Anal Bioanal Chem (2012) 403:1863–1871 DOI 10.1007/s00216-012-5809-x
ORIGINAL PAPER
Visualization of dynamic change in contraction-induced lipid composition in mouse skeletal muscle by matrix-assisted laser desorption/ionization imaging mass spectrometry Naoko Goto-Inoue & Yasuko Manabe & Shouta Miyatake & Shinya Ogino & Ai Morishita & Takahiro Hayasaka & Noritaka Masaki & Mitsutoshi Setou & Nobuharu L. Fujii
Received: 16 November 2011 / Revised: 26 January 2012 / Accepted: 30 January 2012 / Published online: 17 February 2012 # Springer-Verlag 2012
Abstract Lipids in skeletal muscle play a fundamental role both in normal muscle metabolism and in disease states. Skeletal muscle lipid accumulation is associated with several chronic metabolic disorders, including obesity, insulin resistance, and type 2 diabetes. However, it is poorly understood whether the lipid composition of skeletal muscle changes by contraction, due to the complexity of lipid molecular species. In this study, we used matrix-assisted laser desorption/ionization imaging mass spectrometry (MALDI-IMS) to investigate changes in skeletal muscle lipid composition induced by contraction. We successfully observed the reduction of diacylglycerol and triacylglycerol, which are generally associated with muscle contraction. Interestingly, we found the accumulation of some saturated and mono-unsaturated fatty acids and poly-unsaturated fatty acids containing phosphatidylcholine in contracted muscles. Moreover, the distributions of several types of lipid were changed by contraction. Our results show that changes in
Published in the special paper collection Biomedical Mass Spectrometry with guest editors Toyofumi Nakanishi and Mitsutoshi Setou. N. Goto-Inoue : Y. Manabe : S. Miyatake : S. Ogino : A. Morishita : N. L. Fujii (*) Department of Health Promotion Sciences, Graduate School of Human Health Sciences, Tokyo Metropolitan University, 1-1, Minami-Osawa, Hachioji, Tokyo 192-0397, Japan e-mail:
[email protected] N. Goto-Inoue : T. Hayasaka : N. Masaki : M. Setou Department of Cell Biology and Anatomy, Hamamatsu University School of Medicine, 1-20-1, Handayama, Higashi-ku, Hamamatsu, Shizuoka 431-3192, Japan
the lipid amount, lipid composition, and energy metabolic activity can be evaluated in each local spot of cells and tissues at the same time using MALDI-IMS. In conclusion, MALDI-IMS is a powerful tool for studying lipid changes associated with contractions. Keywords Biopolymers/lipids . Imaging (NMR microscopy|electron microscopy) . Mass spectrometry/ICPMS . Biological samples
Introduction Lipids are the major components of cell membranes and are involved in cell signaling, survival, and apoptosis [1]. Lipids comprise a complex range of molecules such as fatty acids (FAs), glycerolipids, glycerophospholipids, and sphingolipids. Each type includes a wide variety of lipid molecular species with different FA compositions. Each lipid species has a specific biological function, and each cell shows a dynamic lipid change in compositions and distributions to perform its particular cell function [2]. Lipids in skeletal muscle play a fundamental role as an energy source both in normal muscle metabolism and in disease states. Excess accumulation of lipid in skeletal muscle is associated with several chronic metabolic disorders, including obesity, insulin resistance, and type 2 diabetes [3]. During muscle contractions, plasma FAs are uptaken and/or stored in muscle cells as an important energy source, then lipolyzed to generate energy. Small changes in total lipid amounts with muscle contraction have been detected by conventional biochemical methods such as thin-layer
1864
chromatography, [4–6], suggesting that dynamic change in lipid composition, rather than lipid amount, occurs during muscle contraction. However, it is poorly understood what type of molecular species is changed by muscle contractions because a valid methodology has not been established. For comprehensive lipid analyses, we use a “lipidomics” approach, which provides further insights into complex metabolic networks of biological systems. Mainly, lipidomics approaches are facilitated by mass spectrometry (MS) to perform full characterization of lipid molecular species. The most widely used MS instrument is MS coupled with liquid chromatography (LC) electrospray ionization (ESI) [7]. LC-ESI-MS can quantify the amounts of lipids as well as identify the structure of lipid molecules. Recently, Hu et al. presented lipidomics analyses during exercise with these systems [5]. However, this technique entails extraction and purification steps, leading to the loss of lipid-distribution information over the biological tissue. Meanwhile, numerous techniques of chemical imaging enable the detection and localization of lipids. Staining with oil red O [8] is a common method to localize the lipid fraction on frozen sections, but once total lipids are stained, it is