Comprehensive Analysis of Short-, Medium-, and Long-Chain Acyl

Subscriber access provided by Kent State University Libraries

Article

Comprehensive analysis of short-, medium- and long-chain acyl-coenzyme As by on-line two dimensional liquid chromatography-mass spectrometry Shuangyuan Wang, Zhichao Wang, Lina Zhou, Xianzhe Shi, and Guowang Xu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b03659 • Publication Date (Web): 03 Nov 2017 Downloaded from http://pubs.acs.org on November 4, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Comprehensive analysis of short-, medium- and long-chain acyl-coenzyme As by on-line two dimensional liquid chromatography-mass spectrometry

Shuangyuan Wang,1,2# Zhichao Wang,1,2# Lina Zhou,1 Xianzhe Shi,1* Guowang Xu1*

1

CAS Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China

2

#

University of Chinese Academy of Sciences, Beijing 100049, China

The two authors contributed equally to this work.

* Correspondence: Prof. Dr. Guowang Xu, e-mail: [email protected], Tel: 0086-411-84379530, Fax: 0086411-84379559 Dr. Xianzhe Shi, e-mail: [email protected], Tel: 0086-411-84379757, Fax: 0086411-84379559

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

Abstract Acyl-coenzyme As (CoAs) are pivotal metabolic intermediates in numerous biological processes. However, comprehensive analysis of acyl-CoAs is still challengeable as the greatly varied properties of acyl-CoAs with different carbon chains. Here, we designed a two-dimensional liquid chromatography coupled with high resolution mass spectrometry (2D LC-HRMS) method to cover all short-, medium- and long-chain acyl-CoAs within one analytical run. Complex acyl-CoAs were separated into two fractions according to their acyl chains by the first dimensional pre-fractionation. Then, two fractions containing short-chain acyl-CoAs or medium- and long-chain acyl-CoAs were further well separated by the two parallel columns in the second dimension. Nineteen representative standards were chosen to optimize the analytical conditions of 2D LCHRMS method. Resolution and sensitivity were demonstrated to be improved greatly and low abundant acyl-CoAs and acyl-CoA isomers could be detected and distinguished. By using the 2D LC-HRMS method, 90 acyl-CoAs (including 21 acyl-dephospho-CoAs) were identified from liver extracts, which indicated that our method was the most powerful approach to obtain comprehensive profiling of acyl-CoAs so far. The method was further employed in the metabolomics study of malignant glioma cells with isocitrate dehydrogenase 1 (IDH1) mutation to explore their metabolic differences. 46 acyl-CoAs (including 2 acyl-dephospho-CoAs) were detected and 12 of them were dysregulated in glioma cells with the IDH1 mutation. These results demonstrated the practicability and the superiority of the established method. Therefore, the 2D LC-HRMS method provides a robust and reproducible approach to the comprehensive analysis of all acyl-CoAs in tissues, cells, and other biological samples. 2


Page 3 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Acyl-coenzyme As (CoAs) belong to the class of fatty acyl lipids, integrating fatty acid groups and free CoA through high-energy thioester bond.1 They are involved in numerous

metabolic

pathways

and

biological

processes,

for

example,

lipid

synthesis/remodeling, ketone body synthesis, fatty acid oxidation, xenobiotic metabolism, etc.2-5 Derived from either glucose or fatty acid catabolism, acetyl-CoA enters the tricarboxylic acid cycle, fuels energy production and relates the following biological activities. Like acetyl-CoA, acyl-CoAs are acyl supplier for epigenetic modification and involved in complex biological processes.6,7 The metabolic network and regulation of acyl-CoAs are sophisticated, and the dysregulation of acyl-CoA metabolism plays important roles during the development of many diseases including cancer, diabetes etc.8,9 Considering the key functions of acyl-CoAs in physiological and pathological processes, qualitative and quantitative analyses of acyl-CoAs are very valuable. Because of the existence of adenosine group in the acyl-CoA structures, early methods were mainly focused on liquid chromatography (LC) coupled with ultraviolet (UV) detector.10,11 However, constrained by the UV detector, these methods are deficient in sensitivity and resolution. Gas chromatography-mass spectrometry (GC-MS) can also be used to analyze acyl-CoAs, but complex derivatization procedures are essential and time consuming.12 Recently, reversed-phase liquid chromatography (RPLC) coupled with MS becomes the most extensively utilized method for acyl-CoAs analysis owing to the high performance of RPLC and high sensitivity of MS.13-15 As their properties varied greatly, 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

