Summary Since the 1980s, major surgical interventions in patients with congenital haemophilia with inhibitors have been performed utilizing bypassing agents for haemostatic coverage. While reports have focused on perioperative management and haemostasis, the US currently lacks consensus guidelines for the management of patients with inhibitors during the surgical procedure, and pre‐ and postoperatively. Many haemophilia treatment centres ( HTC s) have experience with surgery in haemophilia patients, including those with inhibitors, with approximately 50% of these HTC s having performed orthopaedic procedures. The aim of this study was to present currently considered best practices for multidisciplinary care of inhibitor patients undergoing surgery in US HTC s. Comprehensive haemophilia care in the US is provided by ~130 federally designated HTC s staffed by multidisciplinary teams of healthcare professionals. Best practices were derived from a meeting of experts from leading HTC s examining the full care spectrum for inhibitor patients ranging from identification of the need for surgery through postoperative rehabilitation. HTC s face challenges in the care of inhibitor patients requiring surgery due to the limited number of surgeons willing to operate on this complex population. US centres of excellence have developed their own best practices around an extended comprehensive care model that includes preoperative planning, perioperative haemostasis and postoperative rehabilitation. Best practices will benefit patients with inhibitors and allow improvement in the overall care of these patients when undergoing surgical procedures. In addition, opportunities for further education and outcomes assessment in the care of this patient population have been identified.
Introduction: Continuous infusion (CI) of clotting factors as a replacement therapy for perioperative hemostatic protection has been performed for many years, including with factors VIII and IX and recombinant activated factor VII (rFVIIa). This approach provides steady factor levels without requiring frequent administration of bolus doses. Aim: To review safety, efficacy, and dosing data regarding CI of rFVIIa for hemostatic management of patients with congenital hemophilia with inhibitors (CHwI) or congenital factor VII deficiency (C7D). Materials and methods: A literature review identified instances of CI of rFVIIa in patients with CHwI or C7D undergoing surgery or experiencing bleeding episodes. Data regarding safety, efficacy, and dosing were extracted. Results: The safety and efficacy of 50 mcg/kg/h CI of rFVIIa following a 90 mcg/kg bolus injection, vs a standard bolus injection regimen, was reported for 24 patients with CHwI undergoing elective surgery in an open-label, randomized, Phase III trial. Efficacy was similar between CI and bolus injection groups at all postoperative time points assessed. Additionally, a postmarketing surveillance study reported effective (80%) and partially effective (20%) CI of rFVIIa in a Japanese cohort of ten patients with CHwI who underwent 15 surgical procedures. Finally, the safety and dosing of rFVIIa CI in 193 and 26 patients with CHwI and C7D, respectively, were reported in 11 prospective studies, 10 retrospective studies, and 30 case reports. No unexpected safety findings were reported. Conclusion: rFVIIa CI has been performed safely and effectively in patients with CHwI and C7D undergoing surgery and during bleeding episodes in patients with CHwI. Keywords: rFVIIa, continuous infusion, surgery, bleeding
Alternative splicing of a series of 10 contiguous exons present within the CD44 gene can generate a large number of differentially expressed CD44 isoforms that contain additional peptide sequences of varying length inserted into a single site within the extracellular domain of the molecule proximal to the membrane spanning domain. Although distinct functions have been ascribed to certain of these isoforms, the effect of particular inserted domains on the ligand-binding specificity of the CD44 molecule remains unclear. In the present study, we demonstrate that while CD44H, the major CD44 isoform expressed on resting hemopoietic cells, and CD44R1, an alternatively spliced isoform present on transformed epithelial cells and certain activated and/or malignant hemopoietic cell types, can both bind avidly to hyaluronan, only CD44R1 can promote homotypic cellular aggregation when expressed in the CD44-negative murine lymphoma cell line TILL Experiments in which TIL1 cells transduced with different CD44 isoforms were tested for their ability to adhere to one another or to COS7 cells transfected with CD44R1, indicated that CD44R1 can recognize and bind a common determinant present on both CD44H and CD44R1. Monoclonal antibody blocking studies suggest further, that the determinant recognized by CD44R1 is located in a region of the CD44 molecule distinct from that involved in hyaluronan binding. Alternative splicing of a series of 10 contiguous exons present within the CD44 gene can generate a large number of differentially expressed CD44 isoforms that contain additional peptide sequences of varying length inserted into a single site within the extracellular domain of the molecule proximal to the membrane spanning domain. Although distinct functions have been ascribed to certain of these isoforms, the effect of particular inserted domains on the ligand-binding specificity of the CD44 molecule remains unclear. In the present study, we demonstrate that while CD44H, the major CD44 isoform expressed on resting hemopoietic cells, and CD44R1, an alternatively spliced isoform present on transformed epithelial cells and certain activated and/or malignant hemopoietic cell types, can both bind