Lysophosphatidic acid signaling via LPA₆: a negative modulator of developmental oligodendrocyte maturation

Abstract

The developmental process of central nervous system (CNS) myelin sheath formation is characterized by well-coordinated cellular activities ultimately ensuring rapid and synchronized neural communication. During this process, myelinating CNS cells, namely oligodendrocytes (OLGs), undergo distinct steps of differentiation, whereby the progression of earlier maturation stages of OLGs represents a critical step toward the timely establishment of myelinated axonal circuits. Given the complexity of functional integration, it is not surprising that OLG maturation is controlled by a yet fully to be defined set of both negative and positive modulators. In this context, we provide here first evidence for a role of lysophosphatidic acid (LPA) signaling via the G protein-coupled receptor LPA₆ as a negative modulatory regulator of myelination-associated gene expression in OLGs. More specifically, cell surface accessibility of LPA₆ was found to be restricted to the earlier maturation stages of differentiating OLGs, and OLG maturation was found to occur precociously in Lpar6 knockout mice. To further substantiate these findings, a novel small molecule ligand with selectivity for preferentially LPA₆ and LPA₆ agonist characteristics was functionally characterized in vitro in primary cultures of rat OLGs and in vivo in the developing zebrafish. Utilizing this approach, a negative modulatory role of LPA₆ signaling in OLG maturation could be corroborated. During development, such a functional role of LPA₆ signaling likely serves to ensure timely coordination of circuit formation and myelination. Under pathological conditions as seen in the major human demyelinating disease multiple sclerosis (MS), however, persistent LPA₆ expression and signaling in OLGs can be seen as an inhibitor of myelin repair. Thus, it is of interest that LPA₆ protein levels appear elevated in MS brain samples, thereby suggesting that LPA₆ signaling may represent a potential new druggable pathway suitable to promote myelin repair in MS.

Graphical Abstract

INTRODUCTION

Myelination in the central nervous system (CNS) of higher vertebrates has evolved to allow rapid and efficient signal propagation within the physical constraints imposed by the skull and vertebrae (Zalc 2016; Zalc et al. 2008). During development, this process is defined by a tightly regulated differentiation program during which CNS myelinating cells, namely oligodendrocytes (OLGs), undergo a stepwise progression, from lineage committed progenitor cells and through stages of maturing OLGs, to ultimately convert into myelinating and then myelin maintaining OLGs (Sock & Wegner 2021; Elbaz & Popko 2019; Emery & Lu 2015). In this scenario, each of the OLG maturation stages is characterized by a distinct gene expression profile that is determined by both intrinsic mechanisms and extrinsic factors (Adams et al. 2021; Baydyuk et al. 2020; Liu et al. 2016; Mitew et al. 2014; Wheeler & Fuss 2016). Despite the critical importance of this process for the functional acuity of neuronal signal propagation in the CNS, the regulatory molecular pathways are still not fully understood.

Notably, our previous studies identified signaling initiated by the extracellular lipid signaling molecule lysophosphatidic acid (LPA) as a critical regulator of the transcriptional program regulating OLG maturation (Yuelling et al. 2012; Wheeler et al. 2015; Wheeler & Fuss 2016). In general, LPA exerts its biological effects via activation of a family of G protein-coupled receptors with currently six members identified in mammals, namely LPA_1-6 encoded by Lpar1-6 (Kano et al. 2021; Yung et al. 2014; Kihara et al. 2014; Hecht et al. 1996). With the exception of Lpar5, all of these receptors are, at least to some extent, expressed by cells of the mammalian OLG lineage, whereby at more mature stages the expression of Lpar1 and Lpar6 prevails (Marisca et al. 2020; Suckau et al. 2019; Wheeler et al. 2015; Marques et al. 2016; Zhang et al. 2014; Nogaroli et al. 2009; Cahoy et al. 2008; Yu et al. 2004; Stankoff et al. 2002; Weiner et al. 1998). Loss-of-function of Lpar1 has only in the Malága variant of Lpar1 null mice been described to affect CNS myelination by causing defects in protein trafficking without affecting the transcriptional regulation of OLG maturation (Garcia-Diaz et al. 2015; Anliker et al. 2013; Gennero et al. 2011; Estivill-Torrus et al. 2008). Thus, we focused our studies here on LPA₆, the most recently identified member of the LPA receptor family (Yanagida et al. 2009; Pasternack et al. 2008; Shimomura et al. 2008), for which a role in OLG maturation and CNS myelination had, to the best of our knowledge, not yet been explored. Our findings presented here uncover LPA₆ signaling as a new modulator that attenuates gene expression associated with OLG maturation, and they introduce a novel small molecule ligand with preferential selectivity for LPA₆ and functional applicability in an in vitro as well as in vivo model of OLG maturation. Interestingly, our data also provide initial evidence for dysregulated LPA₆ signaling in the major human demyelinating disease multiple sclerosis (MS), in which alterations in OLG gene expression profiles have been associated with pathology and disease progression (Schirmer et al. 2019; Jäkel et al. 2019; Falcão et al. 2018; Duncan et al. 2017; Kuhlmann et al. 2008; Chang et al. 2002). Thus, taken together, our studies identify LPA₆ signaling as a novel mechanism modulating OLG maturation and they suggest LPA₆ as a potential new druggable target for treating CNS diseases associated with myelin deficits.

MATERIALS AND METHODS

Reagents

LPA (LPA 18:1, 1-oleoyl-LPA) was purchased from Avanti Polar Lipids (Cat. No 857130, Alabaster, AL) and stored at −20 °C (10mg/ml in chloroform). (2S)-OMPT (L-sn-1-O-oleoyl-2-O-methylglyceryl-3-phosphothionate) was purchased from Echelon Biosciences (Cat. No L-9418, Salt Lake City UT) and stored at −20°C (10mM in dimethyl sulfoxide (DMSO)). GL-8-28 (Fig. 3) was synthesized by us and stored at −20°C (10mM in DMSO). Unless otherwise stated, all other reagents were purchased from Sigma-Aldrich (St. Louis, MO) or Thermo Fisher Scientific (Waltham, MA). Details to antibodies and PCR primers are shown in Tables 1 and 2.

Fig. 3.

Chemical structures of LPA (18:1; 1-oleyl-LPA), (2S)-OMPT (L-sn-1-O-oleoyl-2-O-methylglyceryl-3-phosphothionate), and GL-8-28. For the design of GL-8-28, a polar head was kept in place to ensure LPA receptor recognition, an amide bond was applied to replace the ester bond present in the natural ligand LPA to achieve higher stability in biological systems, and a long aliphatic tail was maintained.

Table 1.

Antibody RRIDs and Concentrations or Dilutions used in the assays listed.

	RRID	Assay (concentration or dilution)
Primary Antibodies
A2B5 (clone 105)	CVCL_7946	Immunopanning (15 μg/mL) Immunocytochemistry (undiluted hybridoma supernatant)
Anti-APC (CC-1)	AB_2057371	Immunohistochemistry (1:100)
Anti-CD16/CD32 (clone 93)	AB_467133	Flow cytometry (1:100)
Anti-GAPDH	AB_2107445	Western blot (1:10,000)
Anti-LPA4	AB_2340992	Flow cytometry (1:100) Western blot (1:100)
Anti-LPA6	AB_2340993	Flow cytometry (1:100) Western blot (1:100) Immunocytochemistry (1:100)
Anti-MBP (SMI99P)	AB_2314772	Immunocytochemistry (1:100)
O4 (clone O4)	CVCL_Z932	Immunocytochemistry (undiluted hybridoma supernatant)
Anti-Olig2	AB_10807410	Immunohistochemistry (1:100)
Rabbit IgG isotype control	AB_2811130	Flow cytometry (1:100)
Secondary Antibodies
Goat anti-rabbit IgG, AlexaFluor 488 conjugated	AB_143165	Flow cytometry (1:100) Immunocytochemistry (1:500) Immunohistochemistry (1:500)
Goat anti-Mouse IgG2b, AlexaFluor 568 conjugated	AB_2535780	Immunohistochemistry (1:500)
Goat anti-Mouse IgG2b, AlexaFuor 594 conjugated	AB_2535781	Immunocytochemistry (1:500)
Goat anti-Mouse IgM, AlexaFluor 633 conjugated	AB_2535715	Immunocytochemistry (1:500)
Goat anti-Mouse IgG2a, AlexaFluor 633 conjugated	AB_2535775	Immunohistochemistry (1:500)
Goat anti-Mouse IgG. IRDye 680	AB_621840	Western blot (1:10,000)
Goat anti-Rabbit IgG, IRDye 800CW conjugated	AB_10796098	Western blot (1:10,000)

Table 2.

List of primer sequences used for RT-qPCR analysis. actb2, beta actin; cldnk, claudin k; Cnp, 2′,3′ cyclic nucleotide phosphodiesterase; ef1a, elongation factor 1-alpha; Egr1, early growth response 1; Olig2, oligodendrocyte transcription factor 2; Pgk1, phosphoglycerate kinase 1; plp1b, proteolipid protein 1b; Ppia, peptidylprolyl isomerase A (cyclophilin A); Rpl13a, ribosomal protein L13a; Ugt8, UDP glycosyltransferase 8.