not possible to distinguish the distribution of each lipid. Imaging mass spectrometry (IMS), a cutting-edge imaging method, is a two-dimensional MS technique used to visualize the spatial distribution of biomolecules. Several ionization methods, including secondary ion mass spectrometry (SIMS), desorption electrospray ionization [9], laser ablation electrospray ionization, and matrix-assisted laser desorption/ionization (MALDI) [10–12], have been investigated as methods of IMS. Especially, MALDIIMS can detect a wide range of molecules and has the ability to characterize structure by the use of tandem mass spectrometric analysis. There are some reports of IMS of skeletal muscles using time-of-flight (TOF)SIMS. The authors demonstrated the distribution of small lipids (lower than m/z 800) with high spatial resolution in some dystrophic tissue and obese model mice [13, 14]. TOF-SIMS is a superior tool for highspatial-resolution IMS (submicron order) of elements and small molecules at different organelle levels of the cell [15]. However, due to the in-source fragmentation and the lack of tandem mass spectrometric facility, SIMS lacks sensitivity for the mass range over m/z 800 and the ability to identify molecules. In this study, we ventured to apply MALDI-IMS for the visualization of lipids in skeletal muscles. MALDI-IMS has the advantage of detecting more than m/z 800 molecules and identifying them by the use of tandem mass spectrometric analyses on tissue sections directly. The aim of this study was to evaluate changes in lipid molecular species composition induced by muscle contraction in skeletal muscle tissue using MALDI-IMS.
N. Goto-Inoue et al.
Materials and methods Materials The matrices, 2, 5-dihydroxybenzoic acid (DHB) and 9aminoacridine (9-AA), were purchased from Bruker Daltonics (Bremen, Germany) and Merck Schuchardt (Hohenbrunn, Germany). Methanol, ethanol, and ultra pure water (Wako Pure Chemical Industries, Osaka, Japan) were used for the preparation of all buffers and solvents. All chemicals used in this study were of the highest purity available. Somnopentyl for anesthesia was purchased from MSD Animal Health (Hoddesdon, UK). Animals and in situ contraction Care and use of laboratory animals were in accordance with the Experimental Animal Committee of Tokyo Metropolitan University and followed the Guidelines for the Proper Conduct of Animal Experiments established by the Science Council of Japan. In situ contractions were performed as in a previous study [16]. Briefly, male ICR mice weighing 35–40 g were anesthetized with sodium pentobarbital (10 μl/g of body weight, administered intraperitoneally); the sciatic nerves to both legs were surgically exposed, and electrodes (Uchida Denshi, Tokyo, Japan) were attached. One leg was subjected to electrical stimulation using a Electronic Stimulator SEN-3401 (Nihon Kohden, Tokyo, Japan) for 60 min (train rate, 1/s; train duration, 500 ms; pulse rate, 100 Hz; duration, 0.1 ms at 1–5 V), and the other leg served as a sham-operated control. After contraction, the transverse abdominal muscle was immediately dissected and submerged in liquid nitrogen. All experiments were performed in biological triplicate (n03). Lipid analysis Total lipids were extracted from muscle samples with chloroform/methanol (2:1, v/v). Equal amounts of the extract were manually applied as 5-mm-wide spots on silica gel 60 high-performance thin-layer chromatography plates (Merck, Darmstadt, Germany). The plates were developed with a solvent system consisting of methyl acetate/1-propanol/chloroform/methanol/0.25% aqueous potassium chloride (25:25:25:10:9, v/v/v/v/v) for phospholipids, whereas for TAG separation the developing solvent was composed of n-hexane/diethylether/acetic acid (80:30:1, v/v/v). These chromatograms were sprayed with 0.005% (w/v) primuline reagent until completely wet and then air-dried thoroughly. Lipid bands were visualized under ultraviolet light at 366 nm. The relative densities of phosphatidylinositol (PI), phosphatidylethanolamine (PE), phosphatidylcholine (PC), sphingomyelin (SM), diacylglycerol (DAG), and triacylglycerol
Dynamic change in contraction-induced lipid composition
(TAG) were quantitatively determined by Image J software (http://rsbweb.nih.gov/ij/). Histochemical staining Muscles were frozen in liquid nitrogen and stored at −80 °C without fixation. Sections were prepared as described previously [17] but with slight modifications. Consecutive 10-μm sections were cut using a cryostat (CM 1950; Leica Microsystems, Wetzlar, Germany). The serial sections were mounted onto MAS-coated slides (Matsunami, Osaka, Japan) for histochemical staining, and indium-tin oxide (ITO)-coated glass slides (Bruker Daltonics) for imaging mass spectrometry. For morphological observation, sections were stained with hematoxylin–eosin (HE) staining. The serial sections were stained with oil red O for the visualization of lipids.
