Prognostic value of FDG-PET scan imaging in lymphoma patients undergoing autologous stem cell transplantation

We conducted a retrospective analysis of 50 lymphoma patients (Hodgkin's disease and non-Hodgkin's lymphoma) who had an 18F-fluoro-deoxyglucose positron emission tomography (FDG-PET) scan after at least two cycles of salvage chemotherapy and before autologous stem cell transplantation (ASCT) at our institution. The patients were categorized into FDG-PET negative (N = 32) and positive (N = 18) groups. The median follow-up after ASCT was 19 months (range: 3-59). In the FDG-PET-negative group, the median progression-free survival (PFS) was 19 months (range: 2-59) with 15 (54%) patients without progression at 12 months after ASCT. The median overall survival (OS) for this group was not reached. In the FDG-PET-positive group, the median PFS was 5 months (range: 1-19) with only one (7%) patient without progression at 12 months after ASCT. The median OS was 19 months (range: 1-34). In the FDG-PET-negative group, chemotherapy-resistant patients by CT-based criteria had a comparable outcome to those with chemotherapy-sensitive disease. A positive FDG-PET scan after salvage chemotherapy and prior ASCT indicates an extremely poor chance of durable response after ASCT.

Long-term event-free survivors after high-dose therapy and autologous stem-cell transplantation for low-grade follicular lymphoma

Although follicular lymphoma (FL) is generally responsive to conventional-dose chemotherapy, improved survival in patients with this disease has been difficult to demonstrate. High-dose chemo/radiotherapy followed by autologous stem-cell transplantation (ASCT) can improve response rates, although its effects on survival remain controversial. Between 1990 and 2003, we transplanted 49 patients with low-grade FL at our institution. Twenty-two patients (45%) had undergone histologic transformation at the time of ASCT. In all, 44 patients (90%) had relapsed disease and five patients (10%) were resistant to chemotherapy at the time of transplantation. After ASCT, 30 patients (61%) were in complete remission (CR). The median overall survival (OS) has not been reached, while the median event-free survival (EFS) is 2.4 years. At a median follow-up of 5.5 years (longest 12.4 years), a plateau has been reached with 56% of patients remaining alive, and 35% event-free. ASCT was well tolerated except for two (4%) treatment-related deaths. In multivariable analysis, CR after ASCT and age less than 60 years are the best predictors of EFS and OS. ASCT is thus a safe therapeutic approach in FL, resulting in long-term EFS and OS for some patients, even with transformed disease.

Recombinant monoclonal antibody technology

With the development of murine hybridoma technology over a quarter century ago, the ability to produce large quantities of well-characterized monoclonal antibody preparations revolutionized diagnostic and therapeutic medicine. For many applications in transfusion medicine, however, the production of serological reagents in mice has certain biological limitations relating to the difficulty in obtaining murine monoclonal antibodies specific for many human blood group antigens. Furthermore, for therapeutic purposes, the efficacy of murine-derived immunoglobulin preparations is limited by the induction of anti-mouse immune responses. Technical difficulties inherent in human hybridoma formation have led to novel molecular approaches that facilitate the isolation and production of human antibodies without the need for B-cell transformation, tissue culture, or even immunized individuals. These technologies, referred to as 'repertoire cloning' or 'Fab/phage display', involve the rapid cloning of immunoglobulin gene segments to create immune libraries from which antibodies with desired specificities can be selected. The use of such recombinant methods in transfusion medicine is anticipated to play an important role in the development and production of renewable supplies of low-cost reagents for diagnostic and therapeutic applications.