	Forward primer (5’-3’)	Reverse primer (5’-3’)
Mouse Genes
Olig2	ACCGTTAACACGAGGGGCAA	TTAGGAAGCGGCGCAGTACA
Cnp	ATGCCCAACAGGATGTGGTG	AGGGCTTGTCCAGGTCACTT
Ugt8	AGGAGCTCTGGGGAGATTGC	TTTGAATGGCCAAGCAGGTCA
Egr1	CCTGACCACAGAGTCCTTTTCT	AAAGTGTTGCCACTGTTGGG
Mouse Reference Genes
Ppia	GGAGACGAACCTGTAGGACG	GATGCTCTTTCCTCCTGTGC
Pgk1	ATGCAAAGACTGGCCAAGCTAC	AGCCACAGCCTCAGCATATTTC
Rpl13a	GCGCCTCAAGGTGTTGGATG	CGCCCCAGGTAAGCAAACTTTC
Zebrafish Genes
cldnk	TGGCATTTCGGCTCAAGCTCTGGA	GGTACAGACTGGGCAATGGACCTGA
plp1b	TGCCATGCCAGGGGTTGTTTGTGGA	TGGCGACCATGTAAACGAACAGGGC
Zebrafish Reference Genes
ef1a	GTACTACTCTTCTTGATGCCC	GTACAGTTCCAATACCTCCA
actb2	CCCTGTTCCAGCCATCCTT	TTGAAAGTGGTCTCGTGGATACC
rpl13a	TCTGGAGGACTGTAAGAGGTATGC	AGACGCACAATCTTGAGAGCAG

Chemical synthesis and characterization of compound GL-8-28

All reagents were purchased from Sigma-Aldrich or as otherwise stated. Melting points were obtained with a Fisher Scientific micro melting point apparatus and are uncorrected. All infrared (IR) spectra were recorded on a Nicolet Avatar 360 FT-IR Instruments. Proton (300 MHz) and Carbon-13 (75 MHz) nuclear magnetic resonance (NMR) spectra were recorded at ambient temperature with tetramethylsilane as the internal standard on either a Varian Gemini-300MHz “Tesla” spectrometer or Varian Mercury-300MHz NMR spectrometer. Gas chromatography/mass spectrometry (GC/MS) analysis was performed on a Hewlett Packard 6890 (Palo Alto, CA). Thin layer chromatography (TLC) analyses were carried out on Analtech Uniplate F254 plates. Chromatographic purification was carried out on silica gel columns (230~400 mesh, Merck). Yields were not maximized.

Compound GL-8-28 was prepared in four steps. First, 3-trimethylsilanyloxy-propylamine was prepared following a previously reported method (Mormann & Leukel 1988). The mixture of 3-amino-1-propanol (1.50 g, 20 mmol), hexamethyldisilazane (1.78 g, 11 mmol) and trimethylsilane chloride (catalytic amount) was heated at 140 °C for two hrs. After cooled down, the mixture was concentrated under vacuum to give 2.09 g light yellow oil, in 68% yield. ¹H NMR (300MHz, DMSO): δ 3.70(t, J= 7.5Hz, 2 H), 2.80(t, J= 6.6 Hz, 2 H), 1.69(m, 2 H), 0.14 (s, 9 H). Second, octadec-9(Z)-enoic acid (3-trimethylsianyloxy-propyl)-amide was prepared. The mixture of octadec-9(Z)-enoic acid (3.59 g, 12.6 mmol) and thionyl chloride (3.0 g, 25.2 mmol) in chlorobenzene (40 mL) was heated to reflux overnight. After concentrated to remove solvent and the excess thionyl chloride, the residue was dissolved in dry dichloromethane (10 mL). Then the resulting solution was added into a solution of 3-trimethylsilanyloxy-propylamine (1.16 g, 6.3 mmol) in dry dichloromethane (90 mL) at 0 °C. The mixture was stirred overnight. The dichloromethane layer was washed with water (30 mL x 3), brine and dried over sodium sulfate. After filtered and concentrated, the resulting residue was purified by silica gel column, hexane and ethyl acetate (4: 1), to give 1.4 g product in 61% yield. ¹H NMR (300MHz, CDCl₃): δ 5.76 (br, 1 H), 5.37 (m, 2 H), 4.17 (t, J= 6.0Hz, 2 H), 3.33 (t, J= 6.6 Hz, 2 H), 2.34 (t, J= 7.2Hz, 2 H), 2.19 (m, 2 H), 2.05 (m, 2 H), 1.86 (m, 2 H), 1.65 (m, 2 H), 1.29 (m, 16 H), 0.90 (t, J= 6.6Hz, 3 H), 0.13 (s, 9 H); ¹³C NMR (75MHz, CDCl₃) δ: 172.74, 129.59, 129.31, 61.23, 36.47, 35.71, 33.88, 31.49, 31.48, 29.35, 29.34, 29.27, 29.11, 29.10, 28.91, 28.90, 28.73, 28.72, 28.45, 26.80, 25.35, 24.55, 22.27, 13.71. Third, octadec-9(Z)-enoic acid (3-hydroxy-propyl)-amide was prepared. In an ice-water bath, the solution of octadecanoic acid (3-trimethylsilanyloxy-propyl)-amide (500 mg, 1.2 mmol) in ethanol (10 mL) was added 1N HCl (5 mL). The mixture was allowed to stir at room temperature for 5 hrs. Filtration gave 380 mg white powder in 93% yield. ¹H NMR (300MHz, CDCl₃): δ 6.05 (br, 1 H), 5.36 (m, 2 H), 3.63 (m, 2 H), 3.43 (q, J= 6.0 Hz, 2 H), 2.21 (m, 2 H), 2.03 (m, 4 H), 1.70 (m, 4 H), 1.30 (m, 20 H), 0.90 (t, J= 6.9Hz, 3 H); ¹³C NMR (75MHz, CDCl₃) δ: 174.18, 129.57, 129.28, 58.68, 36.29, 35.68, 31.92, 31.47, 29.34, 29.28, 29.09, 28.89, 28.88, 28.84, 28.83, 28.71, 26.79, 26.74, 25.40, 22.26, 13.70. Last, sulfuric acid mono-(3-octadec-9(Z)-enoylamino-propyl) ester (GL-8-28) was prepared based on a reported method (Pogorevc & Faber 2002). To a suspension of sodium hydride (16.3 mg, 0.67 mmol) in dimethylformamide (DMF) (2 mL) was added octadecanoic acid (3-hydroxy-propyl)-amide (105 mg, 0.31mmol). After stirred at room temperature for 1 hour, the resulting mixture was added the solution of sulfur trioxide triethylamine complex (67 mg, 0.37 mmol) in DMF (1 mL). The reaction mixture was stirred overnight. After concentrated under vacuum to remove DMF, the residue was purified by silica gel column, dichloromethane and methanol (10:1) to give 65 mg gum, in 50% yield. ¹H NMR (300MHz, CDCl₃): δ 5.37 (m, 2 H), 4.06 (t, J=6.6Hz, 2 H), 3.23 (m, 2 H), 2.21 (t, J= 7.5Hz, 2 H), 2.04 (m, 2 H), 1.87 (m, 2 H), 1.62 (m, 2 H), 1.34 (m, 20 H), 0.93 (t, J= 6.6Hz, 3 H); ¹³C NMR (75MHz, CD₃OD) δ: 174.34, 128.92, 128.81, 64.75, 46.03, 35.27, 31.14, 28.92, 28.91, 28.69, 28.53, 28.42, 28.41, 28.33, 28.26, 26.21, 25.15, 21.82, 12.57, 7.34; MS (ESI) m/z: 420.3(M+H)⁺.

Experimental Animals

Lpar₆ KO mice (Accession No. CDB0977K: http://www.clst.riken.jp/arg/mutant%20mice%20list.html)) were generated as previously described (Hata et al. 2016); these mice were housed at Akita University (Japan) under temperature-controlled conditions with ad libitum access to food and water in a 12 hr dark/light cycle, and kept on a C57BL/6N background. In total, 17 mice (males and females) were used in these studies. No animals were excluded from the analyses described here, and no exclusion criteria were pre-determined. Zebrafish embryos were obtained through natural matings, raised at 28.5°C, and staged according to morphological criteria and hrs postfertilization (hpf) (Kimmel et al. 1995). Wild-type zebrafish were of the AB or NHGRI-1 strains. Sprague Dawley female rats with early postnatal litters were obtained from Envigo/Harlan Laboratories (Indianapolis, IN). All animal studies were approved by the Institutional Animal Care and Use Committees at Virginia Commonwealth University (IACUC #AM10229 and #AM10189) and Akita University. Unless otherwise stated, n = number of animals used when multiple animals were compared. This study was not preregistered, and no blinding or randomization was performed.

Human Tissue Samples

Human white matter tissue samples from non-MS and MS donors were obtained from the Netherlands Brain Bank, Netherlands Institute for Neuroscience, Amsterdam (open access: www.brainbank.nl). All material has been collected from donors for or from whom a written informed consent for a brain autopsy and the use of the material and clinical information for research purposes had been obtained by the Netherlands Brain Bank. Based on medical history and autopsy evaluation, none of the MS individuals included in this study presented any neurodegenerative disorders other than MS. Brains were classified as MS or non-MS based on medical history and post-mortem pathological analyses. Details to the demographics of the donor patients are shown in Table 3.

