key: cord-103345-v555a2ll authors: Taylor, Adrian; Grapentine, Sophie; Ichhpuniani, Jasmine; Bakovic, Marica title: The novel roles of choline transporter-like 1 and 2 in ethanolamine transport date: 2020-08-28 journal: bioRxiv DOI: 10.1101/2020.08.27.270223 sha: doc_id: 103345 cord_uid: v555a2ll We examined a novel function of mammalian Choline-Transporter-Like proteins CTL1/SLC44A1 and CTL2/SLC44A2 in ethanolamine transport. We established two distinct ethanolamine transport systems of a high affinity (K1 = 55.6 - 66.5 μM), mediated by CTL1, and of a low affinity (K2 = 275 - 299 μM), mediated by CTL2. Both types of transport are Na+-independent and mediated in a pH dependent manner, as expected for ethanolamine/H+ antiporters. Primary human fibroblasts with separate frameshift mutations (M1= SLC44A1 ΔAsp517 and M2= SLC44A1 ΔSer126) are devoid of CTL1 ethanolamine transport but maintain unaffected CTL2 transport. The lack of CTL1 or CTL2 reduced the ethanolamine transport, the flux by the CDP-ethanolamine Kennedy pathway and PE synthesis. Overexpression of CTL1 in SLC44A1 ΔSer126 (M2) cells improved the ethanolamine transport and PE synthesis. The SLC44A1 ΔSer126 cells are reliant on CTL2 function and CTL2 siRNA almost completely abolished ethanolamine transport in the whole cells and mitochondria. Overexpression of CTL1 and CTL2 cDNAs increased ethanolamine transport in control and SLC44A1ΔSer126 cells. CTL1 and CTL2 facilitated mitochondrial ethanolamine uptake, but the transport mediated by CTL1 is predominant in the whole cells and mitochondria. These data firmly established that CTL1 and CTL2 are the first identified ethanolamine transporters in the whole cells and mitochondria, with intrinsic roles in de novo PE synthesis by the CDP-Etn Kennedy pathway and compartmentation of intracellular ethanolamine. Significance The lack of Choline Transporter Like 1 (SLC44A1/CTL1) is the primary cause of a new neurodegenerative disorder with elements of childhood-onset parkinsonism and mitochondrial dysfunction. SLC44A2/CTL2 encodes the human neutrophil antigen 3, causes autoimmune hearing loss and Meniere’s disease, and has been recently identified as the main risk factor for thrombosis-the major cause of death in Covid-19 patients. Our investigation provides insights into the novel functions of CTL1 and CTL2 as intrinsic ethanolamine transporters. CTL1 and CTL2 are high and low affinity transporters, with direct roles in the membrane phospholipid synthesis. The work contributes to new knowledge for CTL1 and CTL2 independent transport functions and the optimization of prevention and treatment strategies in those various diseases. Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) are major components of cellular membranes where they are involved with essential cellular processes (1, 2) . PC and PE are synthesized de novo by CDP-Cho and CDP-Etn branches of the Kennedy pathway in which the extracellular substrates choline (Cho) and ethanolamine (Etn) are actively transported into the cell, phosphorylated and coupled with diacylglycerols (DAG) to form the final phospholipid product. While multiple transport systems have been established for Cho, Etn transport is poorly characterized and there is no single gene/protein assigned a transport function for mammalian Etn. Cho transport for membrane phospholipid synthesis is mediated by Cho transporter like protein CTL1/SLC44A1 (3) . CTL1 is the only well-characterized member of a broader family (CTL1-5/SLC44A1-5) (4, 5) . CTL1/SLC44A1 is a Cho/H + antiporter at the plasma membrane and mitochondria (4, 5) . The role of plasma membrane CTL1 is assigned to Cho transport for PC synthesis, but the exact function of the mitochondrial CTL1 is still not clear. In the liver and kidney, mitochondrial CTL1 transports Cho for oxidation to betaine, the major methyl donor in the one-carbon cycle (6) . In other tissues however, the mitochondrial CTL1 probably maintains the intracellular pools of Cho and as a H + -antiporter and