AQP3 Deficiency in Humans and the Molecular Basis of a Novel Blood Group System, GIL

AQP3 is a water and glycerol channel present on human erythrocytes and in various tissues. By protein and molecular biology analysis, two unrelated probands who developed alloantibodies to the high frequency antigen GIL were found to be AQP3-deficient. The defect is caused by homozygous mutation affecting the 5′ donor splice site of intron 5 of the AQP3 gene. This mutation causes the skipping of exon 5 and generates a frameshift and premature stop codon. Functional studies by 90° light scattering using a stopped-flow spectrometer revealed the absence of facilitated glycerol transport across red cell membranes from the probands, but the water and urea transports were normal. Expression studies into COS-7 cells followed by flow cytometry analysis showed that only cells transfected with AQP3 cDNA strongly reacted with anti-GIL antibodies. These findings represent the first reported cases of AQP3 deficiency in humans and provide the molecular basis of a new blood group system, GIL, encoded by the AQP3 protein. AQP3 is a water and glycerol channel present on human erythrocytes and in various tissues. By protein and molecular biology analysis, two unrelated probands who developed alloantibodies to the high frequency antigen GIL were found to be AQP3-deficient. The defect is caused by homozygous mutation affecting the 5′ donor splice site of intron 5 of the AQP3 gene. This mutation causes the skipping of exon 5 and generates a frameshift and premature stop codon. Functional studies by 90° light scattering using a stopped-flow spectrometer revealed the absence of facilitated glycerol transport across red cell membranes from the probands, but the water and urea transports were normal. Expression studies into COS-7 cells followed by flow cytometry analysis showed that only cells transfected with AQP3 cDNA strongly reacted with anti-GIL antibodies. These findings represent the first reported cases of AQP3 deficiency in humans and provide the molecular basis of a new blood group system, GIL, encoded by the AQP3 protein. Integral membrane proteins that facilitate the transport of water or/and solutes belong to the major intrinsic protein (MIP) 1The abbreviations used for: MIP, major intrinsic protein; AEBSF, 4-(2-aminoethyl)-benzenesulfonyl fluoride; AQP, aquaporin; RBC, red blood cell; GlyAQP, glycosylated aquaporin; RT-PCR, reverse transcription polymerase chain reaction; RFLP, fragment-length polymorphism. family and are involved in many physiological processes and the pathophysiology of several clinical disorders (1Heymann J.B. Engel A. News Physiol. Sci. 1999; 14: 187-194PubMed Google Scholar, 2Verkman A.S. Mitra A.K. Am. J. Physiol. 2000; 278: F13-F28Crossref PubMed Google Scholar, 3Kozono D. Yasui M. King L.S. Agre P. J. Clin. Invest. 2002; 109: 1395-1399Crossref PubMed Scopus (226) Google Scholar, 4Agre P. King L.S. Yasui M. Guggino W.B. Ottersen O.P. Fujiyoshi Y. Engel A. Nielsen S. J. Physiol. 2002; 542: 3-16Crossref PubMed Scopus (932) Google Scholar). They are divided into three subgroups according to their sequences and function properties: (i) the aquaporins (AQP), permeable only to water; (ii) the glycerol facilitators, permeable only to glycerol (not present in mammals); and (iii) the aquaglyceroporins, which present a mixed selectivity. AQP1, the archetype of the MIP family selectively permeable to water (5Preston G.M. Carroll T.P. Guggino W.B. Agre P. Science. 1992; 256: 385-387Crossref PubMed Scopus (1697) Google Scholar), is present in various tissues including red blood cell (RBC) membranes from which it was first purified (6Denker B.M. Smith B.L. Kuhadja F.P. Agre P. J. Biol. Chem. 1988; 263: 15634-15642Abstract Full Text PDF PubMed Google Scholar). The high glycerol permeability of human RBCs is due to aquaglyceroporin AQP3 (7Roudier N. Verbavatz J.M. Maurel C. Ripoche P. Tacnet F. J. Biol. Chem. 1998; 273: 8407-8412Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar, 8Roudier N. Bailly P. Gane P. Lucien N. Gobin R. Cartron J.P. Ripoche P. J. Biol. Chem. 2002; 277: 7664-7669Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar), which is moderately permeable to water, highly permeable to glycerol, and to a lesser extent permeable to urea (9Ma T. Frigeri A. Hasegawa H. Verkman A.S. J. Biol. Chem. 1994; 269: 21845-21849Abstract Full Text PDF PubMed Google Scholar, 10Ishibashi K. Sasaki S. Fushimi K. Uchida S. Kuwahara M. Saito H. Furukawa T. Nakajima K. Yamaguchi Y. Gojobori T. Marumo F. