Rha and Rhag are MGI nomenclature.
Generation and characterisation of Rhd and Rhag null mice
Article first published online: 5 OCT 2009
© 2009 Blackwell Publishing Ltd
British Journal of Haematology
Volume 148, Issue 1, pages 161–172, January 2010
How to Cite
Goossens, D., Trinh-Trang-Tan, M.-M., Debbia, M., Ripoche, P., Vilela-Lamego, C., Louache, F., Vainchenker, W., Colin, Y. and Cartron, J.-P. (2010), Generation and characterisation of Rhd and Rhag null mice. British Journal of Haematology, 148: 161–172. doi: 10.1111/j.1365-2141.2009.07928.x
- Issue published online: 14 DEC 2009
- Article first published online: 5 OCT 2009
- Received 16 June 2009; accepted for publication 30 July 2009
- blood groups;
- intercellular adhesion molecule 4;
- gas transport;
- Top of page
- Materials and methods
- Supporting Information
Mouse Rhd* and Rhag* genes were targeted using insertional vectors; the resulting knockout mice, and double-knockout descendants, were analysed. Rhag glycoprotein deficiency entailed defective assembly of the erythroid Rh complex with complete loss of Rh and intercellular adhesion molecule 4 (ICAM-4), but not CD47, expression. Absence of the Rh protein induced a loss of ICAM-4, and only a moderate reduction of Rhag expression. Double knockout phenotype was similar to that of Rhag targeted mice. Rhd and Rhag deficient mice exhibited neither the equivalent of human Rhnull haemolytic anaemia nor any clinical or cellular abnormalities. Rhd−/− and Rhag−/− erythrocytes showed decreased basal adhesion to an endothelial cell line resulting from defective ICAM-4 membrane expression. There was no difference in recovery from phenylhydrazine-induced haematopoietic stress for double knockout mice as compared to controls, suggesting that ICAM-4 might be dispensable during stress erythropoiesis. Ammonia and methylammonia transport in erythrocytes was severely impaired in Rhag−/− but only slightly in Rhd−/− animals that significantly expressed Rhag, supporting the view that RhAG and Rhag, but not Rh, may act as ammonium transporters in human and mouse erythrocytes. These knockout mice should prove useful for further dissecting the physiological roles of Rh and Rhag proteins in the red cell membrane.
The human Rh (Rhesus) blood group system is of clinical interest due to its role in transfusion medicine, auto-immune anaemia and its implication in materno-fetal incompatibility and haemolytic disease of the newborn (Mollison et al, 1993). Rh deficiency (synonym Rhnull syndrome) is a rare erythrocyte restricted disorder characterised by the lack or severe reduction of Rh antigens associated with a chronic haemolytic anaemia of varying severity, and morphological abnormalities (stomato-spherocytosis), increased osmotic fragility, altered phospholipid asymmetry and cation transport abnormalities (Sturgeon, 1970; Cartron, 1999; Huang et al, 2000). Studies of Rhnull syndrome have been instrumental in our current understanding of the Rh complex. Rh antigens are formed by an oligomeric association in the erythrocyte membrane of two major components, which includes Rh proteins themselves (RhD and RhCcEe, products of the RHD/CE haplotype on chromosome 1p34) and the homologous Rh-associated glycoprotein (RhAG, encoded by a locus on chromosome 6p12–21). Accessory molecules, such as intercellular adhesion molecule 4 (ICAM-4; LW blood group glycoprotein), integrin-associated protein (IAP/CD47), and glycophorin B, are linked to this core by non-covalent bonds (Agre & Cartron, 1991; Avent & Reid, 2000; Avent et al, 2006; Le Van Kim et al, 2006). This Rh complex is also an important interaction site between the membrane lipid bilayer and the spectrin-based skeleton that may regulate the shape, deformability, and mechanical properties of red cells (Nicolas et al, 2003, 2006). Rhnull syndrome is a rare autosomal recessive disorder caused by different mutations that occur at either the RHAG or RHCE locus, respectively, but there is no alteration of the genes encoding the accessory molecules (Cartron, 1999; Huang et al, 2000). As a result, when either the RhAG or the Rh subunit is absent, the Rh complex is missing or severely reduced on Rhnull red blood cells (RBCs).
Rh and RhAG proteins, which compose the core of the Rh complex, display significant sequence similarities (∼36%) and have a similar predicted secondary structure with 12 transmembrane domains. RhAG, but not Rh proteins are glycosylated. They interact together within what is now believed to be a trimeric structure (Conroy et al, 2005; Callebaut et al, 2006). Rh and RhAG are present in other species, and form part of an Rh superfamily in eukaryotes, with homologues expressed in non-erythroid tissues (Huang & Liu, 2001; Huang & Peng, 2005). In the mouse, Rh and Rhag protein homologues exhibit 60% and 79% protein sequence identity to the human proteins and are encoded by a single RH-like gene (Rhd on chromosome 4), and one Rhag gene (on chromosome 17) respectively (Liu & Huang, 1999). The biological function of the proteins in the Rh superfamily is currently the subject of debate. Sequence similarity of the Rh superfamily with the Mep/Amt ammonium transporter family, found in eubacteria, archaebacteria, fungi, plants and invertebrates, as well as functional studies in different systems, suggested a role in ammonium/ammonia transport (Marini et al, 2000; Bakouh et al, 2004; Westhoff et al, 2004; Huang & Peng, 2005; Zidi-Yahiaoui et al, 2005; Ludewig, 2006; Merrick et al, 2006). The study of RBCs with various defects of Rh/RhAG proteins in human and mouse erythrocytes has shown that RhAG/Rhag can effectively mediate facilitated NH3 transport across the membrane (Ripoche et al, 2004). The hypothesis that proteins of the Rh complex could act as a CO2 channel or sensor has also been proposed (Soupene et al, 2002, 2004) [reviewed in Kustu and Inwood (2006), and Burton and Anstee (2008)]. Given that the Rh complex and AE1 (Band 3) are associated in the membrane as a macrocomplex, it has been suggested that these proteins might function as an integrated CO2/O2 gas exchange complex (Bruce et al, 2003). Experimental proof supporting the role of Rh proteins in CO2 transport was obtained recently by measuring the membrane permeability of RBCs to carbon dioxyde and these studies concluded that both RhAG and the water channel AQP1 (aquaporin-1) play an equivalent role (Endeward et al, 2006, 2008). A recent study of patients with over-hydrated hereditary stomatocytosis identified RhAG functional mutations responsible for a large monovalent cation leak, suggesting that RhAG might also act as a cation pore (Bruce et al, 2009).