modulates the electrochemical/proton gradient in the mitochondria (7, 8) . CTL2/SLC44A2 is only indirectly implicated in PC synthesis and its exact function is not firmly established in neither whole cells nor mitochondria (4) . PE is the major inner membrane phospholipid with specific roles in mitochondrial fusion, autophagy and apoptosis (9 -11) . PE is also a valuable source of other phospholipids. PC is produced by methylation of PE while phosphatidylserine (PS) is produced by an exchange mechanism whereby the Etn moiety of PE is replaced with serine and free Etn is released. PC could also produce PS by a similar exchange mechanism, with free Cho being released. The metabolically released Cho and Etn need to be transported in and out of the cytosol and mitochondria or reincorporated into the Kennedy pathway (3 -6) . That mammalian Etn and Cho transport may occur through a similar transport system was implicated from early kinetic studies in bovine endothelial cells, human retinoblastoma cells and glial cells (12 -14) . Here, we demonstrate that CTL1/SLC44A1 and CTL2/SLC44A2 are authentic Etn transporters at the cell surface and mitochondria. We examine the kinetics of Etn transport in CTL1 and CTL2 depleted conditions and overexpressing cells. We characterize Etn transport in human skin fibroblasts that maintain CTL2 but lack CTL1 function due to inherited CTL1/SLC44A1 frameshift mutations (M1= SLC44A1 ΔAsp517 and M2= SLC44A1 ΔSer126 ) (15) . We employ pharmacological and antibody induced inhibition to separate the contributions of the CTL1 and CTL2 to Etn transport and PE synthesis. This study is the first to demonstrate that the CTL1 and CTL2 are high and low to medium affinity cellular and mitochondrial Etn transporters. To our knowledge, this is the first study to demonstrate that as intrinsic Etn transporters, CTL1 and CTL2 regulate the supply of extracellular Etn for the CDP-Etn pathway, redistribute intracellular Etn and balance CDP-Cho and CDP-Etn arms of the Kennedy pathway. To assess the magnitude by which CTL1/2 inhibition affects PE and PC levels, two types of cells were characterized for phospholipid metabolism (MCF-7 and MCF-10) (16, 17) . The cells were treated for 24h with [ 3 H]-glycerol, to label the entire glycerolipid pools (steady-state levels) in the presence and absence of CTL1 transport inhibitor hemicholinium-3 (HC-3) or CTL1 specific antibody ( Fig. 1A and B) . Surprisingly, HC-3 reduced the steady-state levels not only of PC (25-50%) but also of PE (50%) in both cell types. CTL1 antibody similarly reduced PC and PE levels (40-50%), further indicating that CTL1 could be involved in the transport of Etn, in addition to its well-characterized function in Cho transport (15, 18) . µM, Etn, Cho and Etn + Cho were applied respectively (Fig. 1D) . Furthermore, CTL1 antibody inhibited 14 C-Etn transport in a concentration-dependent manner with an IC50 of 50 ng (Fig. 1E ). CTL2 antibody also inhibited both, 14 C-Etn and 3 H-Cho transports in a concentration dependent manner with LC50 50 ng (Fig. 1F, G) . Together, the data showed that Etn is a 3ubstrate for CTL1 and CTL2-mediated transports, in addition to already establish functions in Cho transport. We studied the kinetics of Etn transport in monkey COS-7 cells and control (Ctrl) and CTL1 deficient (M1=SLC44A1 ΔAsp517 and M2=SLC44A1 Ser126 ) primary human fibroblasts. As expected, COS-7 cells and Ctrl fibroblasts expressed CTL1 and CTL2 proteins while CTL1 mutant fibroblasts M1 and M2 only expressed CTL2 protein ( Fig. 2A ). 