different LC conditions are required for acyl-CoAs with different chain lengths to obtain satisfactory separation. Due to the strong polarity, short- and medium-chain acyl-CoAs are commonly analyzed under acidic mobile phases to enhance their retention on RPLC,16,17 but seriously tailed peaks are often observed due to too strong retention of long-chain acyl-CoAs.18 For the analysis of long-chain acyl-CoAs, alkaline mobile phases are often used. Under this condition, the deprotonation of acyl-CoAs occurred, leading to weak retention on RPLC, and peak tailing is avoided.19-21 With these traditional methods, only a limited range of acyl-CoA species was covered in one analysis, usually either short-, medium- or long-chain acyl-CoAs. Therefore, two methods were needed for the analysis of whole acyl-CoA profiling. In recent years, some efforts were made to cover more acyl-CoAs. For example, Zimmermann et al. developed a nontargeted profiling LC-MS method to cover the acylCoAs ranged from free CoA to oleoyl-CoA.22 Ion pairing reagent was utilized to improve the separation of acyl-CoAs, but some short- and medium-chain acyl-CoAs like acetylCoA, malonyl-CoA and glutaryl-CoA were still overlapped and long-chain acyl-CoAs like palmitoyl-CoA and oleoyl-CoA presented serious peak-tailing under their conditions.22 In addition, serious contamination of MS is one of the main drawbacks of these methods with ion pairing reagents.22-24 To improve peak tailing, Li et al. employed mobile phases containing a very high concentration of salt (i.e. ammonium formate at 100 mM).25 However, this approach is harmful to analytical columns and can severely reduce the sensitivity of MS. Magnes et al. established an online SPE (solid phase extraction)-LC-MS method to analyze acyl-CoAs.21

Although the sensitivity was

increased, the separation was relatively poor, especially for short-chain acyl-CoAs, and 4


Page 5 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


only 9 acyl-CoAs were identified in biosamples.21 Recently, Liu et al. optimized methanol/water mobile phases at weak acidic conditions to analyze short-chain to medium-chain acyl-CoAs, and chose acetonitrile/water mobile phases at basic conditions to analyze medium-chain to long-chain acyl-CoAs.26 With the two analyses, over 30 acylCoAs with short-, medium- and long-chains were determined in liver and cell samples.26 Until now, it is extremely difficult for a single LC-MS method to simultaneously cover short-, medium- and long-chain acyl-CoAs with acceptable separation. In the present study, we designed an on-line 2D LC coupled with high resolution mass spectrum (HRMS) method to cover short-chain, medium-chain and long-chain acyl-CoAs simultaneously within a single analytical run. The first dimensional pre-separation and the second dimensional parallel separation were properly combined by three valves. Thus, all acyl-CoAs with different fatty acid chains could be separated under optimized conditions. Resolution and sensitivity were improved greatly, which enabled detection of the low abundant acyl-CoAs and separation of the acyl-CoA isomers. Finally, the method was applied to analyze acyl-CoAs in liver extracts and explore the metabolic differences of malignant glioma cells with/without the isocitrate dehydrogenase 1 (IDH1) mutation (R132H and R132C), which further show the usefulness of 2D LC-HRMS method.

Experimental Section Reagents HPLC grade solvents including acetonitrile and methanol were purchased from Merck (Darmstadt, Germany). Chloroform and ammonium acetate were purchased from SigmaAldrich (St. Louis, USA). Water was prepared using Milli-Q system (Millipore, Bedford, 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

USA). Ammonium hydroxide (25% in water) was purchased from Acros Organics (New Jersey, USA). Acyl-CoA standards including CoA, Malonyl-CoA, Succinyl-CoA, Glutaryl-CoA, Acetoacetyl-CoA,

Acetyl-CoA,

Propionyl-CoA,

2-Butenoyl-CoA,

Butyryl-CoA,

Isobutyryl-CoA, Isovaleryl-CoA, Phenylacetyl-CoA, Decanoyl-CoA, Lauroyl-CoA, Myristoyl-CoA, Palmitoleoyl-CoA, Arachidonoyl-CoA, Oleoyl-CoA and Stearoyl-CoA were from Sigma-Aldrich.

Cell Culture U251 cells were derived from ATCC. IDH1 plasmids (Vector, R132H, and R132C) were bought from Origen (USA), and were transfected into U251 with lipofectamine 2000 (Invitrogen, USA). All the cell lines were cultured in DMEM (Invitrogen, USA) supplemented with 10% CBS (Sigma, USA). Six replications were parallelly processed. For metabolomics analysis, the medium was renewed when U251 cells were grown to 80% confluence. 5% mannitol was used to wash the medium when the cells were harvested, and subsequently quenched with liquid nitrogen.

