Structure-specific, accurate quantitation of plasmalogen glycerophosphoethanolamine

Abstract

Changes in plasmalogen glycerophosphoethanolamine (PE-P) composition (structure and abundance) are a key indicator of altered lipid metabolism. Differential changes in the levels of PE-P have been reported in different disease states, including neurodegenerative diseases. Of particular interest, traumatic brain injury (TBI) has resulted in altered expression of glycerophospholipid profiles, including PE-P. To date, most analytical assays assessing PE-P have focused on general lipidomic workflows to evaluate the relative, semi-quantitative abundance of PE-P during disease progression. This approach provides a broad evaluation of PE-P, yet often lacks specificity and sensitivity for individual PE-P structures which is a necessity for robust quantitative data. The present study highlights the development of a targeted, quantitative method using a HILIC separation and selective reaction monitoring mass spectrometry for the confident identification and accurate quantitation of PE-P. Our innovative method incorporates both the sn-1 alkyl vinyl ether and sn-2 acyl chain as product ion transitions, for specific and sensitive quantitation of 100 PE-P structures. Our method also uniquely allowed for the unambiguous assignment and quantitation of di-unsaturated sn-1 PE-P structures, which to date have not been conclusively quantified. Application of this assay to a TBI mouse model resulted in distinct temporal profiles for plasma PE-P up to 28 days post injury. Plasma PE-P were significantly increased 24 h after induced TBI, followed by a gradual reduction to sham concentrations by day 28. Overall, we established a structure-specific, quantitative assay for identification and quantitation of a comprehensive set of PE-P structures with demonstrated relevance to brain injury.

Graphical Abstract

1. Introduction

Glycerophospholipids (GPs) play a major role in maintaining the integrity of cell membranes, aiding in cell-cell interactions, and regulating cellular signaling [1]. Glycerophosphoethanolamine (PE) is the second most prevalent GP found in cellular membranes and is commonly present as one of three structural variants: diacyl-PE, alkyl ether-PE (PE-O), and alkenyl (vinyl) ether-PE (PE-P). PE-P are also referred to as plasmalogen PE. Diacyl-PE are the most abundant PE structure and uniformly present across cellular membranes, whereas PE-P is preferentially abundant in brain, heart, and spermatozoa [2].

PE-P are comprised of an ether-linkage at the sn-1 position with an adjacent cis double bond [3]. This unique structural arrangement has functional implications in membrane dynamics and integrity. PE-P play a crucial role in decreasing membrane fluidity, forming lipid raft domains, and serving as sacrificial antioxidants [4,5]. In addition, the sn-2 position of PE-P are often reservoirs for polyunsaturated fatty acids (PUFAs), where the enzymatic release of PUFAs is an integral pathway in the production of secondary messengers such as eicosanoids [4]. PUFA acyl chains and the vinyl ether bond of PE-P are susceptible to oxidative damage, due to the low hydrogen bond dissociation energies of neighboring bis-allylic methylene groups [3,6]. Oxidative damage and/or phospholipase activity, during disease progression, can form degradation products such as lysophospholipids, including lysoPE (LPE) and lysoPE-P (LPE-P) [7,8]. Due to the implications of their unique structure, altered levels of PE-P have been identified in Alzheimer’s disease patients, traumatic brain injury (TBI) mouse models, and in peroxisomal disorders such as Zellweger Syndrome [9–15]. Previous work from our group has reported differential expression of PE-P in lysosomes isolated from the brain cortices of mice 1-h after TBI [11]. This data suggested PE-P have the potential to provide unique insight into the pathogenesis of neurodegeneration following TBI. This line of research can be pursued via the accurate quantitation of PE-P structures.

PE-P quantitation has typically fallen under the umbrella of large-scale targeted lipid analysis assays where PE (including diacyl-PE, PE-O, and PE-P) are one of many lipid sub-classes [16,17]. These assays have routinely used mass spectrometry-based techniques employing selective reaction monitoring (SRM) on a tandem quadrupole mass spectrometer and involve either liquid chromatography (LC) or direct infusion (shotgun) approaches [18,19]. The use of LC in quantitative mass spectrometry has traditionally been viewed as providing a necessary layer of selectivity especially for the analysis of complex samples [20,21]. In a recent study, Otoki et al. used a LC-SRM method for the quantitation of PE-P and ether phosphatidylcholine (PC) in human plasma [22]. In this report, twelve abundant diacyl and plasmalogen PC and PE structures were quantified.