Of the other proteins in the Rh complex, CD47 glycoprotein is a broadly expressed member of the immunoglobulin (Ig) superfamily with five membrane-spanning regions (Brown & Frazier, 2001). On the erythrocyte it is notably a phagocytosis-inhibiting ‘marker of self’, interacting with immune inhibitory receptor SIRPα (signal regulatory protein alpha) expressed by macrophages (Oldenborg et al, 2000, 2001; Subramanian et al, 2006a). ICAM-4 is an erythroid specific glycoprotein with two Ig-like domains, belonging to the family of intercellular adhesion molecules (Bailly et al, 1994, 1995). ICAM-4 is a ligand for a large repertoire of integrins present on leucocytes, macrophages, platelets and endothelial cells (Hermand et al, 2000, 2004; Ihanus et al, 2007) and is implicated in a variety of physiological processes under normal and pathological conditions (Toivanen et al, 2008). Red cell adhesion molecules may participate in normal RBC physiology through cellular interactions with counter-receptors present on macrophages (in bone marrow, spleen and liver) as well as in RBC physiopathology, for instance sickle cell disease, through abnormal adhesion to counter-receptors present on the vascular endothelium (Frenette, 2004; Hebbel et al, 2004). Adhesive molecules, such as CD36 and α4β1 integrin (on reticulocytes), CD47, LuBCAM and ICAM-4 have been involved (Spring & Parsons, 2000; Telen, 2005; Cartron & Elion, 2008).
In an attempt to better understand the physiological/biological function of proteins of the Rh complex, and with the aim of developing mouse models of Rhd and Rhag dependent Rhnull syndrome, we targeted each of these genes using insertional vectors (Zheng et al, 1999) and analysed the resulting knockout (KO) mice, as well as their double-KO progeny.
Materials and methods
- Top of page
- Materials and methods
- Supporting Information
Targeted disruption of mouse Rhd and Rhag genes
Insertional vectors were isolated from the mouse 129S5 genomic 3′hprt library (Zheng et al, 1999). Two vectors were selected: (i) the Rhd KO vector Rh3-1, containing a 7 kb genomic fragment of Rhd from which a 688 bp SacI fragment was deleted to produce a ‘gapped’ targeting vector Rh3-1ΔSacI (Goossens et al, 2006); (ii) the Rhag KO vector Rhag2-1, containing a 7·1 kb genomic fragment of Rhag which after mutagenesis and gapping yielded a targeting vector Rhag2-1ΔBglII. Embryonic stem cells derived from 129/Ola mice were transfected with the linearised targeting vectors by electroporation. Gap repair insertional targeting was confirmed by Southern analysis with EcoRI for Rh3-1ΔSacI and EcoRV for Rhag2-1ΔBglII using polymerase chain reaction (PCR) fragments from the gapped region as probes. Targeting resulted in duplication of the whole region of homology, flanking the inserted vector sequence (Fig 1).
Recombinant clones were microinjected into blastocysts of C57BL/6 mice and chimeric mice thus generated were bred with female C57BL/6 mice. For the Rhd gene KO, wild-type (WT), homo- and heterozygous animals were initially distinguished by Southern analysis. NdeI digestion generated an 11·3 kb fragment in the WT allele and a 28·4 kb fragment in the targeted allele. Further screening of both Rhd and Rhag KOs was carried out by a combination of PCRs (i) with primers 5′ and 3′ to the duplicated region, respectively, amplifying for Rhd a 7·5 kb, for Rhag a 7·2 kb fragment in WT and heterozygous mice, but not in homozygotes (the corresponding ∼25 kb fragments were not amplified in the conditions used) and (ii) with one primer in the duplicated region and the other in the vector backbone such that a 0·2 kb fragment was amplified in mice with the insertion, i.e. heterozygotes and homozygotes (Fig 1).
For both targeted genes, heterozygous mating pairs were crossed to obtain homozygous mutants −/−, heterozygous mutants +/−, and WT littermates +/+, on a hybrid 129/Ola-C57Bl/6 background. Double KO Rhd−/−Rhag−/− as well as single KO Rhd−/− or Rhag−/− and WT littermate controls were obtained by crossing Rhd−/− with Rhag−/− mice. Experiments were performed on age- and sex-matched controls. All animals were treated in compliance with French and European Union animal welfare policies.
Mouse phenotypic analysis
Phenotyping tests were performed on the Rhd−/− and Rhag−/− mice, WT (+/+) littermate controls and +/− littermates. Body weight, body mass index and dysmorphology screen of the mice were performed at 9–10 weeks of age. After 1 week, mice were placed individually into metabolic cages for a 24-h urine collection. Urine chemistry and urine protein excretion were evaluated. At the age of 12–13 weeks, blood was collected after 16 h fasting by retro-orbital puncture for basic chemistry and haematology analysis (the latter on a Beckmann Coulter Ac-Tdiff VET™, Villepinte, France). At the end of the study (14–15 week-old mice) animals were sacrificed, and macroscopic examination, body mass index measurements and collection and weighing of spleen, liver, kidney, heart and lung were performed. Histological and bone marrow analysis were performed on two mutant mice and one control of each sex at 4 months of age. Further haematology analyses were carried out on mice at 8 months. A comparison of Rhd−/−, Rhag−/− and double KO Rhd−/−Rhag−/− mice was carried out on a second cohort.
The following monoclonal antibodies were used : rat anti-mouse CD47 (clone miap 301), CD49d (clone R1-2), CD51 (clone RMV-7) and CD147 (clone RL73), hamster anti-mouse CD29 (clone HMβ1-1) and CD61 (clone 2C9.G2), and KO mouse anti-mouse CD36 (clone JC63·1) were from BD Biosciences Pharmingen (Le Pont de Claix, France); human anti-Band 3(HIRO-58-Dib) was a generous gift from M. Uchikawa (Japanese Red Cross Central Blood Centre, Tokyo). Murine Rh and Rhag proteins (mRh and mRhag) were detected by rabbit polyclonal antibodies raised against the C-terminal regions of the human Rh polypeptide (MPC8, cross-reactive with mRh) and of the mouse Rhag glycoprotein, respectively, as previously described (Hermand et al, 1993; Mouro-Chanteloup et al, 2003). Rabbit polyclonal antibody to murine protein 4·2 was a gift of L. Peters (The Jackson Laboratory, ME, USA) (Gwynn et al, 1997).