Sample Preparation Acyl-CoAs in mouse liver tissues were obtained by solvent extraction. Briefly, 20 mg wet liver tissue sample was powered with liquid nitrogen in advance and transferred to a 2 mL Eppendorf tube. Five hundred microliter of methanol was added and mixed with the powered sample. After vortexed for 30 s, 500 µL chloroform was added to the tube and the mixture was vortexed for another 30 s. Then, 200 µL water was added. After 6


Page 7 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


vortexed for 30 s and standing for 10 min, the resulting solution was centrifuged for 10 min at 15,000 rpm, 4 oC. Four hundred and fifty microliters of upper layer solution was transferred and dried at 4 oC and vacuum. The whole process of sample preparation was performed on ice. To obtain acyl-CoAs in glioma cells, cells were scraped down with a scraper (Corning, USA) after adding 1 mL methanol to each plate, and transferred to a 5 mL Eppendorf tube. After vortexed for 30 s, 1 mL chloroform was added to the tube, and followed with 30 s vortex. Then, the addition of 400 µL water was performed and vortexed for 60 s. Tubes were centrifuged with 15,000 g for 15 min after placed on ice for 10 min. Equal aliquots of upper layer solvent from all cell samples were pooled and used as quality control (QC) sample. 600 µL upper layer extracts or QC samples were transferred to an Eppendorf tube and freeze-dried. The whole processes of sample preparation were performed on ice. Before 2D LC-MS analysis, the dried residues were reconstituted with 50 µL methanol/water (1:5, v/v).

2DLC-MS Analysis The separation of extracted acyl-CoAs was performed based on an online 2D LC-MS system. Briefly, the schematic interface of the system is shown in Figure 1. A Shimadzu Prominence system (Shimadzu, Kyoto, Japan) was used in the first dimensional prefractionation. An Acquity BEH C18 pre-column (2.1 × 5 mm, 1.7 µm, Waters, Milford, USA) was used in the first dimension to separate short-chain acyl-CoAs from mediumand long-chain acyl-CoAs. An Acquity HSS T3 column (2.1 × 50 mm, 1.7 µm, Waters) 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

was optimized to analyze short-chain acyl-CoAs. A Shimadzu Nexera system (Shimadzu) was used to separate medium- and long-chain acyl-CoAs. Mobile phases A1 and B1 were water and acetonitrile, respectively. Both A1 and B1 contained 10 mM ammonium acetate and 0.05% ammonium hydroxide. An Acquity BEH C18 column (2.1 × 100 mm, 1.7 µm, Waters) was chosen to analyze medium- and long-chain acyl-CoAs. Mobile phases A2 and B2 were water and acetonitrile, both containing 0.5% ammonium hydroxide. The flow rates of both pump 1 and pump 2 were 0.3 mL/min. An LC-20AD pump was used as pump 3 to replace residual organic solvents in the pre-column. The mobile phase of this flow was water. Specific gradient conditions are listed in Table 1. The temperatures of the autosampler and column oven were set at 4 oC and 30 oC, respectively. The injection volume was 5 µL.

Figure 1. Schematic interface of the 2D LC-MS system. (A) is the flow scheme of position 1 (solid line position) and position 2 (dotted line position). (B) is the flow scheme of position 3 (solid line position) and position 4 (dotted line position).

8


Page 9 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Specific conditions of the 2D LC-MS method. Time Position

Pump 1

Pump 2

Pump 3

(min) 0-1.0 min, 3% B1 1

0-5.5 1.0-13.0 min, 3-15% B1 13.0-14.1, 15-100% B1 0-15 min, 10% B2 14.1-14.7 min, 100% B1

2

5.5-15 14.7-14.71 min, 100-3% B1 14.71-15 min, 3% B1

0-29 min, 15-15.3 min, 10% B2

F = 0 mL/min

15.3-27 min, 10-40% B2 3

27-28 min, 40-100% B2

15-29.3 15-30 min, 3% B1

28-29 min, 100% B2 29-29.01 min, 100-10% B2 4

29.3-30

29-30 min, F=0.5 mL/min

29.01-30 min, 10% B2

* Buffers: B1, acetonitrile with 0.05% ammonium hydroxide. B2, acetonitrile with 0.5% ammonium hydroxide.

A Q-Exactive HF orbitrap mass spectrometer (ThermoFisher Scientific, Rockford, USA) equipped with heated electrospray ionization (HESI) was used to detect acyl-CoAs. Data were acquired in ESI+ mode under the following HESI source parameters: the sheath gas flow rate and the aux gas flow rate were set at 45 and 10 arbitrary unit, respectively; the spray voltage was 3.5 kV; the capillary temperature and the aux gas heater temperature were 300 oC and 350 oC, respectively; the S-lens RF level was set at 50. Full scan experiment was performed from 600 to 1500 m/z and the resolution was chosen at 60,000. For MS2 experiment, we chose data-dependent MS2 (dd-MS2) mode 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

and the top 10 ions were supervised. The resolution was chosen at 15,000 and the isolation window was set at 2.0 m/z. MS2 spectra were acquired with the stepped normalized collision energy (NCE) at 15%, 30% and 45%.

