, as is shown in Figure 7B. Since the d6-cholesterol molecule is not fully deuterated, it is not possible to quantify the uptake as reliably as for deuterated fatty acids. The CD/CH ratio values in d6cholesterol-incubated cells are small, and variation from cell to cell is, in some cases, larger than between the different time points investigated.

are in good agreement with the data obtained from the Raman microscopic measurements, as explained above. Accumulation of Cholesteryl Esters in CholesterolLaden Macrophages. In order to investigate uptake and accumulation of another class of lipids, namely cholesterol or cholesteryl esters, mature macrophages were loaded with d6cholesterol for different incubation times. The presence of serum and serum albumin was used to ensure sufficient in vitro availability allowing a concentration equivalent to 50 μM d6cholesterol. However, in contrast to the rapid uptake of fatty acids, 24 h of incubation was required to allow reproducible detection of CD scattering intensities within the cytoplasm. Figure 7B shows a series of cells incubated with d6-cholesterol between 24 and 72 h; the corresponding Raman spectra are provided in Figure 7A. The distribution of d6-cholesterol, again obtained by employing the N-FINDR algorithm, is depicted in yellow. As expected, cholesterol is stored in lipid dropletlike structures varying from ∼0.5 to 2 μm in diameter. Spectra given in Figure 7A are spectra averaged from these d6-cholesterolcontaining droplets at the three time points indicated. Apart from CD intensities and other bands referring to cholesterol, CH scattering bands, as well as spectral features below 1800 cm−1, are overlaid by Raman signals from fatty acids of cell or serum origin. Thus, the deuterated cholesterol colocalizes with the triglycerides within the lipid droplets. Of particular interest are Raman bands of the CD stretches between 2000 and 2275 cm−1. The symmetric stretches adjacent to the −OH or ester position have shifted from 2081 and 2122 cm−1 to 2090 and 2128 cm−1, respectively, clearly indicating a chemical change close to these bonds. Furthermore, the appearance of the CD stretch of the unsaturated position is altered and no longer observed as an isolated peak at 2240 cm−1. Again, these spectral changes are likely due to changes in bonding in the direct vicinity to the double bond of the B ring of the steran skeleton. It is not clear whether the double bond itself is affected since the CC Raman band at 1654 cm−1 is masked by CC stretches of unsaturated fatty acid chains, which also appear around 1650 cm−1. We were further interested in whether uptake of free cholesterol follows uptake kinetics similar to that of fatty acids.

■

CONCLUSION Common microscopic detection techniques usually do not allow (i) distinguishing between newly formed lipid droplets and the ones existing prior to intervention, such as incubation with serum albumin-complexed fatty acids, or (ii) visualizing individual lipid species. To overcome these obstacles, we used the well-established THP-1 macrophage foam cell formation model to establish a Raman microscopic-based approach to directly visualize lipid droplet formation by detecting individual deuterium-labeled lipid species. Using this approach, it was possible to follow uptake dynamics and intracellular distribution of exogenously added d31-palmitic acid, d33-oleic acid, and d6-cholesterol in human macrophages. All cells exposed to either fatty acids or cholesterol transformed into foam cells during the late stages of incubation. However, in the case of cholesterol loading, the macrophages likely represent an early stage of transformation. From our results, we conclude that the incorporated lipids, to a great extent, are the labeled molecules added to the culture medium. For quantitative comparison of CD stretching intensities, five cells were selected for imaging at each time point. CD/CH ratio values showed little variation at early points and diverged with longer incubation times. The number of cells may not be statistically representative; however, the exponential fitting reflects very well the maximum capacity and progression of the uptake. Incubation times longer than 40 h resulted in the death of the cells, likely due to lipotoxicity.33,34 Palmitic and oleic acid exhibited very similar uptake patterns as 8554