Section 5: Structural/genetic analysis of mAbs to blood group antigens. Coordinator's report

The heavy and light chain immunoglobulin variable region nucleotide sequences for 219 mAbs to human red blood cells were collected from workshop participants, published reports, and Genbank. Information regarding antigen specificity, species of origin, method of cloning, and other relevant serological properties was correlated with the sequence data. Immunoglobulin sequences were analyzed to determine the heavy- and light-chain immunoglobulin genes used and the overall extent of somatic mutation from germline configuration. Approximately 50% of the sequences encoded antibodies with Rh(D) specificity with the remaining sequences encoding mAbs to other Rh-related antigens, antigens of the ABO, MNS, and Kell blood group systems, and several others. Surprisingly, no sequence data were available for mAbs with specificity for a number of common Rh antigens, common Kell antigens, or antigens of the Lewis, Kidd, or Duffy blood group systems. The majority of mAbs were of human origin but included a significant number of macaque mAbs, murine mAbs, and a small number of synthetically-designed recombinant antibodies. Both cellular (EBV-transformation, cell fusion) and molecular (phage display) approaches were used for antibody cloning. Analysis of certain groups of sequences demonstrated patterns of immunoglobulin gene restriction, repertoire shift, and somatic mutation. Analysis of other mAbs demonstrated the value of antibody sequence data for the design and production of novel reagents useful in blood group serology.

Isolation of an IgG anti-B from a human Fab-phage display library

BACKGROUND

ABO incompatibility is a common cause for mild hemolysis in the newborn, ranging from 1 in 30 to 1 in 150 births. Fortunately, hemolysis requiring transfusion is rare and restricted to blood group O mothers, because blood group A and B individuals make poor IgG anti-B and anti-A responses. No human IgG ABO antibody sequences have been reported, in part because of the difficulty in obtaining human IgG hybridomas. Phage-display technology may be able to circumvent these difficulties, but its application to carbohydrate antigens is poorly studied.

STUDY DESIGN AND METHODS

A human IgG1 phage-display Fab library was constructed from splenocytes derived from a nonhyperimmunized blood group O person, and panned against group B RBCs.

RESULTS

After five rounds of panning, essentially all phage bound to group B RBCs. Nucleotide sequence analysis of a single monoclonal IgG1lambda phage, FB5.7, revealed a highly mutated VH4 family heavy chain, and a nearly germline VL7 family lambda light chain. The Fab agglutinated group B, but not group A, random-donor RBCs. However, group B ELISA reactivity could be inhibited by soluble B-trisaccharide, soluble A-trisaccharide, galactose, and N-acetyl galactosamine. Similarly, galactose and N-acetyl galactosamine were able to inhibit group B RBC agglutination.

CONCLUSION

FB5.7 is the first human IgG ABO MoAb described. Alhough it behaves serologically like a group B-specific antibody, it demonstrates interaction with both the A and B epitopes. Phage-display technology can be used to better define the relationship between antibody genotype and phenotype in anti-carbohydrate responses in nonhyperimmunized hosts, and thus to improve our understanding of the composition of the antibody repertoire.

Characterization of a murine monoclonal antibody that mimics heparin-induced thrombocytopenia antibodies

Antibodies to PF4/heparin can be demonstrated in almost all patients with heparin-induced thrombocytopenia/thrombosis (HIT/HITT) and in some persons exposed to heparin who do not have clinical manifestations. The role of anti-PF4/heparin antibodies in the pathogenesis of HIT/HITT has been difficult to establish because the antibodies found in serum are generally polyclonal and polyspecific. To circumvent this problem, we developed a murine monoclonal antibody (mAb) to human (h) PF4/heparin complexes. A monoclonal IgG(2bkappa )antibody (designated KKO) was identified that bound specifically to hPF4/heparin complexes. Maximal binding of KKO to hPF4/heparin complexes occurred at similar molar ratios of PF4:heparin observed for HIT/HITT antibodies. KKO also bound to hPF4 in association with other glycosaminoglycans. Platelet activation by KKO required heparin and was abrogated by blockade of FcgammaRIIA. In the presence of PF4, KKO bound to endothelial cells, but not to CHO cells lacking heparan sulfate proteoglycans. Variants of PF4 complexed to heparin were recognized equally well by KKO and HIT/HITT sera. KKO competes for binding with a subset of HIT/HITT antibodies that are relatively spared by mutations in the 3rd domain of PF4. The nucleotide and predicted amino acid sequences of KKO and RTO, a murine anti-hPF4 mAb that does not require heparin for binding, revealed no obvious relationship in either the heavy- or the light-chain immunoglobulin variable regions. These studies suggest that KKO recapitulates the antigenic and functional specificity of a subset of HIT/HITT antibodies and may, therefore, provide insight into the pathogenesis of thrombocytopenia and thrombosis in affected persons. (Blood. 2000;95:1533-1540)