Accurate quantitation of endogenous PE-P is complicated by a variety of challenging factors: 1) limited number of available authentic PE-P reference standards, 2) limited number of stable, isotopically labeled PE-P internal standards, 3) variability in abundance of endogenous PE-P structures in biological matrices, and 4) selectivity of detection. With these factors in mind, we report the development and validation of a targeted LC-SRM method for the confident identification and accurate quantitation of 100 endogenous PE-P structures in plasma. Our developed method is validated using the FDA’s Bioanalytical Method Validation Guidance [23] and applied for the accurate quantitation of plasma PE-P in a TBI mouse model. The quantitative method described herein provides a robust platform for quantitative assessment of PE-P.

2. Materials and methods

2.1. Materials

Deuterated stable isotope labeled standards, PE(P-18:0/18:1-d9), LPE(18:1-d7), PE(17:0/18:1)-d5, and LPE(17:0)-d5 were purchased from Avanti Polar Lipids (Alabaster, AL). Lipid nomenclature for labeled standards was based on the placement of the deuterium label. For example, the deuterium shown within the parenthesis indicated labeling along the acyl chain, while outside of the parenthesis indicated labeling on the headgroup. UHPLC/MS Optima grade water (H₂O), acetonitrile (ACN), and methanol (MeOH) were purchased from Fisher Scientific (Pittsburg, PA). HPLC grade isopropanol (IPA), chloroform (CHCl₃) and ammonium acetate were also purchased from Fisher Scientific. HPLC grade methyl tert-butyl ether (MTBE) was purchased from Sigma Aldrich (St. Louis, MO). No further purification of solvents was conducted prior to mass spectral analysis. Surrogate mouse plasma was purchased from BioIVT (Westbury, NY).

2.2. Calibration curves

Calibration curves were constructed for authentic reference standards, PE(P-18:0/18:1-d9) and LPE(18:1-d7), in neat solution or the surrogate mouse plasma. Calibration standards ranging from 0.15 nM to 625 nM were used to prepare calibration curves. Secondary internal standards, PE(17:0/18:1)-d5 and LPE(17:0)-d5, were spiked in at 100 nM and were used for the generation of reference calibration curves. Calibration curves were adjusted with a weighting factor of 1/X. Curves constructed in surrogate mouse plasma were used for quantitative analysis. Mouse plasma samples from the TBI mouse model were spiked with 100 nM of the internal standards. Quality control (QC) samples were prepared at three concentrations (5, 20, and 200 nM). Standards and samples were prepared in IPA.

2.3. Traumatic brain injury (TBI) mouse model

All surgical procedures and animal experiments were performed as per the approved protocols and guidelines of the Animal Care and Use Committee of the University of Maryland. Moderate TBI was induced in male C57BL6/J mice by controlled cortical impact (CCI) as described previously [15,24]. At days 1, 3, 7, 14 and 28 after TBI, mice were anesthetized using isofluorane and blood was collected from their heart using potassium-EDTA (5%). Blood was centrifuged at 1500 g for 15 min at 4 °C to collect the plasma. Plasma samples were stored at −80 °C.

2.4. Sample preparation

Total lipid extracts of plasma were prepared using a revised MTBE lipid extraction protocol [25]. Briefly, 400 μL of methanol and 10 μL of internal standards were added to 15 μL of plasma. The mixture was vortexed and incubated on ice for 10 min. 500 μL of MTBE was then added, vortexed, and incubated on ice for 1 h. After incubation, 500 μL of H₂O was added, vortexed, and incubated on ice for 15 min. This mixture was centrifuged at 8000 g for 5 min at 4 °C, and the top organic layer was collected and stored on ice. An additional 200 μL of MTBE was added, vortexed, and incubated on ice for 15 min. This was followed by a second centrifugation step at 8000 g for 5 min at 4 °C. The organic layer was collected and combined with the first. This mixture was dried under N₂ gas and re-suspended in CHCl₃/MeOH (1:1, v/v). Samples were further diluted in IPA prior to analysis.

