Effect of lysosomal storage on bis(monoacylglycero)phosphate

Transcription

1 Biochem. J. (2008) 411, (Printed in Great Britain) doi: /bj Effect of lysosomal storage on bis(monoacylglycero)phosphate Peter J. MEIKLE* 1,3, Stephen DUPLOCK*, David BLACKLOCK*, Phillip D. WHITFIELD* 2, Gemma MACINTOSH* 1, John J. HOPWOOD* and Maria FULLER* *Lysosomal Diseases Research Unit, Department of Genetic Medicine, Children Youth and Women s Health Service, North Adelaide, SA 5006, Australia, and Department of Paediatrics, University of Adelaide, Adelaide, SA 5005, Australia BMP [bis(monoacylglycero)phosphate] is an acidic phospholipid and a structural isomer of PG (phosphatidylglycerol), consisting of lysophosphatidylglycerol with an additional fatty acid esterified to the glycerol head group. It is thought to be synthesized from PG in the endosomal/lysosomal compartment and is found primarily in multivesicular bodies within the same compartment. In the present study, we investigated the effect of lysosomal storage on BMP in cultured fibroblasts from patients with eight different LSDs (lysosomal storage disorders) and plasma samples from patients with one of 20 LSDs. Using ESI-MS/MS (electrospray ionization tandem MS), we were able to demonstrate either elevations or alterations in the individual species of BMP, but not of PG, in cultured fibroblasts. All affected cell lines, with the exception of Fabry disease, showed a loss of polyunsaturated BMP species relative to mono-unsaturated species, and this correlated with the literature reports of lysosomal dysfunction leading to elevations of glycosphingolipids and cholesterol in affected cells, processes thought to be critical to the pathogenesis of LSDs. Plasma samples from patients with LSDs involving storage in macrophages and/or with hepatomegaly showed an elevation in the plasma concentration of the C 18:1 /C 18:1 species of BMP when compared with control plasmas, whereas disorders involving primarily the central nervous system pathology did not. These results suggest that the release of BMP is cell/tissue-specific and that it may be useful as a biomarker for a subset of LSDs. Key words: biomarker, bis(monoacylglycero)phosphate (BMP), electrospray ionization tandem MS (ESI-MS/MS), lysobisphosphatidic acid, lysosomal storage disorder, phosphatidylglycerol. INTRODUCTION Glycerophospholipids are a broad group of biologically important lipids. They consist of a glycerol-3-phosphate backbone whose sn1 and sn2 positions are esterified with fatty acids. The phosphoryl group can be further esterified to one of several alcohol moieties including choline, ethanolamine, serine, inositol and glycerol. PG (phosphatidylglycerol; Figure 1) is an acidic phospholipid containing a glycerol moiety attached to the phosphate and is a minor component of most animal tissues; it is also an essential component of the lung surfactant [1]. BMP [bis(monoacylglycero)phosphate] is a structural isomer of PG, in which each glycerol moiety is esterified through the sn1 position to the phosphate and contains a single fatty acid ester (Figure 1). BMP is located primarily within the endosomal/lysosomal membranes of cells. It is thought to be synthesized from PG in the endosome/lysosome compartment by the action of multiple enzymes including a phospholipase A and a transacylase [2,3]. Within the lysosome, BMP is found almost exclusively in the internal vesicles, where it constitutes over 70 % of phospholipids in a subpopulation of the internal membranes [4]. BMP and other anionic phospholipids act to promote the degradation of glycosphingolipids, targeted to the multivesicular bodies for degradation, by providing a suitable environment for the interaction of the glycosphingolipid hydrolases and their activator proteins with their lipid substrates [5]. BMP has also been shown to regulate cholesterol transport by acting as a collection and distribution device [6]. The acyl groups of BMP are usually enriched in polyunsaturated species [3] and this appears to be a necessary factor for the efficient partitioning of cholesterol in lipid membranes [7] and the subsequent transport of cholesterol out of the internal membranes of the multivesicular bodies [5]. Thus BMP is a critical component of the endosomal/lysosomal network and essential for the correct functioning of this system. The endosomal/lysosomal network is responsible for degrading macromolecules into smaller subunits for re-utilization by the cell. In a group of diseases known as LSDs (lysosomal storage disorders), such degradation is impaired and undegraded material accumulates in the lysosomes of affected cells. The primary storage material is usually the substrate for the deficient enzyme in each LSD and may consist of glycolipids such as glucosylceramide or trihexosylceramide in Gaucher disease and Fabry disease respectively, sphingomyelin or cholesterol in Niemann Pick type A/B or C, GAGs (glycosaminoglycans) in the MPSs (mucopolysaccharidoses) or glycogen in Pompe disease. For many LSDs, the primary storage has been shown to result in impaired lysosomal function (lysosomal dysfunction) leading to the accumulation of secondary metabolites that are not substrates for the deficient enzymes [8]. In some disorders, it has been proposed that these secondarily stored materials may contribute to the pathogenesis of the disease [9,10]. Abbreviations used: BMP, bis(monoacylglycero)phosphate; BMP C 14:0 /C 14:0, sn-(3-myristoyl-2-hydroxy)-glycerol-1-phospho-sn-3 -(1 -myristoyl-2 - hydroxy)-glycerol; BMP C 18:1 /C 18:1, sn-(3-oleoyl-2-hydroxy)-glycerol-1-phospho-sn-3 -(1 -oleoyl-2 -hydroxy)-glycerol; CID, collision-induced dissociation; CV, coefficient of variation; ESI-MS/MS, electrospray ionization tandem MS; GAG, glycosaminoglycan; HDL, high-density lipoprotein; ISTD, internal standard; LC, liquid chromatography; LDL, low-density lipoprotein; LSD, lysosomal storage disorder; MPS, mucopolysaccharidosis; MRM, multiple-reaction monitoring; PC, phosphatidylcholine; PG, phosphatidylglycerol; PG C 14:0 /C 14:0, 1,2-dimyristoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]; PG C 18:1 /C 18:1,1,2- dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]; VLDL, very-low-density lipoprotein. 1 Present address: Baker Heart Research Institute, 75 Commercial Road, Melbourne, VIC 3006, Australia. 2 Present address: Department of Veterinary Preclinical Sciences, Faculty of Veterinary Science, University of Liverpool, Crown Street, Liverpool L69 7ZJ, U.K. 3 To whom correspondence should be addressed ( peter.meikle@baker.edu.au).