avidly to hyaluronan, only CD44R1 can promote homotypic cellular aggregation when expressed in the CD44-negative murine lymphoma cell line TILL Experiments in which TIL1 cells transduced with different CD44 isoforms were tested for their ability to adhere to one another or to COS7 cells transfected with CD44R1, indicated that CD44R1 can recognize and bind a common determinant present on both CD44H and CD44R1. Monoclonal antibody blocking studies suggest further, that the determinant recognized by CD44R1 is located in a region of the CD44 molecule distinct from that involved in hyaluronan binding. CD44 is a broadly distributed cell surface glycoprotein that has been shown to play an important role in many adhesion-dependent cellular processes including lymphocyte recirculation, hemopoiesis, NK-cell-mediated killing, macrophage and lymphocyte activation, and tumor metastasis (1Haynes B.F. Telen M.J. Hale L.P. Denning S.M. Immunol. Today. 1989; 10: 423Abstract Full Text PDF PubMed Scopus (528) Google Scholar, 2Haynes B.F. Liao H.-X. Patton K.L. Cancer Cells. 1991; 3: 347-350PubMed Google Scholar, 3Herrlich P. Zöller M. Pals S.T. Ponta H. Immunol. Today. 1993; 14: 395-399Abstract Full Text PDF PubMed Scopus (270) Google Scholar, 4Lesley J. Hyman R. Kincade P.W. Adv. Immunol. 1993; 54: 271-335Crossref PubMed Scopus (1035) Google Scholar). Although CD44 can function as a receptor for the glycosaminoglycan hyaluronan (5Aruffo A. Stamenkovic I. Melnick M. Underhill C.B. Seed B. Cell. 1990; 61: 1303-1313Abstract Full Text PDF PubMed Scopus (2217) Google Scholar, 6Miyake K. Medina K.L. Hayashi S.L. Ono S. Hamaoka T. Kincade P.W. J. Exp. Med. 1990; 171: 477-488Crossref PubMed Scopus (535) Google Scholar), it is becoming increasingly clear that not all CD44-dependent adhesion events involve recognition of this particular ligand (7Culty M. Miyake K. Kincade P.W. Silorski E. Butcher E.C. Underhill C. J. Cell Biol. 1990; 111: 2765-2774Crossref PubMed Scopus (316) Google Scholar, 8Sugimoto K. Tsurumaki Y. Hoshi H. Kadowaki S. LeBousse-Kerdiles M.C. Smadja-Joffe F. Mori K.J. Exp. Hematol. 1994; 22: 488-494PubMed Google Scholar). CD44 is a very polymorphic molecule, and species ranging in size from 80 to 250 kDa have been detected on various normal and transformed cell types (2Haynes B.F. Liao H.-X. Patton K.L. Cancer Cells. 1991; 3: 347-350PubMed Google Scholar, 4Lesley J. Hyman R. Kincade P.W. Adv. Immunol. 1993; 54: 271-335Crossref PubMed Scopus (1035) Google Scholar). Although some of this heterogeneity can be attributed to differences in the post-translational modification of a common polypeptide core (9Jalkanen S. Jalkanen M. Bargatze R. Tammi M. Butcher E.C. J. Immunol. 1988; 141: 1615-1623PubMed Google Scholar, 10Brown T.A. Bouchard T. St. John T. Wayner E. Carter W.G. J. Cell Biol. 1991; 113: 207-221Crossref PubMed Scopus (321) Google Scholar), this appears not to be the only mechanism involved. In this regard, we and others have demonstrated that the alternative splicing of a series of 10 contiguous exons present within a single copy CD44 gene can produce higher molecular mass CD44 isoforms that contain additional peptide sequences of varying length inserted into a single site within the extracellular domain of the molecule proximal to the membrane spanning domain (10Brown T.A. Bouchard T. St. John T. Wayner E. Carter W.G. J. Cell Biol. 1991; 113: 207-221Crossref PubMed Scopus (321) Google Scholar, 11St. John T. Gallatin W.M. Idzerda R.L. Reg. Immunol. 1989; 2: 300-310PubMed Google Scholar, 12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar, 13Günthert U. Hofmann M. Rudy W. Reber S. Zöller M. Hauβmann I. Matzku S. Wenzel A. Ponta H. Herrlich P. Cell. 1991; 65: 13-24Abstract Full Text PDF PubMed Scopus (1626) Google Scholar, 14Stamenkovic I. Aruffo A. Seed B. EMBO J. 1991; 10: 343-348Crossref PubMed Scopus (535) Google Scholar, 15Jackson D.G. Buckley J. Bell J.I. J. Biol. Chem. 1992; 267: 4732-4739Abstract Full Text PDF PubMed Google Scholar, 16Screaton G.R. Bell M.V. Jackson D.G. Cornelis F.B. Gerth U. Bell J.I. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 12160-12164Crossref PubMed Scopus (996) Google Scholar, 17Tölg C. Hofmann M. Herrlich P. Ponta H. Nucleic Acids Res. 1993; 21: 1225-1229Crossref PubMed Scopus (264) Google Scholar). Interestingly, several recent studies have ascribed unique functional activities to certain of these alternatively spliced CD44 isoforms (13Günthert U. Hofmann M. Rudy W. Reber S. Zöller M. Hauβmann I. Matzku S. Wenzel A. Ponta H. Herrlich P. Cell. 1991; 65: 13-24Abstract Full Text PDF PubMed Scopus (1626) Google Scholar, 14Stamenkovic I. Aruffo A. Seed B. EMBO J. 1991; 10: 343-348Crossref PubMed Scopus (535) Google Scholar, 18Arch R. Wirth K. Hofmann M. Ponta H. Matzku S. Herrlich P. Zöller M. Science. 1992; 257: 682-685Crossref PubMed Scopus (403) Google Scholar), raising the possibility that the inclusion of additional peptide sequences within the extracellular domain of CD44 may alter the ligand-binding specificity of the molecule (14Stamenkovic I. Aruffo A. Seed B. EMBO J. 1991; 10: 343-348Crossref PubMed Scopus (535) Google Scholar). In the present study we demonstrate that while CD44H, the major CD44 isoform present on resting hemopoietic cells, and CD44R1, a differentially expressed isoform containing a 132-amino acid insertion encoded by the alternatively spliced exons v8-vl0 (12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar, 16Screaton G.R. Bell M.V. Jackson D.G. Cornelis F.B. Gerth U. Bell J.I. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 12160-12164Crossref PubMed Scopus (996) Google Scholar, 17Tölg C. Hofmann M. Herrlich P. Ponta H. Nucleic Acids Res. 1993; 21: 1225-1229Crossref PubMed Scopus (264) Google Scholar), can both bind immobilized and soluble hyaluronan when transfected into the CD44-negative murine lymphoma cell line TIL1, only the expression of CD44R1 can induce these cells to homotypically aggregate. Cell mixing studies further suggest that such homotypic aggregation is mediated by the adhesive interaction between a determinant encoded by the inserted region present in CD44R1, and a common region shared by both CD44R1 and CD44H. The SV40-transformed simian fibroblastoid cell line COS7 and the human erythroleukemia cell line K562 were obtained from the American Type Culture Collection (Rockville, MD). CD44 expression was induced on K562 cells by stimulation for 48 h with 10 ng/ml phorbol 12,13-dibutyrate (Sigma). The murine lymphoma cell line TIL1 was derived from a tumor initiated in C3Hf7Sed//Kam mice by the subcutaneous inoculation of syngeneic Fsa-R fibrosarcoma cells that had been genetically engineered through retroviral-mediated gene transfer to express murine interleukin-7 (19McBride W.H. Thacker J.D. Comora S. Economou J.S. Kelly D. Hogge D. Dubinett S.M. Dougherty G.J. Cancer Res. 1992; 52: 3931-3937PubMed Google Scholar). Briefly, the tumor was disaggregated by mincing, and tumor pieces placed in a flask together with approximately 50 ml of Dulbecco's modified Eagle's medium (DME) 1The abbreviations used are: DME, Dulbecco's modified Eagle's medium; FCS, fetal calf serum; FCI, fetal clone I; mAb, monoclonal antibody; PAGE, Polyacrylamide gel electrophoresis; HBSS, Hanks’ balanced salt solution; FITC-HA, fluorescein isothiocyanate-labeled human umbilical cord hyaluronan; FACS, fluorescence-activated cell sorter. (StemCell Technologies Inc., Vancouver, Canada) containing 10% fetal calf serum (FCS) (Hyclone, Logan, UT) (DME + 10% FCS). After approximately 2 weeks in culture, nonadherent cells were transferred to a separate flask and this procedure repeated daily until all adherent fibroblastoid tumor cells were removed. The cell line obtained, designated TIL1, expresses CD4 and CD8, αβ TcR, and LFA-1, but is negative for Mac-1, pl50,95, and CD44. These cells do not contain retroviral vector-derived sequences, nor do they produce interleukin-7. They are, however, malignant and generate a localized ascitic tumor following intraperitoneal injection into syngeneic C3H/HeJ mice. Both K562 and TIL1 cells were maintained in DME + 10% fetal clone I (FCI) (Hyclone, Logan, UT). The generation and characterization of the CD44 mAbs 7F4, 8D8, and 5A4 will be described in detail elsewhere. 2G. J. Dougherty, P. M. Lansdorp, C. Smith, and A. Droll, manuscript in preparation. To generate the additional CD44 mAbs 2B5, 4A4, and 3G12, female C3H/HeJ mice aged 8–10 weeks received at 21-day intervals, three intraperitoneal injections of approximately 107 syngeneic Fsa-NJhCD44Rl tumor cells (see below). Four days after the last injection, spleen cells from the immunized animals were fused with Sp2/0 myeloma cells as described previously (20Harlow E. Lane D. Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1988Google Scholar), and plated directly into ClonaCell hybridoma selection medium, as per the manufacturer's instructions (StemCell Technologies Inc.). After 12 days in culture, individual hybridoma colonies were picked into 100 µl of DME + 10% FCI. 48 h later hybridoma supernatants were screened for differential reactivity with Fsa-NJhCD44Rl and control Fsa-NJneo cells using a cell enzyme-linked immunosorbent assay technique (21Lansdorp P.M. Astaldi G.C.B. Oosterhof F. Janssen M.C. Zeijlemaker W.P. J. Immunol. Methods. 1980; 39: 393-405Crossref PubMed Scopus (85) Google Scholar). Promising mAbs were tested against a panel of human and transfected mouse cell lines by FACS and Western blot analysis to confirm reactivity with CD44. Details concerning the generation and characterization of the CD44 exon v10-specific mAb 2G1 will be given elsewhere. 3G. J. Dougherty, R. K. Chiu, and A. Droll, manuscript in preparation. The Moloney murine leukemia virus-based retroviral vector Jzen.1(22Laker C. Stocking C. Bergholz U. Hess N. de Lamarter J.F. Ostertag W. Proc. Natl. Acad. Sci. U. S. A. 1987; 84: 8458-8462Crossref PubMed Scopus (129) Google Scholar) was used to introduce and express human CD44H and CD44R1 in TIL1 cells. Briefly, full-length CD44H and CD44R1 cDNAs derived, respectively, from the plasmids pCDM8.CD44cl2.7 and pCDM8.CD44cl2.3 (12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar) were inserted into the XbaI site of pTZ19R/Tk-neo (23Hughs P.F.D. Thacker D. Hogge D. Sutherland H.J. Thomas T.E. Lansdorp P.M. Eaves C.J. Humphries R.K. J. Clin. Invest. 1992; 89: 1817-1824Crossref PubMed Scopus (67) Google Scholar). SmaI-HindIII cassettes containing CD44H or CD44R1 together with Tk-neo were isolated and cloned into HpaI-HindIII cut Jzen.1. The plasmids obtained were transfected into the ecotropic packaging line GP+E-86 (24Markowitz D. Goff S. Bank A. J. Virol. 1988; 62: 1120-1124Crossref PubMed Google Scholar) by calcium phosphate precipitation and transfected cells selected in G418 (0.5 mg/ml active weight) (Life Technologies Inc., Grand Island, NY). Supernatants conditioned by the packaging line for 24 h contained >106 colony-forming units/ml. TIL1 cells were infected with cell free JhCD44H or JhCD44Rl viral supernatants containing 5 µg/ml Poly-brene (Sigma), as described previously (23Hughs P.F.D. Thacker D. Hogge D. Sutherland H.J. Thomas T.E. Lansdorp P.M. Eaves C.J. Humphries R.K. J. Clin. Invest. 