2.5. Liquid chromatography tandem mass spectrometry (LC-MS/MS)

LC-MS/MS analysis was conducted on an Ultimate 3000 Ultra High-Performance Liquid Chromatograph (UHPLC) coupled to a Thermo TSQ Altis Tandem Quadrupole Mass Spectrometer (Thermo Scientific, San Jose, CA). Chromatographic separation was achieved with an ACQUITY Amide BEH column (1.7 μm, 2.1 × 150 mm) maintained at 30 °C (Waters, Milford, MO). Mobile phase compositions for solvents A and B consisted of ACN/H₂O (95:5, v/v) and (50:50, v/v) respectively, with 10 mM ammonium acetate. The gradient profile had a flow rate of 0.4 mL min⁻¹ and ramped from 0.1 to 30% B in 5 min, from 30 to 90% B in 0.1 min, held at 90% B for 1 min, dropped from 90 to 0.1% B in 0.4 min, and held 0.1% B for 1.5 min. Total chromatographic run time was 8.0 min. The injection volume was 2 μL. All analytes were detected using positive ionization mode and SRM. Electrospray ionization (ESI) source conditions are included in supporting information (SI) Table S1. Collision energies and RF lens voltage were optimized for each reference standard and shown in SI Table S2 along with the precursor to product ion transitions. Data collection and analysis was done using Xcalibur 4.2 Qual and Quan Browser (Thermo Scientific, San Jose, CA).

2.6. Method validation

Our quantitative method was validated based on the FDA’s Guidance for Industry, Bioanalytical Analytical Method Validation [23]. Validation was accessed based on linearity, specificity, limits of quantitation (LOQ), intraday/interday accuracy and precision, extraction efficiency, matrix effects, and stability.

2.7. Statistics

Statistical analyses were performed using GraphPad Prism 8 (GraphPad, La Jolla, CA). Data obtained from the mouse plasma TBI samples were evaluated based on multiple t-test comparisons using the Holm-Sidak method. Statistical significance was accepted at p < 0.05.

3. Results and discussion

Our primary goal in developing a comprehensive PE-P quantitative assay was to link structure specificity to accurate quantitation. This was achieved using a HILIC separation coupled to SRM mass spectrometry. Structure specificity was defined as confidently assigning the sn-1 and sn-2 positions for each PE-P structure using two highly specific precursor to product ion transitions. Additional specificity was achieved using a HILIC separation, where chromatographic retention of PE-P as a lipid sub-class was isolated to a well-defined retention time window. The combined HILIC-SRM assay provided accurate quantitation of 100 PE-P structures in a single 8-min assay.

3.1. SRM mass spectrometry

LIPIDMAPS [26] was used to generate a comprehensive list of PE-P, LPE-P and LPE, comprising of 100 PE-P, 6 LPE-P and 6 LPE structures as shown in SI Table S2. From this list, we constructed a positive ion mode SRM list consisting of two precursor-to-product ion transitions per lipid structure. The product ions were selected based on literature precedence [27,28], LipidCreator [29], and tandem MS of representative standards. Gas-phase ion dissociation of protonated PE-P yield two abundant product ions that are specific to the sn-1 alkyl vinyl ether and the sn-2 acyl chain [27,28]. Refer to Fig. S1 for a comparison of positive and negative ion mode fragmentation of a representative PE-P structure. In contrast to positive ion mode, negative ion mode fragmentation of deprotonated PE-P is less specific towards acyl chain identity and location, as these product ions are limited to fragmentation associated with only the sn-2 position. PE-P precursor ions in both the positive ion and negative ion modes fragment differently than diacyl-PE and PE-O. The primary and overwhelmingly abundant dissociation channel for protonated diacyl-PE and PE-O precursor ions is the neutral loss of the phosphoethanolamine (PEtn (Fig. 1A, tandem MS spectrum of PE(17:0/18:1)-d5)). This product ion is characteristic of PE lipids in general yet contains limited acyl chain information. As such, diacyl-PE and PE-O structures, are often detected in the negative ion mode, as this provides structural information about the identity of the acyl chains.

Fig. 1.

Tandem mass spectra of PE, PE-P, LPE-P, and LPE in positive ion mode. A) Diacyl-PE showing abundant neutral loss (NL) of 141 Da, characteristic of the phosphoethanolamine (PEtn) headgroup. B) PE-P fragmentation identifying the sn-1 and sn-2 acyl chains. C) LPE-P fragmentation showing product ions associated with neutral loss of the glycerol-phosphate backbone. D) LPE neutral loss of 141 Da (PEtn).

As a representative example, tandem MS of the protonated PE(P-18:0/18:1-d9) produced two abundant product ions which confidently identify and localize the sn-1 alkyl vinyl ether and sn-2 acyl chain (Fig. 1B). An additional product ion corresponding to the neutral loss of PEtn was also present albeit at a much lower abundance. To take advantage of diagnostic product ions, our PE-P SRM list consisted of ion transitions that corresponded to the sn-1 alkyl vinyl ether and sn-2 acyl chain, where the sn-1 was used as the qualifying ion and sn-2 as the quantifying ion.