2 72 P. J. Meikle and others sonication for 20 s. Total cell protein was determined by the method of Lowry et al. [12]. Plasma fractionation Plasma was fractionated into VLDL (very-low-density lipoprotein), LDL (low-density lipoprotein), HDL (high-density lipoprotein) and lipoprotein-depleted plasma by differential ultracentrifugation using potassium bromide [13]. Plasma was diluted with an equal volume of PBS and then centrifuged (10000 g and 30 min) to remove cellular debris. Exosomes were then isolated by ultracentrifugation ( g and 70 min) [14]. Figure 1 Structure and fragmentation of PG and BMP In positive-ion MS/MS, PG (A) undergoes a neutral loss of the ammonium adduct of the glycerophosphate moiety (189.0 Da), and BMP (B) undergoes a neutral loss of the ammonium adduct of the glycerophosphate plus the fatty acid. In order to investigate the effect of lysosomal storage on the structure and function of the endosomal/lysosomal network, we developed a new ESI-MS/MS (electrospray ionization tandem MS)-based method for the specific determination of molecular species of BMP and PG and used this to measure these species in cultured fibroblasts and plasma from patients with a range of LSDs. MATERIALS AND METHODS Materials The ISTDs (internal standards) PG C 14:0 /C 14:0 {1,2-dimyristoylsn-glycero-3-[phospho-rac-(1-glycerol)]}, PG C 18:1 /C 18:1 {1,2- dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]}, BMP C 14:0 / C 14:0 [sn-(3-myristoyl-2-hydroxy)-glycerol-1-phospho-sn-3 -(1 - myristoyl-2 -hydroxy)-glycerol] and BMP C 18:1 /C 18:1 [sn-(3-oleoyl-2-hydroxy)-glycerol-1-phospho-sn-3 -(1 -oleoyl-2 -hydroxy)- glycerol] were obtained from Avanti Polar Lipids (Alabaster, AL, U.S.A.). All solvents were of HPLC grade and were used without further purification. Patient samples The Children, Youth and Women s Health Service, Research Ethics Committee (Adelaide, SA, Australia) approved the use of skin fibroblasts and plasma in these studies. Control plasma and cultured skin fibroblasts were collected from patients, with informed consent, and were sent to the National Referral Laboratory for Lysosomal, Peroxisomal and Related Genetic Diseases (Department of Genetic Medicine, Women s and Children s Hospital, Adelaide, Australia) for testing. Samples used in the present study were de-identified to the researchers. Cell culture Human diploid fibroblasts were established from skin biopsies submitted to the Department of Genetic Medicine for diagnosis [11]. Skin fibroblasts from unaffected individuals and LSD patients were maintained in BME (basal medium Eagle) supplemented with 10% (v/v) foetal calf serum at 37 Cina humidified atmosphere containing 5 % CO 2. Cells were harvested at 2 weeks post-confluence and cell extracts were prepared in 200 μl of 20 mmol/l Tris/HCl and 0.5 mol/l NaCl (ph 7.2) by Extraction of phospholipids Extraction of PG and BMP from fibroblast extracts and plasma was performed by the method of Folch et al. [15]. Briefly, fibroblast extracts (100 μl) or plasma (100 μl) were extracted with 2.0 ml of chloroform/methanol (2:1, v/v) containing 400 pmol of each ISTD (PG C 14:0 /C 14:0 and BMP C 14:0 /C 14:0 ). The mixture was shaken for 10 min and then allowed to stand at room temperature (20 C) for a further 50 min. The samples were partitioned with the addition of 0.4 ml of water and shaken for 10 min. To facilitate phase separation, the mixture was centrifuged (2500 g for 5 min). The lower hydrophobic layer containing the phospholipids was transferred to a clean glass test tube and dried under a gentle stream of nitrogen. This phospholipid extract was reconstituted in 100 μl of 10 mmol/l ammonium formate in methanol. MS analysis of PG and BMP Phospholipid analysis was performed using a PE Sciex API 3000 triple-quadrupole mass spectrometer with a turbo-ionspray source (200 C), a Gilson 233 autosampler and the Analyst 1.1 data acquisition system. N 2 was used as the collision gas at a pressure of torr (1 torr = kpa). In order to identify the molecular species of PG and BMP in cultured skin fibroblasts, flow-injection experiments were performed in negative-ion mode. Samples (20 μl) were injected at a flow rate of 80 μl/min using 10 mmol/l ammonium formate in methanol as the mobile phase. Precursor-ion scans for the mass to charge ratio (m/z) corresponding to the glycerophosphate moiety and for the m/z corresponding to individual fatty acid moieties [i.e. m/z for oleic acid (C 18:1 )] identified individual PG/BMP species. Quantification of individual PG and BMP species was performed in positive-ion mode by LC (liquid chromatography) ESI-MS/MS on an HP 1100 HPLC system (Hewlett Packard). Samples (20 μl) were injected on to an Alltima C18, 3 μm, 50 mm 2.1 mm column and were eluted with an isocratic flow (200 μl/min) of 10 mmol/l ammonium formate in methanol. PG species were quantified by measuring the neutral loss of 189 Da, corresponding to the ammonium adduct of the glycerophosphate (Figure 1). Similarly, BMP species were quantified by measuring the neutral loss corresponding to the monoacylglycerophosphate moiety from each parent molecular species (Figure 1). Multiple PG and BMP species, containing different acyl chains, were measured in each experiment by using MRM (multiple-reaction monitoring); each ion pair was monitored for 50 ms with a resolution of 0.7 a.m.u. (atomic mass units) at half-peak height. PG and BMP concentrations were calculated by relating the peak areas of each species to the peak area of the corresponding ISTD by using Analyst quantification software. PG species were related to PG C 14:0 /C 14:0 and the BMP species were related to BMP C 14:0 /C 14:0.

3 The effect of lysosomal storage on bis(monoacylglycero)phosphate 73 Figure 2 Product-ion spectra for PG and BMP CIDofPGC 14:0 /C 14:0 in negative-ion mode (A) and positive-ion mode (B) was performed with collision energy set to 55 and 25 V respectively. CID of BMP C 14:0 /C 14:0 in negative-ion mode (C) and positive-ion mode (D) was performed by using the same conditions. RESULTS CID (collision-induced dissociation) of PG and BMP CID of PG C 14:0 /C 14:0 and BMP C 14:0 /C 14:0 in negative-ion mode produced similar fragmentation patterns for both compounds (Figures 2A and 2C), making specific determination of either lipid difficult in the presence of the other. Intense product ions were observed for the myristate (C 14:0 ) fatty acid moieties (m/z 227.1) and for the glycerophosphate fragment at m/z Minor differences were observed for the product ions at m/z and corresponding to the neutral loss of the acyl chain + water. One clear distinction between PG and BMP was the ability of PG to lose the glycerol moiety without the loss of a fatty acid. Thus the neutral loss of 74.0 Da is specific for the PG species only, although this represented only a minor fragmentation product and was not suitable for specific quantification of the PG species. In positive-ion mode, the CID patterns for PG and BMP were different (Figures 2B and 2D). The positive-ion ammonium adducts of PG C 14:0 /C 14:0 and BMP C 14:0 /C 14:0 both have precursor ion signals at m/z 684.5, but only the PG species can undergo a neutral loss of Da, corresponding to the ammonium adduct of the glycerophosphate moiety, resulting in the diacylglycerol fragment (m/z 495.5). In contrast, although it is possible for both PG and BMP to produce a monoacylglycerol fragment (m/z 285.3), this fragmentation occurs more readily in the BMP species (Figure 2D). PG species containing different acyl chains gave similar responses for the neutral loss of Da, as determined by the relative responses for PG C 14:0 /C 14:0 and PG C 18:1 /C 18:1 standards. BMP species also gave similar responses for the loss of the monoacylglycerol phosphate moiety, as determined by comparison of the signals from the BMP C 14:0 /C 14:0 and BMP C 18:1 /C 18:1 standards. However, unlike the PG species, fragmentation of BMP to produce the monoacylglycerol ion can occur on either side of the phosphate, leading to two different product ions for asymmetrical BMP species. This reduced the signal from the asymmetric BMP species compared with the symmetrical species. This was confirmed by comparison of the Q1 signals and the precursor-ion signals for a fibroblast extract, where the symmetrical BMP species resulted in product-ion signals that were approx. 24% of the Q1 signal, while asymmetrical species gave product-ion signals that were approx. 12 % of the Q1 signal. To correct for this difference, signals resulting from asymmetric BMP species were multiplied by a factor of two, prior to the calculation of BMP concentrations. Identification of PG and BMP species in cultured skin fibroblasts Precursor-ion scanning in negative-ion mode for the common glycerophosphate/water fragment (m/z 153.0) and for the m/z corresponding to the fatty acid chains allowed us to identify the most abundant PG/BMP species in skin fibroblasts. This was done in the manner described by Han et al. [16] using a two-dimensional ESI-MS/MS approach to identify cross-peak intensities that correspond to the phospholipid species observed in the precursor of m/z spectra. As an example, precursor scanning for m/z identified a signal at m/z that could result from a PG or BMP C 36:1 species. Confirmation