A molecular approach for isolating high-affinity Fab fragments that are useful in blood group serology

BACKGROUND

Multiple mouse hybridoma antibodies recognize the antigens of the MNS blood group system. The Fab fragments of several of these antibodies were expressed on bacteriophage and as soluble proteins. The parental N92 anti-N IgG monoclonal antibody (parental N92 MoAb), but not its monovalent, soluble Fab fragment (N92 Fab fragment), agglutinated antigen-positive red cells by an antiglobulin method. Light-chain shuffling was used to isolate mutant N92 Fab fragments with higher affinity that would function by agglutination.

STUDY DESIGN AND METHODS

Light-chain cDNA libraries, constructed from mice immunized with N-type glycophorin A, were inserted into a recombinant pComb3H vector containing the N92 Fd fragment. The N92 Fd fragment:light-chain libraries were panned on N-type glycophorin A or NN red cells, and antigen-binding clones were isolated. Purified parental N92 MoAb and the Fab fragments were evaluated by enzyme-linked immunosorbent assay and agglutination.

RESULTS

The novel NNA7, C1, and G11 Fab fragments all bound to N-type glycophorin A with higher affinity than did the N92 Fab fragment. The affinity of the library-derived clones was equivalent to that of the parental N92 MoAb. Although their fine specificity differed slightly from the parental N92 MoAb, the clones functioned equivalently by agglutination using an antiglobulin method.

CONCLUSIONS

Light-chain shuffling allowed the isolation of bacterially produced, high-affinity, soluble, monovalent recombinant anti-N Fab fragments that functioned well by agglutination. This approach is useful in obtaining inexpensive serologic reagents that may replace conventional MoAbs produced by tissue culture methods.

Only selected light chains combine with a given heavy chain to confer specificity for a model glycopeptide antigen

The M and N human blood group glycopeptide Ags are carried on RBCs by glycophorin A. Previous results suggested that the murine humoral immune response against the N, but not the M, Ag is restricted. In addition, these results suggested that particular highly homologous heavy chains might be able to combine promiscuously with various light chains to yield anti-N specificity. To examine this, the current study used Fab phage methodology to couple an array of light chains, obtained from cDNA libraries isolated from immunized mice, to single Fd obtained from N61, N92, and 425/2B hybridomas. Interestingly, for the chimeric Fab to retain M or N specificity, the new light chains needed to belong to the same Vk gene family as the light chain from the parental, hybridoma-derived mAb. In some cases the new light chains modified the Fab affinity and fine specificity. For example, library-derived light chains coupled with the N92 Fd yielded chimeric Fab with increased affinity. In particular, the affinity of these univalent chimeric Fab for the N Ag was equivalent to that of the bivalent parental IgG mAb. Taken together, these results demonstrate that particular structures formed by the light chain V region are required to cooperate with a particular heavy chain V region to create a functional binding site for these glycopeptide Ags. They also demonstrate a lack of heavy chain promiscuity in the formation of murine anti-M and anti-N Abs.