In comparison to PE-P, gas-phase ion dissociation of protonated LPE-P is not well documented and to the best of our knowledge these product ions have not been described. The product ions predicted via in silico fragmentation with LipidCreator of protonated LPE-P suggested the neutral loss of PEtn and phosphoric acid would be most likely. Contrary to this, we consistently observed two abundant product ions associated with the neutral loss of the glycerol-phosphate backbone (m/z 294.3 and 312.3, Fig. 1C). To account for these product ions, we propose a gas-phase reaction mechanism whereby the neutral loss of the glycerol-phosphate backbone is accompanied with an alkyl chain/ethanolamine rearrangement, similar to the mechanism for PE-P proposed by Zemski Berry and Murphy [27] (SI Scheme S1). These product ions were not observed in PE-P tandem mass spectra. An additional product ion associated with only the neutral loss of the glycerol backbone was observed in the LPE-P tandem mass spectrum (m/z 392.3, Fig. 1 C). This product ion is analogous to the sn-1 product ion observed in PE-P fragmentation here within and previously reported [27]. The two most abundant product ions, those associated with the neutral loss of the glycerol-phosphate backbone, were used for constructing the LPE-P SRM list. The SRM list also included a series of LPE structures. Fragmentation of protonated LPE results in an abundant PEtn neutral loss and a much lower abundant glycerol-phosphate neutral loss (Fig. 1D). These product ions were used for LPE quantitation.

3.2. Chromatographic separation

PE-P, LPE-P, and LPE structures were chromatographically separated using a BEH Amide column over an 8-min gradient (Fig. 2). The BEH Amide column provided a hydrophilic interaction liquid chromatography (HILIC) separation that effectively isolated PE subspecies and allowed for co-elution of endogenous lipids with their respective reference and internal standards.

Fig. 2.

Extracted ion chromatograms (EICs) of standards and endogenous PE species in extracted mouse plasma. HILIC separation provided lipid class separation, with PE-P structures eluting around 2.00 min (A–D), and LPE/LPE-P structures eluting around 4.30 min (E–G).

The reference standard PE(P-18:0/18:1-d9) and internal standard PE(17:0/18:1)-d5 were used for PE-P quantitation. Likewise, two standards (LPE(18:1-d7) (reference) and LPE(17:0)-d5 (internal)) were used for LPE-P and LPE quantitation. Representative extracted ion chromatograms (EICs) detailing co-elution of PE-P, LPE-P, and LPE structures with corresponding standards are displayed in Fig. 2. Across all PE-P structures included in our SRM list, the elution profile ranged between 1.71 and 2.20 min. The solvent composition over this gradient ranged from 10.3 to 13.3% solvent B, indicating the co-elution of PE-P structures and spiked standards sufficiently allowed for ionization of PE-P and standards under similar solvent composition. The co-elution of LPE-P and LPE plus their respective standards was also achieved with our HILIC separation. The LPE-P and LPE structures had an elution profile that ranged between 3.90 and 4.42 min. The solvent composition over this gradient ranged from 23.4 to 26.5% solvent B, indicating the co-elution of LPE-P, LPE and spiked standards sufficiently allowed for ionization under similar solvent composition. Of note, the HILIC separation provided baseline resolution of LPE positional isomers (Fig. 2E–G). These structures would be representative of the hydroxyl group positioned at either the sn-1 or sn-2 position, where the LPE reference standard is predominately observed as the sn-1 isomer indicating the later eluting peak corresponds to the sn-1 isomer [30].

3.3. Method validation

The proposed method for accurate quantitation of PE-P, LPE-P, and LPE was validated in mouse plasma. Validation included an evaluation of linearity, specificity, limit of quantitation (LOQ), intraday/interday accuracy and precision, extraction efficiency, matrix effects, and stability. A consequential decision when pursing accurate quantitation of endogenous analytes in general and PE-P specifically is the choice of calibration (reference) and internal standards. The reference standard is typically an authentic standard of the analyte itself or a closely related analogue. This standard is used to establish the instrument response obtained from the analyte in an unknown sample to that of a known concentration via construction of a calibration curve. In contrast, an internal standard is added prior to the analysis to account for extraction efficiency, matrix suppression, and instrument variability. The first choice for an internal standard is a stable, isotopically labeled version of the target analyte (or a closely related analogue). In circumstances where authentic or isotopically labeled standards of the target analyte are not available (or not feasible), care should be taken to ensure reference and internal standards match as best as possible the physiochemical properties of the target analytes. This presents a challenge for accurate quantitation of PE-P where it is not feasible to have reference and internal standards that represent endogenous PE-P structural diversity. This is a common scenario for quantitative lipidomics and routinely handled by selection of one or a few isotopically labeled standards that represent the target lipid (sub)-class [31].