of the presence of the C 18:0 /C 18:1 species was made by conducting precursor-ion scans for the fatty acid products C 18:0 (m/z 283.3) and C 18:1 (m/z 281.3) respectively and identifying the parent ion at m/z Fatty acid species identified on the PG/BMP species included C 16:0,C 16:1,C 18:0,C 18:1,C 18:2,C 20:3,C 20:4,C 22:4,C 22:5 and C 22:6. While most combinations of fatty acids could be identified, many were minor components. No distinction was made between PG and BMP species in these analyses. The PG/BMP species identified in negative-ion mode were used to define MRM experiments for both PG and BMP species in positive-ion mode. Subsequent analyses of skin fibroblast extracts in positive-ion mode identified the major PG species in control skin fibroblasts as PG C 36:1, which represented species containing primarily C 18:0 and C 18:1 fatty acids, and PG C 34:1,which represented a mixture of PG species containing either C 16:1 and C 18:0 or C 16:0 and C 18:1 fatty acids. Other minor species include PG C 34:2,PGC 36:2,PGC 36:3,PGC 38:5 and PG C 40:5. PG species with two long-chain polyunsaturated fatty acids were present in relatively minor amounts. In contrast, the prominent BMP species of control fibroblasts were BMP C 18:1 /C 22:6,BMPC 22:5 /C 22:6 and

4 74 P. J. Meikle and others Figure 3 Analysis of PG C 36:2 and BMP C 18:1 /C 18:1 by LC-MS/MS Lipid extracts of plasma samples from a control individual and a Gaucher patient were analysed as described in the Materials and methods section. The signals corresponding to PG C 36:2 (MRM=m/z 792.5/603.5) were normalized to the signal from the ISTD PG C 14:0 /C 14:0 (MRM=m/z 684.5/495.5) (A). The signals corresponding to BMP C 18:1 /C 18:1 (MRM=m/z 792.5/339.4) from each sample were normalized to the signal from the ISTD BMP C 14:0 /C 14:0 (MRM=m/z 684.5/285.3) (B). BMP C 22:6 /C 22:6.BMPC 18:0 /C 18:1 constituted less than 1% of the total BMP in the control fibroblast extracts. Quantification of PG and BMP by LC ESI-MS/MS Analysis of PG and BMP, in cell or plasma extracts, by flow injection in positive-ion mode, resulted in greater than 98% signal suppression. To overcome this suppression effect, an LC step was incorporated into the analysis to separate the PG and BMP from other more easily ionized lipids and salts; this reduced the suppression to approx. 50 % for both PG and BMP and resulted in a limit of detection in plasma of 2 and 6 nmol/l for PG and BMP respectively. The most likely source of the signal suppression was thought to be the abundant PC (phosphatidylcholine) species; monitoring of the PC species by precursor ion scanning of m/z 184 demonstrated that the major species (PC C 32:1,C 34:1,C 34:2 and C 36:2 ) were eluted after the PG and BMP species and so had no influence on signal suppression when the LC step was used (results not shown). Retention times of the PG and BMP species on the LC column were in the range of min. Figure 3 shows chromatograms of PG C 36:2 and BMP C 18:1 /C 18:1 species in control and Gaucher plasma extracts with PG C 14:0 /C 14:0 and BMP C 14:0 /C 14:0 ISTDs. The double peaks observed for both BMP C 14:0 /C 14:0 and BMP C 18:1 /C 18:1 (Figure 3B) were thought to result from structural isomers of BMP with different acylation of the glycerol moieties. The PG C 14:0 /C 14:0 ISTD shows a single peak since only one structural isomer can exist. However, the PG C 36:2 MRM chromatogram shows multiple peaks, which represent several PG species, which could include PG C 18:1 /C 18:1, PG C 18:0 /C 18:2 and PG C 16:0 /C 20:2. The neutral loss of the ammonium adduct of the glycerophosphate moiety, which was used to monitor the PG species in positive-ion mode, cannot occur in the BMP species as both glycerol structures are acylated. However, analysis of the BMP C 14:0 /C 14:0 ISTD did show a small signal corresponding to the neutral loss of a glycerophosphate moiety. The elution time of this signal (2.1 min) coincided with the BMP C 14:0 /C 14:0 ISTD and so precluded the possibility of a contamination of the BMP ISTD with PG C 14:0 /C 14:0, which had an elution time of 1.9 min. This signal represented approx. 2% of the BMP signal and is thought to result from in-source rearrangement of an acyl group to form PG C 14:0 /C 14:0 (results not shown). We also observed a small signal from the PG C 14:0 /C 14:0 ISTD, corresponding to the monoacylglycerol product determined by the MRM (m/z 685/285). The collision energy used for the quantification of BMP species was subsequently optimized to minimize the intensity of the signal from the PG species while maintaining the signal intensity for the BMP species. The resulting signal from the PG species using the BMP-specific MRM was approx. 