1865
Results Figure 1a shows HE staining images of muscle to identify the morphological differences between basal and contracted tissues (contraction). There are no obvious differences in myofiber conditions between them. Even though skeletal muscles contain many kinds of lipids such as the phospholipids including PI, PE, phosphatidylserine, PC and SM, sphingolipids including ceramides, and neutral lipids including DAG and TAG, we did not see any oil-droplets in either type of tissue section stained by oil red O (Fig. 1b), probably due to the lipid level in the tissue being below the detection limit of the methods. On the other hand, using mass spectrometry, many peaks of lipids were successfully detected in the mass range of m/z 700 to 900. Representative mass spectra of the total lipids extracted from basal and contracted muscle tissues are shown in Fig. 1c.
Imaging mass spectrometry
Tandem mass spectrometry Tandem mass spectrometric analysis was performed using a QSTAR Elite high-performance, hybrid quadrupole time-offlight mass spectrometer (Applied Biosystems, Foster City, CA) with MALDI ion source according to a previously described procedure [2]. Data analysis
(a)
100 um
100 um
(b)
400 um
400 um
(c) 2500
basal
1500 500 0 2500
contraction
1500 500 0 600
Statistical analysis was performed using StatView 5.0 (SAS Institute, Tokyo, Japan). Energy charge was assigned as ([ATP]+0.5[ADP])/([ATP]+[ADP]+[AMP]) [18], and calculated using SIMenergy software (Shimadzu co., Kyoto, Japan) currently under development.
contraction
HE stain
basal
Oil Red O stain
For positive-ion mode, DHB of 50 mg/mL in methanol/ water (7/3, v/v) was uniformly sprayed over the sections with a 0.2-mm nozzle caliber airbrush (Procon Boy FWA Platinum; Mr. Hobby, Tokyo, Japan). Mass spectrometry (MS) was performed with a MALDI TOF/TOF-type instrument, the Ultraflex II (Bruker Daltonics), which was equipped with a 355-nm Nd/YAG laser with a repetition rate of 200 Hz. For negative-ion mode, 9-AA was used as a matrix and MS analyses were performed. The laser irradiated 200 times per position. All pixel sizes of imaging were 25 μm. The MS parameters were set to obtain the highest sensitivity with m/z values in the range of 400–1,000 in positive-ion mode and 200–1,000 in negative-ion mode. Automatic acquisition of the spectra and reconstruction of the ion images were performed using FlexImaging software (Bruker Daltonics). Normalization by total ion current was performed using the same software.
650
700
750
800
850
900
Fig. 1 The optical images of muscle tissue sections. HE stained serial sections used for IMS analyses are shown in (a). Scale bar is 100 μm. Oil red O staining images are shown in (b). Scale bar is 400 μm. We shot the laser randomly to the section. Averaged mass spectra of both tissue sections are shown in (c)