Table 3.

Demographics of non-multiple sclerosis and multiple sclerosis patients. MS: Multiple Sclerosis, SPMS: Secondary Progressive Multiple Sclerosis, PPMS: Primary Progressive Multiple Sclerosis, PMI (hrs): Postmortem Interval in hours.

	MS type	Age (yrs)/Sex	PMI (hrs)
Control 1	Non-MS	90/F	7.25
Control 2	Non-MS	84/M	-
Control 3	Non-MS	51/M	7.5
Control 4	Non-MS	87/F	8
Control 5	Non-MS	86/F	13.3
MS 1	SPMS	49/M	8
MS 2	MS	44/M	10
MS 3	MS	57/F	8.75
MS 4	SPMS	75/M	7.5
MS 5	SPMS	59/F	4.75
MS 6	MS	68/F	8.25
MS 7	SPMS	45/M	7.75
MS 8	PPMS	57/F	5.75
MS 9	PPMS	74/F	5.5
MS 10	SPMS	48/F	11.75
MS 11	MS	60/F	10.5
MS 12	MS	56/F	8.25
MS 13	SPMS	78/M	8.75
MS 14	MS	53/M	10

Cell Culture

Primary OLG progenitors were isolated from postnatal day 2 (P2) rat brains by A2B5 immunopanning as described previously (Barres et al. 1992; Martinez-Lozada et al. 2014; Wheeler et al. 2015). For each individual experiment, brains from all animals (8 to 12; male and female pups) of an entire litter were pooled. For immunopanning, supernatants of A2B5 hybridoma cells (ATCC, Manassas, VA) cultured in Dulbecco’s Modified Eagle’s Medium (DMEM; Cat. No. 11995040, Gibco/Thermo Fisher Scientific, Waltham, MA) containing 10% fetal bovine serum (FBS; Cat. No. SH30071.03, HyClone, Cytvia, Marlborough, MAre) and 1% antibiotic-antimycotic (Cat. No. 15240062, Gibco/Thermo Fisher Scientific, Waltham, MA) were used directly or A2B5 antibodies were purified by ammonium sulfate precipitation (Andrew et al. 2009) and used at a concentration of 15 μg/mL as determined by ELISA (Mouse IgM ELISA Kit, Cat. No. ab133047, Abcam, Cambridge, MA). Isolated OLG progenitors were plated onto fibronectin (10 μg/ml; Cat. No 3416355MG or 341668500UG, MilliporeSigma, Burlington, MA)-coated tissue culture dishes, glass coverslips, or 96-well plates and cultured at 37°C and 5% CO₂ in serum-free proliferation medium [DMEM containing N2 supplement (Cat. No. 17502001, Thermo Fisher Scientific, Waltham, MA), 5 ng/mL human platelet-derived growth factor (PDGF, Cat No. 300-176P, Gemini Bio-Products, West Sacramento, CA) and 5 ng/mL basic fibroblast growth factor (bFGF, Cat. No. 300-113P, Gemini Bio-Products, West Sacramento, CA)] for 24 hrs, after which cells were allowed to differentiate in serum-free differentiation medium [DMEM containing N2 supplement (Thermo Fisher Scientific, Waltham, MA) and 40 ng/mL tri-iodo-thyronine (T3; Cat. No. T2877, Sigma, St Louis, MO)] over the time periods indicated. Differentiating OLGs were collected for RNA isolation, and supernatants [phenol red-free DMEM, Cat. No. 31053036, Gibco/Thermo Fisher Scientific, Waltham, MA with added L-glutamine (Cat. No 25030, Gibco/Thermo Fisher Scientific, Waltham, MA) and sodium pyruvate (Cat. No. 11360070 Gibco/Thermo Fisher Scientific, Waltham, MA)] were collected for determining cell viability using the Thermo Scientific Pierce LDH Cytotoxicity Assay Kit (Cat. No. 88954, Thermo Fisher Scientific, Waltham, MA) and a PHERAstar plate reader (BMG LABTECH Inc, Cary, NC). For all primary OLG cell culture studies, at least three independent experiments were performed, whereby an independent experiment refers to an experiment in which cells were isolated from a separate P2 rat litter at an independent time point (day) and treated separately from all other independent experiments.

For the determination of LPA receptor selectivity, the PRESTO-Tango assay was used (Kroeze et al. 2015). HTLA cells, (an HEK293 cell line stably expressing a tTA-dependent luciferase reporter and a β-arrestin2-TEV fusion gene) were kindly provided by Dr. Wesley Kroeze and maintained in DMEM supplemented with 10% FBS, 2 μg/ml puromycin (Cat. No. A1113803, Gibco/Thermo Fisher Scientific, Waltham, MA) and 100 μg/ml hygromycin B (Cat. No. 10687010, Thermo Fisher Scientific, Waltham, MA) in a humidified atmosphere at 37°C in 5% CO₂. This cell lines is not listed as commonly misidentified cell line by the International Cell Line Authentication Committee (ICLAC). TANGO-ized plasmid constructs for LPAR1, 2 and LPAR4-6 (Kroeze et al. 2015) were gifts from Dr. Bryan Roth and obtained through Addgene (LPAR1-Tango, RRID: Addgene_66418; LPAR2-Tango, RRID: Addgene_66419; LPAR4-Tango, RRID: Addgene_66420; LPAR5-Tango, RRID: Addgene_66421; LPAR6-Tango, RRID: Addgene_66422). The LPAR3-TANGO plasmid construct was designed analogous to all other TANGO-ized G protein-coupled receptor plasmid constructs and synthesized through Thermofisher Scientific’s Synthetic Biology and Genome Engineering service. Plasmid constructs were confirmed by sequencing and purified using an EndoFree plasmid maxi prep kit (Cat. No. 12362, Quiagen Inc, Valencia, CA). For transfection, HTLA cells were plated at a density of 1x10⁶ cells per well into 12-well cell culture plates and transfected the following day (day 2) with 4 μg (LPA_1-5) or 8 μg (LPA₆) of ‘TANGO-ized’ LPA receptor plasmid construct per well using Lipofectamine 2000 (Cat. No. 11668019, Thermo Fisher Scientific, Waltham, MA) in serum-free Opti-MEM medium (Cat. No. 31985070, Gibco/Thermo Fisher Scientific, Waltham, MA) without antibiotics. On day 3, cells were re-plated at 1x10⁵ cells per well into poly-L-lysine coated 96-well white, clear flat bottom microplates (Cat. No. CLS3610, Corning Life Sciences, Tewksbury, MA) in Opti-MEM medium containing 10% charcoal-stripped FBS (Cat. No. 12676029, Gibco/ThermoFisher Scientific, Waltham, MA). On day 4, 5x drug stimulation solutions were prepared in filter-sterilized assay buffer, which consisted of 20 mM HEPES in Hank’s Balanced Salt Solution (HBSS, Cat. No. 14170112, Gibco/Thermo Fisher Scientific, Waltham, MA), pH 7.4, and 20 μL were added to each well containing 80 μL of medium. Constitutive activity controls received only the assay buffer. On day 5 (after about 18 hrs of incubation), luminescence, as a measure for receptor activation, was determined using the Bright-Glo Luciferase Assay System (Cat. No. E2610, Promega, Madison, WI) and a PHERAstar plate reader (BMG LABTECH Inc, Cary, NC).

Treatment with Pharmacological Compounds

Zebrafish embryos (at 24 hpf) and primary cultures of differentiating OLGs (at 24 hrs, i.e. at the time of switch to differentiation medium) were treated with LPA, (2S)-OMPT or GL-8-28 at the concentrations indicated. LPA was dissolved in DMEM containing 0.1% fatty acid-free bovine serum albumin (BSA; Cat. No. A8806, Sigma, St. Louis MO), while (2S)-OMPT and GL-8-28 were dissolved in dimethyl sulfoxide (DMSO), resulting in a final experimental concentration of 0.1% DMSO. Vehicle treatments were used as controls.

Immunocytochemistry, Immunohistochemistry, RNAscope, and Confocal Microscopy

Differentiating OLGs, plated on fibronectin-coated coverslips, were immunostained live (O4 or A2B5 hybridoma supernatants, anti-LPA₆ antibodies) or after fixation in 4% paraformaldehyde/PBS (anti-MBP, anti-LPA₆ antibodies). In combined double staining procedures, live staining was performed first, followed by fixation. Fixed cells were permeabilized using 0.5% Triton X-100/PBS, and nonspecific binding sites were blocked by incubation in 10% FCS/DMEM (live cells) or 2% BSA/PBS (fixed cells) for 30 min at room temperature. Primary antibodies were diluted in 10% FCS/DMEM (live cells) or 1% BSA/PBS (fixed cells), and cells were incubated for 1 hr at room temperature followed by incubation with secondary antibodies diluted in 1% BSA/PBS for 30 min or 1 hr at room temperature. Nuclei were counterstained using Hoechst 33342 (Cat. No. 14533; MilliporeSigma, Burlington MA), and sections were mounted using Vectashield (Cat. No. H-1000-10, Vector Laboratories, Burlingame, CA).