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6269-6273Crossref PubMed Scopus (534) Google Scholar, 11Echevarria M. Windhager E.E. Tate S.S. Frindt G. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10997-11001Crossref PubMed Scopus (268) Google Scholar). AQP3 is present in rat RBCs but absent from mouse RBCs (8Roudier N. Bailly P. Gane P. Lucien N. Gobin R. Cartron J.P. Ripoche P. J. Biol. Chem. 2002; 277: 7664-7669Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar). It is encoded by a single-copy gene composed of six exons distributed over 6 kilobases of DNA located on human chromosome 9p13 (12Inase N. Fushimi K. Ishibashi K. Uchida S. Ichioka M. Sasaki S. Marumo F. J. Biol. Chem. 1995; 270: 17913-17916Abstract Full Text Full Text PDF PubMed Scopus (75) Google Scholar). The predicted protein of 292 residues is organized into six bilayer-spanning domains with the NH2- and COOH termini located intracellularly (13Preston G.M. Jung J.S. Guggino W.B. Agre P. J. Biol. Chem. 1994; 269: 1668-1673Abstract Full Text PDF PubMed Google Scholar). On human RBCs, AQP3 presumably is glycosylated on all subunits at Asn-141, the putative N-glycosylation site located in loop “c” at the external face of the cell membrane (14Ishibashi K. Sasaki S. Saito F. Ikeuchi T. Marumo F. Genomics. 1995; 27: 352-354Crossref PubMed Scopus (62) Google Scholar). So far, AQP1 and AQP3 are the only two known proteins of the aquaporin family identified in human RBCs. As AQP1 express Colton blood group antigens (15Smith B.L. Preston G.M. Spring F.A. Anstee D.J. Agre P. J. Clin. Invest. 1994; 94: 1043-1049Crossref PubMed Google Scholar, 16Agre P. Smith B.L. Preston G.M. Transfus. Clin. Biol. 1995; 2: 303-308Crossref PubMed Scopus (20) Google Scholar), the question was raised as to whether AQP3 might also be encoded by a blood group gene. We reasoned that if AQP3 carries a blood group specificity, the corresponding antigen should be of high frequency because AQP3 is a common protein of human RBCs. Accordingly, we performed an immunostaining analysis of human RBCs proteins from patients who had developed alloantibodies against high frequency antigens that caused delayed or severe hemolytic transfusion reactions. This approach led us to identify the first examples of AQP3null individuals and to define a new blood group system called GIL. 2The GIL blood group system was designated as No. 29 by the International Society of Blood Transfusion (Workshop Party on Blood Group Nomenclature, Vancouver, Canada, August, 2002). Proband 1 is a white French woman born in 1925. She had 10 pregnancies before 1979 when an antibody against a high frequency antigen (GIL) reacting with all human RBCs except her own was identified following a hemolytic reaction that occurred during orthopedic surgery (17Daniels G.L. Delong E.N. Hare V. Johnson S.T. Le Pennec P.Y. Mallory D. Marshall M.J. Oliver C. Spruell P. Immunohematology. 1998; 14: 49-52PubMed Google Scholar). Her RBCs were phenotyped O, GIL-negative. Blood samples from proband 1 and her family were collected after informed consent at the Centre National de Référence pour les Groupes Sanguins (CNRGS, Paris, France). Proband 2 is an unrelated white American woman born in 1956. She had no previous history of blood transfusion. In 1979, the RBCs from her first child had a weakly positive direct antiglobulin test, but there was no clinical sign of hemolytic disease of the newborn (17Daniels G.L. Delong E.N. Hare V. Johnson S.T. Le Pennec P.Y. Mallory D. Marshall M.J. Oliver C. Spruell P. Immunohematology. 1998; 14: 49-52PubMed Google Scholar). Her serum contains an anti-GIL antibody and she was phenotyped A, GIL-negative. Serum and DNA from proband 2 came from the frozen collection of the Bristol Institute for Transfusion Sciences (Bristol, UK). Modifying enzymes came from New England Biolabs (Hertfordshire, UK). Expand High Fidelity PCR, Titan One Tube RT-PCR systems, and N-glycosidase F (PNGase-F, 50 units/mg) were from Roche Molecular Biochemicals. Nucleotide primers were purchased from Genset (Paris, France). For primer designation, nucleotide (nt) position +1 was taken as the first nucleotide of the initiation codon of the human AQP3 cDNA clone (NCBI accession NM_004925). All reagents and peptide came from Sigma. AEBSF protease inhibitor was from Interchim (Montluçon, France). Rabbit polyclonal sera against human AQP1 (total protein) and the COOH-terminal region of rat AQP3 (residues 263–292) were characterized previously (7Roudier N. Verbavatz J.M. Maurel C. Ripoche P. Tacnet F. J. Biol. Chem. 1998; 273: 8407-8412Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar, 18Laizé V. Rousselet G. Verbavatz J.M. Berthonaud V. Gobin R. Roudier N. Abrami L. Ripoche P. Tacnet F. FEBS Lett. 1995; 373: 269-274Crossref PubMed Scopus (40) Google Scholar). RBC membrane proteins were prepared by hypotonic lysis in the presence of 0.5 mm AEBSF as described (19Steck T.L. Kant J.A. Methods Enzymol. 