1866
N. Goto-Inoue et al.
To identify the lipid subclasses, we separated these lipids by thin-layer chromatography (TLC). Figure 2a shows thinlayer chromatograms of the extracted lipids. PE, PI, PC, and SM were detected with phospholipid separation, and DAG and TAG were detected with neutral lipid separation. Each detected spot was quantified, and the amount between the conditions was compared. As shown in Fig. 2b, there was no difference between basal and contraction levels of PI, PE, SM, PC, and DAG. The amount of TAG tended to be reduced in contraction (p<0.1). To visualize the localization of each lipid, we performed MALDI-IMS analyses with frozen sections of muscle. Figure 3a represents ion images detected in positive-ion mode. There has been no report in other MALDI-IMS analyses of peaks under m/z 700 such as m/z 661, 683, and 685. Interestingly, all small molecules were drastically reduced by muscle contraction. Since the molecules are ionized as [M+H]+, [M+Na]+, and [M+K]+ in the positive-ion mode of MALDI-IMS, it is difficult to identify the molecules by comparing their molecular mass alone. Therefore, we tried a tandem mass spectrometric analysis to identify the molecules. First, we obtained an MS spectrum of standard DAG (18:1/18:1), then detected [M+Na]+ and [M+K]+ ions at m/z 643 and 659, respectively. Performing tandem mass spectrometric analyses, only the [M+K]+ ion at m/z 659 could get the fragment ion at m/z 321, which is derived from the neutral loss of oleinic acid (282 Da; Fig. 3b). Previous study has demonstrated that the sn-1 position of DAG is easy to fragment [19]. Therefore, by comparison with the tandem mass spectrum of standard DAG (18:1/ 18:1), the ions at m/z 661, 683, and 685 could be identified as DAG (18:1/18:0), DAG (18:1/20:1), and DAG (18:1/ Fig. 2 (a) Representative thinlayer chromatogram of phospholipids and neutral lipids extracted from muscles. (b) Detected spots were visualized by primuline after total lipids were separated by TLC. Comparisons between basal and contraction conditions were made by the Student’s t test (n03)
20:0), respectively. However, it is difficult to distinguish whether these ions derived from the native DAG molecules and/or the fragmentation of TAG molecules. As shown in TLC analyses, the major lipids in the muscle tissues are PCs (Fig. 2a). Therefore, we could detect PCrelated ions predominantly at more than m/z 700. By tandem mass spectrometric analyses, m/z 794, 796, 798, 806, 820, 822, 830, 834, 844, 846, 848, 868, 870, and 872 are assigned as PC (diacyl-34:3), PC (diacyl-34:2), PC (diacyl-34:1), PC (diacyl-38:6), PC (diacyl-36:4), PC (diacyl-36:3), PC (diacyl-38:5), PC (diacyl-38:3), PC (diacyl-38:6), PC (diacyl-38:5), PC (diacyl-38:4), PC (diacyl-40:8), PC (diacyl-40:7), and PC (diacyl-40:6), respectively. Representative tandem mass spectrum of m/z 844 and 848 is shown in Fig. 3b. The adduct ions are assigned as potassium ions because m/z 163 corresponding to [(CH2)2PO4H+K]+ was detected. Moreover, an [M-C20:4+K]+ ion at m/z 540 was detected in the tandem mass spectrometric spectrum of m/z 844. Therefore, we identified that PC (diacyl38:6) is PC (diacyl-18:2/20:4). The molecular ion at m/z 741 was assigned as SM (d18:1/ 16:0). The SM distribution and amount did not change by muscle contraction (Fig. 3a). TAG is also a major lipid type in muscles and is easy to detect with this method. A representative ion at m/z 889 was assigned as TAG (16:1/16:1/ 20:4), and the amounts were lowered by muscle contraction. This result corroborates our TLC analysis (Fig. 2b). To confirm the different distribution of lipids in basal and contraction conditions, we looked at the three ions at m/z 683, 844, and 848, and merged these images (Fig. 3c). The signal intensity of the ion at m/z 683, assigned as DAG (18:1/20:1), was lower in the contraction than basal
(a)
PE
(b) basal
TAG contraction
PI
PC
100
100
100
80
80
80
60
60
60
40
40
40
20
20
20
0
0 PI
SM
100
0 PE
150
150
80
DAG
SM
100
60 40
100
50
*
50
20 0
0 basal contraction basal contraction
PC
0 DAG
TAG
* p <0.1