Analytical Chemistry was described by Lokesh and Wrann.35 Time frame and maximum storage capacity were essentially the same. For both fatty acids, the storage process was not reversible after incubation with the serum- and fatty acid-free culture medium. No significant spectral changes were observed after incorporation of the compounds into the lipid compartments. For conformational changes, shifts on the order of several wavenumbers would be expected, and the conformational orientation of isolated palmitic and oleic acid at room temperature is random. Since no shifts in band positions were observed, it is reasonable to assume that the aliphatic side chains of triglycerides within lipid droplets also do not have distinct orientations. The Raman band associated with the double bond of oleic acid is apparently also not affected after storage. Incubation experiments with deuterated cholesterol also resulted in a steady uptake and storage deposition within lipid droplets. The effective concentration was lower than that of fatty acids because of the apparently lower complexation affinity of cholesterol. Cholesterol or cholesterol ester molecules were found to be stored, along with other lipids of serum origin, in lipid dropletlike structures. After 24 h of incubation, the composition of lipid droplets no longer changed in favor of cholesterol deposition. This early saturation, compared to deuterated fatty acids, may be a consequence of the lower concentration applied or the incubation model chosen, as this was also observed by others previously.36 However, it is also possible that the storage capacity for cholesterol is lower than that for fatty acids, or that a simultaneous upregulation of cholesterol efflux pathways results in equilibrium. Interestingly, incubation with serum-complexed cholesterol did not lead significantly to cell death within 72 h. In contrast to the uptake experiments with fatty acids, significant and reproducible spectral changes in CD stretching regions of the deuterated cholesterol did occur. The shifts of the symmetric stretches of CD2 moieties at positions 2 and 4 likely indicate esterification of the hydroxy group. Changes in the appearance of the CD stretch of the double bond, however, may indicate intracellular oxidation at position 7 of the B ring. This likely represents the formation of either 7-ketocholesterol, 7β-hydroxycholesterol, 5,6-epoxycholesterol, or 3β,5α,6β-trihydroxycholestane by a reactive oxygen species because macrophages do not express cholesterol 7α-hydroxylase (CYP7A1) which would catalyze oxidation of cholesterol at position 7 to 7α-hydroxycholesterol.37 In conclusion, we have successfully applied Raman microspectroscopy in combination with the use of deuterated lipids for visualizing and investigating the formation of cytosolic lipid droplets in macrophages. This approach may be useful to better trace and locate the intracellular fate (for example, oxidation, chain extension, desaturation, cellular localization) and uptake dynamics of different lipid species and other compounds, such as metabolites, drugs, or nutrients.

■

■

ACKNOWLEDGMENTS

■

REFERENCES

Article

We are grateful to Maria Braun for her excellent technical assistance and Marten Maeß for critically reading the manuscript. Financial support from the European Union via the “Europäischer Fonds für Regionale Entwicklung (EFRE)” and the “Thüringer Ministerium für Bildung, Wissenschaft und Kultur (TMBWK)” (Project B714-07037) is greatly acknowledged.