Tissue sections for immunohistochemistry and RNAscope were prepared in principle as described previously (Dupree et al. 1999; Benusa et al. 2017). Briefly, mice were anesthetized by intraperitoneal injection of 0.8ml/20 g (of mouse body weight) 2-2-2 tribromoethanol (20g/ml; Cat. No. T48402, Sigma-Aldrich, St. Louis, MO) and then transcardially perfused with 4% paraformaldehyde in 0.1 M Millonig’s phosphate buffer (Karlsson & Schultz 1965), brains were removed, post-fixed for 24 hrs in perfusion fixative, cryoprotected by immersion in 30% sucrose in PBS for 48 hrs, and then embedded and frozen in Tissue-Tek O.C.T. compound (Cat. No. 4583, Sakura Finetek USA, Torrance, CA). For immunohistochemistry, serial coronal sections (40-μm) were prepared using a Leica CM 1850 cryostat (Leica Biosystems, Buffalo Grove, IL) and stored at −80°C. Tissue sections were permeabilized for 10 minutes in ice-cold acetone and blocked for 1 hr at room temperature using PBS containing 1% Triton X-100 and 5% cold fish gelatin (Cat. No. 50-259-35, Electron Microscopy Science, Hatfield, PA). Primary antibodies were diluted in blocking solution, and sections were incubated for 48 hrs at 4°C followed by incubation with secondary antibodies for 90 min at room temperature. Nuclei were counterstained using Hoechst 33342 (Cat. No. 14533; MilliporeSigma, Burlington MA), and sections were mounted using Vectashield (Cat. No. H-1000-10, Vector Laboratories, Burlingame, CA). For RNAscope, coronal sections (15-μm) were prepared using a Leica CM 1850 cryostat (Leica Biosystems, Buffalo Grove, IL), sections were mounted on Leica Bond Plus slides (Cat. No. S21.2113.A, Leica Biosystems, Deer Park, IL) and post-fixed as follows: 60°C for 30 min, 4% paraformaldehyde in PBS for 15 min at 4°C, 50% ethanol for 5 min at room temperature, 70% ethanol for 5 min at room temperature, 100% ethanol for 5 min at room temperature (twice), air dried for 5 min, and stored at −80°C. RNAScope was performed using a Leica Biosystems Bond RX automated immunohistochemistry/in situ hybridization staining system (Leica Biosystems, Deer Park, IL) located within VCU’s Tissue and Data Acquisition and Analysis Core and the following RNAscope 2.5 LS Probe probes (all from Advanced Cell Diagnostics, Inc., Newark, CA): Mm-Lpar6 (Cat. No. 318358), Mm-Olig2-C2 (Cat. No. 447098-C2), Mm-Plp1-C4 (Cat. No. 428188-C4); for fluorescent detection, RNAscope LS Fluorescent Reagent and RNAscope LS 4-Plex Ancilliary kits (Cat Nos. 322800 and 322830, Advanced Cell Diagnostics, Inc., Newark, CA) were used in combination with the following Opal fluorophores (Akoya Biosciences, Marlborough, MA): Opal 690 (Cat. No. PN FP1497001KT), Opal 620 (Cat. No. PN FP1495001KT), Opal 520 (Cat. No. PN FP1487001KT).

All confocal images were collected using a Zeiss LSM 710 or LSM 880 confocal laser scanning microscope (Carl Zeiss Microscopy, LLC, Thornwood, NY) located within VCU’s Microcopy Shared Resource. For immunostained cells, confocal z-stacks, each spanning the optical distance of the entire cell, were collected at 0.48 μm intervals using a 63x oil-immersion objective with a numerical aperture of 1.4 and the following settings: a pinhole size of one Airy unit, a dimension of 1912x1912 pixels and 4 times line averaging. For immunostained sections, confocal z-stacks, each spanning an optical distance of 15μm, were collected at 1 μm intervals using a 40x oil-immersion objective with a numerical aperture of 1.3 and the following settings: a pinhole size of one Airy unit, a dimension of 1248x1248 pixels and 4 times line averaging. For sections labeled by RNAscope, confocal z-stacks, each spanning an optical distance of 15μm, were collected at 0.5 μm intervals using a 63x oil-immersion objective with a numerical aperture of 1.4 and the following settings: a pinhole size of one Airy unit, a dimension of 1200x1200 pixels and 4 times line averaging. For image acquisition and the generation of maximum intensity projections ZEN imaging software (Carl Zeiss Microscopy, LLC, Thornwood, NY; RRID:SCR_013672) was used.

Western Blot Analysis

For Western blot analysis, human tissue samples were homogenized in lysis buffer (10 μL per mg tissue; phosphate-buffered saline (PBS), 2 mM EDTA) containing 1x Halt protease and phosphatase inhibitor cocktail (Cat. No. 78430, Thermo Fisher Scientific, Waltham, MA). After centrifugation at 15,000 rpm for 15 min at 4°C, protein concentrations of the supernatants were determined using the BCA protein assay kit (Cat. No. 23225, Pierce/Thermo Fisher Scientific, Waltham, MA). Samples were heat-denatured (90°C for 5 min) in Laemmli sample buffer (Cat. No. S3401, MilliporeSigma, Burlington, MA), and equal amounts of denatured protein (60 μg) were separated by electrophoresis through 4-20% gradient sodium dodecyl sulfate (SDS)-polyacrylamide gels (Cat. No. 4561094, Bio-Rad Laboratories, Hercules, CA). Separated proteins were subsequently electroblotted onto Immobilon-P polyvinylidene difluoride (PVDF) membranes (Cat. No. IPVH00010, MilliporeSigma, Burlington, MA). Total protein levels were determined using Revert 700 Total Protein Stain (Cat. No. 926-11011, LI-COR Biosciences, Lincoln, NE) and membranes were subsequently incubated in blocking buffer (Cat. No. 927-70001, LI-COR Biosciences, Lincoln, NE) for 1 hr at room temperature prior to incubation with primary antibodies diluted in blocking buffer (48 hrs at 4°C). Bound primary antibodies were detected using IRDye 680RD- or 800CW-conjugated secondary antibodies (LI-COR Biosciences, Lincoln, NE) diluted in blocking buffer (3 hrs at room temperature). For quantification, membranes were imaged using an Odyssey infrared imaging system and analyzed using Image Studio and Empiria Studio software packages (LI-COR Biosciences, Lincoln, NE).

Flow Cytometry

Differentiating OLGs were collected after 48 hrs culture in serum-free differentiation medium using Accutase cell detachment solution (Cat. No. A6964, MilliporeSigma, Burlington, MA). Cells were counted, and 10⁶ cells per sample were immunostained for analysis. Cells-only controls were placed immediately into flow cytometry buffer [0.5% bovine serum albumin (BSA) and 2mM EDTA in phosphate-buffered saline (PBS)]. All immunostaining procedures were performed on ice in PBS containing 0.5% BSA and 2mM EDTA. Fc receptors were blocked by incubation with anti-mouse CD16/32 for 15 minutes, and cells were incubated for 30 min first with anti-LPA receptor antibodies or their respective isotype controls and then with AlexaFluor 488-conjugated secondary antibodies. Cells were washed twice and resuspended in flow cytometry buffer. Immediately prior to flow analysis, 7-AAD viability stain (Cat. No. A1310, Invitrogen/eBioscience/Thermo Fisher Scientific, Waltham, MA) was added, and samples were run on a LSRFortessa-X20 flow cytometer (BD Biosciences, San Jose, CA) at room temperature using a 488 nm laser (>20mW power, Coherent solid-state) and 530/30 nm and 647/20 nm bandpass filters. Settings were carefully determined empirically and exactly reproduced in each experiment. Gates were demarcated to count 488-positive (7-AAD-negative) cells up to 10,000 events. FACSDIVA software (BD Biosciences, San Jose, CA) was used for acquisition, and data were analyzed using FCS Express Flow Cytometry software (DeNovo Software, Pasadena, CA).

RNA Isolation and RT-qPCR

RNA was purified using RNeasy Mini (cultured OLGs) or Midi (zebrafish embryos) Kits (Cat Nos. 74104 and 75144, Qiagen LLC, Germantown, MD). For the isolation of RNA from cultured rodent OLGs, cells were plated on fibronectin-coated 12 well plates (1x10⁶ cells per well) and collected in RTL lysis buffer (Qiagen LLC, Germantown, MD) containing 1% 2-mercaptoethanol. For RNA isolation from zebrafish embryos at 48 hpf, embryos were anesthetized using 0.015% tricaine methanesulfonate (Syncaine, Syndel/Western Chemical, Ferndale, WA), chorions were removed using 2mg/mL Pronase (Cat. No. 11459643001, Sigma, St Louis, MO), and pools of about 20 embryos were homogenized in RTL lysis buffer (included in RNeasy kits, Qiagen LLC, Germantown, MD) containing 1% 2-mercaptoethanol. All RNA samples were treated with DNase (DNA-Free Kit; Cat. No. AM1906, Applied Biosystems/Thermo Fisher Scientific, Waltham, MA). RNA concentrations and purity were determined using a 2100 Bioanalyzer and RNA 6000 Pico kits (Cat. No. 5067-1513, Agilent, Santa Clara, CA); samples with an RNA integrity number above 7 were used for further analysis. Oligo-dT- and random hexamer-primed cDNAs were synthesized from 50 ng to 1 μg of RNA using Omniscript or Sensiscript Reverse Transcription Kits (Cat. Nos. 205113 and 205213, Qiagen LLC, Germantown, MD) according to the guidelines of the manufacturer. RNA samples were normalized to the same approximate concentration, and the same amount of RNA was used for all conditions of an individual independent RT-qPCR experiment.