Rabbit polyclonal anti-murine ICAM-4 antibody L267 was raised in rabbit against a synthetic peptide (sequence SVTSAPFWVRLNPELEAV) coupled to keyhole limpet haemocyanin (NeoMPS, Strasbourg, France). The peptide sequence corresponded to the mouse counterpart of the human ICAM-4 domain interacting with αVβ3 integrins (Hermand et al, 2004; Mankelow et al, 2004). Antibody reactivity in immunoblot was specifically blocked after preincubation of the antibody with the peptide immunogen.
Flow cytometry analysis of erythrocyte antigens
Indirect immunofluorescence of mouse RBCs for cell-surface or intracellular epitope analysis was performed with a FACScalibur flow cytometer (BD Biosciences Pharmingen) as previously described (Gane et al, 2001).
Red cell membrane protein analysis
Membrane proteins from whole ghost lysates were prepared by hypotonic lysis, loaded at 40 μg per lane, separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis(10% acrylamide) as previously described (Mouro-Chanteloup et al, 2002) and immunoblotted with the relevant antibodies.
Osmotic fragility assay
Osmotic fragility was measured in male Rhd−/−, Rhag−/−, Rhd−/−Rhag−/− and WT control erythrocytes according to a procedure adapted from Gilligan et al (1999) and Godal et al (1979). Briefly, blood was collected into EDTA by retro-orbital puncture either tested directly or after incubation for 15 h at 37°C, 100 μl were washed twice in 1ml of mouse isotonic saline and resuspended at a final haematocrit of 50%. Lysis medium of different osmolalities was prepared from a stock solution of NaCl (in mmol/l:1·54, Na2HPO4 38·14, an NaH2PO4 20·25). Ten microlitre of erythrocyte suspension were added to 290 μl of lysis medium in microplates and incubated at 20°C for 45 min. After centrifugation, supernatant was removed and absorbance measured at 540 nm. The percentage of erythrocyte lysis was calculated and a fragility curve generated.
NH3 and CH3NH2 uptake by red cell ghosts
RBC ghost preparation. All preparation steps, except resealing (37°C), were carried out at 4°C. Each volume of fresh mouse blood was washed three times in phosphate-buffered saline (PBS) and resuspended in 80 volumes of hypotonic lysis buffer (in mmol/l:K gluconate 3·5, HEPES/KOH 10, pH 7·2) for 40 min on ice with gentle agitation. After 15 min centrifugation at 27 000 × g, pellets were resuspended in resealing buffer (in mmol/l:K gluconate 100, HEPES 10, pH 7·2 with KOH) containing 1 mmol/l Mg gluconate and, in stopped-flow experiments, the fluorescent pH-sensitive dye, pyranine (1-hydroxypyrene-3,6,8-trisulfonic acid, Sigma-Aldrich, Saint Quentin Fallavier, France) 0·15 mmol/l. The mixture was incubated at 37°C for 1 h with gentle shaking. After three washes in incubation medium without the fluorescent probe and Mg gluconate (three centrifugations at 27 000 × g), ghosts were kept on ice before assays in the same medium.
pHi determination by stopped-flow analysis. Ghost pellets were diluted at 1–2% cytocrit in the incubation medium and mixed (vol/vol) with ammonium or methylammonium buffer (in mmol:K gluconate 80, ammonium or methylammonium gluconate 20, HEPES 10, pH 7·2 with KOH), generating an inwardly directed 10 meq NH4+ or CH3NH3+ gradient with a stopped-flow instrument (SFM3, Bio-Logic, Grenoble, France), as describe previously (Ripoche et al, 2004, 2006). Transport was measured in iso-osmotic conditions (200 mOsmol/Kg H2O) at 15°C. The excitation wavelength was 465 nm and the emitted light was filtered with a 520 nm cut-on filter. The intracellular pH-dependent fluorescence changes were followed and analysed, as fluorescence increase corresponded to pH elevation. Over the pH range used (6·8–7·8) the relative fluorescence of the dye was proportional to pH, as determined by titration on ghosts incubated in 2 ml incubation medium in the presence of valinomycin 10−6 mol/l and FCCP 10−5 mol/l (Sigma-Aldrich) and submitted to pH change by step additions of 2 μl KOH (1N). Data from five to eight time courses were averaged and the rates of alkalinisation were fitted to a single exponential function by using the Simplex procedure of the Biokine software (Bio-Logic). Kinetic rate constants (k) were compared as different values of k correspond to different ghost permeabilities (Ripoche et al, 2004).
Phenylhydrazine-induced stress erythropoiesis
To analyse the stress response to phenylhydrazine-induced haemolysis, 60 mg/kg phenylhydrazine (PHZ) was injected intraperitoneally on days 0 and 1 to double KO Rhd−/−Rhag−/− mice and WT (Rhd+/+Rhag+/+) littermate controls. Seven mice of each type were included. Blood was drawn on days 0–4–6–8–12 and cell count was performed on a Melet Schloesing MS-9 counter (using the mouse programme). Reticulocyte content was measured by flow cytometry after labelling by Retic-count (BD Biosciences, CA, USA). To evaluate haematopoietic potential, clonogenic assays were performed on bone marrow and spleen from double KO and control mice in basal state, or on day 4 after PHZ treatment, as described (Foudi et al, 2006). Cells were plated respectively at 5 × 104 (bone marrow) and 105 (spleen) in methylcellulose (StemCell Technologies, Grenoble, France) according to the manufacturer’s instructions. All experimental samples were performed in duplicate for each animal. Erythroid burst-forming units (BFU-E), granulocyte-macrophage (CFU-GM) and mixed lineage (CFU-GEMM) colony-forming units were counted on day 8, using an inverted microscope (Zeiss TELAVAL 3; AFT Micromécanique, Fillinges, France). Results were expressed as ‘erythroid’ colonies (combined counts for BFU-E, E-M and CFU-GEMM) and ‘myeloid’ colonies (CFU-GM and CFU-M).