Data Processing Peak areas of all the identified acyl-CoAs were obtained using Tracefinder (version 3.2, ThermoFisher Scientific) and corrected according to total area. Statistical analysis was performed based on the new peak table. Before multivariate statistical analysis, the data were performed with unit variance (UV) scaling. Partial least squares-discriminate analysis (PLS-DA) was proceeded using SIMCA (version 13.0, Umetrics, Umea, Sweden). Student t-test was performed by Statistical Package for the Social Sciences (SPSS, version 22.0, SPSS Inc., Chicago, USA). Acyl-CoAs with p < 0.05 were chosen as differential metabolites.

Results and Discussion Design of 2D LC-MS system This study aimed at designing a 2D LC-MS system to realize simultaneous coverage of short-, medium- and long-chain acyl-CoAs with satisfactory separation. As shown in Figure 1, the 2D LC system was ingeniously constructed by using a heart-cutting strategy with three switching valves, which effectively connected pre-separation in the first dimension with parallel separation in the second dimension. By controlling the three valves, the analytical operation was performed in the following 4 steps as listed in Table 1. 10


Page 11 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


At the beginning, the three valves were switched to position 1 (Figure 1A, solid line position). At this position, pre-separation occurred in the first dimensional pre-column and acyl-CoAs were divided into two fractions according to their hydrophobicity. A medium-chain acyl-CoAs (C6:0-CoA) was chosen as the boundary of the two fractions, and the cut time was optimized at 5.5 min. Short-chain acyl-CoAs had stronger hydrophilicity and retained weaker than C6:0-CoA. Hence, they were firstly eluted out from the pre-column and included in the fraction 1. As the column 1 was directly coupled with the pre-column at this position, the fraction 1 was transferred straightly into the column 1 for further analysis. After C6:0-CoA was completely eluted out at 5.5 min, the valve A was switched to position 2 (Figure 1A, dotted line position). At this position, short-chain acyl-CoAs were further finely separated in column 1, meanwhile fraction 2 including medium- and longchain acyl-CoAs was still retained in the pre-column. Next, the three valves were switched to position 3 (Figure 1B, solid line position). At this position, the first dimensional pre-column was directly coupled to column 2. Then, the retained medium- and long-chain acyl-CoAs were transferred into column 2 for further detailed analysis by pump 2. When this procedure was finished, the pre-column had been filled with strong basic mobile phases and it was unsuitable for next injection. Thus, solvent replacement was necessary. After the three valves were switched to position 4 (Figure 1B, dotted line position), this operation was implemented by pump 3 and water was used to replace the residual basic solvents. Through the above described 4 steps, short-, medium- and long-chain acyl-CoAs were simultaneously analyzed by 2D LC-MS system. 11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

Optimization and validation of 2D LC-MS method In order to obtain satisfactory peak shapes and resolution, we further optimize the detailed separation conditions, and 19 representative standards including short-, mediumand long-chain acyl-CoAs were analyzed by the 2D LC-MS method. As described above, the pH of mobile phases had a great influence on the retention of acyl-CoAs in a reversed-phase column. In the present study, the optimal separation of both short-chain and medium- and long-chain acyl-CoAs could be realized under different pH conditions owing to pre-fractionation in the first dimension and two parallel separations in the second dimension. Specifically, short-chain acyl-CoAs were separated under weakly basic conditions, and medium- and long-chain acyl-CoAs were analyzed at strongly basic conditions. Here, the separation of short-chain acyl-CoAs was not performed under the commonly used acidic mobile phases in the constructed 2D LC system because they could be seriously residual in the first dimensional pre-column. In addition, the T3 column was chosen to separate the short-chain acyl-CoAs as it presented better retention and resolution compared with the C18 column. After the gradient conditions were optimized, the final 2D LC conditions were listed in Table 1. Under these conditions, almost all acyl-CoA standards were baseline separated with reasonable retention times and good peak shapes, as shown in Figure 2 (A). Figure 2 (B) further displays the good resolution of the short-chain acyl-CoAs including malonyl-CoA, succinyl-CoA, free CoA, glutaryl-CoA, acetoacetyl-CoA, and acetyl-CoA. In addition, the established 2D LC-MS method presented excellent ability in distinguishing acyl-CoA isomers. For example, isobutyryl-CoA and butyryl-CoA were well separated (Figure 2 (C)). 12


Page 13 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. (A) EIC chromatogram of the 19 standards. (B) Retention of the 6 weakly retained acylCoAs. (C) Separation of acyl-CoA isomers. Standards: 1, Malonyl-CoA; 2, Succinyl-CoA; 3, free CoA; 4, Glutaryl-CoA; 5, Acetoacetyl-CoA; 6, Acetyl-CoA; 7, Propionyl-CoA; 8, 2-Butenoyl-CoA; 9, Isobutyryl-CoA; 10, Butyryl-CoA; 11, Isovaleryl-CoA; 12, Phenylacetyl-CoA; 13, Decanoyl-CoA; 14, Lauroyl-CoA; 15, Myristoyl-CoA; 16, Palmitoleoyl-CoA; 17, Arachidonoyl-CoA; 18, Oleoyl-CoA; 19, Stearoyl-CoA.