For all RT-qPCR experiments the Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) guidelines were followed (Bustin et al. 2009). Briefly, RT-qPCR primers (Table 3) were designed and in silico tested for specificity using the National Center for Biotechnology Information’s basic local alignment search tool (Primer-BLAST) (Ye et al. 2012). All primers were designed to amplify all known splice variants, and for all primer pairs melting curves were used to ensure specificity. cDNA reactions without reverse transcriptase were performed for all samples to ensure no-reverse-transcriptase quantitation cycle (C_q) numbers of at least five cycles below the lowest C_q for any of the experimental samples. RT-qPCR reactions with three technical replicates per sample were performed on a CFX96 real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA) using the iTaq Universal SYBR Green Supermix (Cat. No. 1725121, Bio-Rad Laboratories, Hercules, CA). PCR conditions were as follows: 95°C for 3 min, followed by 40 cycles of 95°C for 15 s, 58°C for 30 s, and 95°C for 10 s. Expression levels were determined using the ΔΔCT method relative to the geometric mean of the three reference genes (Livak & Schmittgen 2001).

Homology Modeling and Molecular Docking

The published zebrafish LPA₆ crystal structure was utilized as the template protein to build the 3D conformations of human LPA₆ (Taniguchi et al. 2017). The amino acid sequence for human LPA₆ was downloaded from UniProtKB (primary accession number P43657). The sequence alignment between target (human LPA₆) and template (zebrafish LPA₆) sequences were performed by the program ClustalX 2.1 (Larkin et al. 2007) with default parameters. Based on the sequence alignments, 100 3D models of LPA₆ were built by MODELLER 9.19. The 3D models with the highest DOPE assessment score were selected as the optimal homology model for human LPA₆ (Shen & Sali 2006).

GL-8-28 was sketched in Sybyl-X 2.0 and further energy minimized to a gradient of 0.05 with Gasteiger-Hückel charges assigned under the Tripos force field (TAFF) (Ballesteros & Weinstein 1995). The molecular docking study was performed by GOLD 2020 (Cambridge Crystallographic Data Centre, CCDC, Cambridge, UK) to obtain the ligand-receptor complex. The putative docking sites of the human LPA₆ model were defined by 10 Å around the α-carbon atom of V^5.39 of LPA₆, which formed directly hydrophobic interaction with the LPA species in the crystal structure of the zebrafish LPA₆ (Taniguchi et al. 2017). Except for the above parameters, molecular docking studies were conducted with standard default settings. Fifty docking solutions were generated in the respective optimal homology model of LPA₆. The docking solutions of GL-8-28 with the highest CHEM-PLP score within the fifty docking solutions were chosen as the optimal docking poses in the homology model of human LPA₆.

Statistical Analysis

GraphPad Prism (GraphPad Software Inc., San Diego, CA; RRID:SCR_002798) was used for all statistical analyses. Prior sample calculations were not performed; sample size was estimated based on previous studies of a similar nature (Dennis et al. 2008; Lafrenaye & Fuss 2011; Martinez-Lozada et al. 2014; Thomason et al. 2022; Waggener et al. 2013; Wheeler et al. 2015). Data were assessed for normality using the Shapiro-Wilk normality test prior to analysis. Data compared with a set control value lacking variability were analyzed using the one-sample t test (Dalgaard 2008; Skokal & Rohlf 1995) and presented in graphs showing individual data points plus means with SEM. For comparing multiple groups one-way ANOVA or in case of multiple data points per animal nested one-way ANOVA was used. PRESTO-Tango assays were analyzed using dose-response curves, and data are presented as means ± SEMs for each agonist concentration plus dose-response curves. p≥0.05 was used as threshold for significant for all statistical tests used. EC₅₀ values were calculated from the dose-response curves, and the percent maximum activity was calculated using the following formula: % maximum activity = (maximum response of GL-8-28 − constitutive activity)/(maximum response of full agonist − constitutive activity) x 100.

RESULTS

LPA₆ is expressed by OLG lineage cells and uniquely localized at the surface of earlier maturation stages

In order to gain a deeper understanding of the contribution of individual LPA receptors to the regulation of OLG maturation, we focused our studies here on the LPA receptor LPA₆, which has previously been shown to be expressed by OLG lineage cells (Suckau et al. 2019; Wheeler et al. 2015). To better define the expression and surface localization of LPA₆, enriched cultures of rat brain-derived differentiating OLGs were immunolabeled using anti- LPA₆ antibodies recognizing the second extracellular loop region of LPA₆ in combination with the following antibodies marking selective OLG maturation stages: A2B5 (OLG progenitor cells; (Duchala et al. 1995), O4 (maturing OLGs(Duchala et al. 1995), anti-MBP (mature OLGs(Dubois-Dalcq et al. 1986). As shown in Fig. 1A, upon fixation and permeabilization of cultured cells, LPA₆ protein could be detected at all stages of the OLG lineage. Similarly, in the 3-week-old mouse corpus callosum, Lpar6 mRNA was detected in all cells expressing Olig2, a marker for all stages of the OLG lineage in rodents (Zhou et al. 2000; Lu et al. 2000; Wegner 2001), including later stages of maturing OLGs as identified by the presence of Plp1 mRNA (Dubois-Dalcq et al. 1986; Duchala et al. 1995) (Fig. 1C). These findings are consistent with previously published mRNA profiling data (Suckau et al. 2019; Marques et al. 2016; Zhang et al. 2014). Interestingly, however, when using a live staining protocol, surface localization of LPA₆ was found to be restricted to earlier maturation stages of differentiating OLGs (Fig. 1B). Taken together, these data indicate that while LPA₆ is expressed throughout the OLG lineage, its functional roles are likely restricted to earlier maturation stages of differentiating OLGs.

Fig. 1.

LPA₆ is expressed throughout the OLG lineage but surface localization is restricted to earlier maturation stages of differentiating OLGs. A. and B. Representative confocal images of cultured rat brain-derived OLG lineage cells immunolabeled with antibodies detecting an extracellular surface epitope of LPA₆ in combination with the following antibodies marking selective maturation stages: A2B5 (OLG progenitor cells) (Duchala et al. 1995), O4 (maturing OLGs) (Duchala et al. 1995), anti-MBP (mature OLGs) (Dubois-Dalcq et al. 1986). Cells were labeled after fixation and permeabilization (A) or live (B). Scale bar: 10 μm. C. Representative confocal images of the 3-week-old mouse corpus callosum, triple labeled using RNAScope for mRNAs encoding Lpar6, Olig2 to mark all OLG lineage cells (Zhou et al. 2000; Lu et al. 2000; Wegner 2001) and Plp1 to mark later OLG maturation stages (Dubois-Dalcq et al. 1986; Duchala et al. 1995). Stars indicate nuclei of Olig2-positive cells. Scale bar: 20 μm.

During development, LPA₆ functions as a negative modulator of OLG maturation

After having established that LPA₆ is expressed by OLGs, we next assessed potential functional effects of a loss of LPA₆ on OLG maturation via the use of Wt and Lpar6 KO mice (Hata et al. 2016). More specifically, OLG maturation was analyzed in the corpus callosum and cortex at 2 and 3 weeks of age, a developmental time point that coincides with a rapid phase of myelination (Bergles & Richardson 2015; Sturrock 1980) associated with active proliferation of OLG lineage cells (Spitzer et al. 2019; Young et al. 2013; Sturrock 1983). As shown in Fig. 2B, based on immunostaining with anti-Olig2 antibodies, known to mark all OLG lineage cells in rodents (Wegner 2001; Zhou et al. 2000; Lu et al. 2000), no changes in the total number of OLG lineage cells were noted. Co-immunostaining with CC1 antibodies, labeling maturing OLGs (Bin et al. 2016; Fuss et al. 2000), on the other hand, revealed an increased percentage of maturing OLG lineage cells at 3 but not 2 weeks of age (Fig. 2C, D). Thus, there is a precocious appearance of maturing OLGs during the rapid phase of myelination in the corpus callosum and cortex of Lpar6 KO mice. This alteration appears largely due to an acceleration of OLG maturation rather than an overall increase in the number of OLG lineage cells, and it provides first evidence for a role of LPA₆ in modulating the timing of OLG maturation in a negative regulatory, or restraining, fashion.

Fig. 2.