Cell culture media and reagents were from Gibco® (Invitrogen, France). Except when otherwise mentioned, reagents were purchased from Sigma-Aldrich (Saint Louis, MO, USA). The osmolality of PBS and Hank’s balanced salt solution (HBSS) was adjusted to 320 mOsm/kg H2O by NaCl addition, to prevent haemolysis of mouse RBCs. HBSS contained 1 mmol/l CaCl2 and 1 mmol/l MgCl2. In some experiments, the endothelial monolayer was incubated with a synthetic peptide T(91)REATARI (T-8-I) derived from one of the αV-binding domains of murine ICAM-4 and a control peptide A (76)WNSLAHC (A-8-C). These murine counterparts of peptides described by Kaul et al. (2006) were synthesised, purified by reverse phase HPLC, and analysed by matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (Genosphere Biotechnologies, Paris, France).
Preparation and culture of endothelial cells. Mouse endothelioma cell line bEnd.3 (American Type Culture Collection, Manassas, VA, USA), established from polyoma-induced hemangioma in brain (Montesano et al, 1990), was chosen as the endothelial system. Cells were grown in Iscove’s modified Dulbecco’s medium (IMDM) containing 10% heat inactivated fetal calf serum (FCS), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0·25 μg/ml amphotericin. In all experiments, the bEnd.3 cells were used at passages 26–28. In some experiments, endothelial monolayers were treated with 200μmol/l peptide for 30 min before RBC injection.
Preparation and treatment of RBCs. Blood samples (50–150 μl) were collected by retro-orbital bleeding. Purified RBCs were recovered after centrifugation on Ficoll-Paque Plus (Amersham, Saclay, France) following the elimination of platelets and white cells. RBCs were fluorescently labelled with either PKH67 (green) or PKH26 (red), according to the manufacturer’s instructions.
Erythrocyte adhesion to endothelial cells under flow conditions. Endothelial cells were cultured in flat glass capillaries according to a method previously described (Cooke et al, 1993). A suspension of bEnd3 cells (6 × 106/ml IMDM with 10% FCS) was injected into gelatine (2%)-coated flat glass capillaries (microslides, Camlab, Cambridge, UK) and left to adhere for 2 h. Cells were then grown to confluence for 48 h, with replacement of the capillary medium every 2 h. Microslides were then mounted on the stage of a video camera-equipped stereomicroscope (Leica Microsystems, Nanterre, France). One end of the microslide was connected to a syringe pump (Harvard Apparatus, Holliston, MA) to control flow rate, which determined shear stress levels. The other end was connected to an electronic valve allowing switching between RBC suspension and cell-free washing buffer (HBSS). The differently labelled experimental and control RBCs were mixed in HBSS at a final 0·2% haematocrit each, before injection into microslides. The RBC mixture was injected into the microslide at a flow rate achieving a wall shear stress of 0·02 Pa for 5 min. RBCs that had not adhered to the endothelial monolayer during the infusion time were removed by infusion of cell-free HBSS at the same shear stress for 5 min, and then stepwise at increasing shear stress of 0·04, 0·1, 0·2, and 0·4 Pa, every 5 min. At the end of each wash period, images of nine consecutive fields of known surface area along the centre line of the microslide were recorded under fluorescent philtres adequate for discriminating experimental from control RBCs (Optimas G.S., Silver Spring, MD). The number of RBCs in each field was counted and averaged over the nine fields. Adhesion assays were conducted either under basal conditions or after incubation of the endothelial monolayer with synthetic peptides T-8-I and A-8-C (250 μmol/l), for 30 min at room temperature. Results are expressed as means ± standard error of the mean (SEM) for four different mice. Data were analysed using analysis of variance (anova), with the post hoc Dunnett test using StatView software (SAS Institute, Cary, NC, USA). Statistical significance was considered for P values <0·05.
- Top of page
- Materials and methods
- Supporting Information
Generation of mice with inactivated Rhd or Rhag genes
Mouse Rhd and Rhag genes were disrupted by gap repair insertional targeting using the method described by Zheng et al (1999). KO status of the animals was determined by Southern blot or PCR as exemplified in Fig 1. Breeding of heterozygous F1 mice led to WT (+/+), heterozygous (+/−), and homozygous KO (−/−) offspring with classical Mendelian ratios. Double KO mice and controls were obtained by crossing homozygous Rhd and Rhag KO.
The Rhd or Rhag nullizygous mice and the double KO exhibited normal growth, development and fertility. In these mice no significant change in body weight and body mass index was observed. Macroscopic examination and the weight of selected organs (spleen, liver, kidney, heart and lung) showed no significant differences between control and targeted mice in either the Rhd or the Rhag KO. There were no significant differences in plasma chemistry parameter levels of KO mice as compared to the controls. Notably, there were no signs of haemolysis (bilirubin; see Table SIA). No significant abnormalities in kidney function were detected by blood and urine measurements (data not shown).
In a comparative study at 17 weeks, iron levels were slightly increased in both male and female Rhd−/− and Rhag−/− mice while ferritin levels showed a tendency toward decrease in both male and female Rhag−/− and in female Rhd−/−. A trend towards a decrease in transferrin levels was noted in Rhd−/− and Rhag−/−, but this was not statistically significant except in male Rhag−/−. However, in the double KO mice, no significant changes were observed in iron, transferrin, or in ferritin levels (Table SIB).
Rh complex in RBCs from targeted mice with Rhd or Rhag inactivation
The protein components of the Rh complex on RBCs from KO animals were investigated by flow cytometry and Western blot analysis. Flow cytometric analysis of RBCs from Rhd−/−, Rhd+/−, and Rhd+/+ mice showed a loss of Rh protein expression in Rhd KO mice but no significant differences between heterozygote and control mice (Fig 2, top). Rhag glycoprotein expression was reduced in Rhd−/− mice to approximately 70% of controls, while CD47 expression remained unchanged. AE1 (Band 3) expression was not modified. RBCs from Rhag targeted mice showed a loss of Rhag expression in homozygote and a ∼50% decrease in heterozygotes as compared to littermate controls (Fig 2, bottom). Rh protein expression was also lost in Rhag−/− mice, and reduced to approximately 50% in heterozygotes. CD47 expression remained unchanged in all mice, as was AE1.