13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

Table 2. Calibration curves and LODs of the 19 acyl-CoA standards.

Acyl-CoAs

Calibration curve

Linearity

LLOD

range (ng/mL)

(ng/mL)

r2

Malonyl-CoA

y = 7458.3x - 576478

0.9974

5-5000

2

Succinyl-CoA

y = 5272x - 898754

0.9935

5-10000

1

CoA

y = 8209.3x - 2E+06

0.991

10-10000

0.2

Glutaryl-CoA

y = 8206.6x - 1E+06

0.9959

5-15000

0.5

Acetoacetyl-CoA

y = 6376.6x - 459035

0.9962

5-5000

2

Acetyl-CoA

y = 14153x - 468985

0.9989

1-5000

0.2

Propionyl-CoA

y = 16770x - 2E+06

0.9995

2-20000

0.5

2-Butenoyl-CoA

y = 21974x - 2E+06

0.9994

0.5-20000

0.2

Isobutyryl-CoA

y = 20996x - 578030

0.9999

0.5-20000

0.2

Butyryl-CoA

y = 20473x - 904210

0.9999

1-20000

0.2

Isovaleryl-CoA

y = 35441x - 976540

0.9999

0.5-20000

0.5

Phenylacetyl-CoA

y = 11718x - 959794

0.9997

1-20000

1

Decanoyl-CoA

y = 33585x + 2E+06

0.9992

0.5-15000

0.2

Lauroyl-CoA

y = 66783x + 841252

0.9994

0.2-5000

0.2

Myristoyl-CoA

y = 52502x - 463711

0.9998

0.2-5000

0.2

Palmitoleoyl-CoA

y = 51961x + 703758

0.9981

0.2-5000

0.2

Arachidonoyl-CoA

y = 35441x + 709731

0.9961

0.5-5000

0.2

Oleoyl-CoA

y = 7069x - 201196

0.9982

2-2500

2

Stearoyl-CoA

y = 11034x - 369044

0.9949

2-2500

2

*Five µL standards were injected to establish calibration curves and evaluate LLODs.

14


Page 15 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Subsequently, the linearity and sensitivity of the 2D LC-MS method were further investigated by using the 19 standards. As listed in Table 2, the intensity of the 19 acylCoAs was linearly increased with the concentration at 3 to 5 orders of magnitude, their calibration curves are shown in Figure S1 (Supporting Information). Because of the extremely low noise level for high resolution mass spectrometer, the sensitivity of the 2D LC-MS method was evaluated using lower limit of detections (LLODs), which was defined as the lowest concentration that can be detected in the used conditions. As a result, LLODs of almost all acyl-CoA standards were at 0.2 to 2 ng/mL (when 5 µL standards were injected), indicating that the 2D LC-MS method possesses high sensitivity and is capable of detecting low abundant acyl-CoAs. Then, the established 2D LC-MS method was ready for the simultaneous analysis of short-chain and medium- and longchain acyl-CoAs in biological samples.

Metabolic profiling of acyl-CoAs in liver tissue The liver is one of the most important organs for metabolism, and many metabolic processes such as fatty acid metabolism and lipid metabolism are carried out in liver. The pool of acyl-CoA intermediates plays highly active roles during these processes. Thus, we first employed the 2D LC-MS method to profile all acyl-CoAs in the mouse liver tissues. Base peak chromatogram (BPC) of acyl-CoAs in liver tissue is shown in Figure S2 (Supporting Information). The broad coverage of the 2D LC method is beneficial for achieving comprehensive information of acyl-CoA profiling. In order to identify these

15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

acyl-CoAs definitely, the cleavage rules of acyl-CoAs were explored in advance by studying MS fragmentation patterns of acyl-CoAs. Taking acetyl-CoA as an example, Figure 3 (A) shows the molecular structure and MS2 spectra of acetyl-CoA in ESI+ mode. The predominant ion in MS2 spectra (i.e. m/z 303.14) was produced by neutral loss of adenosine diphosphate (507 Da). Moreover, the fragment ion at m/z 428.04, which originated from the CoA moiety, was also observed as another typical product ion of acyl-CoAs. Based on the detected molecular ions and cleavage rules, 69 acyl-CoAs were identified in mouse liver.

Figure 3. Molecular structures and MS2 spectra of (A) Acetyl-CoA and (B) Acetyl-dephospho-CoA.