1974; 31: 172-180Crossref PubMed Scopus (895) Google Scholar). After SDS-polyacrylamide gel electrophoresis, immunoblotting was performed with rabbit anti-hAQP1 (1:4000) or anti-rAQP3 (1:500) antisera as described (8Roudier N. Bailly P. Gane P. Lucien N. Gobin R. Cartron J.P. Ripoche P. J. Biol. Chem. 2002; 277: 7664-7669Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar). N-glycosidase F treatment of RBC membrane proteins was performed overnight at 37 °C according to manufacturer's instructions. Total blood RNA extracted by the acid-phenol-guanidinium method (20Lozano M.E. Grau O. Romanowski V. Trends Genet. 1993; 9: 296Abstract Full Text PDF PubMed Scopus (11) Google Scholar) was used for the first PCR in the Titan One Tube reverse transcription polymerase chain reaction (RT-PCR) system between primers SP1 (nt −52 to −36) and ASP1 (nt 932–913) according to the manufacturer's instructions with annealing temperature of 60 °C and elongation time of 45 s. The second PCR was performed with one fiftieth of the first reaction in the same condition, except that the first step of the first cycle was omitted and the elongation temperature was 72 °C, using primers SP2 (nt −23 to −7) and ASP2 (nt 907–889) and Expand High Fidelity system. The PCR products were subcloned and sequenced on both strands using the Big Dye Terminator Cycle Sequencing ready Reaction Kit and analyzed on an ABI-Prism 310 Genetic Analyzer (Applied Biosystems, Foster City, CA). Direct PCR amplification was carried out on genomic DNA between primers SP3 (5′-aacagagtctcaggccctcc-3′, nucleotides −171 to −152 upstream exon 4) and ASP3 (nucleotides 807–788 in exon 6) using the Expand High Fidelity system according to manufacturer's instructions with annealing temperature of 62 °C and elongation time of 1 min. PCR products were subcloned and sequenced as above. The AQP3(Δ5) mutation was detected by PCR restriction fragment-length polymorphism (PCR-RFLP). The 1027-bp PCR product was digested with 5 units ofPmlI endonuclease and analyzed on a 1% (w/v) agarose gel. Each PCR reaction was performed in a total volume of 50 μl containing 100 ng of genomic DNA extracted from blood cells using a Wizard Genomic DNA purification kit from Promega (Madison, WI). Glycerol influx was induced by mixing a fresh or thawed RBC suspension (1.5% hematocrit) with an equal volume of a hyperosmotic solution of glycerol to produce a 100 mosmol/kg H2O inwardly directed osmotic glycerol gradient. The kinetics of RBC volume changes were followed by 90° light scattering using a stopped-flow spectrophotometer Biologic SFM3 instrument (Claix, France) as described (7Roudier N. Verbavatz J.M. Maurel C. Ripoche P. Tacnet F. J. Biol. Chem. 1998; 273: 8407-8412Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). Glycerol efflux was induced by equilibrating the RBC suspension in 200 mosmol/kg H2O glycerol for 1 h prior analysis (to insure that RBCs would be loaded with glycerol even if a glycerol proteic pathway were absent) and mixing them with a 200 mosmol/kg H2O mannitol solution in isoosmotic conditions to produce a 100 mosmol/kg H2O outwardly directed glycerol gradient. Inhibition studies were performed by incubating RBCs for 10 min with 0.1 mm CuCl2, and in some cases, 1 mm Gly-Gly-His peptide was added to remove the copper effect. Data from 5–10 time courses were averaged and fitted to single exponential functions by using the Simplex procedure of the Biokine software (Biologic, France) to determine the glycerol uptake or efflux rate constants (k, in s−1). Transient expression of AQP3 in COS-7 cells obtained from American Type Culture Collection (Manassas, VA) was performed by DEAE-dextran transfection with the full-length hAQP3 cDNA (kindly provided by Dr. Ishibashi) subcloned into the pcDNA3 expression vector (Invitrogen, Leek, The Netherlands) as described P. M. Gane P. Cartron J.P. Bailly P. J. Biol. Chem. 2000; Full Text Full Text PDF PubMed Scopus Google Scholar). after cells in with 1% (w/v) serum and serum were at 37 °C with sera at in serum After the cells were and with of cytometry analysis was performed as before (20Lozano M.E. Grau O. Romanowski V. Trends Genet. 