(1) Bogdan, C. In Encyclopedia of Life Sciences; John Wiley & Sons Ltd.: Chichester, West Sussex, England, 2008. (2) Lorkowski, S.; Cullen, P. Atherosclerosis: Pathogenesis, Clinical Features and Treatment; John Wiley & Sons, Ltd.: Chichester, West Sussex, England, 2007. (3) Cullen, P.; Rauterberg, J.; Lorkowski, S. Handb. Exp. Pharmacol. 2005, 170, 3−70. (4) Rader, D.; Daugherty, A. Nature 2008, 451, 904−913. (5) Robenek, H.; Buers, I.; Hofnagel, O.; Robenek, M.; Troyer, D.; Severs, N. Biochim. Biophys. Acta 2009, 1791, 408−418. (6) Minor, L.; Rothblat, G.; Glick, J. J. Lipid Res. 1989, 30, 189−197. (7) Uzunbajakava, N.; Lenferink, A.; Kraan, Y.; Volokhina, E.; Vrensen, G.; Greve, J.; Otto, C. Biophys. J. 2003, 84, 3968−3981. (8) Uzunbajakava, N.; Otto, C. Opt. Lett. 2003, 28, 2073−2075. (9) Krafft, C.; Knetschke, T.; Funk, R.; Salzer, R. Anal. Chem. 2006, 78, 4424−4429. (10) Matthäus, C.; Chernenko, T.; Newmark, J.; Warner, C.; Diem, M. Biophys. J. 2007, 93, 668−673. (11) Neugebauer, U.; Clement, J. H.; Bocklitz, T.; Krafft, C.; Popp, J. J. Biophotonics 2010, 3, 579−587. (12) Rösch, P.; Harz, M.; Peschke, K. D.; Ronneberger, O.; Burkhardt, H.; Popp, J. Biopolymers 2006, 82, 312−316. (13) Baranska, M.; Schulz, H.; Rösch, P.; Strehle, M. A.; Popp, J. Analyst 2004, 129, 926−930. (14) Chan, J.; Lieu, D. J. Biophotonics 2009, 2, 656−668. (15) Schie, I. W.; Wu, J.; Weeks, T.; Zern, M. A.; Rutledge, J. C.; Huser, T. J. Biophotonics 2011, 4, 425−434. (16) Saha, A.; Arora, R.; Yakovlev, V.; Burke, J. J. Biophotonics 2011, 4, 805−813. (17) Kakita, M.; Kaliaperumal, V.; Hamaguchi, H. J. Biophotonics 2011, 5, 20−24. (18) Robinson, I.; Ochsenkühn, M.; Campbell, C.; Giraud, C.; Hossack, W.; Arlt, J.; Crain, J. J. Biophotonics 2010, 3, 138−146. (19) Matthäus, C.; Kale, A.; Chernenko, T.; Torchilin, V.; Diem, M. Mol. Pharmacol. 2008, 5, 287−293. (20) van Manen, H. J.; Lenferink, A.; Otto, C. Anal. Chem. 2008, 80, 9576−9582. (21) Matthäus, C.; Bergner, G.; Krafft, C.; Dietzek, B.; Lorkowski, S.; Popp, J. Proc. SPIE 2010, 7715, 771511−771511. (22) Gaber, B.; Yager, P.; Peticolas, W. Biophys. J. 1978, 22, 191−207. (23) van Manen, H.; van Apeldoorn, A.; Roos, D.; Otto, C. Proc. SPIE 2006, 6093, 122−130. (24) Schnoor, M.; Buers, I.; Sietmann, A.; Brodde, M. F.; Hofnagel, O.; Robenek, H.; Lorkowski, S. J. Immunol. Methods 2009, 344, 109− 115. (25) Schnoor, M.; Cullen, P.; Lorkowski, J.; Stolle, K.; Robenek, H.; Troyer, D.; Rauterberg, J.; Lorkowski, S. J. Immunol. Methods 2008, 180, 5707−5719. (26) Jepson, C.; Harrison, J.; Kraemer, F.; Yeaman, S. Biochem. J. 1996, 318, 173−177. (27) Miljković, M.; Chernenko, T.; Romeo, M.; Bird, B.; Matthäus, C.; Diem, M. Analyst 2010, 135, 2002−2013. (28) Hedegaard, M.; Matthäus, C.; Hassing, S.; Krafft, C.; Diem, M.; Popp, J. Theor. Chem. Acc. 2011, 130, 1249−1260. (29) Winter, M. E. Proc. SPIE 1999, 3753, 266−275. (30) Du, Q.; Raksuntorn, N.; Younan, N. H.; King, R. L. Appl. Opt. 2008, 47, F77−F84.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions ⊥

Both authors contributed equally.

Notes

The authors declare no competing financial interest. 8555



Article

(31) Krafft, C.; Neudert, L.; Simat, T.; Salzer, R. Spectrochim. Acta, Part A 2005, 61, 1529−1535. (32) Spiro, T. G. Biological Applications of Raman Spectroscopy; Mercer University Press: New York, 1987. (33) Aronis, A.; Madar, Z.; Tirosh, O. Nutrition 2008, 24, 167−176. (34) Hosek, J.; Zavalova, V.; Kollar, P. Immunopharmacol. Immunotoxicol. 2010, 32, 462−465. (35) Lokesh, B.; Wrann, M. Biochim. Biophys. Acta 1984, 792, 141− 148. (36) Chen, J.; Greenberg, A.; Tseng, Y.; Wang, S. J. Cell. Biochem. 2001, 83, 187−199. (37) Björkhem, I. J. Clin. Invest. 2002, 110, 725−730.

8556


Noninvasive Imaging of Intracellular Lipid Metabolism in Macrophages

Recommend Documents