Developmental maturation of OLGs is accelerated in Lpar6 KO mice. A. Coronal brain section marking brain regions (Image credit: Allen Mouse Brain Atlas, Allen Institute). B. Graph depicting the number of Olig2-positive OLG lineage cells per mm² in sections of 15 μm depth from 3-week-old Wt and Lpar6 KO mice as determined by immunostaining and subsequent image analysis. C. Representative confocal images of the corpus callosum in 3-week-old Wt and Lpar6 KO mice double-immunostained with an anti-Olig2 antibody to mark all OLG lineage cells (Zhou et al. 2000; Lu et al. 2000; Wegner 2001) and with the CC1 antibody to identify mature OLGs (Bin et al. 2016; Fuss et al. 2000); arrows mark CC1-Olig2 double-positive cells. Scale bar: 50 μm. D. Graphs depicting the percentage of mature CC1-Olig2 double-positive OLGs in the midcaudal corpus callosum (left) and motor cortex (right) of Wt and Lpar6 KO mice at 2 weeks and 3 weeks of age as determined by immunostaining and subsequent image analysis. D. Dots in all graphs represent 3 fields of view (40x objective) from non-adjacent sections for each individual animal. ***p ≤ 0.001, **p ≤ 0.01 ^ns p>0.05, 3 fields of view from 5 (3-week-old Wt mice) or 4 (2-week-old Wt and Lpar6 KO mice, 3 week Lpar6 KO mice) animals, nested two-tailed t-Test, t=0.8708 and df=7 (B, corpus callosum), t=0.6845 and df=7 (B, cortex), t=0.8294 and df=13 (D, two weeks), t=4.517 and df=25 (D, three weeks), t=0.5011 and df=5 (E, two weeks), t=3.680 and df=25 (three weeks).

The small molecule GL-8-28 represents a dual LPA_4/6 ligand with preferential agonist activity at LPA₆

While the data thus far support a negative modulatory role of LPA₆ in the timing of OLG maturation, they do not directly address LPA₆ signaling in OLGs and the potential of pharmacological accessibility for LPA₆ initiated signaling. Hence, we turned our attention to a library of small molecules that had been designed and prepared by us as potential LPA receptor ligands. We first used the PRESTO-Tango GPCR assay system (Kroeze et al. 2015) to examine agonist activity at the six known LPA receptors (LPA_1-6). For LPA_1-5, the well-established endogenous agonist LPA (18:1) served as positive control, while, due to the much lower affinity of LPA (18:1) for LPA₆, the LPA phosphorothioate analog (2S)-OMPT was used for assaying LPA₆ ligands (Jiang et al. 2013; Yanagida et al. 2009; Inoue et al. 2011; Inoue et al. 2012). As shown in Fig. 4, one of the compounds from our library, designated here as GL-8-28 (Fig. 3), was found to partially activate both LPA₄ and LPA₆. Notably, based on estimated EC₅₀ calculations, the potency of GL-8-28 appears to be approximately five times higher for LPA₆ compared to LPA_4. Neither of the remaining known LPA receptors were activated by GL-8-28 above a level of 20% of control (Fig. 4B), thus defining GL-8-28 as a dual LPA_4/6 partial agonist with reasonable selectivity to LPA₆ over LPA₄ as well as other LPA receptors. This observation was further corroborated by molecular modeling and ligand-receptor docking using the published zebrafish LPA₆ crystal structure (Taniguchi et al. 2017) as a template to build the 3D conformations of human LPA₆ bound to the ligand GL-8-28. As shown in Fig. 5, the polar head group of GL-8-28 interacts with the conserved positively charged residues K^1.31, R^2.60, R^6.62, and R^7.32, previously shown to be particularly important for ligand recognition and receptor activation (Taniguchi et al. 2017). Agonist binding to the LPA₆ receptor has been proposed to induce a conformational inward shift of the transmembrane helices 6 (TM6) and 7 (TM7) resulting in direct interaction with all of the above conserved positively charged residues with the ligand headgroup; this shift is not depicted in Fig. 5. The long aliphatic chain of GL-8-28 is positioned within a hydrophobic environment formed at the TM4-TM5 interfaces; this hydrophobic cleft has been implicated in binding of the acyl chain of LPA (Taniguchi et al. 2017). Thus, GL-8-28 displays ligand-receptor interactions that involve binding pocket residues predicted to be crucial for binding of the endogenous agonist LPA and subsequent receptor activation.

Fig. 4.

GL-8-28 functions as an LPA_4/6 dual ligand with preferential LPA₆ agonist activity. A. Graphs depicting dose response curves as determined by the PRESTO-Tango GPCR assay system (Kroeze et al. 2015). Note, that LPA (18:1) was used as previously established agonist for LPA_1-5, while (2S)-OMPT was used for LPA₆ (Jiang et al. 2013). Means ± SEM are shown for at least five independent experiments performed in triplicates. Significant agonist activity for GL-8-28 is defined as above 20% of control (see B). Note that the estimated potency for GL-8-28 is much higher for LPA₆ compared to LPA₄. B. Graph depicting the percentage of GL-8-28 activity compared to control. Data points represent independent experiments (n=5 (LPA₅), 6 (LPA₂), 7 (LPA₃), 9 (LPA_1,4), 10 (LPA₆)) performed in triplicates. *p≤0.05, **p≤0.01, one-sample two tailed t-test, theoretical mean set to 20% for LPA_{1,2,4 and 6}; to 2% for LPA₃, t=0.3839 and df=8 (LPA₁), t=0.1326 and df=5 (LPA₂), t=0.2239 and df=6 (LPA₃), t=3.515 and df=8 (LPA₄), t=0.5633 and df=4 (LPA₅), t=3.230 and df=9 (LPA₆).

Fig. 5.

GL-8-28 displays ligand-receptor interactions that involve binding pocket residues predicted to be crucial for binding of the endogenous agonist LPA and receptor activation. The binding pocket of LPA₆ is shown as cartoon model (lavender) and the ligand GL-8-28 is depicted as stick model (green carbon atoms). Key amino acid residues binding with GL-8-28 are shown as sphere model. Arrowheads (gray) point to conserved positively charged residues implicated in head group binding and receptor activation (Taniguchi et al. 2017)}; positively charged residues (pink), hydrophobic residues (orange), oxygen atoms (red), nitrogen atoms (blue), sulfonate atoms (yellow).

Pharmacological activation of LPA₆ signaling attenuates OLG maturation

In order to directly assess the effect of LPA₆ activation on OLG maturation, we treated primary cultures of differentiating rat brain-derived OLGs with the LPA₆ preferential agonist GL-8-28. In agreement with our live staining experiments (see Fig. 1) surface localization was detected by flow cytometry on a subpopulation of OLG lineage cells (approximately 43%; see Fig. 6 A’). In contrast, only very few cells (approximately 6%) displayed detectable levels of surface localization for LPA₄ (Fig. 6 A’). Thus, any effects exerted by GL-8-28 are mediated primarily by its LPA₆ agonist activity and are directed at earlier maturation stages of differentiating OLGs (see Fig. 1). As shown in Fig. 6D, in the presence of the highest concentration tested, i.e. 10 μM, GL-8-28 treatment led to significantly reduced mRNA levels for OLG expressed genes associated with OLG maturation and CNS myelination, namely 2,′3′-cyclic nucleotide 3′-phosphodiesterase (Cnp) (Kurihara et al. 1992; Gravel et al. 1998) and UDP glycosyltransferase 8 (Ugt8) (Gard & Pfeiffer 1990; Bosio et al. 1996). This outcome was not found to be associated with a cytotoxic effect (Fig. 6B) or a change in the expression of Olig2 (Fig. 6C), a transcription factor expressed throughout the OLG lineage in rodents (Zhou et al. 2000; Lu et al. 2000; Wegner 2001). Additionally, at the 10 μM concentration, GL-8-28 significantly increased the expression of the transcriptional repressor Egr1 (Swiss et al. 2011; Sock et al. 1997), a known inhibitor of the gene expression profile associated with OLG maturation and CNS myelination. In this context, it is important to note that downregulated Egr1 expression concomitant with OLG differentiation was previously shown to occur upon thyroid hormone application (Swiss et al., 2011), thus demonstrating that in differentiating OLGs upregulation of Egr1 is not generally associated with a stimulus evoked response as for example seen in the context of neuronal activity and plasticity (Duclot & Kabbaj 2017).

Fig. 6.

The preferential LPA₆ agonist GL-8-28 attenuates maturation in primary cultures of rat OLGs. A. Scheme depicting the experimental paradigm; the inset in A’ represents the percent of LPA₄ (left) and LPA₆ (right) positive OLGs at 48 h, as determined by flow cytometry. B-E. Graphs depicting the extracellular LDH levels (B) and mRNA levels as determined by RT-qPCR analysis (C-E). Levels for control (vehicle-treated) cells were set to 100% (horizontal dotted line) and relative levels were calculated accordingly. Individual data points represent independent experiments, means ± SEM are shown as horizontal lines with error bars. *p≤0.05 and ****p≤0.0001, one sample two-tailed t-test, B: n=6, t=1.897 and df=5; C: 1 nM: n=4, p=0.7220, t=0.3909, df=3, 10nM: n=6, p=0.0523, t=2.534, df=5, 100nM: n=7, p=0.6686, t=0.4498, df=6, 1μM: n=5, p=0.0508, t=2.760, df=4, 10μM: n=5, p=0.2441, t=1.365, df=4; D: 1 nM: n=4, p=0.0542, t=3,079, df=3 (CNP); n=5, p=0.2653, t=1.294, df=4 (Ugt8) 10nM: n=7, p=0.0606, t=2.306, df=6 (CNP); n=6, p=0.1412, t=1.746, df=5 (Ugt8), 100nM: n=7, p=0.5995, t=0.5543, df=6 (CNP); n=6, p=0.2538, t=1.289, df=5 (Ugt9), 1μM:n=5, p=0.6630, t=0.4697, df=4 (CNP); n=5, p=0.4583, t=0.8188, df=4 (Ugt8), 10μM:n=5, p˂0.0001, 4t=21.56, df=4 (CNP), n=5, p˂0.0001, t=18.93, df=4, E: 1 nM: n=5, p=0.5163, t=0.7111, df=4, 10nM: n=7, p=0.4297, t=0.8465, df=6, 100nM: n=7, p=0.8491, t=0.1986, df=6, 1μM: n=5, p=0.4981, t=0.7442, df=4, 10μM: n=5, p=0.0300, t=3.299, df=4.