1993; 9: 296Abstract Full Text PDF PubMed Scopus (11) Google Scholar). the cells were first in 1% for min at temperature and with 1% before with anti-rAQP3 and with a second RBCs from individuals who developed alloantibodies directed against high frequency antigens were analyzed by immunoblotting with the anti-rAQP3 which with human AQP3 protein (7Roudier N. Verbavatz J.M. Maurel C. Ripoche P. Tacnet F. J. Biol. Chem. 1998; 273: 8407-8412Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). RBC membranes from one of proband the and glycosylated AQP3 protein in RBCs Proband 1 the called and her serum an (17Daniels G.L. Delong E.N. Hare V. Johnson S.T. Le Pennec P.Y. Mallory D. Marshall M.J. Oliver C. Spruell P. Immunohematology. 1998; 14: 49-52PubMed Google Scholar). RBCs from a second were also found 1 As all RBCs showed a expression of AQP1 protein (not determine the molecular basis of AQP3 AQP3 from proband 1 and a were by RT-PCR and several were called which from the of exon was identified in proband 1 Genomic analysis that the skipping of exon 5 was caused by a that the nucleotide of the splice site of intron 5 The same was found in the genomic DNA from proband 2 (not The AQP3(Δ5) would a of residues with a new COOH-terminal region by a frameshift and premature The predicted AQP3(Δ5) would by exon the and domains and the loop of the hAQP3 protein 2 and in the proband 1 family and in blood and because the to a with a restriction site we were to a 2 analysis that the 1027-bp product from individuals was into of and 2 and In the PCR product from proband 1 and also from proband 2 (not the AQP3(Δ5) the of mutation in both of the proband 1 family for the AQP3(Δ5) mutation that her and 10 to were as a of digested of and was analysis with a rabbit that the RBC of AQP3 is in in the homozygous there is some between samples which might be in to the of protein loaded on the gel. of the same gel for the AQP1 protein revealed a in all samples of the family some gene with transport function as The glycerol permeability of RBCs from and from probands 1 and 2 the AQP3 protein was by stopped-flow analysis In hyperosmotic conditions directed glycerol RBCs a with a first followed by a of the light to the of RBCs due to water efflux and the of RBCs from the glycerol influx with a water flow only a in light was detected with RBCs from probands, that the glycerol permeability due to the facilitated pathway was constants of the cell volume due to the glycerol uptake were and at °C for proband 1 and proband 2 RBCs, the glycerol permeability of two and was that of the with rate constants of and (not in with the AQP3 expression protein detected on an the glycerol transport of RBCs with CuCl2, an inhibitor of the pathway A. Physiol. PubMed Scopus Google Scholar), was and was to proband RBC glycerol when the copper peptide Gly-Gly-His was added after with CuCl2, the glycerol transport across RBCs The performed at us to the of the glycerol transport which were and for and proband 1 RBCs, glycerol channel in human RBCs. Glycerol efflux was followed in isoosmotic conditions loaded with glycerol, an outwardly directed glycerol with and proband 1 RBCs. The efflux of glycerol a of RBCs by an of the light In the light with proband 1 RBCs and with RBCs first with the glycerol transport in the after the of peptide that water transport and urea transport of proband RBCs were (not both probands were and had alloantibodies to the GIL antigen in their serum (17Daniels G.L. Delong E.N. Hare V. Johnson S.T. Le Pennec P.Y. Mallory D. Marshall M.J. Oliver C. Spruell P. Immunohematology. 1998; 14: 49-52PubMed Google Scholar), flow cytometry analysis was performed to that proband sera reacted with RBCs O, but with proband 1 RBCs as in As proband sera reacted strongly with RBCs even at and when the RBCs were or (not no of the was with proband 1 RBCs, which the AQP3 protein. the of in proband sera after transfusion or as as the absence of AQP3 in their RBC membranes by strongly that both were provide the that AQP3 carried the blood group GIL specificity, expression of AQP3 protein in COS-7 cells was performed by AQP3 cDNA into the human expression cytometry analysis of AQP3 was performed with rabbit anti-rAQP3 serum and anti-GIL antibodies. The anti-rAQP3 reacted with the AQP3 but with COS-7 of cells reacted with a rabbit serum (not immunoblotting with the anti-rAQP3 showed that the