2% of the signal observed in the PG-specific MRM (neutral loss of Da; results not shown). The response of BMP in the PG-specific MRM and the PG in the BMP-specific MRM did not affect quantification for most species as the chromatographic step provided adequate separation of the corresponding PG and BMP species. However, LC analysis of PG C 40:8,PGC 42:10,PGC 44:10,PGC 44:11 and PG C 44:12 did not provide complete separation from the corresponding BMP species as they all eluted between 1.7 and 2.0 min. Consequently, for quantification, each of these PG species was corrected for the contribution of the corresponding BMP species and each BMP species was corrected for the contribution of the corresponding PG species. Assay performance The PG/BMP assay was linear within the range pmol/ml (R 2 > 0.999). Intra- and inter-assay CVs (coefficients of variation) were determined for each analyte using 100 μg of cell lysate per assay. Intra-assay CVs were calculated from five repeats of the same sample and inter-assay CVs were calculated from 30 repeats performed over 6 days. Intra-assay CVs were less than 10% and inter-assay CVs less than 15% for all PG and BMP species present at concentrations greater than 2.0 pmol/100 μg of cell protein (30 of 36 species). PG and BMP amounts in cultured skin fibroblasts Analysis of fibroblast extracts from LSD patients showed relatively minor changes in the amount of PG (Table 1). In addition, the relative amounts of each PG species (PG profile) in most LSD cell lines were almost identical with the control fibroblasts (compare Figures 4A and 4G), the exception being Pompe fibroblasts (Figure 4E), which had a decreased amount of PG C 36:1 compared with control cells and corresponding increased concentrations of PG C 34:1 and PG C 36:2. LSD fibroblasts also showed only a modest increase (2 3-fold) in total BMP amounts compared with control fibroblasts (Table 1), although this was greater in Fabry cell lines (5 9-fold). However, in all cell lines, with the exception of the MPS IIIA cell line, there was a substantial increase in the amount of BMP C 18:1 /C 18:1 (3 20-fold). In contrast with the BMP C 18:1 /C 18:1 species, most cell lines showed no increase in the major BMP species, BMP C 22:6 /C 22:6, with several cell lines showing a decrease in this species (Table 1). Thus, with the exception of Fabry disease, all LSD cell lines showed a different BMP profile compared with the control cell lines (compare Figures 4B with 4H); in particular, Niemann Pick type C cell lines had almost no BMP species containing two polyunsaturated fatty acids (Figure 4F).

5 The effect of lysosomal storage on bis(monoacylglycero)phosphate 75 Table 1 PG and BMP in control and LSD skin fibroblasts Values are means (S.D.) and are expressed as pmol/mg of cell protein. Significantly different from the control (P < 0.05, Student s t test, equal variance not assumed). Total Total BMP BMP Disorder PG BMP 18:1/18:1 22:6/22:6 Control (n = 7) 135 (41) 602 (110) 20 (6) 146 (34) Fabry (n = 3) 211 (32) 4392 (1159) 231 (75) 898 (224) Gaucher (n = 2) 299 (10) 430 (172) 85 (42) 35 (11) MPS I (n = 3) 325 (69) 1430 (241) 310 (101) 103 (18) MPS II (n = 2) 424 (87) 1426 (306) 326 (85) 99 (8) MPS IIIA (n = 1) Niemann Pick type A/B (n = 3) 219 (75) 1514 (1018) 249 (122) 133 (100) Niemann Pick type C (n = 2) 88 (23) 342 (51) 95 (10) 2(3) Pompe (n = 1) observed in the cultured cells (Table 2; Figure 5B). The major species in plasma was BMP C 18:1 /C 18:1, with significant amounts of the shorter-chain species BMP C 16:0 /C 16:0 and BMP C 16:0 /C 18:1, while the major long-chain species observed in fibroblasts (BMP C 22:6 /C 22:6 ) was present in only minor amounts. Only the Niemann Pick type A patients showed a significant increase in the concentration of plasma PG as determined by the Mann Whitney U test (Table 2). In contrast, five of the 20 LSDs showed a significant increase in the concentration of plasma BMP, and examination of the BMP profile for Niemann Pick type A/B showed that this was primarily an increase in the concentrations if BMP C 18:1 /C 18:1 and BMP C 18:1 /C 18:2 (Figure 5D). When the Mann Whitney U test was applied to the BMP C 18:1 /C 18:1 data, ten LSDs show a significant increase (P < 0.05) compared with the control group (Table 2). PG and BMP concentrations in plasma Plasma samples from 20 control individuals and 64 LSD-affected individuals representing 20 LSDs were analysed for ten PG species and ten BMP species by the LC ESI-MS/MS method. The profile of PG in plasma was different from that observed in fibroblasts. In control plasma, the most abundant species were PG C 34:1,PGC 36:1 and PG C 36:2 (Figure 5A). The BMP profile from the control individuals was also substantially different from that BMP localization in plasma Fractionation of plasma samples from control individuals (n = 3) and Gaucher patients (n = 3) into VLDL, LDL, HDL and lipoprotein-deficient plasma followed by quantification of BMP C 18:1 / C 18:1 in each fraction showed that, in control plasma, BMP was present in all plasma compartments with approx. 