We also observed some acyl-dephospho-CoAs, which were from dephosphorylation of acyl-CoAs. They displayed similar cleavage rules to those of acyl-CoAs. Figure 3 (B) shows the molecular structure and MS2 spectra of acetyl-dephospho-CoA. The characteristic fragment ion at m/z 303.14 was yielded by neutral loss of 427 Da rather than 507 Da for acetyl-dephospho-CoA because of a phosphate group loss of acyldephospho-CoAs at the 3’-ribose moiety. Similarly, another common fragment ion was detected at m/z 348.07 for acyl-dephospho-CoAs which originated from the dephospho16


Page 17 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


CoA moiety.25 Moreover, acyl-dephospho-CoAs retained stronger than the corresponding acyl-CoAs. Finally, 21 acyl-dephospho-CoAs were identified in mouse liver extracts. The specific information of all the 69 acyl-CoAs and 21 acyl-dephospho-CoAs is listed in Table S1 (Supporting Information). To our knowledge, this is the maximum number of acyl-CoAs detected from liver tissue compared with the previously reported studies.25,26 These identified acyl-CoAs possess not only various acyl chain length, but also different acyl chain structures with saturated, unsaturated, and hydroxylated functionalities etc., which can help understanding of metabolic pathways with CoA intermediates. The extracted ion current (EIC) chromatogram of all the acyl-CoAs and acyldephospho-CoAs is shown in Figure 4 (A). Since the polarity of acyl-CoAs mainly depends on their fatty acyl chains, the retention of acyl-CoAs was closely related to the length of fatty acyl chains. Figure 4 (B) and (C) show the retention of short-chain and medium- and long-chain acyl-CoAs with saturated fatty chains. The results elucidated that the retention of saturated acyl-CoAs was positively correlated with the carbon number of fatty acyl chain. As shown in Figure 4 (D), the retention times of acyl-CoAs with the same double bond number were linearly increased with the increase of carbon number of fatty acyl chains. Furthermore, the retention times of acyl-CoAs with the same carbon number of fatty acyl chains were also linearly increased with the decrease of double bond number (Figure 4 (E)). These observations can help to predict and identify more acyl-CoAs.

17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

Figure 4. EIC chromatograms of (A) all identified acyl-CoAs; (B) saturated short-chain acyl-CoAs; (C) saturated medium- and long-chain acyl-CoAs in liver extracts. The correlation between the retention times of acyl-CoAs and (D) carbon number or (E) double bond number of acyl chains. DB: double bond number, CN: carbon number.

Application of the 2D LC-MS method in studying metabolic differences of glioma cells Malignant glioma is the most frequently occurring primary brain tumor, and related patients usually have overall survival less than one year after prognosis.27,28 Mutant IDH1 often occurs in WHO grade II and III gliomas and glioblastomas that derived from low grade glioma.29 Based on the published literatures, many metabolic processes were 18


Page 19 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


highlighted and altered in IDH1 mutant glioma.30,31 As acyl-CoAs are crucial metabolic intermediates, the detection of acyl-CoAs may help revealing how IDH1 mutation affects metabolism. Thus, we explored metabolic differences of acyl-CoAs in glioma cells with mutant IDH1. The BPC chromatogram of acyl-CoA in glioma cells is shown in Figure S3 (Supporting Information). Based on the above-mentioned qualitative rules, 44 acyl-CoAs and 2 acyl-dephospho-CoAs were determined from glioma cells. Reproducibility was evaluated using QC samples which were inserted in the 2D LC-MS sequence every six samples. Over 95% of the acyl-CoAs were detected with RSD (relative standard deviation) values less than 20%, which accounts for 93% of the total peak areas of acylCoAs (Figure S4, Supporting Information). These data meet the requirement of metabolomics studies. Subsequently, the multivariate statistical analysis was conducted and PLS-DA model was established to visualize the overall differences between glioma cells with and without the IDH1 mutation. As shown in Figure 5 (A), the complete distinction of the vector group and two mutant IDH1 groups indicated that significant differences of acyl-CoAs were caused by IDH1 mutation. These differential acyl-CoAs (p < 0.05) were defined using the Student-t test. Figure 5 (B) presents the heatmap of all the differential acylCoAs. Many short-chain acyl-CoAs (C3:0 CoA, C4:0 CoA, iso-C4:0 CoA, hydroxy-C4:0 CoA, C5:0 CoA, iso-C5:0 CoA, C5:1 CoA), medium-chain acyl-CoA (C6:1 CoA, C12:0 CoA) and long-chain acyl-CoAs (C14:3 CoA, C22:0 CoA) significantly reduced in IDH1 mutant cells, while only C16:1 CoA increased. A lower concentration of short-chain acyl-

19



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

CoA in IDH1 mutant U251 cells may indicate the reduced activity of short-chain acylCoA synthetase (ACSS) in IDH1 mutant U251 cells.

Figure 5. (A) PLS-DA model of glioma cells with and without the IDH1 mutation. (B) Heatmap of all the differential acyl-CoAs. (C) Relative levels of FFA C12:0, C12:0-CoA and carnitine C12:0 in glioma cells with and without the IDH1 mutation. (D) Metabolic pathway of acyl-CoAs.