Our quantitative assay included two deuterium stable-labeled standards per PE subclass. One standard was used as a reference standard and the other as an internal standard. The rationale for using two stable-labeled standards per PE subclass was two-fold: 1.) a reference standard that was structurally representative of the PE subclass could be used to construct calibration curves in the sample matrix, and 2.) a secondary internal standard could be used to normalize area counts to compensate for sample preparation and analytical variation. Importantly, all standards were non-endogenous to biological matrices. We purposely chose commercially accessible standards to make the developed assay readily transferable to other laboratories. As such, we chose PE(P-18:0/18:1-d9) as the reference standard and PE(17:0/18:1)-d5 as the internal standard for all PE-P SRM transitions. The use of a diacyl-PE for the internal standard was a consequence of feasibility, availability, and similar chromatographic response to PE-P structures. LPE-P and LPE were grouped together and had the same reference and internal standards. The reference standard was LPE(18:1-d7) and the internal standard was LPE(17:0)-d5. Analogous to the rationale for including diacyl-PE standard for PE-P, the same constraints were present to obtaining stable-labeled LPE-P standards.

Calibration curves of PE(P-18:0/18:1-d9) and LPE(18:1-d7) were constructed in neat solution and mouse plasma, and spiked with the respective internal standard. Linear regression analysis was applied by plotting the ratio of peak areas for the reference standard to the internal standard against concentration (Fig. 3). Calibration curves linearity spanned over three orders of magnitude, ranging from 0.15 nM to 625 nM for both reference standards (Fig. 3). Correlation coefficients for mouse plasma calibration curves were >0.998. Limit of Quantitation (LOQ) for PE(P-18:0/18:1-d9) and LPE(18:1-d7) were 0.6 nM and 5 nM, respectively. To ensure robustness of the LOQ, these were defined by a signal-to-noise ratio (S/N) > 10, within ±20% of the nominal concentration, and less than 15% coefficient of variation (CV). A comparison of calibration curves constructed in mouse plasma versus neat solution showed a profound effect of the matrix on the reference standards (Fig. 3), and thus calibration curves were constructed in mouse plasma. Specificity was validated through unique SRM transitions including characteristic fragments of both the sn-1 and sn-2 positions. LPE-P and LPE structures included specific quantifying and qualifying ions to minimize interference from co-eluting species.

Fig. 3.

Calibration curves for reference standards in neat solution and mouse plasma. Peak area response was peak area of reference standard divided by peak area of internal standard.

Intraday/interday accuracy and precision were assessed across three independent batches at the LOQ and three quality control (QC) concentrations (low, 5 nM; middle, 20 nM; high, 200 nM) within the surrogate mouse plasma. Intraday experiments evaluated the QC levels across an 8-h period, while interday experiments measured the QCs across three individual days. Acceptance criteria for these assessments were defined as QCs within ±25% of the nominal concentration (30% for the LOQ), and less than 25% CV. Calculated concentrations reported as the mean and % CV at each quality control level are presented in SI Table S3.

Extraction efficiency and matrix effects were evaluated in surrogate mouse plasma at low, middle, and high QC concentrations. At the low QC level, mean extraction efficiencies for PE(P-18:0/18:1-d9) and LPE(18:1-d7) were 87.9 ± 14.1 and 93.8 ± 8.3, respectively. Mean matrix effects for PE(P-18:0/18:1-d9) and LPE(18:1-d7) were 31.0 ± 8.7 and 48.3 ± 12.9. These results emphasized the necessary preparation of calibration curves in the surrogate matrix for subsequent quantitation. A complete table of mean percent extraction efficiencies and matrix effects at all three concentrations are listed in SI Table S4.

Stability of storage and sample preparation conditions were assessed following experiments of consecutive freeze-thaw cycles, short-term/benchtop stability, and long-term benchtop and auto-sampler stability. Each experiment was conducted at low and high QC levels. Reported values for each experimental condition are detailed in SI Table S5. Percent differences for benchtop and auto-sampler stability at designated time points were <15%. Freeze-thaw stability cycles showed a decrease in reported concentrations, specifically at the low QC value, indicating care should be taken for planning experiments where freeze/thaw cycles are necessary.