Dynamic change in contraction-induced lipid composition
1867
(a) Positive ion mode Imaging DAG
SM
m/z 661 m/z 683
m/z 685
PC
m/z 741
m/z 794 m/z 796
m/z 798
m/z 806
m/z 820
basal
100 %
contraction
400 um 400 um
0%
(18:1/18:0)
(18:1/20:1) (18:1/20:0)
(d18:1/16:0) (diacyl-34:3) (diacyl-34:2) (diacyl-34:1) (diacyl-38:6) (diacyl-36:4)
TAG
PC m/z 822
m/z 830 m/z 834
m/z 844
m/z 846 m/z 848
m/z 868
m/z 870 m/z 872
m/z 889 100 %
0%
(diacyl-36:3) (diacyl-38:5) (diacyl-38:3) (diacyl-38:6) (diacyl-38:5) (diacyl-38:4) (diacyl-40:8) (diacyl-40:7) (diacyl-40:6)(16:1/16:1/20:4)
(b) Tandem mass spectrum
(c) Merged ion images
DAG(18:1/18:1)
basal contraction
m/z 683
100 % [M-C18:1-H2O+H]+
[M-C18:1-H2O+H]+ 345.2 24.0
60
22.0
Intensity, counts
50 45 40 35 30 25 20 15
339.2
10
200
300
400 500 m/z, Da
600
16.0 14.0 12.0
[M+K]+
10.0
683.5
339.3
8.0
400 um
299.2
6.0
700
800
900
100
200
300
m/z 844
400 500 m/z, Da
600
700
800
900
m/z 848 [M-183+K]+ 661.4
18.0
400 um
0%
merged basal contraction
2.0
659.4
100
18.0
4.0
[M+K]+
5
20.0
162.9
28
m/z 844
Intensity, counts
55
0
m/z 683
321.2
65
26 16.0
24
12.0 10.0
[M+K]+ 844.5
8.0
[M-59+K]+ 785.4
6.0
184.0
4.0
665.5
20 18
[M+K]+ 848.6
184.0
16 14 12 10 8 6
[M-C20:4+K]+
4
540.2
2.0
[M-183+K]+
22
m/z 848
162.9
Intensity, counts
Intensity, counts
14.0
502.8
2 0.0
100
200
300
400
500
600
700
800
900
100
200
300
400 500 m/z, Da
600
700
800
900
Fig. 3 (a) Optical image of muscle before IMS analysis and ion images of each m/z value in positive-ion mode. (b) Tandem mass spectrum of standard diolein, m/z 683, 844, and 848 on tissue sections
is shown. (c) Ion images of m/z 683 in red, m/z 844 in green, and m/z 848 in blue, and their merged ion images are shown. Scale bar is 400 μm
condition. Conversely, those of the ion at m/z 844, assigned as PC (diacyl-18:2/20:4), and the ion at m/z 848, assigned as
PC (diacyl-38:4), were higher in the contraction than basal condition. However, the distributions of the latter two in the
1868
N. Goto-Inoue et al.
muscle tissue were completely different; while the accumulation of ion signals of m/z 844 was ubiquitous throughout muscle myofibers, m/z 848 was distributed specifically inside each myofiber. The distribution is specific to contracted tissue; other ions such as m/z 868, 870, and 872 showed the same tendency of localization inside the myofiber in contraction condition. Figure 4a demonstrates the IMS results in negative-ion mode. Only [M−H]− ions were detected in the mode, and they were FA-related ions. However, these FAs were derived from not only free FAs but also included fragmented ions from other phospholipids. The ions detected at m/z 255, 279, 281, 283, 301, 303, and 327 were confirmed as C16:0, C18:2, C18:1, C18:0, C20:5, C20:4, and C22:6, respectively, by the information from the previous study [20]. The signal intensities of saturated FA (m/z 255 and 283) and mono-unsaturated FA (MUFA; m/z 281) after contraction were higher than the basal values, suggesting that these FAs accumulate in muscle during contraction. Poly-unsaturated FAs (PUFA) were
detected in both basal and contraction conditions at almost the same signal intensities. The ion image of m/z 885, assigned to the PI, also showed no significant difference in signal intensity or distribution between the two conditions. All detected molecules are listed in Table 1. To evaluate change in the energy charge situation in skeletal muscles by electrical pulse-evoked contraction, we examined adenosine phosphate-related ions. The energy charge was calculated by a formula proposed by Atkinson et al. [18]. Generally, the normal energy charge is around 0.8, and a lower energy charge represents up-regulations of metabolism. As shown in Fig. 4b, ATP, ADP, and AMPrelated ions were detected. The energy charge was calculated in each pixel and showed a lower tendency in contracted than in basal muscle. The histogram shows average energy charges of 0.849 in basal muscle and 0.785 in contracted muscle, respectively. These data indicated that ATP was generated at a high degree in contraction samples compared with basal samples.