14 C-Etn transport rates (V) plotted against [Etn] produced a series of saturation curves, as expected for protein mediated transports ( Fig. 2B) . Vmax values were nearly identical in Ctrl and COS-7 cells (Vmax = 26.9 and 26.3 nmol/mg protein/min) and M1 and M2 cells had reduced but similar Vmax = 20.6 -21.2 nmol/mg protein/min ( Fig. 2B) , apparently caused by the absence of the CTL1 transport component. Indeed, the Eadie-Hofstee plots derived from the saturation curves were biphasic in Ctrl fibroblasts and COS-7 cells and linear for M1 and M2 cells (Fig. 2C ). This type of behavior indicated the presence of two distinct transport systems in Ctrl and COS-7 cells with two binding constants, of high and low affinity for Etn, and one transport system of a lower affinity in M1 and M2 cells. As further shown in Fig. 2C , Ctrl fibroblasts, high affinity (K1 = 66.5 ± 8.5 µM) and low (K2 = 299.0 ± 13.1 µM) affinity Etn bindings were similar to COS-7 cells bindings (K1 = 55.6 ± 14.8 µM and K2 = 277.3 ± 7.9 µM). On the other hand, M1 and M2 cells are characterized by a single transport with a binding constant for Etn of 275.4 -279.6 µM which is the second (K2), low affinity, binding constant as determined in Ctrl and COS-7 cells (Fig. 2C ). M1 and M2 cells only express CTL2 and at levels similar to Ctrl and Cos 7 cells, and do not have a functional CTL1 protein ( Fig. 2A,D) , strongly implicating CTL2 as responsible for the low affinity Etn transport. Indeed, CTL2 depletion by siRNA knockdown in Ctrl cells completely abolished the low affinity transport component while the high affinity component remained intact (Fig. 2E ). This analysis also confirmed that the high affinity transport (K1) which is absent in M1 and M2 cells and remained intact in CTL2 siRNA treated Ctrl is CTL1-mediated Etn transport. Since the effects of pH and [Na + ] ions on choline transport is well established (19) , their effects on Etn transport were also investigated ( Fig. 2F and G). Etn transport in Ctrl (CTL1 + CTL2 transport) and M2 fibroblasts (CTL2 transport) (Fig. 2F ) was reduced when extracellular pH was lowered from 7 to 5.5 and stimulated when pH was increased to 8.5. Additionally (Fig. 2G ), as expected, the rate of Etn transport in Ctrl cells was higher than in M2 cells but the rates were not modified when Na + ions were replaced by Li + ions in the uptake buffer. Altogether, the data established that CTL1 and CTL2 acts as Etn/H + antiporters, driven by a proton gradient and they are both independent of Na + , as in case of Cho transport (19) . state levels) showed unchanged PC, reduced PE, PS and DAG and increased triglycerides (TAG) in M2 cells ( Fig. 3C and D) . Therefore, reduced Etn transport, slower P-Etn and CDP-Etn formation, and reduced DAG levels, collectively slowed the CDP-Etn pathway (Fig. 3A ,B) and reduced PE levels (Fig. 3C ) in M2 cells. The CDP-Etn formation from PEtn is usually the rate-regulatory step in the Kennedy pathway and is controlled by Pcyt2 (CTP: phosphoethanolamine cytidylyltransferase) (11) . Indeed, in accordance with reduced CDP-Etn formation above, the activity and expression of Pcyt2 were also reduced in M2 cells (Fig. 3 E,F) . The expression of Etn kinase (EK) was similarly decreased by 25% (Fig. 3F) , explaining why the formation of P-Etn was reduced in M2 cells (Fig. 3A , B). In addition to PE, PS levels were reduced in M2 cells but the expression of PS synthesis (PS syntase1/2-PSS1/2) and PS degradation (PS decarboxylase-PSD) genes ( Fig.3F ) were unaltered ( Fig 3C) . The expression of the PC synthesis genes Pcyt1 (CTP: phosphocholine cytidylyltransferase) was decreased by 70% and choline kinase (CK) by 25% in M2 cells (Fig. 3F ), yet unexpectedly PC levels were unchanged (Fig. 3C ). We previously established (15) that the constant PC levels in CTL1 deficient M1 and M2 cells are maintained by reduced PC turnover and increased formation from other phospholipids (PC is made at the expanse of PE and PS), as the main mechanism to maintain PC as a source of choline in a new neurodegenerative disorder caused by frame-shift mutations in the CTL1 gene M1=SLC44A1 ΔAsp517 and M2=SLC44A21 Ser126 (15) . These data collectively provided strong genetic and metabolic evidence that CTL1 and CTL2 independently contributes not only to the CDP-Cho but also to the CDP-Etn Kennedy pathway. CTL2 is not over expressed in deficient M1 and M2 cells ( Fig.2A ) and as such it cannot compensate for the absence of CTL1 in those