3.4. Application

Following validation, our newly developed quantitative assay was applied to plasma collected from a TBI mouse model for the identification and accurate quantitation of endogenous PE-P, LPE-P, and LPE. Plasma collected from sham mice and those subjected to TBI were evaluated across different time points, to assess changes in PE-P, LPE-P, and LPE profiles during injury progression. Briefly, C57BL6/J mice were exposed to a moderate cortical injury and plasma samples were collected 1-, 3-, 7-, 14-, and 28-days post injury. Sham samples underwent the same experimental procedure without the mechanical injury. Ten biological replicates were included in the sham samples, and five biological replicates were included for each time point.

Of the 112 individual structures screened, 54 structures were detected in plasma samples collected from the TBI mouse model (data not shown). From these structures, 32 were quantified above the LOQ (Table S6). This included 23 PE-P, 6 LPE, and 3 LPE-P structures, with sham concentrations ranging from 0.6 nM to 139.7 nM (Table S6). Fig. 4 represents the general shift in PE-P concentrations 24-h post injury, focusing on the respective sn-1 alkyl vinyl ether or sn-2 acyl chain. PE-P structures with a saturated sn-1 chain were not only highly abundant, but also represented the PE-P structures that were significantly elevated 24-h post TBI. Note, saturation of the sn-1 chain for PE-P does not include the double bond at the vinyl ether location. When considering the sn-2 acyl chain, PE-P structures containing 18:2, 20:4, and 22:6 were the most abundant and those with 18:1, 18:2, 18:3, 20:4, and 22:6 had the most profound increase in concentration following TBI (Fig. 4).

Fig. 4.

Changes in PE-P concentration based on the sn-1 or sn-2 chain across mouse plasma from sham and TBI (24-h post injury) mice. Concentrations are reported in nM. PE-P concentrations for sham samples (n = 10) and day 1 TBI samples (n = 5) were reported as mean ± SEM. Significance was determined using a multiple t-test, *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.0001, ns = non-significant (p > 0.05).

Of the structures identified, seventeen were found to have a significant change in concentration 24 h post-TBI (Table 1). As noted in Fig. 4, PE-P structures containing a saturated sn-1 alkyl vinyl ether were most significantly changed from sham to day 1. PE-P and LPE-P structures all showed an increase in concentration 1-day post injury, with nearly all PE-P structures subsequently dropping below the sham concentration by day 7 and returning to sham concentrations by day 28 (Table 1). Temporal changes in LPE concentration did not show a defined trend, with LPE(18:0) and LPE(18:1) displaying opposite shifts in abundance within the first 24 h. Fig. 5 presents four significantly changed structures displaying all investigated time-points after TBI onset. Each structure gradually returned towards the sham concentration after an immediate increase (PE-P/LPE-P) or decrease (LPE) in abundance.

Table 1.

Structures that were significantly changed across sham and TBI sample sets.