Immunoblot analysis of Rhd−/− erythrocyte membrane proteins confirmed the complete absence of Rh protein and mICAM4 and the presence of Rhag and CD47. AE1 and protein 4-2 (the latter associated withCD47) were also present at normal levels (Fig 3). In RBC membranes of Rhag KO mice, Rh protein and mICAM4 were absent while CD47, AE1 and protein 4-2 were present (Fig 3). Expression of these proteins for Rhd−/−Rhag−/− mice was the same as for single Rhag KO animals (data not shown).
Ammonium and methylammonium transport is reduced in red cell ghosts from Rhag−/− mutant mice
The pH-dependent fluorescent changes of ghosts from Rhag+/+ and Rhag−/− RBCs submitted to a 10 meq inwardly directed ammonium and methylammonium gradient in iso-osmotic conditions are shown in Fig 4. As fluorescence increase corresponded to intracellular pH elevation, there was clearly an alkalinisation that could be accounted for by the uptake of NH3 or CH3NH2 and their protonation within the cells until the equilibrium was reached.
The calculated alkalinisation rate constants (k) of RBC ghosts from WT control cells were 15·5 ± 0·4/s and 1·1 ± 0·07/s for Rhag+/+ (n = 5), and 14·5 ± 1·2/s and 0·98 ± 0·04/s for Rhd+/+ (n = 5), for NH3 and CH3NH2 respectively. The alkalinisation rate constants were slightly higher than the values observed in previous experiments (Ripoche et al, 2004). This is due to the anion used in this study, which is gluconate, reducing the interference between the alkalinisation caused by the NH3 or CH3NH2 uptake and the acidification caused by AE1 (Band 3) activity. In comparison, k values for ghosts from Rhag−/− mice were reduced by 70–85% (4·4 ± 0·1/s and 0·15 ± 0·004/s for NH3 and CH3CH2 respectively, n = 5, P < 0·001). These last values probably correspond to the permeability across lipids or other non-identified proteins. Similar experiments performed with ghosts from homozygous Rhd−/− animals revealed a 30–40% reduction of k values for NH3 (8·4 ± 0·21/s) and CH3NH2 (0·7 ± 0·04/s) transport as compared to WT control cells (Fig 4), correlating well with the 30% decrease of Rhag expression in these animals detected by flow cytometry (see above).
Defective ICAM-4 dependent adhesion of Rhd and Rhag KO RBCs to endothelial cells
The absence of ICAM-4 on Rhd and Rhag KO RBCs (Fig 3) and the lack of Lu/BCAM on mouse RBCs (Rahuel et al, 2008), prompted us to examine the adhesive interactions of Rhd or Rhag KO RBCs with bEnd3 endothelial cells. As αVβ3 (CD51/CD61) is the counter receptor of ICAM-4 of endothelium vessels (Hermand et al, 2004; Mankelow et al, 2004), its presence was first confirmed on bEnd.3 cells (data not shown). Rhd−/− and Rhag−/− RBCs were shown to exhibit CD36, CD47 and CD147 levels similar to that of their respective WT littermates (Fig S1). Integrins α4β1 and αVβ3 could not be detected at the surface of either WT or KO RBCs (not shown). Basal adhesion to the bEnd.3 endothelial cell line in flow conditions of both Rhd−/− and Rhag−/− RBCs was significantly (P < 0·05) lower than that of WT RBCs. This difference was found irrespective of shear stress applied (Fig 5 left panel). At the highest shear stress used, 0·4 Pa, adhesion levels were lower by about 50% for Rhd−/− and 25% for Rhag−/− cells. To determine whether defective ICAM-4 expression on Rhd−/− and Rhag−/− RBCs was indeed responsible for the decreased adhesion to endothelial cells, a bEnd.3 monolayer was pre-incubated either with the synthetic octapeptide T-8-I derived from one of the αV-binding domains of ICAM-4 or with the irrelevant A-8-C control octapeptide. As shown in Fig 5 (right panel), adhesion in the presence of the latter was lower for Rhd−/− RBCs than for WT (Rhd+/+) RBCs, both at 0·1 and 0·4 Pa shear stress. On the other hand, adhesion to endothelia pretreated with ICAM-4 peptide T-8-I remained unchanged for Rhd−/− RBCs, while for WT RBCs it was reduced to the same level.
Rhd and Rhag KO mice show no major alterations in erythrocyte parameters, blood cell count, blood cell morphology or histology of spleen and bone marrow
In two successive studies of Rhd−/− male and female mice aged 12–13 weeks, RBC counts were not significantly modified, but the mice displayed a trend toward decreased haemoglobin and haematocrit levels, a slight microcytosis accompanied by a slight hypochromia of RBCs. A similar trend was observed for double KO mice, but not for Rhag−/− mice, the results for which were much more heterogeneous (Table SII). Microscopic analyses did not show abnormalities in RBC morphology. At 8 months a slightly increased reticulocytosis was observed in Rhag−/− mice. A further study of mice at 1 year showed no significant differences between KO mice and their WT controls. No significant variations were observed in the other blood cell lineages (data not shown). When tested for resistance to osmotic stress, RBCs from Rhd−/−, Rhag−/−, and Rhd−/−Rhag−/− did not differ significantly from that of controls (Fig S2). Haematoxylin/eosin staining performed on histological sections of the liver, the spleen and the humerus bone of two mutant mice and one control of each sex at 4 months of age showed no differences between mutant and control mice (data not shown).
Stress erythropoiesis was not modified in Rhd/Rhag double KO mice
To explore stress erythropoiesis, double KO animals received a PHZ dose of 60 mg/kg on days 0 and 1, which induced intense anaemia with a >50% fall in erythrocyte count by day 4, accompanied by an elevation in production of reticulocytes and concomitant rise in platelets and white blood cells. A return to normal levels was observed by day 12 post-injection. The kinetics of stress-induced erythropoiesis did not differ significantly between Rhd−/−Rhag−/− mice and WT littermates (Fig 6A). Quantification of haematopoietic precursors from bone marrow and spleen in semi-solid cultures showed no significant differences between control and double KO mice, either in basal conditions (not shown) or in PHZ-acute anaemia (Fig 6B).
- Top of page
- Materials and methods
- Supporting Information
This study generated, by insertional targeting, the first mouse strains in which Rhd and Rhag genes were respectively or simultaneously inactivated. The aim was to develop a mouse model for Rhnull haemolytic anaemia and to further dissect the biological role of the Rh membrane complex.