In an attempt to extend these findings into an in vivo system and to gain initial insight into the in vivo use of GL-8-28, we took advantage of the following characteristics of the developing zebrafish. First, in this model, pharmacological agents can be taken up from the surrounding water source via passage through the skin (Rombough 2002). Second, a lack of a functional blood-brain-barrier (BBB) prior to 3 days of age (Li et al. 2017; Fleming et al. 2013) allows diffusion into the developing CNS prior to completion of myelination (Buckley et al. 2008; Jung et al. 2010; Yoshida & Macklin 2005; Brosamle & Halpern 2002), for which key regulatory mechanisms are highly conserved during the development of mammals and zebrafish (Ackerman & Monk 2016; Lyons & Talbot 2014; Preston & Macklin 2015). In this context, the expression of proteolipid protein, present in both maturing rodent (Plp1) and zebrafish (plp1a/plp1b) OLGs (Siems et al. 2021), as well as myelin protein zero (mpz) and claudin K (cldnk), found more specifically in maturing zebrafish OLGs, has been reported to significantly increase between 24 and 48 hrs post fertilization (hpf) in the CNS of the developing zebrafish embryo (Preston et al. 2019; Wheeler et al. 2015; Munzel et al. 2012; Takada & Appel 2010; Schweitzer et al. 2006; Yoshida & Macklin 2005; Brosamle & Halpern 2002). In the presence of GL-8-28, mRNA levels for all these three zebrafish genes known to be associated with OLG maturation were significantly reduced (Fig. 7). At the same time, gross anatomical features remained unchanged up to a concentration of 1 μM (Fig. 7B). Importantly, based on recently published gene expression profiling data (Marisca et al. 2020), LPA receptor expression in zebrafish OLGs appears to be limited to differentiating rather than progenitor stages of the lineage and restricted to the receptors lpar1 and lpar6a. Furthermore, equivalents of several residues found to be important for ligand binding to human LPA₆ were also found to affect binding in models based on the crystallized structure of the zebrafish receptor, thus providing confidence of equivalent binding characteristics in all species analyzed here (Taniguchi et al. 2017; Raza et al. 2014). Thus, the OLG maturation attenuating effects of GL-8-28 in the developing zebrafish are likely to be mediated, at least in part, via activation of LPA₆ in differentiating OLGs. However, these effects of GL-8-28 were observed at much lower concentrations as seen in the rat primary cultures, possibly due to receptor desensitization in the culture system and/or cumulative effects in the in vivo system. Nevertheless, the above data demonstrate an important functional role of LPA₆ signaling in regulating the timely appearance of mature OLGs in both rodent cultures of differentiating OLG and the developing zebrafish.

Fig. 7.

The preferential LPA₆ agonist GL-8-28 attenuates OLG maturation in the developing zebrafish. A. Scheme depicting the experimental paradigm. B. Representative images of control and Gl-8-28 (1μM)-treated zebrafish embryos at 48 hpf (hours post fertilization). C-E. Graphs depicting mRNA levels as determined by RT-qPCR analysis. Levels for control (vehicle-treated) zebrafish embryos were set to 100% (horizontal dotted line) and relative levels were calculated accordingly. Individual data points represent independent experiments, means ± SEM are shown as horizontal lines with error bars. plp1b: proteolipid protein, mpz: myelin protein zero, cldnk: claudin K. *p≤0.05 and **p≤0.001, one sample two-tailed t-test, C: 100pM: n=4, p=0.0084, t=6.226, df=3, 1nM: n=4, p=0.0033, t=8.621, df=3, 10nM: n=4, p=0.0087, t=6.142, df=3, 100nM: n=4, p=0.0432, t=3.377, df=3, 1μM: n=4, p=0.4916, t=0.7814, df=3 D: n=4, p=0.0004, t=17.23, df=3, 1nM: n=4, t=9.185, df=3, 10nM: n=4, p=0.0085, t=6.198, df=3, 100nM: n=4, p=0.0067, t=6.738, df=3, 1μM: n=4, p=0.9009, t=0.1353, df=3, E: 100pM: n=5, p=0.0051, t=5.573, df=4, 1nM: n=5, p=0.0109, t=4.491, df=4, 10nM: n=5, p=0.0131, t=4.258, df=4, 100nM: n=5, p=0.1138, t=0.1138, t=2.017, df=4, 1μM: n=5,p=0.1684, t=1.679, df=4.

LPA₆ protein levels appear elevated in MS white matter lesions

In the context of development, the maturation attenuating functional role of LPA₆ likely serves to ensure timely coordination of circuit formation and myelination (Fletcher et al. 2021). Under pathological conditions, however, persistent LPA₆ expression and signaling could be seen as an inhibitor of efficient OLG maturation and myelin repair. In this regard, it is of particular interest that our analysis of human MS lesion samples (Fig. 8) revealed, on average, elevated LPA₆ protein levels. Interestingly, the variance seen within the MS sample pool was significantly broader than the one for the control sample pool, an observation that is consistent with the increasingly recognized heterogeneity in MS pathology between patients (Smets et al. 2021; Patrikios et al. 2006; Lucchinetti et al. 2000). Curiously, in addition to a somewhat diffuse band with the expected apparent molecular weight of approximately 50 kDa, a sharp lower molecular weight band of approximately 40 kDa was observed in some of the MS samples, possibly reflecting differences in N-glycosylation (Suckau et al. 2019). Taken together, these data suggest that LPA₆ protein levels may be increased in at lest some types of MS white matter lesions, an observation that is consistent with recent transcriptome profiling data (Elkjaer et al. 2019; Falcão et al. 2018).

Fig. 8.

LPA₆ protein levels are increased in multiple sclerosis (MS) white matter lesions. A. Representative Western blot depicting LPA₆ and total protein between molecular weight markers 70 kDa (top) and 38 kDa (bottom) in control (CTRL) white matter and multiple sclerosis (MS) white matter lesions; total protein per lane was used for normalization. B. Graph depicting the percentage of LPA₆ protein levels in control (CTRL) white matter and MS white matter lesions (MS) (left y-axis) and the mean difference (right y-axis) as determined by Western blot analysis. Tissues were obtained from the Netherlands Brain Bank (for details see Table 2). The mean control value, normalized to total protein, was set to 100% and all other values were adjusted accordingly (average values from two independent Western blots are shown). ***p≤0.001 Welch’s two-tailed t-test (^Variancep≤0.05), Welch-corrected t=3.997, df=17.00. CI, confidence interval.

DISCUSSION

In this study we identified signaling via the LPA receptor LPA₆ as a novel modulator of the transcriptional regulation of OLG maturation. More specifically, our data revealed a precocious developmental appearance of OLGs expressing a marker for more mature stages of the lineage in Lpar6 KO mice. Thus, initial evidence was obtained that activation of LPA₆ signaling in OLGs negatively modulates the transcriptional program associated with OLG maturation. This interpretation could be substantiated via the use of a novel LPA₆ selective ligand with agonist activity, here referred to as GL-8-28; both in vitro in cultures of differentiating OLGs and in vivo in the developing zebrafish, treatment with GL-8-28 was found to impede the expression of genes known to be associated with OLG maturation. During development, such a functional role for LPA₆ signaling is likely transient and serves to ensure timely coordination of circuit formation and myelination (Fletcher et al. 2021). This point of view may be supported by the apparent lack of an overt myelination phenotype in Lpar6 KO mice (Hata et al. 2016). However, future studies are needed to further substantiate such a proposed developmentally transient functional role of LPA₆ signaling in maturing OLGs. Under pathological conditions accompanying white matter lesions in MS, LPA₆ signaling appears to be persistent, at least in some patients. Hence, our studies reveal a novel molecular mechanism modulating OLG maturation, and they uncover a novel and conceivably druggable signaling pathway with the potential for future developments toward innovative therapeutic interventions stimulating myelin repair.