cells and affected individuals (15) . To demonstrate that CTL1 and CTL2 are both Etn and Cho transporters, the cells were transiently transfected with CTL1 cDNA or CTL2 cDNA and the protein expression and transport determined after 48h. As shown in Fig. 4A , in M2 cells, that completely lack CTL1 protein, CTL1 cDNA It is well known that various organic cations can inhibit CTL1 mediated Cho transport (20) . We assessed the inhibitory effect of organic cations and CTL2 knockdown on Etn transport in Ctrl and M2 cells (Fig. 5 ). This helped us understand the magnitude at which CTL1 and CTL2 contributed to Etn transport. Total (CTL1+CTL2) transport ( showing that is not a CTL1/2 related transport. Finally, comparison of all transport velocities (Fig. 5E ) showed a general order of contributions, from the high affinity CTL1 (Ctrl + CTL2 siRNA), low affinity CTL2 (M2), and the residual very low affinity (M2+CTL2 siRNA) transports for Etn. CTL1(Ctrl) contributed 80%, CTL2 (M2) 12.5%, and the unrelated residual transport (M2 + CTL2 siRNA) accounted for 7.5% to the total transport. The Ctrl and M2 cells express OCT and OCTN and other unspecific Cho transporters that could be contributing to this residual transport (15) . CTL1 and CTL2 mediate Etn transport to mitochondria. CTL1 and CTL2 are present in the mitochondria and are involved in mitochondrial Cho transport (4, 5) . We used COXIV as a marker of mitochondria isolated from Ctrl and M2 cells and CTL2 mRNA expression for the siRNA knockdown of CTL2 transport (Fig. 6A) . By comparing the contributions of all mitochondria transport components (Fig. 6B) , CTL1+CTL2 (Ctrl) contributed 70%, CTL2 (M2) 20%, and the residual unrelated transport (M2 + CTL2 siRNA) contributed 10% to the total mitochondria Etn transport. We also compared the rates of [ 14 C]-Etn transport in the isolated mitochondria and the whole cells and in the presence and absence of the specific inhibitor HC-3. As expected, the CTL1 and CTL2 inhibitor HC-3 blunted Etn uptake in a time dependent manner in the mitochondria (Fig. 6C , E) and the whole cells (Fig. 6 D, F) . In the absence of HC-3, the rate of Ctrl mitochondria transport was similar to the whole cell Ctrl transport (0.04 and 0.05 µmol/mg/min, respectively); M2 mitochondrial transport was also similar to the whole cell transport (0.008 and 0.01 µmol/mg/min respectively) (Fig. 6E, F) , demonstrating that the same proteins are responsible for the transports in the whole cell and mitochondria. In addition, the M2 mitochondrial (CTL2 only) transport was 5-fold slower that the total (CTL1+CTL2) transport of the Ctrl mitochondria. Taken together, these data established that CTL1 and CTL2 mediate mitochondrial Etn transport with the same kinetic properties as in the whole cells. PE and PC are bilayer forming phospholipids involved in fundamental membrane processes, growth, survival and cell signaling (11) . PE and PC are similarly synthesized by CDP-Etn and CDP-Cho Kennedy pathway, in which the extracellular substrates Cho and Etn are actively transported into the cell, phosphorylated, and coupled with diacylglycerols (DAG) to form the final phospholipid product. Cho and Etn released from PC and PE also need to be transported in and out of the cytosol and mitochondria or reincorporated into the Kennedy pathway. The plasma membrane CTL1 is firmly assigned to Cho transport for PC synthesis (15) , yet the exact function of the mitochondrial CTL1 is still not clear. In the liver and kidney mitochondria, Cho is specifically oxidized to betaine, the major methyl donor in the one-carbon cycle (6) . Since broadly expressed, it is proposed that the mitochondrial CTL1 could maintain the intracellular pools of Cho and as + H-antiporter could regulate the electrochemical/proton gradient in the mitochondria (5, 15, 18) . CTL2 is only indirectly implicated in Cho transport and until this work the exact CTL2 substrate binding and transport mechanism were not firmly established, in neither the whole cells nor