Lipid	Average Concentration (nM)						Significance (p-value)
	SHAM	Day 1	Day 3	Day 7	Day 14	Day 28	Sham Day 1	Sham Day 3	Sham Day 7	Sham Day 14	Sham Day 28
PE(P-16:0/18:1)	3.5 ± 0.1	4.3 ± 0.3	3.1 ± 0.2	2.6 ± 0.1	3.4 ± 0.3	3.2 ± 0.2	* (0.0132)	ns	*** (0.0007)	ns	ns
PE(P-16:0/18:2)	12.2 ± 0.7	17.7 ± 1.0	13.2 ± 0.5	9.6 ± 0.4	14.8 ± 1.0	11.7 ± 1.0	*** (0.0004)	ns	* (0.0224)	* (0.0473)	ns
PE(P-16:0/18:3)	0.7 ± 0.04	1.0 ± 0.1	0.6 ± 0.03	0.5 ± 0.1	0.8 ± 0.1	0.6 ± 0.1	** (0.0055)	ns	* (0.0325)	ns	ns
PE(P-16:0/20:4)	5.7 ± 0.1	6.7 ± 0.5	5.0 ± 0.2	4.3 ± 0.1	4.9 ± 0.2	5.2 ± 0.2	* (0.0250)	** (0.0064)	**** (<0.0001)	** (0.0052)	ns
PE(P-16:0/22:6)	19.7 ± 0.5	23.5 ± 1.3	17.3 ± 1.0	13.7 ± 0.7	17.7 ± 0.9	17.8 ± 1.1	** (0.0059)	* (0.0243)	**** (<0.0001)	ns	ns
PE(P-18:0/18:1)	1.6 ± 0.1	2.6 ± 0.1	1.8 ± 0.05	1.2 ± 0.1	1.8 ± 0.1	1.7 ± 0.1	**** (<0.0001)	* (0.0338)	*** (0.0008)	ns	ns
PE(P-18:0/18:2)	9.6 ± 0.4	17.7 ± 1.2	10.7 ± 0.4	7.9 ± 0.2	11.6 ± 0.8	9.0 ± 0.5	**** (<0.0001)	ns	* (0.0104)	* (0.0243)	ns
PE(P-18:0/18:3)	0.6 ± 0.03	1.1 ± 0.1	0.6 ± 0.04	0.5 ± 0.03	0.7 ± 0.1	0.6 ± 0.1	**** (<0.0001)	ns	* (0.0493)	ns	ns
PE(P-18:0/20:4)	7.0 ± 0.1	10.2 ± 0.7	6.5 ± 0.2	5.1 ± 0.2	6.7 ± 0.5	6.4 ± 0.3	**** (<0.0001)	ns	**** (<0.0001)	ns	ns
PE(P-18:0/22:6)	20.0 ± 0.7	31.9 ± 2.1	21.1 ± 1.2	15.0 ± 1.2	20.3 ± 1.0	18.0 ± 1.2	**** (<0.0001)	ns	*** (0.0021)	ns	ns
PE(P-20:0/18:2)	1.5 ± 0.1	2.5 ± 0.1	1.5 ± 0.1	1.3 ± 0.1	1.7 ± 0.1	1.4 ± 0.1	**** (<0.0001)	ns	ns	ns	ns
PE(P-20:0/20:4)	2.1 ± 0.1	2.9 ± 0.1	1.9 ± 0.1	1.5 ± 0.05	2.2 ± 0.2	1.8 ± 0.1	**** (<0.0001)	ns	*** (0.0002)	ns	* (0.0340)
PE(P-20:0/22:6)	4.0 ± 0.1	5.8 ± 0.3	4.4 ± 0.2	3.4 ± 0.2	4.7 ± 0.2	3.9 ± 0.1	**** (<0.0001)	ns	* (0.0206)	* (0.0257)	ns
LPE(18:0)	48.2 ± 2.5	59.9 ± 3.0	46.8 ± 3.3	47.0 ± 4.6	54.7 ± 3.8	44.2 ± 4.5	* (0.0136)	ns	ns	ns	ns
LPE(18:1)	39.8 ± 1.7	32.6 ± 2.4	39.8 ± 1.7	38.8 ± 3.3	45.1 ± 2.3	36.5 ± 5.3	* (0.0286)	ns	ns	ns	ns
LPE(P-18:0)	2.3 ± 0.2	3.7 ± 0.2	2.5 ± 0.2	2.2 ± 0.2	2.8 ± 0.3	1.6 ± 0.1	*** (0.0005)	ns	ns	ns	* (0.0277)
LPE(P-20:0)	1.0 ± 0.1	1.6 ± 0.1	1.3 ± 0.2	1.1 ± 0.2	1.1 ± 0.1	0.7 ± 0.1	*** (0.0003)	ns	ns	ns	* (0.0455)

Concentrations were reported as mean ± SEM.

Significancewas determined using amultiple t-test,

^*p < 0.05,

^**p < 0.01,

^***p < 0.005,

^****p < 0.0001,

ns = non-significant (p > 0.05).

All PE-P structures listed showed a significant increase within the first 24-h of induced TBI.

Fig. 5.

Altered abundance of specific PE-P, LPE-P, and LPE structures in mouse plasma from sham and TBI-induced mice. Endogenous concentrations for sham samples (n = 10) and TBI samples (n = 5 for all time points) were reported as mean ± SEM. Significance was determined using a multiple t-test, *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.0001.

PE-P structures predominately consist of a 16:0, 18:0, and 18:1 alkyl vinyl ether in the sn-1 position [4]. To date, there have been limited reports of the presence of di-unsaturated PE-P structures from biological matrices [32,33]. Our initial screening for sn-1 di-unsaturated PE-P was conducted in the surrogate mouse plasma (Table S2), to which these were further evaluated and quantitated in plasma from TBI-induced mice. Two sn-1 di-unsaturated PE-P structures, PE(P-18:2/18:2) and PE(P-18:2/22:6), were identified and quantitated in plasma from the TBI mouse model. Sham concentrations for these two structures were 1.3 nM and 0.9 nM respectively (Table S6). The di-unsaturated PE-P structures were confirmed via a HILIC separation coupled to high-resolution tandem mass spectrometry (SI Fig. S2 and associated description in SI).