Genetic and immunological properties of phage-displayed human anti-Rh(D) antibodies: implications for Rh(D) epitope topology

Understanding anti-Rh(D) antibodies on a molecular level would facilitate the genetic analysis of the human immune response to Rh(D), lead to the design of therapeutically useful reagents that modulate antibody binding, and provide relevant information regarding the structural organization of Rh(D) epitopes. Previously, we described a Fab/phage display-based method for producing a large array of anti-Rh(D) antibodies from the peripheral blood lymphocytes of a single alloimmunized donor. In the current study, we present a detailed analysis of 83 randomly selected clones. Sequence analysis showed the presence of 28 unique gamma1 heavy chain and 41 unique light chain gene segments. These paired to produce 53 unique Fabs that had specificity for at least half of the major Rh(D) epitopes. Surprisingly, despite this diversity, only 4 closely related heavy chain germline genes were used (VH3-30, VH3-30.3, VH3-33, and VH3-21). Similarly, nearly all Vkappa light chains (15/18) were derived from one germline gene (DPK9). lambda light chains showed a more diverse VL gene usage, but all (23/23) used the identical Jlambda2 gene. Several Fabs that differed in epitope specificity used identical heavy chains but different light chains. In particular, 2 such clones differed by only 3 residues, which resulted in a change from epD2 to epD3 specificity. These results suggest a model in which footprints of anti-Rh(D) antibodies are essentially identical to one another, and Rh(D) epitopes, as classically defined by panels of Rh(D) variant cells, are not discrete entities. Furthermore, these data imply that the epitope specificity of an anti-Rh(D) antibody can change during the course of somatic mutation. From a clinical perspective, this process, which we term epitope migration, has significance for the design of agents that modulate antibody production and for the creation of mimetics that block antibody binding in the settings of transfusion reactions and hemolytic disease of the newborn.

The human immune response to red blood cell antigens as revealed by repertoire cloning

A major goal of current immunologic research is to develop specific therapeutic strategies by which the enormous diversity in immune response can be enhanced, attenuated, or eliminated, depending on the particular disease process. For nearly a century, the human immune response to red blood cell antigens has served as a paradigm for understanding the pathophysiology of autoimmune disorders and alloimmune reactions to foreign cells and tissues. Recent developments in molecular biology have facilitated the expression of immune repertoires in the form of immunoglobulin Fab fragments on the surface of filamentous bacteriophage. Such approaches have provided powerful means for producing monoclonal antibodies for research, clinical, and therapeutic applications. Our laboratory has combined these techniques with novel cell-surface selection methods to isolate extraordinarily large arrays of human antibodies to the clinically relevant red blood cell Rh(D) antigen. Our results have provided a comprehensive genetic and serologic analysis of anit-Rh(D) antibodies within single alloimmunized individuals thereby offering new insights into the development of human immune repertoires.

Isolation of cell surface-specific human monoclonal antibodies using phage display and magnetically-activated cell sorting: applications in immunohematology

A method is described for the isolation of filamentous phage-displayed human monoclonal antibodies directed at unpurifiable cell surface-expressed molecules. To optimize the capture of antigen-specific phage and minimize the binding of irrelevant phage antibodies, a simultaneous positive and negative selection strategy is employed. Cells bearing the antigen of interest are pre-coated with magnetic beads and diluted into an excess of unmodified antigen-negative cells. Following incubation of the cell admixture with a Fab/phage library, the antigen-positive cell population is retrieved using magnetically-activated cell sorting and antigen-specific Fab/phage are eluted and propagated in bacterial culture. Utilizing this protocol with magnetically-labeled Rh(D)-positive and excess unlabeled Rh(D)-negative human red blood cells and a Fab/phage library constructed from human peripheral blood lymphocytes, dozens of unique clinically-useful gamma 1 kappa and gamma 1 lambda anti-Rh(D) antibodies were isolated from a single alloimmunized individual. This cell-surface selection method is readily adaptable for use in other systems, such as for the identification of putative tumor-specific antigens and provides a rapid (< 1 month), high-yield approach for isolating self-replicative antibody reagents directed at novel or conformationally-dependent cell-surface epitopes.