For Rhd, Rhag, or double inactivated mice, the KO animals were indistinguishable from their WT littermates at a gross phenotypic level. No differences in basic plasma and urine chemistry were observed. The mice were not anaemic and showed no obvious anomalies of standard haematological parameters or of erythrocyte morphology or fragility. The slight increase in iron levels observed in Rhd KO mice and Rhag KO mice could suggest a low level haemolysis or a decreased/ineffective erythropoiesis. However, the absence of increased bilirubinaemia and the decrease in ferritin levels were in contradiction with this. Furthermore, in the double KO mice no iron increase was observed. The fact that the slight trends observed were not consistently seen in the different types of KO would seem to argue against their significance. These observations contrast with the stomato-spherocytosis and compensated haemolytic anaemia seen in Rhnull individuals (Sturgeon, 1970; Cartron, 1999), underlining that the KO mouse models are not equivalent to human Rhnull syndrome, and that data from mouse models cannot be directly extrapolated to human pathology, but contribute to a better understanding of the systems involved.
In the Rhag-targeted mouse, there is no Rh protein membrane expression. This is in keeping with published data demonstrating that RhAG is a strictly required post-transcriptional factor for Rh expression (Mouro-Chanteloup et al, 2002). In Rhd KO mice, expression of the Rhag glycoprotein is only slightly reduced, as opposed to the severe reduction in human Rhnull erythrocytes of the amorph type. A third element of the complex, ICAM-4, is also absent in both KO. However, there is no impact in either Rhd−/− or Rhag−/− mice on the expression of CD47, confirming CD47 to be independent of the Rh complex in the mouse. A previous indication that mouse CD47 protein interactions were not identical to human was the respectively altered and unaltered CD47 expression in 4·2-deficient patients and 4·2−/− mice, (Mouro-Chanteloup et al, 2003). Our observations are also consistent with the high mobility of mouse erythrocyte CD47 observed by fluorescence-imaged micro-deformation (Subramanian et al, 2006b). Thus, it is clear that in the mouse the relationship between the core components Rh and Rhag, as well as other members of the Rh complex, differs from that in human erythrocytes.
Rh- and Rhag-deficient mice provided us with a tool for further dissecting NH3/CH3NH2 transport by RBCs. Our results with KO mice confirm the conclusions of studies with human ghosts, particularly human Rhnull variants (RhAG versus RhAGnull of the regulator type) and murine variants with Rh/Rhag protein alteration described earlier (ankyrin or spectrin deficient mice with a reduction in Rh and Rhag) (Ripoche et al, 2004). The finding that ammonia transport is severely altered in mouse Rhag−/− RBC which lack the Rhag protein, but only partially reduced (30–40%) in Rhd−/− RBC, which express substantial amounts of Rhag (70% of WT controls) indicates that ammonia transport is directly proportional to the amount of Rhag protein present on RBCs, as proposed earlier (Ripoche et al, 2004). These findings support the view that RhAG and Rhag are major ammonium transporters in human and mouse erythrocytes, and the lack of a significant role for the Rh protein component in this transport process.
Rhd and Rhag KO RBCs also showed a decreased basal adhesion to the endothelial cell line bEND3 in vitro, for which defective ICAM-4 expression was responsible, as demonstrated by a substantial and significant inhibition using a peptide derived from the interaction domain of ICAM-4 with its counter-receptor αVβ3. Given that αVβ3 is also the counter receptor of ICAM-4 on macrophages (Lee et al, 2006), we expected to observe a difference in sensitivity to PHZ-induced stress haematopoiesis between WT and KO animals. We found no difference in the recovery of double KO mice as compared to controls, when estimated either by RBC count, reticulocyte production, or regeneration of erythroid and myeloid progenitor cells. Lee et al (2006), in their study of mice with targeted disruption of ICAM-4, observed neither anaemia nor any particular red cell phenotype. This report described a defect in the formation of erythroblastic islands that is probably due to a loss of adhesive interaction with the αV-integrin of central macrophages, suggesting that a blunted response might be expected in stress conditions. Observations from our model would seem to indicate that ICAM-4 might be dispensable during stress erythropoiesis. It remains to be determined whether the loss of ICAM-4 can be compensated in vivo by interactions through other proteins ensuring the integrity of erythroblastic islands. It should be noted that the rare known human subjects with an ICAM-4 deficiency exhibit no phenotypic abnormality (Cartron, 2008). As a reduced adhesion of RBCs to endothelium is a critical factor in pathological models, such as sickle cell disease (Frenette, 2004; Hebbel et al, 2004), Rhd and Rhag KO mice may represent a useful complement to analyse the impact of ICAM-4 in this disease (Kaul et al, 2006; Zennadi et al, 2007).
The Rhd, Rhag and double KO mice will also prove useful for more detailed understanding of the physiological roles of Rh and Rhag proteins and further comparison with human Rh or RhAGnull erythrocytes, particularly regarding erythrocyte cation transport which has been found altered in human Rhnull cells (Lauf & Joiner, 1976; Ballas et al, 1984; Lauf et al, 2009), or the membrane stability and organisation of the Rh complex in mouse. Additionally, studies on CO2 gas transport by human RBC, thought to be mediated by RhAG and the water channel AQP1 (Endeward et al, 2006, 2008; Bruce, 2008), might be extended to a similar analysis of murine RBCs from Rhag KO, Aqp1 KO animals and double KOs to estimate their physiological impact.
- Top of page
- Materials and methods
- Supporting Information
This work was supported by grants from the Association Recherche Transfusion (Contracts 25-2001 and 34-2003), GIS-Maladies Rares (support for part of the analysis at the Institut Clinique de la Souris, Illkirch), SESAME 2007 (F-08-1104/R) and Cancéropole Ile de France. The authors would like to thank Pr. Allan Bradley (Wellcome Trust Sanger Institute, Cambridge UK) for the 3′hprt vector library. They also wish to thank Dr. Thérèse Cynober (Hôpital Kremlin-Bicêtre) for her help with haematological analysis, Viviane Bony who participated in initial ES experiments and Pierre Gane (INTS, Paris) for the flow cytometry; Marie-Paule Wautier for putting the adhesion equipment at our disposal, Isabelle Mouro-Chanteloup and Cécile Rahuel (Inserm U665, Paris) for helpful discussions.