(a) Negative ion mode Imaging m/z 255
m/z 283
m/z 281
m/z 279
m/z 301
m/z 303
m/z 327
m/z 885
basal
100 %
contraction
400 um
400 um
0%
C16:0 C18:0 saturated FA
C18:1 MUFA
C18:2
C20:5 C20:4 PUFA
C22:6
PI
(b) Energy charge imaging basal
contraction
ATP
100 %
400 um
0% 100 %
1 0.8 0.6 0.4 0.2 0
0.25
ADP
400 um
contraction
Energy charge
basal
AMP
0% 100 %
frequency
0.2
basal contraction
0.15
0.1
0.05
0 0.58
0%
Fig. 4 a Optical image of muscle before IMS analysis and ion images of each m/z value in negative-ion mode. Scale bar is 400 μm. b In negative-ion mode, we visualized ATP, ADP, and AMP distribution in
0.68
0.78
0.88
0.98
energy charge
the tissue. Energy charge was calculated and visualized with SIMenergy software (Shimadzu Co., Japan). Scale bar is 400 μm. Histogram was calculated examining each pixel, and the result is shown
Dynamic change in contraction-induced lipid composition Table 1 List of detected molecules
Mode Positive-ion mode
Negative-ion mode
a
Identified by tandem mass spectrometric analysis
1869
Observed m/z
Adduct
Annotation
Change after contraction
661
[M+K]+
683 685 741 794 796 798 806 820 822 830 834 844 846
+
[M+K] [M+K]+ [M+K]+ [M+K]+ [M+K]+ [M+K]+ [M+H]+ [M+K]+ [M+K]+ [M+Na]+ [M+Na]+ [M+K]+ [M+K]+
DAG(18:1/18:0)a DAG(18:1/20:1)a DAG(18:1/20:0)a SM (d18:1/16:0)a PC (diacyl-34:3)a PC (diacyl-34:2)a PC (diacyl-34:1)a PC (diacyl-38:6)a PC (diacyl-36:4)a PC (diacyl-36:3)a PC (diacyl-38:5)a PC (diacyl-38:3)a PC (diacyl-18:2/20:4)a PC (diacyl-38:5)a
↓ ↓ ↓ – ↑ – – – – – ↑ ↑ ↑ ↑
848
[M+K]+
868 870 872 889 255 279 281 283 301 303 327 346 426 506 885
+
PC (diacyl-38:4)a PC (diacyl-40:8)a PC (diacyl-40:7)a PC (diacyl-40:6)a TAG (16:1/16:1/20:4)a FA 16:0 FA 18:2 FA 18:1 FA 18:0 FA 20:5 FA 20:4 FA 22:6 AMP ADP ATP PI (diacyl-38:4)a
↑ ↑ ↑ ↑ ↓ ↑ ↑ ↑ ↑ – – – ↑ ↑ ↓ –
Discussion In the present study, we used MALDI-IMS to detect and identify a wide range of lipids changes associated with contraction in skeletal muscles. Previous reports utilizing LC-MS analyses have succeeded in showing lipid changes during exercise at the molecular-species level [5]. However, these data lacked lipid distribution information in muscles. Some reports using TOF-SIMS revealed the lipid dynamic changes in human dystrophic muscle [13] or ob/ob mice [14], though TOF-SIMS could only speculate about these molecules based on their mass. Furthermore, since the mass range of TOF-SIMS is limited to m/z 800 or below, PUFA-containing phospholipids and TAG, which have masses greater than m/z 800, were hardly detected. It is possible that MALDI-IMS will be one of the leading methods to overcome these deficits in the current situation.