Except for short chain acyl-CoAs, medium chain C12:0-CoA was also decreased significantly in the IDH1 mutant U251 cells (Figure 5 (C)). Further data showed that the corresponding carnitine C12:0 also decreased in the same proportion while no significant difference was observed for fatty acid C12:0. All these metabolites are included in the mitochondria β-oxidation pathway, which is one of the main approaches for fatty acid degradation. Fatty acids need to be transferred to the corresponding acyl-CoAs before entering mitochondria. Then, acyl-CoAs are catalyzed to acyl-carnitines by Carnitine palmitoyltransferase I (CPT1) and transported to mitochondria (Figure 5 (D)). In the present study, we found that β-oxidation was significantly inhibited in IDH1 mutant 20


Page 21 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


U251 cells, which was indicated by a great decrease of total acyl-carnitines (Figure S5, Supporting Information). Thus, the decrease of C12:0-CoA and carnitine C12:0 in the same proportion may indicate that the inhibition of β-oxidation is due to the decrease of corresponding acyl-CoAs, but not the decrease of fatty acids or CPT1 inhibition. In summary, the profiling of acyl-CoAs can provide some helpful clues of cell status and significant changes of enzyme activity, which influences cancer malignancy and progression.

Conclusions In the present study, we established a 2D LC-HRMS method to obtain comprehensive information of short-, medium- and long-chain acyl-CoAs and acyl-dephospho-CoAs. The method possessed several meaningful innovation and promotion. First, the isolation and transfer of short-chain acyl-CoA fraction from medium- and long-chain acyl-CoA fraction were implemented by introducing a pre-fractionation column and a heart-cutting strategy with three switching valves. Second, short-chain acyl-CoAs and medium- and long-chain acyl-CoAs were respectively analyzed under their optimal LC conditions. High resolution and good peak shape were obtained for all acyl-CoAs. Third, not only labor and cost but also sample can be saved by using the 2D LC-MS method. Forth, the established method possesses a wide linearity range, high sensitivity, and good reproducibility. Based on the 2D LC-HRMS method, 90 and 46 acyl-CoAs were detected from mouse liver tissues and glioma cells, respectively. Compared with the previous studies, our method significantly extends the coverage of acyl-CoA ranging from free CoA to C22:0-CoA in biological samples. The 2D LC-MS method was further applied in 21



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

investigating metabolic differences of acyl-CoAs in glioma cells. Twelve acyl-CoAs were significantly changed in glioma cells with IDH1 mutation. We believe that the developed 2DLC-HRMS method will be a valuable approach for profiling the full-scale acyl-CoAs and exploring the relevant metabolic pathways.

Acknowledgements This research was supported by the foundations (81472374, 21505132 and 21575142) and key foundation (21435006) from the National Natural Science Foundation of China and the National Key Research and Development Program of China (2017YFC0906900).

Supporting information available Additional information as noted in the text. This information is available free of charge via the Internet at http://pubs.acs.org/. Figures and tables include raw data of calibration curves, BPC chromatogram of acyl-CoAs in liver extracts and glioma cells, reproducibility of identified acyl-CoAs in glioma cells, relative level of total acyl-carnitines in glioma cells with and without the IDH1mutation and all identified acyl-CoAs in liver and glioma cells.

References (1) Fahy, E.; Subramaniam, S.; Murphy, R. C.; Nishijima, M.; Raetz, C. R.; Shimizu, T.; Spener, F.; van Meer, G.; Wakelam, M. J.; Dennis, E. A. J. Lipid Res. 2009, 50 Suppl, S9-14. (2) Kind, S.; Becker, J.; Wittmann, C. Metab. Eng. 2013, 15, 184-195. (3) Li, L. O.; Klett, E. L.; Coleman, R. A. Biochim. Biophys. Acta 2010, 1801, 246-251. 22