glycosylated and of the AQP3 protein were present in AQP3 but in COS-7 cells studies with human sera from both probands reacted with but with COS-7 as the of cells were and and for probands 1 and In a human serum of anti-GIL antibody used as reacted with AQP3 COS-7 with COS-7 These findings that the directed the of a AQP3 cell protein by sera containing developed by probands 1 and 2 who were AQP3null in This is the first of human AQP3 it was by with a rabbit serum against the AQP3 RBCs from two unrelated patients who developed alloantibodies against high frequency blood group antigens following transfusion or The two probands belong to the blood group called found cases previously (17Daniels G.L. Delong E.N. Hare V. Johnson S.T. Le Pennec P.Y. Mallory D. Marshall M.J. Oliver C. Spruell P. Immunohematology. 1998; 14: 49-52PubMed Google Scholar). both probands, who the same molecular were it be whether might a common and genomic analysis that the of AQP3 protein was caused by a single nucleotide at the 5′ donor represent the of splice site of intron 5 of which in exon 5 and premature of that the probands were homozygous for the AQP3(Δ5) the and 10 of proband 1 were In there was a gene as two in who were both as carried a of AQP3 protein on RBCs and a glycerol transport that of as by stopped-flow AQP3null RBCs a of glycerol permeability as from analysis of facilitated glycerol These that the AQP3(Δ5) is or at the RBC The of water and urea in proband RBCs be because AQP1 and were present and for at of the water and urea across RBCs, water permeability detected in which AQP1, was previously to AQP3 (7Roudier N. Verbavatz J.M. Maurel C. Ripoche P. Tacnet F. J. Biol. Chem. 1998; 273: 8407-8412Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). The high of glycerol permeability found in AQP3null cells the absence of glycerol RBCs are in AQP3 expression showed a glycerol permeability to protein transport pathway T. Verkman A.S. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). of studies is the that AQP3 carries blood group GIL (i) AQP3null individuals belong to the (ii) the rabbit antibody on with and AQP3 present in but (iii) the glycerol permeability of RBCs is in sera reacted by flow cytometry with but AQP3null cells reacted with anti-GIL present in the serum of findings provide the to define the molecular basis of a new blood group system, GIL, the Accordingly, it be that the GIL is located on the of chromosome 9p13 and is the second blood group on chromosome with the located on the at As AQP3 protein a in the cells of and and in a in water and cell volume S. King L.S. B.M. Agre P. Am. J. Physiol. 273: PubMed Google Scholar, A. Verkman A.S. Proc. Natl. Acad. Sci. U. S. A. 1995; PubMed Scopus Google Scholar, A. A. J. 2000; PubMed Scopus Google Scholar, T. M. R. Verbavatz J.M. Verkman A.S. J. Biol. Chem. 2002; 277: Full Text Full Text PDF PubMed Scopus Google Scholar), it is that the AQP3null individuals identified or the reported (17Daniels G.L. Delong E.N. Hare V. Johnson S.T. Le Pennec P.Y. Mallory D. Marshall M.J. Oliver C. Spruell P. Immunohematology. 1998; 14: 49-52PubMed Google Scholar), clinical which transport glycerol, might be in PubMed Scopus Google Scholar). The absence of clinical disorders in conditions also in patients who are for AQP1 G.M. Smith B.L. Agre P. Science. 1994; PubMed Scopus Google Scholar, S. Smith B.L. Preston G.M. N. M. Agre P. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar). It is only conditions that a L.S. M. Cartron J.P. Agre P. N. J. PubMed Scopus Google Scholar) and a in permeability L.S. Nielsen S. Agre P. Proc. Natl. Acad. Sci. U. S. A. 2002; PubMed Scopus Google Scholar) detected in clinical studies of AQP3null with should provide on the of the AQP3 protein. of AQP3null revealed a and to of a T. Verkman A.S. J. Biol. Chem. Full Text Full Text PDF PubMed Scopus Google Scholar, T. Y. A. Verkman A.S. Proc. Natl. Acad. Sci. U. S. A. 2000; PubMed Scopus Google Scholar). was also detected in AQP3null AQP3 might be involved in as a defect of described in T. M. R. Verbavatz J.M. Verkman A.S. J. Biol. Chem. 2002; 277: Full Text Full Text PDF PubMed Scopus Google Scholar). that from studies is the of AQP3null mutation in humans and the of the blood group system GIL. These findings should in RBCs and various to the physiological of aquaporin and to the of channel selectivity. We the proband 1 family for for and Tacnet for the of rabbit sera and to stopped-flow

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.