[M+K] [M+K]+ [M+K]+ [M+K]+ [M−H]− [M−H]− [M−H]− [M−H]− [M−H]− [M−H]− [M−H]− [M−H]− [M−H]− [M−H]− [M−H]−
In positive-ion mode, m/z 661, 683, and 683 were detected and assigned as DAG by tandem mass spectrometric analyses. This is the first report to reveal DAG distribution in muscles with structural characterization. Since these signals were not found in the report of whole-body MALDIIMS of mouse embryos [2], they are considered to be skeletal muscle-specific DAG species. Additionally, all of these signal intensities were lower in the contraction condition than the basal condition, suggesting that muscle contraction decreased all these DAGs in skeletal muscles. TLC analyses did not show any significant difference of the amount of DAG. We thought that DAG bands contained other kinds of lipids, such as cholesterol, because the RF value of DAG and cholesterol is quite similar. Previous reports have pointed out that the accumulated DAG in skeletal muscle cells may contribute to insulin resistance by the activation of protein kinase C family [21, 22]. Therefore, the decreased muscle DAGs by contraction implies
1870
contribution to improving insulin sensitivity. TAGs, as well as DAGs, had a tendency to be lowered by muscle contraction. Due to the characteristic MALDI ionization pattern, it is difficult to distinguish whether these ions derived from the native DAG molecules and/or the fragmentation of TAG molecules. However, this finding is convincing with the previous report using LC-ESI-MS [5]. Moreover we confirmed that other tissues, for example liver and fatty cell tissue, which contains much amount of TAGs, are also easy to detect TAG. However, fragmented DAG from TAG is hardly detected. Therefore, we realized that our highly detected DAG molecules are really DAG itself. FA oxidation is accelerated in skeletal muscle during contraction; therefore, TAGs and DAGs in muscle cells are supposed to be used to produce free FAs by lipase. However, we have to take into account that some part of DAG might include TAG-derived ions because TAGs are easily fragmented to DAGs by laser irradiation. On the contrary, some PUFA-containing PC ions were significantly higher in contracted muscles compared to resting muscles. Interestingly, localizations of some PUFAcontaining PCs in the muscle tissue were drastically changed by contraction. We speculated that this relocation was derived from the remodeling process of FAs with lysophosphatidylcholine acyltransferase (LPCAT). Saturated FAs are easily oxidized in cells; in other words, PUFAs will remain in cells. We demonstrated that in negative-ion mode, the signal intensities of saturated FAs and MUFAs in the skeletal muscle were significantly higher in contraction while PUFA levels remain unchanged. Plasma membranes remodel the FAs of phospholipids by LPCAT. If there are many PUFA series in cells, LPCAT might use these FAs without selectivity. A previous report showed that not only PUFA but also saturated FA signal intensities were higher in the skeletal muscle of obese ob/ob mice than in that of lean wild-type mice [14]. It is reasonable that these changes contribute to decreased insulin sensitivity because the abnormal composition of FAs of phospholipids in skeletal muscle leads to changes in the membrane fluidity [23], which is one of the causes of abnormal insulin actions [24]. Matsuzaka et al. showed that insulin resistance is improved by the modification of hepatic FA composition, even in cases of persistent obesity [25]. They generated knockout mice deficient for Elovl6, which catalyzes the conversion of palmitate to stearate. The mice showed marked protection from high-fat dietinduced hyperinsulinemia, hyperglycemia, and hyperleptinemia. These data demonstrate that FA composition, rather than FA amount, is a dominant for physiological abnormalities. Muscle contraction (or physical exercise) is a potent stimulus that enhances insulin sensitivity, though the mechanism is still unknown [26]. The changes in muscle lipid composition by contraction observed in this study underlie
N. Goto-Inoue et al.
this mechanism. Further progress in this line of inquiry could be made using our method. In the current trial, we could not verify that these FAs were derived from free FA and/or fragmentation of other lipids. For further analyses, it is necessary to verify the FA accumulation from total lipids by the use of gas-chromatography. Energy charge was calculated from pixels of ATP, ADP, and AMP and showed as images (Fig. 4b). Energy charge is an index of metabolic activity, and low signals of the index indicate up-regulation of the metabolism [18]. Therefore, local energy metabolism in cells or tissues can be evaluated by the imaging of the index. In the current study, energy charges were higher in basal than contraction condition, indicating that the muscle contraction activated energy metabolism in the tissues. This is a good tool to investigate local metabolic activity with contraction in skeletal muscle. Our results show that changes in the lipid amount, lipid composition, and energy metabolic activity can be evaluated in each local spot of cells and tissues at the same time using MALDI-IMS and tandem mass spectrometric analyses.
Conclusion We showed dynamic change in lipid composition of murine skeletal muscle during contraction using the MALDI-IMS technique. Although oil red O staining and TLC did not show significant change in lipid amounts from basal to contraction conditions, we proved that the lipid composition, i.e., the relative amounts of FA, PC, DAG, and TAG, was actually changed by contraction. Moreover, some molecules show characteristic distributions after contraction that could be related with contraction-induced lipid dynamics. In the near future, we will follow these molecules to confirm lipid-induced insulin sensitivity or a glucose-uptake mechanism in skeletal muscles. Acknowledgments We are grateful to Yukiko Sugiyama for her assistance. This work was supported by SENTAN, JST to T. H., a grant-in-aid for Young Scientists B (21780110) to N. G-I, Machineries of bioactive lipids in homeostasis and disease from Japan Society for the Promotion of Science to M. S., and a grant-in-aid by the Funding Program for WorldLeading Innovative R&D on Science and Technology by the Council for Science and Technology Policy to N.L.F.