To the memory of Charles Babinet (Institut Pasteur, Paris) who counselled and encouraged this project.
- Top of page
- Materials and methods
- Supporting Information
- 1991) Molecular biology of the Rh antigens. Blood, 78, 551–563. & (
- 2000) The Rh blood group system: a review. Blood, 95, 375–387. & (
- 2006) Molecular biology of Rh proteins and relevance to molecular medicine. Expert Reviews in Molecular Medicine, 8, 1–20. , , , , & (
- 1994) The LW blood group glycoprotein is homologous to intercellular adhesion molecules. Proceedings of the National Academy of Sciences of the United States of America, 91, 5306–5310. , , , , , & (
- 1995) The red cell LW blood group protein is an intercellular adhesion molecule which binds to CD11/CD18 leukocyte integrins. European Journal of Immunology, 25, 3316–3320. , , , & (
- 2004) NH3 is involved in the NH4+ transport induced by the functional expression of the human Rh C glycoprotein. Journal of Biological Chemistry, 279, 15975–15983. , , , , , & (
- 1984) Red cell membrane and cation deficiency in Rh null syndrome. Blood, 63, 1046–1055. , , , , , , & (
- 2001) Integrin-associated protein (CD47) and its ligands. Trends in Cell Biology, 11, 130–135. & (
- 2008) Red cell membrane transport abnormalities. Current Opinion in Hematology, 15, 184–190. (
- 2003) A band 3-based macrocomplex of integral and peripheral proteins in the RBC membrane. Blood, 101, 4180–4188. , , , , , , , & (
- 2009) The monovalent cation leak in overhydrated stomatocytic red blood cells results from amino acid substitutions in the Rh-associated glycoprotein. Blood, 113, 1350–1357. , , , , , , , , & (
- 2008) Structure, function and significance of Rh proteins in red cells. Current Opinion in Hematology, 15, 625–630. & (
- 2006) Hydrophobic cluster analysis and modeling of the human Rh protein three-dimensional structures. Transfusion Clinique et Biologique, 13, 70–84. , , , , , , & (
- 1999) RH blood group system and molecular basis of Rh-deficiency. Bailliere’s Best Practice & Research. Clinical Haematology, 12, 655–689. (
- 2008) Blood group antigens structure and function: recent advances. Hematology Education: Education Programme for the Annual Congress of the European Hematology Association, 2008, 158–174. (
- 2008) Erythroid adhesion molecules in sickle cell disease: effect of hydroxyurea. Transfusion Clinique et Biologique, 15, 39–50. & (
- 2005) Modelling the human rhesus proteins: implications for structure and function. British Journal of Haematology, 131, 543–551. , , & (
- 1993) A simplified method for culture of endothelial cells and analysis of adhesion of blood cells under conditions of flow. Microvascular Research, 45, 33–45. , , & (
- 2006) Evidence that aquaporin 1 is a major pathway for CO2 transport across the human erythrocyte membrane. FASEB Journal, 20, 1974–1981. , , , , , , , , & (
- 2008) RhAG protein of the Rhesus complex is a CO2 channel in the human red cell membrane. FASEB Journal, 22, 64–73. , , & (
- 2006) Reduced retention of radioprotective hematopoietic cells within the bone marrow microenvironment in CXCR4−/− chimeric mice. Blood, 107, 2243–2251. , , , , , , , & (
- 2004) Sickle cell vasoocclusion: heterotypic, multicellular aggregations driven by leukocyte adhesion. Microcirculation, 11, 167–177. (
- 2001) Flow cytometric analysis of the association between blood group-related proteins and the detergent-insoluble material of K562 cells and erythroid precursors. British Journal of Haematology, 113, 680–688. , , , , , , & (
- 1999) Targeted disruption of the beta adducin gene (Add2) causes red blood cell spherocytosis in mice. Proceedings of the National Academy of Sciences of the United States of America, 96, 10717–10722. , , , , & (
- 1979) The osmotic fragility of red blood cells: a re-evaluation of technical conditions. Scandinavian Journal of Haematology, 23, 55–58. , & (
- 2006) Generation of mice with inactivated Rh or Rhag genes. Transfusion Clinique et Biologique, 13, 164–166. , , , & (
- 1997) The gene encoding protein 4.2 is distinct from the mouse platelet storage pool deficiency mutation pallid. Genomics, 42, 532–535. , , , & (
- 2004) The endothelial biology of sickle cell disease: inflammation and a chronic vasculopathy. Microcirculation, 11, 129–151. , & (
- 1993) Immunochemical characterization of rhesus proteins with antibodies raised against synthetic peptides. Blood, 82, 669–676. , , , , , , & (
- 2000) Binding sites of leukocyte beta 2 integrins (LFA-1, Mac-1) on the human ICAM-4/LW blood group protein. Journal of Biological Chemistry, 275, 26002–26010. , , , , , , & (
- 2004) Integrin receptor specificity for human red cell ICAM-4 ligand. Critical residues for alphaIIbeta3 binding. European Journal of Biochemistry, 271, 3729–3740. , , , , & (
- 2001) New insights into the Rh superfamily of genes and proteins in erythroid cells and nonerythroid tissues. Blood Cells, Molecules, and Diseases, 27, 90–101. & (
- 2005) Evolutionary conservation and diversification of Rh family genes and proteins. Proceedings of the National Academy of Sciences of the United States of America, 102, 15512–15517. & (
- 2000) Molecular biology and genetics of the Rh blood group system. Seminars in Hematology, 37, 150–165. , & (
- 2007) Red-cell ICAM-4 is a ligand for the monocyte/macrophage integrin CD11c/CD18: characterization of the binding sites on ICAM-4. Blood, 109, 802–810. , , , & (
- 2006) Peptides based on alphaV-binding domains of erythrocyte ICAM-4 inhibit sickle red cell-endothelial interactions and vaso-occlusion in the microcirculation. American Journal of Physiology. Cell Physiology, 291, C922–C930. , , , , , , , , & (
- 2006) Biological gas channels for NH3 and CO2: evidence that Rh (Rhesus) proteins are CO2 channels. Transfusion Clinique et Biologique, 13, 103–110. & (