Direct NMR measurement of folding kinetics of a trimeric peptide

Direct NMR measurements of the folding kinetics are performed on a collagen-like triple helical peptide. The triple helical peptide was designed to model a biologically important region of collagen and has the sequence (POG)3ITGARGLAG(POG)4. Triple helical peptides were synthesized with specifically labeled 15N amino acid residues in key positions, and the kinetics of folding of the individual residues were monitored directly by measuring the loss of monomer intensity and the increase in trimer intensity. The residues at the terminal ends and central region could be followed independently and quantitated directly. Residues located at the terminal ends have rates and kinetics of folding that are distinct from residues in the central region of the peptide. This allows the monitoring of different steps in the folding mechanism and the postulation of the existence of a kinetic intermediate. The NMR data are consistent with a mechanism of association/nucleation and propagation. Hereditary connective tissue diseases are associated with mutations that result in abnormal folding of collagen, and the NMR folding experiments on a collagen-like peptide provide a basis for characterizing the molecular defect in folding mutations.

Purification of a secreted form of recombinant rabies virus glycoprotein: comparison of two affinity tags

Expression of recombinant eukaryotic proteins in transfected mammalian cell lines has become an important approach for the characterization of the structure and function of these proteins. However, it is often difficult to recover and purify the recombinant proteins. Therefore, the use of fusion proteins incorporating epitope or affinity tags has become more widespread. In this paper, we directly compare two affinity tags, the hexahistidyl tag and the biotin peptide mimetic, Strep-tag, for use in purification of a recombinant soluble form of rabies virus glycoprotein secreted by transfected Chinese hamster ovary fibroblasts. The recombinant rabies virus glycoproteins are denoted RGP(WT)T441his and RGP(WT)T443s-tag, respectively. These affinity tags were chosen because the chromatographic matrices (Ni(II)-NTA-agarose and recombinant core streptavidin-agarose, respectively) were readily available and these methods offered the possibility of a one-step purification using mild elution conditions. However, in our hands, neither method allowed for a one-step purification protocol. Nonetheless, it was possible to purify RGP(WT)T441his to homogeneity from crude conditioned medium using a combination of metal-chelate affinity chromatography and immunoaffinity chromatography. In contrast, although the Strep-tag has been useful for purifying recombinant proteins expressed in bacteria, we were not able to effectively purify RGP(WT)T443s-tag from conditioned medium using chromatography on recombinant core streptavidin-agarose.

Acid destabilization of a triple-helical peptide model of the macrophage scavenger receptor

Electrostatic interactions were studied in a triple-helical peptide, (POG)3PKGQKGEKG(POG)4, which contains a lysine-rich 9 residue sequence from the collagen-like domain of the macrophage scavenger receptor (MSR). This peptide adopts a stable triple-helical conformation only when the pH is higher than 4.5, corresponding to ionization of the Glu side chain. Modeling shows Glu forms ion pairs with one of the Lys residues, stabilizing the structure. Previously studied collagen-like peptides show relatively small contributions of electrostatic interactions to stability. The large magnitude of the pH mediated structural changes seen for this peptide suggests that specific placement of charged residues in the triple-helix conformation can generate strong electrostatic interactions.

Construction of bacteriophage expressing mouse monoclonal Fab fragments directed against the human MN glycophorin blood group antigens

BACKGROUND

The MN human blood group antigens are complex glycopeptide antigens at the amino terminus of glycophorin A. Many different mouse monoclonal antibodies to these antigens have been produced and characterized. The construction of combinatorial immunoglobulin libraries displaying antibody Fab fragments on the surface of bacteriophage (Fab-phage) represents a novel approach for developing monoclonal reagents, for exploring the diversity of the immune response to specific antigens, and for understanding the molecular basis of the interaction of an antibody with its antigen. However, it is necessary to determine whether Fab fragments displayed on bacteriophage surfaces retain immunologic characteristics similar to the intact antibodies.