LPA signaling plays diverse roles in the regulation of OLG maturation and CNS myelination

A key step in the biosynthetic pathway(s) generating extracellular LPA is represented by the conversion from lysophosphatidylcholine (LPC) via the enzymatic activity of secreted autotaxin, also known as extracellular nucleotide pyrophosphatase-phosphodiesterase 2 (ENPP2) or lysophospholipase D (lysoPLD) (Aoki et al. 2008; Tanaka et al. 2006; van Meeteren et al. 2006; Gijsbers et al. 2003; Tokumura et al. 2002; Umezu-Goto et al. 2002). In our previous studies, we had identified OLG expressed autotaxin as a positive regulator of OLG maturation (Wheeler et al. 2015; Yuelling et al. 2012; Dennis et al. 2012; Nogaroli et al. 2009; Dennis et al. 2008; Fox et al. 2004; Fox et al. 2003). Importantly, autotaxin was found to stimulate the transcriptional expression profile leading to OLG differentiation via its enzymatic activity generating LPA (Wheeler et al. 2016; Wheeler et al. 2015; Yuelling et al. 2012). In addition, autotaxin-dependent activation of specifically LPA₆ has recently been proposed to exert functional effects on cells other than OLGs (Matas-Rico et al. 2021; Okasato et al. 2021; Masago et al. 2018). Thus, a maturation attenuating role of LPA₆ signaling on OLGs, as described here, may appear puzzling. However, despite the simple basic structure of LPA, i.e. a glycerol backbone connected to a phosphate head group at the sn-3 position and a fatty acid chain linked to the sn-1 or sn-2 position, variations in fatty acid side chain length, saturation and backbone position generate diversity that impacts receptor selectivity and downstream signaling outcomes (Aikawa et al. 2015; Hayashi et al. 2001; Okudaira et al. 2014; Yung et al. 2014; Okudaira et al. 2010; Yoshida et al. 2003; Baker et al. 2001; Bandoh et al. 2000). Furthermore, structure-function analyses suggest a highly localized delivery of LPA to the respective receptor to be activated, thereby refining the selectivity for LPA receptor subtype activation (Taniguchi et al. 2017; Hausmann et al. 2013; Houben et al. 2013; Fulkerson et al. 2011; Moolenaar & Perrakis 2011; Hausmann et al. 2011; Nishimasu et al. 2011; Kanda et al. 2008). Regarding LPA₆, preference for 2-acyl rather than 1-acyl LPA species confer somewhat unique characteristics within the LPA receptor family (Aikawa et al. 2015; Nishimasu et al. 2011; Inoue et al. 2011; Inoue et al. 2012; Yanagida et al. 2009; Morishige et al. 2007; Tokumura et al. 2002). Thus, fine-tuned regulation of the LPA receptor profile present on OLG surfaces in combination with the availability and local delivery of specific LPA species may determine the balance between LPA-mediated maturation promoting and attenuating outcomes. Interestingly, a counter-regulatory function of LPA₆, compared to one or more of the remaining known LPA receptors, has also been described for cell types other than OLGs (Matas-Rico et al. 2021; Takahashi et al. 2017; Ishii et al. 2015).

The small molecule GL-8-28 represents a novel chemical compound with preferential LPA₆ agonist activity

Small molecule modulators of the LPA family of G protein-coupled receptors are emerging as promising drugs for the treatment of a variety of diseases including those affecting the nervous system (Liu et al. 2021; Geraldo et al. 2021; Yanagida & Valentine 2020; Herr et al. 2020; Stoddard & Chun 2015; Kihara et al. 2015; Yung et al. 2015). Importantly, recent progress has been made in developing ligands with high selectivity for LPA receptor subtypes, in particular LPA_1-3 (Liu et al. 2021). In addition, despite the high sequence homology between LPA₄ and LPA₆ (Yanagida & Ishii 2011; Pasternack et al. 2008; Noguchi et al. 2003), unique ligand preferences have been described (Kano et al. 2019; Yanagida et al. 2013), thereby supporting a long-term prospect of highly selective LPA₆ ligands. Intriguingly, a novel compound with selective LPA₆ antagonist activity has recently been reported (Gnocchi et al. 2020).

The preferential LPA₆ agonist GL-8-28 attenuates OLG maturation in both the developing zebrafish and primary cultures of differentiating OLGs.

In both the in vitro cell culture and the in vivo zebrafish model, GL-8-28 was found to attenuate the gene expression profile associated with OLG maturation. However, much higher ligand concentrations were needed in the in vitro studies. In this context, it has been well-established that LPA receptors, similar to other G protein-coupled receptors, can be subject to receptor desensitization (Solís et al. 2021; Alcántara-Hernández et al. 2015; Avendaño-Vázquez et al. 2005; Zhao et al. 2021; Urs et al. 2005). Interestingly, next to homologous desensitization by the agonist there is evidence for agonist-independent heterologous desensitization (Alcántara-Hernández et al. 2015; Castillo-Badillo et al. 2015). Thus, it is conceivable that the culture conditions used in the in vitro experiments regulate extracellular LPA levels differently than seen in vivo and/or more significantly contribute to heterologous LPA₆ desensitization, thereby requiring higher GL-8-28 concentrations to trigger measurable functional outcomes. In addition, in the in vivo studies using the developing zebrafish, functional consequences of GL-8-28 on cell types other than OLGs could contribute to the observed attenuation of OLG maturation. In particular, LPA₆ signaling has been implicated in the regulation of vascular development (Okasato et al. 2021; Yasuda et al. 2019; Kano et al. 2019), and OLG-vascular crosstalk has been associated with a modulation of developmental OLG differentiation (Tsai et al. 2016; Yuen et al. 2014; De La Fuente et al. 2017). Further studies will be necessary to dissect the contributions of receptor desensitization and LPA₆-mediated non-OLG target effects to the differences in GL-8-28 sensitivity observed in the in vitro versus in vivo systems used here. Nevertheless, both the in vivo and in vitro data provide compelling support for a negative modulatory role of LPA₆ signaling on the expression of genes associated with OLG maturation.

LPA₆ signaling emerges as a novel and conceivably druggable signaling pathway with the potential toward promoting myelin repair

Changes in transcriptional regulation are thought to be linked to modifications in OLG heterogeneity (Jäkel et al. 2019; Falcão et al. 2018) and inefficiencies in OLG maturation and myelin repair (Duncan et al. 2017; Kuhlmann et al. 2008; Chang et al. 2002). As cause for such dysregulated gene expression in OLGs, recent evidence points to the actions of extrinsic factors rather than intrinsic OLG defects (Saraswat et al. 2021; Golan et al. 2021; Mozafari et al. 2020; Starost et al. 2020; Kirby et al. 2019). Thus, targeting druggable signaling pathways with the capacity to modulate gene expression profiles associated with OLG maturation has emerged as a promising approach toward promoting repair of the myelin sheath in MS (Rajendran et al. 2021; Angelini et al. 2021; Jeffries et al. 2021; Jeffries et al. 2016; Mausner-Fainberg et al. 2021; Roggeri et al. 2020; Lecca et al. 2020; Wang et al. 2020; Skinner & Lane 2020; Thümmler et al. 2019; Göttle et al. 2019; Welliver et al. 2018; Petersen et al. 2021; Green et al. 2017; Gaesser & Fyffe-Maricich 2016; Mierzwa et al. 2013). Despite an increasing number of potential pathways with promyelinating characteristics, however, clinical translation has, up until now, been below expectation (Lubetzki et al. 2020; Abu-Rub & Miller 2018). In light of the evidence for increased levels of LPA₆ in MS white matter lesions, targeting LPA₆ signaling may, therefore, represent a promising novel approach for stimulating OLG maturation in the context of myelin repair.

Abbreviations:

actb2

beta actin

BSA

bovine serum albumin

cldnk

claudin k

Cnp

2′,3′ cyclic nucleotide phosphodiesterase

CNS

central nervous system

DMEM

Dulbecco’s modified Eagle’s medium

DMSO

dimethyl sulfoxide

DMF

dimethylformamide

EAE

experimental autoimmune encephalomyelitis

EDTA

ethylenediaminetetraacetic acid

ef1a

elongation factor 1-alpha

Egr1

early growth response 1

ENPP2

ectonucleotide pyrophosphatase/phosphodiesterase 2

FBS

fetal bovine serum

bFGF

basic fibroblast growth factor

GC/MS

gas chromatography/mass spectrometry

HBSS

Hank’s balanced salt solution

hpf

hours postfertilization

hrs

hours

IR

infrared

LPA

lysophosphatidic acid

LPC

lysophosphatidylcholine

lysoPLD

lysophospholipase D

MHz

megahertz

MS

multiple sclerosis

NMR

nuclear magnetic resonance

OLG

oligodendrocyte

Olig2

oligodendrocyte transcription factor 2

(2S)-OMPT

L-sn-1-O-oleoyl-2-O-methylglyceryl-3-phosphothionate

PA

phosphatidic acid

PBS

phosphate-buffered saline

PDGF

platelet-derived growth factor

Pgk1

phosphoglycerate kinase 1

plp1b

proteolipid protein 1b

Ppia

peptidylprolyl isomerase A (cyclophilin A)

Rpl13a

ribosomal protein L13a

RT-qPCR

quantitative reverse transcription polymerase chain reaction

SDS

sodium dodecyl sulfate (SDS)

T3

tri-iodo-thyronine

TAZ

transcriptional coactivator with PDZ-binding motif

TLC

thin layer chromatography

Ugt8

UDP glycosyltransferase 8

YAP

Yes-associated protein

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.