isolated mitochondria. Based on our extensive work on CTL1 and Cho transport (3) (4) (5) (6) (7) (8) 15 ) and known similarities between Cho and Etn transports in various conditions (12) (13) (14) we postulated that CTL1 could be that long-searched for Etn/Cho transporter and the last missing link between CDP-Cho and CDP-Etn pathways for phospholipids synthesis. We conducted an extensive number of kinetic, metabolic, and genetic experiments to solidify this hypothesis. We established that CTL1 mediated a high affinity Etn transport with K1 = 56-67µM and that CTL2 mediated a low Etn affinity transport with K2 = 275-299 µM in primary human fibroblasts and monkey Cos7 cells. Importantly, the CTL1 affinity constant for Etn binding is in the range of physiological Etn concentration in rat and humans (10-75µM) (13) , and explains why CTL1 contributed the most (70-80%) of the Etn transport in the whole cells and mitochondria. We recently described the first human disorder caused by homozygous frame-shift mutations in the CTL1 gene SLC44A1: M1= SLC44A1 ΔAsp517 , M2= SLC44A1 ΔSer126 and M3= SLC44A1 ΔLys90 (15) . After an extensive characterization of transport and metabolism in patient's fibroblasts it was apparent that diminished Cho transport is the primary cause of this new neurodegenerative disorder with elements of childhood-onset parkinsonism and MPAN (mitochondrial membrane protein-associated neurodegeneration)-like abnormalities. Paradoxically, although Cho transport and CDP-Cho Kennedy pathway were diminished, PC remained preserved in the cerebrospinal fluid and skin fibroblasts of the affected individuals (15) . The cell membranes were however drastically remodeled and depleted of PE and PS, apparently as a homeostatic response to preserve PC and prevent Cho deficiency in the affected individuals (15) . could maintain the intracellular pools of Cho and Etn and since they are both proton antiporters, they could be significant regulators of the proton gradient in the mitochondria. Recent studies in CTL2 knockout mice established that that CTL2 mediated mitochondrial Cho transport is critical for ATP and ROS production, platelet activation and thrombosis (31) . CTL2 gene SLC44A2 is well-established the human neutrophil antigen (32) , and genetic risk factor for hearing loss, Meniere's disease and venous thrombosis (33) . Neutrophil CTL2 could interact directly with platelets' integrin αIIbβ3, and induce neutrophil extracellular trap (NETosis) that than promote thrombosis (34) . In studies with CTL1 and CTL2 antibodies 8.0 x 10 4 cells/well were seeded in 6-well plates and grown for 24h. After 24h of growth, various amounts of antibodies were added to cells and incubated for 24h and Cho and Etn uptake were then conducted. radiolabeling with 3H-glycerol as previously described (40) . Pcyt2 Activity Assay. The assay was conducted as described (40 PSD, PSS1 and PSS2 was determined by PCR using the primers and conditions as before (15) . Reactions were standardized by amplifying glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and relative band intensity was quantified using ImageJ software (NIH, Bethesda, MD, USA). Mitochondrial Isolation. Mitochondria were isolated as initially described (18) . In brief, cells were incubated for 20 minutes in ice-cold RSB swelling buffer and homogenized; 19 ml MS buffer was added, and the cell homogenate was centrifuged at 2500 rpm for 5 minutes. This step was repeated twice, and the final supernatant was centrifuged at 12,500 rpm. The resulting pellet was resuspended in MS buffer. The mitochondria purity was determined using COXIV mitochondria marked and β-tubulin as a whole cell control. Statistical Analysis. All measurements are expressed as means from quadruplets ± SEM. Statistical analysis was performed using GraphPad Prism software (GraphPad, Inc.). Data were subjected to students T-test. Differences were considered statistically significant at *p < 0.05. All study data are included in the article analyzed data and wrote