40% associated with lipoproteins (10 % in VLDL, 22 % in LDL and 9% in HDL) and 60% remaining in the lipoprotein-deficient plasma. In contrast, plasma from Gaucher patients had approx. 80% of Figure 4 PG and BMP profiles from control and LSD skin fibroblasts Control skin fibroblasts (n = 7) and LSD skin fibroblasts (n = 17) were analysed for PG and BMP species by LC ESI-MS/MS as described in the Materials and methods section. The relative amount of each species of PG and BMP is shown as a percentage of the total with error bars indicating one S.D. Control PG and BMP are shown in (A)and(B) respectively. PG and BMP species present in Fabry skin fibroblasts are shown in (C) and(d) respectively. PG species in Pompe skin fibroblasts are shown in (E) and BMP species in Niemann Pick type C skin fibroblasts in (F). PG species present in the remaining LSD skin fibroblasts [Gaucher (n = 2), MPS I (n = 3), MPS II (n = 2), MPS IIIA (n = 1), Niemann Pick type A/B (n = 3) and Niemann Pick type C (n = 2)] are shown in (G) and BMP in the remaining skin fibroblasts [Gaucher (n = 2), MPS I (n = 3), MPS II (n = 2), MPS IIIA (n = 1), Niemann Pick type A/B (n = 3) and Pompe (n = 1)] are shown in (G).

6 76 P. J. Meikle and others Figure 5 PG and BMP profiles for control and Niemann Pick type A/B plasma samples Plasma samples from control individuals (n = 20; A, B) and Niemann Pick type A/B patients (n = 6; C, D) were analysed for PG and BMP species by LC ESI-MS/MS as described in the Materials and methods section. The amount of each species of PG and BMP is shown as a percentage of the total with the error bars indicating one S.D. Table 2 PG and BMP concentrations in control and LSD plasma Values are means (S.D.) and are expressed as nm. G M1,G M1 gangliosidosis; MLD, metachromatic leukodystrophy; ML II/III, mucolipidosis type II/III; JNCL, juvenile neuronal ceroid lipofuscinosis; LINCL, late infantile neuronal ceroid lipofuscinosis; MSD, multiple sulfatase deficiency; M W U, value from a Mann Whitney U test for each group compared with the control group. P < Patient group n Total PG M W U Total BMP M W U BMP 18:1/18:1 M W U Control (278) 197 (80) 47 (12) Fabry (110) (60) (11) 40 Gaucher (166) (615) (362) 0 G M (14) (85) (60) 1 Krabbe (111) (123) (13) 6 MLD (66) (10) 9 38 (10) 17 MLII/III (134) (43) (18) 22 MPS I (138) (104) (86) 4 MPS II (127) (72) (63) 5 MPS IIIA (156) (39) (21) 30 MPS IV (175) (68) (19) 22 MPS VI MPS VII N P A (369) (4606) (3274) 0 N P B (378) (3605) (1963) 0 N P C (181) (77) (51) 0 Sandhoff (280) (52) (10) 1 Tay Sachs (79) (31) (16) 7 JNCL (65) 4 81 (18) 0 10 (1) 0 LINCL (34) (21) (12) 19 MSD BMP associated with the lipoprotein fraction (55 % in VLDL, 15% in LDL and 10% in HDL) and only 20% remaining in the lipoprotein-deficient plasma. No detectable BMP could be isolated in exosomes from either control individuals or Gaucher patients. DISCUSSION To investigate the effect of lysosomal storage on BMP, we sought to develop a rapid sensitive method for the specific quantification of individual BMP species in complex mixtures that also contain its structural isomer, PG. Analysis of molecular species of BMP has previously been performed by fast-atom bombardment MS after separation of the PG and BMP species by TLC [17]. More recently, Kakela et al. [18] used LC ESI- MS in negative-ion mode to quantify molecular species of BMP. This method relied on the separation of the PG and BMP by LC as they were not differentiated by MS in negative-ion mode; consequently, each analysis required greater than 30 min to complete. In an analysis of lipid remodelling in essential fatty acid-deficient mice, Duffin et al. [19] used the neutral loss of Da in positive-ion mode to identify the ammonium adducts of PG species. In the present study, we performed ESI-MS/MS analysis of both PG and BMP species in positive-ion mode, which resulted in different CID patterns and enabled quantification of both PG and BMP species without the need for prior separation. We combined this with a short (10 min) LC step, to reduce signal suppression and improve the sensitivity of the quantification.