References 1. Hannun YA, Linardic CM (1993) Biochim Biophys Acta 1154:223–236 2. Hayasaka T, Goto-Inoue N, Zaima N, Kimura Y, Setou M (2009) Lipids 44:837–848 3. Pan DA, Lillioja S, Kriketos AD, Milner MR, Baur LA, Bogardus C, Jenkins AB, Storlien LH (1997) Diabetes 46:983–988 4. Tsalouhidou S, Petridou A, Mougios V (2009) Exp Physiol 94:362–370
Dynamic change in contraction-induced lipid composition 5. Hu C, Hoene M, Zhao X, Haring HU, Schleicher E, Lehmann R, Han X, Xu G, Weigert C (2010) PLoS One 5:e13318 6. Blachnio-Zabielska A, Zabielski P, Baranowski M, Gorski J (2011) Lipids 46:229–238 7. Taguchi R, Houjou T, Nakanishi H, Yamazaki T, Ishida M, Imagawa M, Shimizu T (2005) J Chromatogr B Anal Technol Biomed Life Sci 823:26–36 8. van Goor H, Gerrits PO, Grond J (1986) Histochemistry 85:251–253 9. Wiseman JM, Ifa DR, Venter A, Cooks RG (2008) Nat Protoc 3:517–524 10. Caprioli RM, Farmer TB, Gile J (1997) Anal Chem 69:4751–4760 11. Chaurand P, Caprioli RM (2002) Electrophoresis 23:3125–3135 12. Goto-Inoue N, Hayasaka T, Zaima N, Setou M (2011) Biochim Biophys Acta 1811:961–969 13. Tahallah N, Brunelle A, De La Porte S, Laprevote O (2008) J Lipid Res 49:438–454 14. Magnusson YK, Friberg P, Sjovall P, Malm J, Chen Y (2008) Obesity (Silver Spring) 16:2745–2753 15. Yang HJ, Ishizaki I, Sanada N, Zaima N, Sugiura Y, Yao I, Ikegami K, Setou M (2010) Med Mol Morphol 43:158–164 16. Fujii N, Hirshman MF, Kane EM, Ho RC, Peter LE, Seifert MM, Goodyear LJ (2005) J Biol Chem 280:39033–39041
1871 17. Goto-Inoue N, Hayasaka T, Zaima N, Setou M (2009) Glycobiology 19:950–957 18. Atkinson DE (1968) Biochemistry 7:4030–4034 19. Bowden JA, Albert CJ, Barnaby OS, Ford DA (2011) Anal Biochem 417:202–210 20. Hayasaka T, Goto-Inoue N, Zaima N, Shrivas K, Kashiwagi Y, Yamamoto M, Nakamoto M, Setou M (2010) J Am Soc Mass Spectrom 21:1446–1454 21. Schmitz-Peiffer C (2002) Ann N Y Acad Sci 967:146–157 22. Li Y, Soos TJ, Li X, Wu J, Degennaro M, Sun X, Littman DR, Birnbaum MJ, Polakiewicz RD (2004) J Biol Chem 279:45304– 45307 23. Tepperman HM, Tepperman J (1985) Proc Nutr Soc 44:211–220 24. Borkman M, Storlien LH, Pan DA, Jenkins AB, Chisholm DJ, Campbell LV (1993) N Engl J Med 328:238–244 25. Matsuzaka T, Shimano H, Yahagi N, Kato T, Atsumi A, Yamamoto T, Inoue N, Ishikawa M, Okada S, Ishigaki N, Iwasaki H, Iwasaki Y, Karasawa T, Kumadaki S, Matsui T, Sekiya M, Ohashi K, Hasty AH, Nakagawa Y, Takahashi A, Suzuki H, Yatoh S, Sone H, Toyoshima H, Osuga J, Yamada N (2007) Nat Med 13:1193– 1202 26. Goodyear LJ, Kahn BB (1998) Annu Rev Med 49:235–261