1992; 89: 1817-1824Crossref PubMed Scopus (67) Google Scholar). Infected cells were selected and maintained in medium containing G418 (0.3 mg/ml active weight). To ensure that any heterogeneity present within the starting TIL1 cell population was retained, at least 200 clones were pooled and expanded for further study. Transduced cells were regularly examined by FACS analysis (see below) to confirm continued high level expression of the introduced genes. FACS analysis of CD44 expression on retrovirally transduced TIL1 cells was carried out as described previously (19McBride W.H. Thacker J.D. Comora S. Economou J.S. Kelly D. Hogge D. Dubinett S.M. Dougherty G.J. Cancer Res. 1992; 52: 3931-3937PubMed Google Scholar). Transfected TIL1 cells were resuspended at 2 × 107 cells/ml in PBS containing 1% (v/v) Nonidet P-40, 5 mM EDTA, and 10 mM phenylmethylsulfonyl fluoride. Lysates were incubated on ice for 15 min, microcentrifuged for 5 min to pellet nuclei and other insoluble cell debris, and stored at –70 °C until required. Aliquots were rapidly thawed, added to an equal volume of nonreducing sample buffer containing 125 mM Tris, 20% (v/v) glycerol, 4.6% (w/v) SDS, pH 6.8, and incubated at 100 °C for 5 min. Total cellular proteins were separated on SDS-PAGE gels and transferred onto nitrocellulose membranes (Schleicher & Schuell, Inc.) as described previously (25Hakomori S. Kannagi R. Weir D.M. Herzenberg L.A. Blackwell C. Herzenberg L.A. Handbook of Experimental Immunology. Blackwell Scientific Publications, Edinburgh1986: 9.1-9.39Google Scholar). Nonspecific binding sites were blocked by incubating the filters at 4 °C overnight in PBS containing 5% (w/v) milk protein. After washing in HBSS, filters were incubated at room temperature for 4 h with mAb tissue culture supernatant, washed with HBSS, and incubated for a further 1 h at room temperature with a 1:100 dilution of horseradish peroxidase-conjugated goat anti-mouse IgG (Dako, Carpinteria, CA) in DME + 10% FCI. After extensive washing in HBSS, the reaction was developed in PBS containing 0.06% (w/v) 3–3′-diaminobenzidine (Sigma) and 0.012% (v/v) hydrogen peroxide (Sigma). The wells of a 96-well tissue culture plate were coated with hyaluronan by adding 100 µl of a solution of potassium hyaluronate purified from human umbilical cord (Sigma) (5 mg/ml in PBS) to each well and incubating the plate at 4 °C overnight. Unbound hyaluronan was removed by extensive washing with HBSS and the wells blocked by incubation for a further 1 h with 100 µl of PBS containing 1% (w/v) bovine serum albumin (Sigma). Infected TIL1 cells were fluorescently labeled by incubating for 30 min at 37 °C in HBSS containing 5 µg/ml Calcein (Molecular Probes, Eugene, OR). After extensive washing in DME + 10% FCI, approximately 1 × 104 labeled TIL1 cells in a final volume of 100 µl of DME + 10% FCI were added to each well. In some experiments, the cells were incubated in mAb tissue culture supernatants for 30 min on ice and washed extensively at room temperature before being added to hyaluronan-coated wells. Plates were incubated at room temperature for 15 min and any unattached cells removed by sealing the plate with plastic film and centrifuging inverted for 5 min at 400 rpm. The number of cells bound to each well was quantitated using a fluorescence plate reader (Millipore, Malborough, MA). Fluorescein-labeled human umbilical cord hyaluronan (FITC-HA), prepared as described previously (26De Belder A.N. Wik K.O. Carbohydr. Res. 1975; 44: 251-257Crossref PubMed Scopus (175) Google Scholar), was a generous gift from Dr. Saghi Ghaffari, Terry Fox Laboratory. To quantitate the binding of soluble hyaluronan to retrovirally transduced TIL1 cells, log phase cultures were harvested by vigorous pipetting, the cells washed extensively with HBSS + 2% FCS, and resuspended in HBSS + 2% FCS containing 2 µg/ml FITC-HA. Following incubation at room temperature for 1 h, the cells were washed 3 times with HBSS + 2% FCS, resuspended in HBSS containing 1 µg/ml propridium iodide, and analyzed for fluorescence intensity on a FACSort (Becton Dickinson, Mountain View, CA.) 5 × 105 transduced TIL1 cells in a final volume of 0.5 ml of PBS were added to the wells of a 24-well plate (Nunc, Life Technologies Inc., Burlington, Ontario, Canada). Plates were taped to the deck of a Mistral Multi-Mixer platform shaker (Lab-Line Instruments Inc., Melrose Park, IL) and agitated at one-third maximal speed for 30 min at room temperature. In some experiments the cells were first incubated for 30 min at 0 °C with mAb tissue culture supernatants and washed extensively at room temperature before being added to the wells. In others, equal numbers (2.5 × 105) of Calcein-labeled and unlabeled cells were mixed together before being added to the wells and allowed to aggregate as described above. The cellular clusters produced were photographed under both visible and UV light. In order to confirm the results obtained using the above assay, an alternative approach was employed in which aliquots of TILJneo, TILJhCD44H, and TILJhCD44Rl cells were labeled with red (PKH26) or green (PKH2) fluorescent cell linkers (Sigma) as per the manufacturer's instructions. Both linkers are aliphatic molecules that stably integrate into cell membrane by selective partitioning. After washing three times in DME + 10% FCI at room temperature, cells were resuspended in PBS and equal numbers (5 × 106) of PKH2-labeled (green) and PKH26-labeled (red) cells were mixed together, agitated on a shaker for 15 min, and cellular aggregation (i.e. the number of green and red double stained events) determined by analyzing the samples on a FACSort (Becton Dickinson). COS7 cells were transfected with the plasmid pCDM8.CD44cl2.3 encoding CD44R1 as described previously (27Dougherty G.J. Cooper D.L. Memory J.F. Chiu R.K. J. Biol. Chem. 1994; 269: 9074-9078Abstract Full Text PDF PubMed Google Scholar). 