7 The effect of lysosomal storage on bis(monoacylglycero)phosphate 77 The major species of BMP in cultured skin fibroblasts contained primarily mono-unsaturated (C 18:1 ) and polyunsaturated (C 22:5 and C 22:6 ) fatty acids with BMP C 22:6 /C 22:6 being the most abundant. These species were also identified by Kakela et al. [18] in brain tissue of patients with infantile and juvenile neuronal ceroid lipofuscinoses, both LSDs. This fatty acid composition is unusual in that while other phospholipids contain polyunsaturated species it is usually in combination with a saturated fatty acid [18]. In contrast with BMP, the major PG species contained primarily mono-unsaturated (C 18:1 and C 16:1 ) and saturated (C 18:0 and C 16:0 ) fatty acids. We observed relatively small changes in the total BMP amounts within LSD-affected cells, although in addition to BMP increases in Niemann Pick type A/B cells, reported previously [20], we also observed increases in MPS I, MPS II and Fabry disease cells. More importantly, all cell lines, with the exception of Fabry disease, showed an altered BMP profile regardless of the degree of BMP elevation. The change in the BMP profile was consistent for most LSD cell lines (Figure 4H) but was most striking in the Niemann Pick type C cell lines, which had the greatest reduction in the amount of polyunsaturated BMP species (Figure 4F). Altered lysosomal function associated with these changes in BMP composition as is evident from the reports of both glycosphingolipid and cholesterol accumulation in a range of LSD types, which include not only lipid storage disorders but also disorders such as the MPS where the primary storage is GAG [8,21]. The observed reduction in the amount of polyunsaturated BMP species is likely to alter the availability of glycolipids for interaction with saposins and the glycosidases required for their degradation [5,22] and will affect cholesterol partitioning within these membranes [7] thereby reducing their ability to remove cholesterol from the lysosome. These changes in BMP composition resulting from lysosomal storage may therefore provide the mechanistic link between the primary storage and the secondary accumulation of lipids and cholesterol, which appears to be a critical event in the pathogenesis of many LSDs. There are a number of possible mechanisms whereby lysosomal storage may induce changes in the BMP profile. Lysosomal storage may result in alterations in lipid and protein trafficking within the endosomal/lysosomal compartment, thereby influencing the availability of either enzymes or substrates involved in BMP synthesis. Alternatively, alterations in endosomal/lysosomal ph resulting from lysosomal storage may affect the specificity of the enzymes involved in BMP synthesis. In any event, this would appear to be a general result of endosomal/lysosomal storage and not dependent on the material stored. The exception to this observation is Fabry disease where the accumulation of trihexosylceramide resulted in the largest increase in total BMP (5 7-fold) but did not affect the BMP profile (Figure 4D). Interestingly, while many LSDs, including Gaucher disease, Niemann Pick disease types A and C and the MPSs, result in secondary accumulation of gangliosides in neurons, neuronal storage in Fabry disease appears limited and does not involve alterations in ganglioside expression [8]. In light of the ubiquitous nature of the alterations of BMP in LSD, we quantified PG and BMP in plasma from control and LSD-affected individuals to further investigate the nature of the changes in BMP associated with LSD. The differences in the plasma BMP species compared with the fibroblasts suggests that secretion of BMP into the circulation may be fatty aciddependent with a preference for saturated and mono-unsaturated fatty acids. The LSDs resulting in the highest elevation of plasma BMP C 18:1 /C 18:1 include Gaucher disease, Niemann Pick type A/B and Niemann Pick type C; this suggests that, at least in part, BMP is coming from macrophages, which are the primary site of storage in these disorders. Other disorders, including MPS I and MPS II, also showed a significant elevation of BMP C 18:1 /C 18:1, suggesting that other cell types and tissues may also be contributing to the plasma BMP, in these disorders. Fractionation of plasma demonstrated that approx. 40 % of BMP C 18:1 /C 18:1 was associated with lipoprotein particles in control plasma and that this was greater (80%) in Gaucher disease. This suggests that the transport of BMP out of liver cells in the form of newly synthesized VLDL is likely to be a major source of plasma BMP. In addition, the presence of BMP in HDL indicates that release of BMP from macrophages, which are known to release cholesterol and phospholipids into HDL via a number of mechanisms [23 25], may also contribute to plasma BMP. This work was supported by the Women s and Children s Hospital Research Foundation (Australia), the National Health and Medical Research Council (Australia) and Genzyme. REFERENCES 1 Postle, A. D., Heeley, E. L. and Wilton, D. C. (2001) A comparison of the molecular species compositions of mammalian lung surfactant phospholipids. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 129, Amidon, B., Brown, A. and Waite, M. (1996) Transacylase and phospholipases in the synthesis of bis(monoacylglycero)phosphate. Biochemistry 35, Heravi, J. and Waite, M. (1999) Transacylase formation of bis(monoacylglycerol)phosphate. Biochim. Biophys. Acta 1437, Kobayashi, T., Beuchat, M. H., Chevallier, J., Makino, A., Mayran, N., Escola, J. M., Lebrand, C., Cosson, P. and Gruenberg, J. (2002) Separation and characterization of late endosomal membrane domains. J. Biol. Chem. 277, Kolter, T. and Sandhoff, K. (2005) Principles of lysosomal membrane digestion: stimulation of sphingolipid degradation by sphingolipid activator proteins and anionic lysosomal lipids. Annu. Rev. Cell Dev. Biol. 21, Kobayashi, T., Beuchat, M. H., Lindsay, M., Frias, S., Palmiter, R. D., Sakuraba, H., Parton, R. G. and Gruenberg, J. (1999) Late endosomal membranes rich in lysobisphosphatidic acid regulate cholesterol transport. Nat. Cell Biol. 1, Wassall, S. R., Brzustowicz, M. R., Shaikh, S. R., Cherezov, V., Caffrey, M. and Stillwell, W. (2004) Order from disorder, corralling cholesterol with chaotic lipids. The role of polyunsaturated lipids in membrane raft formation. Chem. Phys. Lipids 132, Walkley, S. U. (2004) Secondary accumulation of gangliosides in lysosomal storage disorders. Semin. Cell Dev. Biol. 15, McGlynn, R., Dobrenis, K. and Walkley, S. U. (2004) Differential subcellular localization of cholesterol, gangliosides, and glycosaminoglycans in murine models of mucopolysaccharide storage disorders. J. Comp. Neurol. 480, Liour, S. S., Jones, M. Z., Suzuki, M., Bieberich, E. and Yu, R. K. (2001) Metabolic studies of glycosphingolipid accumulation in mucopolysaccharidosis IIID. Mol. Genet. Metab. 72, Hopwood, J. J., Muller, V., Harrison, J. R., Carey, W. F., Elliott, H., Robertson, E. F. and Pollard, A. C. (1982) Enzymatic diagnosis of the mucopolysaccharidoses: experience of 96 cases diagnosed in a five-year period. Med. J. Aust. 1, Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, Havel, R. J., Eder, H. A. and Bragdon, J. H. (1955) The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J. Clin. Invest. 34, Laulagnier, K., Motta, C., Hamdi, S., Roy, S., Fauvelle, F., Pageaux, J.-F., Kobayashi, T., Salles, J.-P., Perret, B., Bonnerot, C. and Record, M. (2004) Mast cell- and dendritic cell-derived exosomes display a specific lipid composition and an unusual membrane organization. Biochem. J. 380, Folch, J., Lees, M. and Sloane Stanley, G. H. (1957) A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226, Han, X., Yang, J., Cheng, H., Ye, H. and Gross, R. W. (2004) Toward fingerprinting cellular lipidomes directly from biological samples by two-dimensional electrospray ionization mass spectrometry. Anal. Biochem. 330, Holbrook, P. G., Pannell, L. K., Murata, Y. and Daly, J. W. (1992) Bis(monoacylglycero)phosphate from PC12 cells, a phospholipid that can comigrate with phosphatidic acid: molecular species analysis by fast atom bombardment mass spectrometry. Biochim. Biophys. Acta 1125, Kakela, R., Somerharju, P. and Tyynela, J. (2003) Analysis of phospholipid molecular species in brains from patients with infantile and juvenile neuronal-ceroid lipofuscinosis using liquid chromatography electrospray ionization mass spectrometry. J. Neurochem. 84,

8 78 P. J. Meikle and others 19 Duffin, K., Obukowicz, M., Raz, A. and Shieh, J. J. (2000) Electrospray/tandem mass spectrometry for quantitative analysis of lipid remodeling in essential fatty acid deficient mice. Anal. Biochem. 279, Besley, G. T. and Elleder, M. (1986) Enzyme activities and phospholipid storage patterns in brain and spleen samples from Niemann Pick disease variants: a comparison of neuropathic and non-neuropathic forms. J. Inherit. Metab. Dis. 9, Walkley, S. U., Thrall, M. A., Haskins, M. E., Mitchell, T. W., Wenger, D. A., Brown, D. E., Dial, S. and Seim, H. (2005) Abnormal neuronal metabolism and storage in mucopolysaccharidosis type VI (Maroteaux Lamy) disease. Neuropathol. Appl. Neurobiol. 31, Salvioli, R., Tatti, M., Ciaffoni, F. and Vaccaro, A. M. (2000) Further studies on the reconstitution of glucosylceramidase activity by Sap C and anionic phospholipids. FEBS Lett. 472, Takahashi, Y. and Smith, J. D. (1999) Cholesterol efflux to apolipoprotein AI involves endocytosis and resecretion in a calcium-dependent pathway. Proc. Natl. Acad. Sci. U.S.A. 96, Lewis, G. F. and Rader, D. J. (2005) New insights into the regulation of HDL metabolism and reverse cholesterol transport. Circ. Res. 96, Arakawa, R., Abe-Dohmae, S., Asai, M., Ito, J. I. and Yokoyama, S. (2000) Involvement of caveolin-1 in cholesterol enrichment of high density lipoprotein during its assembly by apolipoprotein and THP-1 cells. J. Lipid Res. 41, Received 1 August 2007/22 November 2007; accepted 6 December 2007 Published as BJ Immediate Publication 6 December 2007, doi: /bj