STUDY DESIGN AND METHODS

Fab-phage were constructed from three anti-N (AH7, N61, and N92) and two anti-M (425/2B and M2A1) murine hybridomas. The Fab-phage and parental hybridomas were compared by enzyme-linked immunosorbent assay, Western blotting, and flow cytometry.

RESULTS

In each case, the Fab-phage and its parental hybridoma antibody had similar immunologic characteristics. In particular, their dependence on the pH of the buffer and on sialylation of the target antigen was similar.

CONCLUSION

These results suggest that Fab-phage may provide novel reagents with applications in immunohematology and may be useful in the study of the immune response to human blood group antigens.

Expression and characterization of recombinant anti-Rh(D) antibodies on filamentous phage: a model system for isolating human red blood cell antibodies by repertoire cloning

The production of human anti-red blood cell (RBC) Igs in vitro from immunized individuals would greatly facilitate the genetic analysis of the human immune response to RBC antigens and also provide useful serologic reagents. Technical difficulties inherent in human B-cell immortalization have led to the development of molecular approaches that bypass the need for cell transformation. By cloning human Ig gene segments into bacterial expression vectors, libraries are created of filamentous phage particles displaying Fab fragments on their surfaces. Libraries have been screened with purified, soluble antigen and selected clones genetically manipulated in Escherichia coli to produce soluble Fab fragments. Our goal has been to adapt this technique to the study of RBC autoantibodies and alloantibodies that have specificities against unpurifiable membrane-bound antigens. To test the feasibility of this approach, two sets of phage were created, one set expressing a human anti-Rh(D) Ig and the other expressing a human antitetanus toxoid Ig. After verifying the presence of functional phage-displayed Fabs through biochemical, flow cytometric, and electron microscopic analyses, a model library was constructed comprising one anti-Rh(D)-expressing phage per 10(4) antitetanus toxoid-expressing phage. A method was developed for screening the library with intact Rh(D)-positive RBCs. After four rounds of panning, anti-Rh(D) specificity was enriched more than 10,000-fold to a final frequency of approximately 100%. Plasmid DNA derived from anti-Rh(D) phage was used to produce milligram quantities of soluble recombinant anti-Rh(D) Fabs purified by nitrogen cavitation and nickel-chelation affinity chromatography. The authenticity of the Fabs was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotting, which showed bands with molecular weights of approximately 50 kD and 26 kD under nonreducing and reducing conditions, respectively. Binding of recombinant anti-Rh(D) Fabs to Rh(D)-positive RBCs was demonstrated by flow cytometry and by an agglutination assay. Our results suggest that repertoire cloning by cell-surface enrichment may have broad application to the study of the human immune response to erythroid antigens in addition to membrane-bound antigens present on other hematopoietic cells.

Inappropriate testing for diarrheal diseases in the hospital

To assess the degree to which routine stool cultures, ova and parasite examinations, and Clostridium difficile toxin assays may be inappropriately ordered on hospitalized patients, we conducted a retrospective study to determine the relative yield of these tests on specimens collected from outpatients and inpatients as a function of time after admission. During a 3-year period, only 1 of 191 positive stool cultures and none of the 90 ova and parasite examinations with positive results were from the group of patients who had stool specimens submitted after 3 days of hospitalization. Analysis of laboratory work load for a 1-year period showed that specimens from this patient group contributed nearly 50% of the more than 3000 specimens received each year. In contrast, approximately 25% (range, 17% to 33%) of samples, regardless of admission status, were positive for C difficile toxin. Eliminating routine stool cultures and ova and parasite examinations on hospitalized patients would significantly reduce hospital and patient costs without altering patient care. Nationwide, such a policy might achieve a cost savings of +20 to +30 million per year.