48 h later the transfected cells were harvested and 5 × 104 added to each well of a 24-well plate. Following overnight incubation, the medium was removed and 5 × 105 Jneo-, JhCD44H-, or JhCD44Rl-transduced TIL1 cells in a final volume of 0.5 ml were added to each well. The plates were incubated at room temperature for 15 min and any nonadherent cells removed by washing 3 times with PBS. Adherent cell monolayers were then fixed in methanol for 5 min at room temperature, air dried, and stained for CD44 expression using mAb 8d8 as described previously (12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar). The percentage of CD44-positive COS7 cells binding 3 or more TIL1 cells was determined by counting on an inverted phase microscope. All experiments were repeated at least three times. Where appropriate mean, S.E. and p values (Student's t test) were calculated using the InStat Biostatistics Program (GraphPad Software, San Diego, CA). As shown in Fig. 1, TIL cells transduced with either JhCD44H or JhCD44Rl expressed high, and approximately equal levels of CD44 as determined by reactivity with mAb 3G12. As expected, only TILJhCD44Rl cells reacted with the CD44 exon v10-specific mAb 2G1, while TILJneo cells were unreactive with both mAb 3G12 and 2G1. All three cell lines were unreactive with mAb TIB240 directed against murine CD44 (data not shown). Western blot analysis confirmed that the CD44 species expressed by TILJhCD44H and TILJhCD44R1 cells were appropriately processed and similar in size to the CD44H and CD44R1 molecules present on phorbol 12,13-dibutyrate-stimulated K562 cells (Fig. 2). No species reactive with mAb 3G12 were detected in unstimulated K562 cells or TILJneo cells (Fig. 2). Similarly, no species were detected in any lane when blots were probed with the isotype matched control mAb TIB 191 (ATCC) directed against the hapten TNP (data not shown).FIG. 2Western blot analysis of CD44 expression on TILJneo, TILJhCD44H, and TILJhCD44R1 cells. Approximately 2 × 105 cell equivalents were run in each lane of a 5% SDS-PAGE gel, transferred to nitrocellulose, and probed with mAb 3G12 followed by a peroxidase-conjugated rabbit anti-mouse IgG second antibody. The reaction was developed in PBS containing 0.06% (w/v) 3–3′-diaminobenzidine and 0.012% (v/v) hydrogen peroxide. PDBu, phorbol 12,13-dibutyrate.View Large Image Figure ViewerDownload Hi-res image Download (PPT) TILJneo, TILJhCD44H, and TILJhCD44R1 cells were next tested for their ability to adhere to hyaluronan-coated plastic. While CD44-ve TILJneo cells almost completely lacked hyaluronan binding activity, TILJhCD44H and TILJhCD44R1 bound avidly to plastic surfaces coated with hyaluronan (Fig. 3). This adhesion was mediated by CD44 and could be inhibited by some, but not all, mAbs directed against this molecule. All CD44 mAbs that blocked the adhesion of TILJhCD44H cells to hyaluronan also blocked the binding of TILJhCD44R1 cells. In addition, some partial inhibition of TILJhCD44R1 binding to hyaluronan was obtained with the CD44 exon v10-specific mAb 2G1. This mAb, of course, had no effect on the adhesion of TILJhCD44H cells. None of the mAbs tested had any effect on the binding of TILJneo cells to hyaluronan. Transduced TIL1 cells were also tested for their ability to bind soluble hyaluronan. As shown in Fig. 4, both CD44H- and CD44R1-transduced cells were able to bind FITC-HA and did so to a similar extent. Little or no binding of FITC-HA to TILJneo cells was evident. Taken together, these data confirm that CD44H and CD44R1 are both able to bind immobilized and soluble hyaluronan when expressed in TIL1 cells. As shown in Fig. 5, while TILJneo and TILJhCD44H cells remain largely monodisperse when incubated in PBS, TILJhCD44R1 cells spontaneously form large homotypic aggregates. Similar results were obtained using an alternative approach in which equal numbers of red or green fluorescently labeled TILJneo, TILJhCD44H, or TILJhCD44R1 cells were mixed together and aggregates identified on the FACS by virtue of their co-expression of both labels (Fig. 6). Since both TILJhCD44H and TILJhCD44R1 cells can bind immobilized and soluble hyaluronan, it is unlikely that the homotypic aggregation of TILJhCD44R1 cells involves the recognition of hyaluronan on the surface of interacting cells. In support of this conclusion, rather than inhibiting aggregation of TILJhCD44R1 cells, the hyaluronan adhesion-blocking mAbs 4A4 and 3G12 appeared to potentiate it (Fig. 5). This enhanced aggregation did not seem to involve the simple cross-linking of CD44 molecules on adjacent cells, as neither of these two mAbs induced the aggregation of TILJhCD44H cells. Taken together with the results presented above, these data strongly suggest that CD44R1 promotes the homotypic aggregation of transduced TIL1 cells by recognizing and binding a ligand present on the surface of adjacent cells that is distinct from hyaluronan.FIG. 6Homotypic aggregation of TILJneo, TILJhCD44H, and TILJhCD44R1 cells. Transduced TIL1 cells were labeled with red or green fluorescent cell linkers washed three times in DME + 10% FCI at room temperature, resuspended in PBS, and equal numbers (5 × 105) of red- and green-labeled cells were mixed together, agitated on a shaker for 15 min, and analyzed on the FACS. Monodisperse nonaggregating cells labeled with the red cell linker are represented by the cell population with high FL2 fluorescence (y axis) and low FL1 fluorescence (x axis). Similarly, monodisperse nonaggregating cells labeled with the green cell linker are represented by the cell population with high FL1 fluorescence (x axis) and low FL-2 fluorescence (y axis). Cellular aggregates containing both red-labeled cells (high FL-2, y axis) and green-labeled cells (high FL-1, × axis) are represented by the population present within the upper right-hand quadrant of each