the paper The relationship between phospholipids and insulin resistance: From clinical to experimental studies Phosphatidylethanolamine deficiency in mammalian mitochondria impairs oxidative phosphorylation and alters mitochondrial morphology Choline transport for phospholipid synthesis Choline transport for phospholipid synthesis: An emerging role of choline transporter-like protein 1 The solute carrier 44A1 is a mitochondrial protein and mediates choline transport Betaine and choline improve lipid homeostasis in obesity by participation in mitochondrial oxidative demethylation Choline supplementation restores substrate balance and alleviates complications of Pcyt2 deficiency Palmitic acid and oleic acid differentially regulate choline transporter-like 1 levels and glycerolipid metabolism in skeletal muscle cells De novo synthesis of phospholipids is coupled with autophagosome formation Ethanolamine and phosphatidylethanolamine: partners in health and disease Regulation of phosphatidylethanolamine homeostasis -the critical role of CTP:phosphoethanolamine cytidylyltransferase (Pcyt2) Ethanolamine and choline transport in cultured bovine aortic endothelial cells Effect of ethanolamine on choline uptake and incorporation into phosphatidylcholine in cells. human Y79 retinoblastoma Uptake of ethanolamine in neuronal and glial cell cultures Choline transporter-like 1 deficiency causes a new type of childhoodonset neurodegeneration Isoform-specific and protein kinase C-mediated regulation of CTP:phosphoethanolamine cytidylyltransferase phosphorylation Stimulation of the human CTP:phosphoethanolamine cytidylyltransferase gene by early growth response protein 1 Mechanism of choline deficiency and membrane alteration in postural orthostatic tachycardia syndrome primary skin fibroblasts Functional expression of choline transporter like-protein 1 (CTL1) and CTL2 in human brain microvascular endothelial cells Molecular and functional characterization of choline transporter in human colon carcinoma HT-29 cells Molecular and functional characterization of choline transporter-like proteins in esophageal cancer cells and potential therapeutic targets Identification and functional analysis of choline transporter in tongue cancer: A novel molecular target for tongue cancer therapy Ethanolamine inhibits choline uptake in the isolated hamster heart Effects of amino acids and Ethanolamine on choline uptake and phosphatidylcholine biosynthesis in baby hamster kidney-21 cells Functional analysis of [methyl-(3)H]choline uptake in glioblastoma cells: Influence of anti-cancer and central nervous system drugs Ethanolamine ameliorates mitochondrial dysfunction in cardiolipin-deficient yeast cells Ethanolamine enhances the proliferation of intestinal epithelial cells via the mTOR signaling pathway and mitochondrial function Looking Beyond Structure: Membrane Phospholipids of Skeletal Muscle Mitochondria Cardiolipin and mitochondrial cristae organization Effects of lipids on mitochondrial functions The choline transporter Slc44a2 controls platelet activation and thrombosis by regulating mitochondrial function Molecular genetics of the human neutrophil antigens SLC44A2 single nucleotide polymorphisms, isoforms, and expression: Association with severity of Meniere's disease? Bercu Activated αIIbβ3 on platelets mediates flow-dependent NETosis via SLC44A2. Elife Characterization of hemostasis in mice lacking the novel thrombosis susceptibility gene Slc44a2 SLC44A2 deficient mice have a reduced response in stenosis but not in hypercoagulability driven venous thrombosis SLC44A2-A novel therapeutic target for venous thrombosis? Identification and expression of a mouse muscle-specific CTL1 gene A rapid method of total lipid extraction and purification Developmental and metabolic effects of disruption of the mouse CTP:phosphoethanolamine cytidylyltransferase gene (Pcyt2) We thank Christina Fagerberg (Odense University, Denmark) and Felix Distelmaier The authors declare no competing interests.