Discriminating rejection from CMV infection in renal allograft recipients using flow cytometry

The ability to distinguish among rejection, cytomegalovirus (CMV) infection, and cyclosporin toxicity in the symptomatic renal allograft recipient remains one of the major issues in clinical transplantation. The practical application of immunologic monitoring of peripheral blood lymphocytes through the use of fluorescently labeled monoclonal antibodies and single-color flow cytometry has been limited by the inability to demonstrate significant correlations between the levels of specific T-cell subset populations and the cause of impaired renal function. In the present study using two-color analysis, we monitored the expression of interleukin-2 receptor (IL-2R) and HLA-DR antigen on the T-cells of a group of 51 renal cadaveric allograft recipients receiving cyclosporin, azathioprine, and prednisone for an average of 4 months after transplantation. We found that the proportion of CD3+ cells coexpressing IL-2R increased above baseline during 12 out of 14 rejection episodes that took place during the course of the study (P less than 10(-6)). Alternatively, we found that the proportion of cells coexpressing HLA-DR antigen on CD2+ cells increased above baseline during 11 out of 11 CMV infections (P less than 10(-6)). There was no correlation between the level of IL-2R+CD3+ cells and CMV infection or between the level of CD2+DR+ cells and rejection. These relationships showed a high degree of sensitivity and specificity when used to discriminate among possible etiologies for decreased renal function in the symptomatic patient.

Partial purification and characterization of an actin-bundling protein, band 4.9, from human erythrocytes

Band 4.9 (a 48,000-mol-wt polypeptide) has been partially purified from human erythrocyte membranes. In solution, band 4.9 polypeptides exist as trimers with an apparent molecular weight of 145,000 and a Stokes radius of 50 A. Electron microscopy shows that the protein is a three-lobed structure with a radius slightly greater than 50 A. When gel-filtered rabbit muscle actin is polymerized in the presence of band 4.9, actin bundles are generated that are similar in appearance to those induced by "vinculin" or fimbrin. The bundles appear brittle and when they are centrifuged small pieces of filaments break off and remain in the supernatant. At low band 4.9 to actin molar ratios (1:30), band 4.9 lowers the apparent steady-state low-shear falling ball viscosity by sequestering filaments into thin bundles; at higher ratios, the bundles become thicker and obstruct the ball's movement leading to an apparent increase in steady-state viscosity. Band 4.9 increases the length of the lag phase and decreases the rate of elongation during actin polymerization as measured by high-shear Ostwald viscometry or by the increase in the fluorescence of pyrene-labeled actin. Band 4.9 does not alter the critical actin monomer concentration. We hypothesize that band 4.9, together with actin, erythrocyte tropomyosin, and spectrin, forms structures in erythroid precursor cells analogous to those formed by fimbrin, actin, tropomyosin, and TW 260/240 in epithelial brush borders. During erythroid development and enucleation, the actin filaments may depolymerize up to the membrane, leaving a membrane skeleton with short stubs of actin bundled by band 4.9 and cross-linked by spectrin.

The effect of endogenous proteases on the spectrin binding proteins of human erythrocytes

We have demonstrated that in human erythrocyte ghosts endogenous proteolytic activity is responsible for the digestion of the spectrin binding proteins (bands 2.1 to 2.6). The pH optimum, cofactor requirements and inhibitor sensitivity have been established. Our results indicate that proteolysis of bands 2.1 to 2.6 and the formation of 3', a fragment containing an active spectrin binding site, can occur through two enzymatic pathways: a cascade of consecutive proteolytic cleavages of the spectrin binding proteins inhibited by phenylmethylsulfonyl fluoride or a Ca2+-stimulated, phenylmethylsulfonyl fluoride-insensitive, EDTA-inhibited cleavage of band 2.1 to band 2.3, followed by digestion to band 3' by phenylmethylsulfonyl fluoride-inhibitable enzymes. These findings may provide the techniques necessary to prevent proteolysis of the spectrin binding proteins during purification and reconstitution experiments and provide insight into how they are formed in vivo.