graph (high FL-1 and high FL-2). a, TILJneo-TILJneo; b, TILJhCD44H-TILJhCD44H; c, TILJhCD44R1-TILJhCD44R1; d, TILJhCD44H(FL2)-TILJhCD44RHFLl).View Large Image Figure ViewerDownload Hi-res image Download (PPT) In order to determine whether CD44R1 promotes cellular aggregation by homotypically interacting with other CD44R1 molecules, or whether it recognizes some other ligand expressed on the surface of transduced TIL1 cells, TILJhCD44R1 cells were fluorescently labeled with Calcein and mixed with an equal number of unlabeled TILJneo, TILJhCD44H, or TILJhCD44R1 cells. The cellular aggregates that formed were examined for the presence of both fluorescently labeled and unlabeled cells. As shown in Fig. 7, the aggregates that formed when TILJneo and TILJhCD44R1 cells were mixed contained only fluorescently labeled TILJhCD44R1 cells. These data indicate that CD44R1 does not promote the cellular aggregation of TILJhCD44R1 cells by recognizing and binding a ligand that is also present on the surface of control TILJneo cells. Interestingly, when TILJhCD44R1 cells were mixed with TILJhCD44H cells, the cellular aggregates that formed contained both labeled TILJhCD44R1 cells and a small number of unlabeled TILJhCD44H cells. This finding strongly suggests that CD44R1 can recognize and bind CD44H. This conclusion is supported by the FACS data shown in Fig. 6D, in which small numbers of double labeled aggregates appear to be produced when equal numbers of PKH26 (red) labeled TILJhCD44R1 and PKH2 (green) labeled TILJhCD44H cells were mixed together. Finally, when Calcein-labeled and unlabeled TILJhCD44R1 cells were mixed, the aggregates that formed contained approximately equal numbers of labeled and unlabeled cells (Fig. 7), confirming that CD44R1 molecules can also bind to one another. To further confirm these results and to more easily quantitate the interaction between the various CD44 isoforms, TILJhCD44H and TILJhCD44R1 cells were also tested for their ability to adhere to COS7 cells transiently transfected with the CD44R1 cDNA. As shown in Table I, CD44H- and CD44R1-transduced, but not control Jneo-transduced cells were able to bind to transfected CD44R1-positive COS7 cells. Moreover, in agreement with the cellular aggregation data presented in FIG. 6, FIG. 7, a significantly higher proportion (p < 0.05) of CD44R1-positive COS7 cells bound TILJhCD44R1 cells than bound TILJhCD44H cells.Table IBinding of transduced TIL1 cells to CD44R1 transfected COS7 cellsCD44R1+ve COS7 cells binding TIL1%TILJneo<1TILJhCD44H27.6 ± 5.53TILJhCD44R144.5 ± 0.7 Open table in a new tab Although the widely distributed cell surface glycoprotein CD44 has been shown to function as a receptor for the glycosaminoglycan hyaluronan (5Aruffo A. Stamenkovic I. Melnick M. Underhill C.B. Seed B. Cell. 1990; 61: 1303-1313Abstract Full Text PDF PubMed Scopus (2217) Google Scholar, 6Miyake K. Medina K.L. Hayashi S.L. Ono S. Hamaoka T. Kincade P.W. J. Exp. Med. 1990; 171: 477-488Crossref PubMed Scopus (535) Google Scholar), there is increasing evidence that not all CD44-dependent cellular adhesion events involve recognition of this particular ligand. For example, while the anti-CD44 mAb Hermes-3 can block lymphocyte attachment to high endothelial venules in frozen sections of human mucosal lymphoid tissues (28Jalkanen S.T. Bargatze R.F. de los Toyos J. Butcher E.C. J. Cell Biol. 1987; 105: 983-990Crossref PubMed Scopus (511) Google Scholar), this same antibody had no effect on the hyaluronan-dependent binding of a B-cell line transfected with human CD44 to cultured rat endothelial cells (14Stamenkovic I. Aruffo A. Seed B. EMBO J. 1991; 10: 343-348Crossref PubMed Scopus (535) Google Scholar), nor did it inhibit the binding of [3H]hyaluronan to detergent extracts of human CD44-positive cells (7Culty M. Miyake K. Kincade P.W. Silorski E. Butcher E.C. Underhill C. J. Cell Biol. 1990; 111: 2765-2774Crossref PubMed Scopus (316) Google Scholar). These data are consistent with the observation that the epitope recognized by mAb Hermes-3 maps to a region of the CD44 molecule close to the membrane spanning domain (29Goldstein L.A. Zhou D.F.H. Picker L.J. Minty C.N. Bargatze R.F. Ding J.F. Butcher E.C. Cell. 1989; 56: 1063-1072Abstract Full Text PDF PubMed Scopus (448) Google Scholar) and some distance from the amino-terminal region implicated in hyaluronan binding (30Peach R.J. Hollenbaugh D. Stamenkovic I. Aruffo A. J. Cell Biol. 1993; 122: 257-264Crossref PubMed Scopus (328) Google Scholar). Taken together these findings suggest strongly that CD44 can recognize and bind a ligand present on endothelial cells that is distinct from hyaluronan. Moreover, since mAb Hermes-3 had little or no effect on the CD44-dependent binding of lymphocytes to endothelial cells in sections of synovial and peripheral tissues (28Jalkanen S.T. Bargatze R.F. de los Toyos J. Butcher E.C. J. Cell Biol. 1987; 105: 983-990Crossref PubMed Scopus (511) Google Scholar) it appears that this putative ligand is differentially expressed on endothelial cells in different tissues. At present the molecular nature of this alternative CD44 ligand(s) remains undefined. Recently, it has become clear that alternative splicing events can generate a large number of differentially expressed CD44 isoforms that contain additional peptide sequences of varying length inserted into a single site within the extracellular domain of the molecule proximal to the membrane spanning domain (15Jackson D.G. Buckley J. Bell J.I. J. Biol. Chem. 1992; 267: 4732-4739Abstract Full Text PDF PubMed Google Scholar, 16Screaton G.R. Bell M.V. Jackson D.G. Cornelis F.B. Gerth U. Bell J.I. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 12160-12164Crossref PubMed Scopus (996) Google Scholar, 17Tölg C. Hofmann M. Herrlich P. Ponta H. Nucleic Acids Res. 1993; 21: 1225-1229Crossref PubMed Scopus (264) Google Scholar). At present, however, the effect of such inserted regions on the ligand-binding specificity of the CD44 molecule remains unclear. We have shown previously (27Dougherty G.J. Cooper D.L. Memory J.F. Chiu R.K. J. Biol. Chem. 1994; 269: 9074-9078Abstract Full Text PDF PubMed Google Scholar), and have confirmed in this study, that CD44R1, the major alternatively spliced CD44 isoform expressed by various transformed hemopoietic and non-hemopoietic cell types (12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar), retains the ability to recognize and bind immobilized hyaluronan. Moreover, although the CD44 molecules expressed by certain cell types have been reported to differ in their ability to bind immobilized versus soluble hyaluronan (4Lesley J. Hyman R. Kincade P.W. Adv. Immunol. 1993; 54: 271-335Crossref PubMed Scopus (1035) Google Scholar), both CD44H and CD44R1 also appeared equally capable of binding this latter ligand when expressed in the CD44-negative murine T-cell lymphoma cell line TIL1. There are, however, differences in the ligand-binding specificity of these two molecules. In particular, expression of CD44R1, but not CD44H, caused transduced TIL1 cells to homotypically aggregate. Cell mixing experiments indicated further that JhCD44R1 transduced TIL1 cells could also aggregate with CD44H-transduced cells, but not control TILJneo cells, suggesting that CD44R1 can recognize and bind a common determinant present on both CD44H and CD44R1. Since a significantly higher percentage of COS7 cells transfected with CD44R1 bound TILJhCD44R1 cells than bound TILJhCD44H cells, it would seem that the reciprocal interaction between two CD44R1 molecules is of a higher affinity and/or avidity than the interaction between CD44H and CD44R1. In a previous study, St. John et al. (31St. John T. Meyer J. Idzerda R.L. Gallatin W.M. Cell. 1990; 60: 45-52Abstract Full Text PDF PubMed Scopus (119) Google Scholar) demonstrated that murine L cells transfected with the baboon CD44H cDNA exhibited a new adhesive phenotype characterized by the formation of large cell aggregates. In this instance, however, CD44H transfected cells were also able to aggregate with nontransfected parental cells. Moreover, such aggregation could be readily inhibited by mAbs directed against the hyaluronan-binding domain of the CD44 molecule. Since many fibroblastoid cell lines produce hyaluronan and are surrounded by hyaluronidase-sensitive cell coats (32McBride W.H. Bard J.B.L. J. Exp. Med. 1979; 149: 507-515Crossref PubMed Scopus (157) Google Scholar, 33Bard J.B.L. McBride W.H. Ross A.R. J. Cell Sci. 1983; 62: 371-383Crossref PubMed Google Scholar, 34Laurent T. Fraser J. FASEB J. 1992; 6: 2397-2404Crossref PubMed Scopus (2104) Google Scholar), it is likely that the aggregation noted in this study involved the interaction between the transfected CD44 molecules and hyaluronan on the surface of opposing cells. In contrast, as both CD44H and CD44R1 can bind hyaluronan, but only CD44R1 promotes the homotypic aggregation of transduced TIL1 cells, it is unlikely that the aggregation response observed in the present study involves the recognition of hyaluronan. Indeed lymphoid cells appear to produce little if any hyaluronan (34Laurent T. Fraser J. FASEB J. 1992; 6: 2397-2404Crossref PubMed Scopus (2104) Google Scholar). Moreover, treatment of CD44R1-transduced TIL1 cells with either mAbs that block cellular attachment to hyaluronan (Fig. 5) or with Type × leech hyaluronidase (Sigma), an enzyme that specifically cleaves the β-glucuronate-(1Haynes B.F. Telen M.J. Hale L.P. Denning S.M. Immunol. Today. 1989; 10: 423Abstract Full Text PDF PubMed Scopus (528) Google Scholar, 2Haynes B.F. Liao H.-X. Patton K.L. Cancer Cells. 1991; 3: 347-350PubMed Google Scholar, 3Herrlich P. Zöller M. Pals S.T. Ponta H. Immunol. Today. 1993; 14: 395-399Abstract Full Text PDF PubMed Scopus (270) Google Scholar)-GlcNAc glycosidic bonds in hyaluronan, 4A. Droll and G. J. Dougherty, manuscript in preparation. had little or no effect on the aggregation of these cells. The molecular nature of the determinant present on CD44R1 recognized by other CD44R1 molecules and by CD44H remains to be determined. In addition to binding hyaluronan, CD44H has also been reported to recognize and bind chondroitin-4-sulfate (30Peach R.J. Hollenbaugh D. Stamenkovic I. Aruffo A. J. Cell Biol. 1993; 122: 257-264Crossref PubMed Scopus (328) Google Scholar, 35Naujokas M.F. Morin M. Anderson M.S. Peterson M. Miller J. Cell. 1993; 74: 257-268Abstract Full Text PDF PubMed Scopus (203) Google Scholar). Indeed, Toyama-Sorimachi and Miyasaka (36Toyama-Sorimachi N. Miyasaka M. Int. Immunol. 1994; 6: 655-660Crossref PubMed Scopus (52) Google Scholar) have recently demonstrated that murine CTLL-2 cells express a high molecular weight (gp600) molecule extensively modified by chondroitin sulfate side chains that can be recognized and bound by CD44H. However, since TIL1 cells transfected with CD44H do not homotypically aggregate or bind control Jneo transduced TIL1 cells it must be assumed that this particular ligand is not present on the TIL1 cell line. Recently we have found that CD44R2, an alternatively spliced CD44 isoform containing only exon v10 (12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar), also promotes cellular adhesion,4 indicating that the functionally important determinant present within CD44R1 is encoded by exon v10. Although exon v10 contains a potential site of chondroitin sulfate attachment (12Dougherty G.J. Lansdorp P.M. Cooper D.L. Humphries R. K J. Exp. Med. 1991; 174: 1-5Crossref PubMed Scopus (156) Google Scholar), there is