Our analytical assay included several LPE-P and LPE structures. These structures were included in the assay for both analytical and biological considerations. LPE-P and LPE co-elute on our HILIC separation making use of a single set of reference and internal standards, advantageous especially since isotopically labeled LPE-P standards are not feasibly available. Further, protonated LPE-P and LPE both provided two specific precursor to product ion transitions enabling highly selective detection. The combined chromatographic and tandem MS attributes of LPE-P and LPE made their inclusion not only convenient but analytically sound. In addition to the analytical rationale, the inclusion of LPE-P and LPE provides a pathway for monitoring potential by-products resulting from phospholipase activity. Caution should be applied for directly inferring enzymatic activity to the presence of lysophospholipids when enzymatic activity is not directly monitored. Yet, it does provide supporting information in the context of a well-characterized model system where increased phospholipase activity is well-documented [15,34,35].

The strength of our newly developed quantitative assay lies in the direct connection between confident PE-P structure identification and accurate quantitation. This is highly relevant to models of brain injury where literature precedent has established PE-P structures were differentially expressed following TBI [36]. Nearly all plasma PE-P structures identified showed an increase in concentration 24-h following TBI. This was consistent with literature, as studies evaluating plasma have found significant elevations in PE-P levels within the first hour after brain injury [37]. Our results corroborated the notion that plasma PE-P were indicative of a systemic response to the initial brain injury. Moreover, the accurate quantitation of specific PE-P structures allowed us to consider the role structure specificity at the sn-1 and sn-2 position plays in systemic response to acute brain injury. For instance, saturated vinyl ether alkyl chains (P-16:0, P-18:0, and P-20:0) at the sn-1 position and specific sn-2 acyl chains (e.g., 18:2, 18:3, 20:4, and 22:6) were preferentially elevated at 24-h post brain trauma. This would indicate that PE-P structure was differentially impacted during the initial response to the brain injury. Ultimately, each PE-P structure gradually returned to near sham concentrations over the 28-days of monitoring the systemic response. This observation indicated a re-equilibration of circulating PE-P levels after brain injury and a return towards a normal, homeostatic state.

Our data suggests that plasma PE-P show great potential as circulating markers that are indicative of an initial systemic response to acute brain trauma. This method is directly transferable to other biological matrices (e.g., tissue, cells) and as such provides the foundation for further investigation into how PE-P structure and abundance contribute to increasing our fundamental knowledge of injury progression following acute brain trauma.

4. Conclusions

Our newly developed quantitative assay involving confident structure assignment and accurate quantitation is uniquely suited for comprehensive plasmalogen PE detection. The combined use of HILIC separation, diagnostic product ions, and inclusion of isotopically labeled standards allowed us to quantitatively monitor over 100 PE-P, LPE-P, and LPE structures in mouse plasma. Our quantitative assay was validated according to the FDA’s Guidance for Bioanalytical Method Validation and applied to quantifying plasma PE-P in a TBI mouse model. We identified several PE-P structures that were significantly elevated 24-h post TBI exposure. The initial abundance of circulating PE-P robustly correlated to brain trauma and over time, out to 28 days post-exposure, PE-P abundance relaxed to near sham levels. Our results indicated there was an early time frame where brain trauma is potentially detected by circulating PE-P abundance. Taken together, our innovative approach to anchor structure specificity to abundance for comprehensive PE-P quantitation has high potential to substantially progress our knowledge of the plasmalogen PE following acute brain injury. Moreover, the knowledge gained from the accurate quantitation of structure-specific PE-P promises to provide mechanistic insight not only for brain injury progression but also may be informative for injury diagnosis and potentially instructive for long term cognitive impact resulting from an acute injury.

HIGHLIGHTS.

• Accurate quantitation of structure-specific plasmalogen PE.

• Quantitatively measure over 100 plasmalogen PE and LPE structures in plasma.

• Identification and quantitation of di-unsaturated sn-1 plasmalogen PE structures.

• Plasma plasmalogen PE significantly elevated 24-h after acute brain trauma.

Abbreviations:

PE-P

plasmalogen glycerophsphoethanolamine

LPE-P

lysoplasmalogen PE

LPE

lysophosphatidylethanolamine

GPs

glycerophospholipids

PE

glycerophosphoethanolamine

PE-O

alkyl-ether PE

PUFAs

polyunsaturated fatty acids

TBI

traumatic brain injury

SRM

selective reaction monitoring

PC

phosphatidylcholine

H₂O

water

ACN

acetonitrile

MeOH

methanol

MTBE

methyl tert-butyl ether

IPA

isopropanol

CHCl₃

chloroform

CCI

controlled cortical impact

PEtn

phosphoethanolamine headgroup

NL

neutral loss