- 1976) Increased potassium transport and ouabain binding in human Rhnull red blood cells. Blood, 48, 457–468. & (
- 2009) Role of Rh (rhesus) complex in K-Cl cotransport (KCC) regulation in red blood cells (RBCs) of Rh and Rh-associated glycoprotein (RhAg) knockout (KO) mice. Faseb Journal, 23, (abstract) 797.795. , , & . (
- 2006) Rh proteins: key structural and functional components of the red cell membrane. Blood Reviews, 20, 93–110. , & (
- 2006) Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythroblastic island formation. Blood, 108, 2064–2071. , , , , , , , , & (
- 1999) The mouse Rhl1 and Rhag genes: sequence, organization, expression, and chromosomal mapping. Biochemical Genetics, 37, 119–138. & (
- 2006) Ion transport versus gas conduction: function of AMT/Rh-type proteins. Transfusion Clinique et Biologique, 13, 111–116. (
- 2004) Identification of critical amino-acid residues on the erythroid intercellular adhesion molecule-4 (ICAM-4) mediating adhesion to alpha V integrins. Blood, 103, 1503–1508. , , , , , & (
- 2000) The human Rhesus-associated RhAG protein and a kidney homologue promote ammonium transport in yeast. Nature Genetics, 26, 341–344. , , , , & (
- 2006) The Escherichia coli AmtB protein as a model system for understanding ammonium transport by Amt and Rh proteins. Transfusion Clinique et Biologique, 13, 97–102. , , , , , , & (
- 1993) Blood Transfusion in Clinical Medicine. Blackwell Science Ltd., London. , & (
- 1990) Increased proteolytic activity is responsible for the aberrant morphogenetic behavior of endothelial cells expressing the middle T oncogene. Cell, 62, 435–445. , , , , & (
- 2002) Cell-surface expression of RhD blood group polypeptide is posttranscriptionally regulated by the RhAG glycoprotein. Blood, 100, 1038–1047. , , , , , , & (
- 2003) Evidence that the red cell skeleton protein 4.2 interacts with the Rh membrane complex member CD47. Blood, 101, 338–344. , , , , , , , , & (
- 2003) Rh-RhAG/ankyrin-R, a new interaction site between the membrane bilayer and the red cell skeleton, is impaired by Rh(null)-associated mutation. Journal of Biological Chemistry, 278, 25526–25533. , , , , , & (
- 2006) Functional interaction between Rh proteins and the spectrin-based skeleton in erythroid and epithelial cells. Transfusion Clinique et Biologique, 13, 23–28. , , , , , , , & (
- 2000) Role of CD47 as a marker of self on red blood cells. Science, 288, 2051–2054. , , , , & (
- 2001) CD47-signal regulatory protein alpha (SIRPalpha) regulates Fcgamma and complement receptor-mediated phagocytosis. Journal of Experimental Medicine, 193, 855–862. , & (
- 2008) Genetic inactivation of the laminin alpha5 chain receptor Lu/BCAM leads to kidney and intestinal abnormalities in the mouse. American Journal of Physiology. Renal Physiology, 294, F393–F406. , , , , , , , , & (
- 2004) Human Rhesus-associated glycoprotein mediates facilitated transport of NH(3) into red blood cells. Proceedings of the National Academy of Sciences of the United States of America, 101, 17222–17227. , , , , & (
- 2006) Role of RhAG and AQP1 in NH3 and CO2 gas transport in red cell ghosts: a stopped-flow analysis. Transfusion Clinique et Biologique, 13, 117–122. , , , , , & (
- 2002) Rhesus expression in a green alga is regulated by CO(2). Proceedings of the National Academy of Sciences of the United States of America, 99, 7769–7773. , , , , , & (
- 2004) Lack of the Rhesus protein Rh1 impairs growth of the green alga Chlamydomonas reinhardtii at high CO2. Proceedings of the National Academy of Sciences of the United States of America, 101, 7787–7792. , & (
- 2000) Erythroid cell adhesion molecules. Transfusion Medicine Reviews, 14, 351–363. & (
- 1970) Hematological observations on the anemia associated with blood type Rhnull. Blood, 36, 310–320. (
- 2006a) Species- and cell type-specific interactions between CD47 and human SIRPalpha. Blood, 107, 2548–2556. , , , & (
- 2006b) The ‘metabolon,’ CD47, and the ‘phagocytic synapse’: molecular co-localization and species divergence. Transfusion Clinique et Biologique, 13, 31–38. , & (
- 2005) Erythrocyte adhesion receptors: blood group antigens and related molecules. Transfusion Medicine Reviews, 19, 32–44. (
- 2008) Importance of molecular studies on major blood groups--intercellular adhesion molecule-4, a blood group antigen involved in multiple cellular interactions. Biochimica et Biophysica Acta, 1780, 456–466. , , , & (
- 2004) Mechanism of genetic complementation of ammonium transport in yeast by human erythrocyte Rh-associated glycoprotein (RhAG). Journal of Biological Chemistry, 279, 17443–17448. , , & (
- 2007) Epinephrine-induced activation of LW-mediated sickle cell adhesion and vaso-occlusion in vivo. Blood, 110, 2708–2717. , , , , , , , & (
- 1999) A system for rapid generation of coat color-tagged knockouts and defined chromosomal rearrangements in mice. Nucleic Acids Research, 27, 2354–2360. , & (
- 2005) Human Rhesus B and Rhesus C glycoproteins: properties of facilitated ammonium transport in recombinant kidney cells. Biochemical Journal, 391, 33–40. , , , , , , , & (
- Top of page
- Materials and methods
- Supporting Information
Fig S1. Adhesion molecules of RBCs from Rh−/− and Rhag−/− mice.
Fig S2. Osmotic fragility of RBCs from Rh and Rhag targeted mice.
Tables SI-AB. -A bilirubin and -B ferritin, iron and transferrin values in KO Rh, Rhag and double KO and controls.
Table SII. KO Rh, Rhag and double KO mice show no major alterations in erythrocyte parameters.
Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
|BJH_7928_sm_fig S1.tif||13162K||Supporting info item|
|BJH_7928_sm_fig S2.tif||14056K||Supporting info item|
|BJH_7928_sm_Tables SIA, SIB.doc||56K||Supporting info item|
|BJH_7928_sm_Table SII.doc||64K||Supporting info item|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.