no evidence based on the apparent molecular mass of the molecule on SDS-PAGE, that a significant proportion of CD44R1 molecules expressed by TILJhCD44R1 cells are modified in this way. It should be noted, however, that an alternative possibility, not excluded by the studies reported in this article, is that the expression of CD44R1 may switch on a second gene product which may in turn serve as a ligand for CD44R1 and/or CD44H. Recently, we have demonstrated that while resting T cells are CD44R1-negative, stimulation with the mitogen phytohemagglutinin or with a combination of OKT3 and phorbol 12,13-dibutyrate, dramatically up-regulated expression of the CD44R1 isoform without significantly altering the total level of CD44 expressed by these cells.3 CD44R1 also appears to be expressed by transformed endothelial cell lines and by non-transformed vascular endothelial cells stimulated with tumor necrosis factor-α. 5J. F. Dirks and G. J. Dougherty, unpublished observation. Thus the differential expression of alternatively spliced CD44 isoforms on activated lymphocytes and endothelial cells may help regulate the adhesive interaction between these two cell types under both steady state conditions and in response to inflammatory stimuli. Alternatively spliced CD44 isoforms are also differentially expressed on various malignant cell types (37Matsumura Y. Tarin D. Lancet. 1992; 340: 1053-1058Abstract PubMed Scopus (459) Google Scholar, 38Tanabe K.K. Ellis L.M. Saya H. Lancet. 1993; 341: 725-726Abstract PubMed Scopus (278) Google Scholar) and the adhesive interaction between these molecules and CD44 proteins present on endothelial cells and/or various immune cell types may help explain the suggested involvement of CD44 tumor metastasis.
Abstract Introduction Levels of pain and dysfunction appear to differ among people with hemophilia despite similar levels of joint disease. Objective To determine patient characteristics that influence pain and function independent of joint status. Methods US adults with hemophilia completed a survey that included information on clinical characteristics, demographics, and patient‐reported outcome instruments assessing pain (Brief Pain Inventory v2 Short Form [ BPI ]), functional impairment (Hemophilia Activities List [ HAL ]), and health status ( EQ ‐5D‐5L). Additionally, physiotherapists optionally completed a clinical joint evaluation (Hemophilia Joint Health Score [ HJHS ]). Associations were examined using simple and multiple regression models. Results Of 381 adults enrolled, 240 had complete HJHS scores (median age, 32 years). After controlling for HJHS and opiate use, anxiety/anxiolytic use was significantly associated with worse pain severity and interference scores. After controlling for HJHS , the most significant predictors of functional impairment were older age, unemployment, more severe hemophilia, and greater pain. EQ ‐5D‐5L pain/discomfort was associated with worse outcomes on most HAL scores. Conclusion Unemployment, anxiety, and depression were each associated with both greater pain and functional disability after controlling for joint status. Continued attention to pain and psychosocial issues will be important in improving clinical care and research efforts in the hemophilia population.
Abstract Unlike tumors that develop because of somatic genetic alterations such as frequent coding mutations or chromosomal deletions, amplifications or translocations, some tumors are addicted to expression of certain genes for their sustained proliferation and survival - for example, LIM domain only 1 (LMO1) expression in neuroblastoma (NB). Numerous genome-wide association studies (GWAS) have identified significant associations between germline single-nucleotide polymorphisms (SNPs), many in the non-coding genome, and cancer. A robust genome-wide association has been reported previously between an intronic LMO1 SNP and NB susceptibility. By further sequencing and epigenetic fine mapping of the SNP locus, it was demonstrated that the causal SNP is part of a transcription factor binding site within an enhancer element, regulates expression of LMO1 and in turn increases the tumorigenic potential. To identify additional examples of regulatory SNPs as cancer drivers, we overlaid published genome-wide significant cancer associations with active chromatin marks from Encyclopedia of DNA Elements and searched for SNPs that resided within gene regulatory elements. To map these SNPs to candidate genes and determine direction of effect, we co-localized GWAS signals with expression quantitative trait (eQTL) signals from the Genotype-Tissue Expression (GTEx) Consortium database. Lastly, we checked for gene amplification and/or overexpression of the mapped genes in the Cancer Genome Atlas (TCGA) data to identify a set of target genes that not only exhibit significant cancer association in GWAS, but also have evidence for epigenetic regulation and propensity for amplification and/or overexpression in tumors. We identified more than 25 novel cancer-target pairs with strong germline, regulatory and somatic evidence. A look up through synthetic lethality screen data available in-house suggested that several of these targets are self-lethal, further underscoring their importance for cancer cell proliferation and survival. Additional in vitro experiments are being planned to further validate the targets. Citation Format: Diptee A. Kulkarni, Karl Guo, Junping Jing, Mugdha Khaladkar, Kijoung Song, Coco Dong, David Cooper, Benjamin Schwartz. Identification of novel cancer target genes by combining data from the cancer genome-wide association studies (GWAS), regulatory DNA elements and The Cancer Genome Atlas (TCGA) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 236.