Page 23 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(4) Ellis, J. M.; Li, L. O.; Wu, P.-C.; Koves, T. R.; Ilkayeva, O.; Stevens, R. D.; Watkins, S. M.; Muoio, D. M.; Coleman, R. A. Cell Metab. 2010, 12, 53-64. (5) Tillander, V.; Arvidsson Nordström, E.; Reilly, J.; Strozyk, M.; Van Veldhoven, P. P.; Hunt, M. C.; Alexson, S. E. H. Cellular Mol. Life Sci. 2014, 71, 933-948. (6) Peleg, S.; Feller, C.; Forne, I.; Schiller, E.; Sévin, D. C.; Schauer, T.; Regnard, C.; Straub, T.; Prestel, M.; Klima, C.; Schmitt Nogueira, M.; Becker, L.; Klopstock, T.; Sauer, U.; Becker, P. B.; Imhof, A.; Ladurner, A. G. EMBO Rep. 2016, 17, 455-469. (7) Fan, J.; Krautkramer, K. A.; Feldman, J. L.; Denu, J. M. ACS Chem. Biol. 2015, 10, 95-108. (8) Huang, Q.; Tan, Y.; Yin, P.; Ye, G.; Gao, P.; Lu, X.; Wang, H.; Xu, G. Cancer Res. 2013, 73, 4992-5002. (9) Janzer, A.; German, N. J.; Gonzalez-Herrera, K. N.; Asara, J. M.; Haigis, M. C.; Struhl, K. Proc. Natl. Acad. Sci. USA 2014, 111, 10574-10579. (10) Deutsch, J.; Rapoport, S. I.; Rosenberger, T. A. Neurochemical Res. 2002, 27, 1577-1582. (11) Shurubor, Y. I.; D'Aurelio, M.; Clark-Matott, J.; Isakova, E. P.; Deryabina, Y. I.; Beal, M. F.; Cooper, A. J. L.; Krasnikov, B. F. Molecules 2017, 22, 1388-1400. (12) Kasumov, T.; Martini, W. Z.; Reszko, A. E.; Bian, F.; Pierce, B. A.; David, F.; Roe, C. R.; Brunengraber, H. Anal. Biochem.2002, 305, 90-96. (13) Kasuya, F.; Oti, Y.; Tatsuki, T.; Igarashi, K. Anal. Biochem. 2004, 325, 196-205. (14) Yang, X.; Ma, Y.; Li, N.; Cai, H.; Bartlett, M. G. Anal. Chem. 2017, 89, 813-821. (15) Basu, S. S.; Mesaros, C.; Gelhaus, S. L.; Blair, I. A. Anal. Chem. 2011, 83, 1363-1369. (16) Frey, A. J.; Feldman, D. R.; Trefely, S.; Worth, A. J.; Basu, S. S.; Snyder, N. W. Anal. Bioanal. Chem. 2016, 408, 3651-3658. (17) Neubauer, S.; Chu, D. B.; Marx, H.; Sauer, M.; Hann, S.; Koellensperger, G. Anal. Bioanal. Chem. 2015, 407, 6681-6688. (18) Snyder, N. W.; Basu, S. S.; Zhou, Z.; Worth, A. J.; Blair, I. A. Rapid Commun. Mass Spectrom. 2014, 28, 1840-1848. 23



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

(19) Morin-Rivron, D.; Christinat, N.; Masoodi, M. Rapid Commun. Mass Spectrom. 2017, 31, 344-350. (20) Blachnio-Zabielska, A. U.; Koutsari, C.; Jensen, M. D. Rapid Commun. Mass Spectrom. 2011, 25, 2223-2230. (21) Magnes, C.; Suppan, M.; Pieber, T. R.; Moustafa, T.; Trauner, M.; Haemmerle, G.; Sinner, F. M. Anal. Chem. 2008, 80, 5736-5742. (22) Zimmermann, M.; Thormann, V.; Sauer, U.; Zamboni, N. Anal. Chem. 2013, 85, 8284-8290. (23) Purves, R. W.; Ambrose, S. J.; Clark, S. M.; Stout, J. M.; Page, J. E. J. Chromatogr. B 2015, 980, 1-7. (24) Seifar, R. M.; Ras, C.; Deshmukh, A. T.; Bekers, K. M.; Suarez-Mendez, C. A.; da Cruz, A. L.; van Gulik, W. M.; Heijnen, J. J. J. Chromatogr. A 2013, 1311, 115-120. (25) Li, Q.; Zhang, S.; Berthiaume, J. M.; Simons, B.; Zhang, G. F. J. Lipid Res. 2014, 55, 592602. (26) Liu, X.; Sadhukhan, S.; Sun, S.; Wagner, G. R.; Hirschey, M. D.; Qi, L.; Lin, H.; Locasale, J. W. Mol. Cell. Proteomics 2015, 14, 1489-1500. (27) Wen, P. Y.; Kesari, S. New Eng. J. Med. 2008, 359, 492-507. (28) Cuddapah, V. A.; Robel, S.; Watkins, S.; Sontheimer, H. Nat. Rev. Neurosci. 2014, 15, 455465. (29) Yan, H.; Parsons, D. W.; Jin, G.; McLendon, R.; Rasheed, B. A.; Yuan, W.; Kos, I.; BatinicHaberle, I.; Jones, S.; Riggins, G. J.; Friedman, H.; Friedman, A.; Reardon, D.; Herndon, J.; Kinzler, K. W.; Velculescu, V. E.; Vogelstein, B.; Bigner, D. D. New Eng. J. Med. 2009, 360, 765-773. (30) Bogdanovic, E. Biochimica et. biophysica Biophys. acta Acta 2015, 1850, 1781-1785. (31) Esmaeili, M.; Hamans, B. C.; Navis, A. C.; van Horssen, R.; Bathen, T. F.; Gribbestad, I. S.; Leenders,

W.

P.;

Heerschap,

A.

Cancer

Res.

2014,

74,

4898-4907.

24


Page 25 of 25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For TOC only.

25


Comprehensive Analysis of Short-, Medium-, and Long-Chain Acyl

Recommend Documents