Correct disulfide pairing and efficient refolding of detergent-solubilized single-chain Fv proteins from bacterial inclusion bodies.

In vitro folding of denatured proteins has remained an inefficient and empirical process that has limited the use of bacterially expressed recombinant proteins. In this paper we show that in vitro folding of recombinant single-chain Fv (sFv) proteins is markedly facilitated when disulfide bonds are formed in detergent solution. sFv proteins from three different antibodies were expressed as bacterial cytoplasmic inclusion bodies and solubilized in the weak ionic detergent, sodium lauroylsarcosine (SLS). Upon oxidation in air in the presence of metal ion catalysts, all three sFvs quantitatively formed intrachain disulfide bonds which ran as a single band in SDS-polyacrylamide gel electrophoresis under non-reducing conditions. By contrast, oxidation from 6 M urea gave large amounts of disulfide linked aggregates, and three closely spaced bands of monomeric protein. Detergent was removed from the oxidized sFvs by addition of 6 M urea and absorption with an ion exchange resin. After dialysis and gel filtration in non-denaturing solution, moderate to high yields of monomeric sFv were obtained, depending upon the sFv. All three sFvs gave single bands on isoelectric focussing and SDS-PAGE gels and had similar or identical binding specificities and affinities as the parental Fabs, implying that the final products contained correctly paired disulfide bonds. The correct disulfide pairing suggests that the disulfide loops within the detergent-solubilized sFvs adopt a native-like structure.

A single-chain bispecific Fv2 molecule produced in mammalian cells redirects lysis by activated CTL.

Single-chain Fv (sFv) molecules consist of the two variable domains of an antibody (Ab) connected by a polypeptide spacer and contain the binding activities of their parental antibodies (Abs). In this paper we have attached the C-terminus of 2C11-sFv (anti-mouse CD3 epsilon-chain) to the N-terminus of OKT9-sFv (anti-human transferrin receptor [TfR]) through a 23 amino acid inter-sFv linker consisting primarily of CH1 region residues from 2C11, to form a single-chain bispecific Fv2 [bs(sFv)2] molecule. The bs(sFv)2 was expressed in COS-7 cells, and was secreted at the same rate as the two parental sFvs. The secreted protein had both anti-CD3 and anti-TfR binding activities. Essentially all of the secreted bs(sFv)2 molecules bound TfR and the binding affinity of the bs(sFv)2 was comparable to that of OKT9 sFv and Fab. Thus, the attachment of the inter-sFv linker to the N-terminus of OKT9-sFv did not impair its binding function. The bs(sFv)2 retained both binding specificities after long-term storage at 4 degrees C or overnight incubation at 37 degrees C. It redirected activated mouse CTL to specifically lyse human TfR+ target cells at low (ng/ml) concentrations and was much more active than a chemically cross-linked heteroconjugate prepared from the same parental mAbs. Because bs(sFv)2 molecules secreted by mammalian cells are homogeneous proteins containing two binding sites in a single polypeptide chain, they hold great promise as an easily obtainable, economic source of a bispecific molecule suitable for in vivo use.

On the origin of the FcRIII (CD16)-containing vesicle population in human neutrophil granulocytes.

Human blood neutrophils bear two types of Fc receptors that recognize the Fc portion of immunoglobulin G: FcRII and FcRIII. In earlier studies we found that neutrophils not only express FcRIII on their plasma membrane but also contain a large population of FcRIII-containing vesicles mainly located in the juxtanuclear area. To find out whether these vesicles derive from the plasma membrane, we used electron microscopic techniques to study compartments involved in ligand-independent endocytosis in human neutrophil granulocytes. The endocytic compartments were labeled with BSA-gold. This marker entered the cell through non-coated invaginations of the plasma membrane as well as via coated pits. After internalization, BSA-gold was present in numerous electron-lucent vesicles in the juxtanuclear area and in the trans-Golgi reticulum, endosomes, and lysosome-like structures. FcRIII also occurred in the BSA-gold-containing electron-lucent vesicles in the juxtanuclear area, as shown by postembedding immunocytochemical labeling of FcRIII in cells already containing BSA-gold. Quantification showed that 29% of all FcRIII-containing vesicles also bear BSA-gold while the remaining 71% contain only the receptor. In sum, our findings show that one third of the FcRIII-containing electron-lucent vesicles in neutrophil granulocytes derive from the plasma membrane and are involved in ligand-independent endocytosis of FcRIII. The majority of these vesicles, however, are not of an endocytic origin and might constitute an "internal pool" of receptors in these cells.

Targeting of anti-tumor responses with bispecific antibodies.

T cells can be induced to specifically lyse tumor cells with bispecific antibodies containing anti-T cell receptor mAbs crosslinked to anti-tumor mAbs. Such "targeted cytolysis" requires that the target cell be bound directly to the cytotoxic cell. In addition, targeted T cells mediate a second activity, the secretion of factors that can block the growth of both tumor target cells and bystander tumor cells. When given to nude mice bearing intraperitoneal human ovarian carcinoma, targeted human T cells cause the rapid removal of most tumor cells from the peritoneum, and markedly prolong the times of survival of treated mice. The efficacy of targeted T cells for treating human cancer is currently being tested in clinical trials.

The human neutrophil FcRIII (CD16) on the plasma membrane can be replenished from an intracellular source.

The internalization of peroxidase-antiperoxidase (PAP) immune complexes by human neutrophil granulocytes was studied at an ultrastructural level. PAP initially bound to the plasma membrane at 4 degrees C accumulated in endosomes within 5 min of internalization. By that time, all of the ligand bound to the plasma membrane had been removed from the cell surface and the cells were able to bind newly added PAP. Pre-embedding labelling of FcRIII on these cells showed that this receptor was present on the cell surface, indicating involvement of FcRIII in the rebinding of PAP in neutrophils. The origin of FcRIII present on the plasma membrane after Fc receptor-mediated internalization of PAP was investigated in another series of experiments. Incubation of the cells with pronase eliminated the epitope on the Fc receptor recognized by anti-FcRIII. After the pronase treatment hardly any Fc receptors were detected on the plasma membrane. However, incubation of the cells for only 5 min in a protease-free medium after the pronase treatment led to an abundance of FcRIII on the plasma membrane of the neutrophils. These findings support the hypothesis that FcRIII on the plasma membrane of human neutrophil granulocytes is replenished from an internal source of free Fc receptors and suggest that at least some of the receptors present on the cell surface after the binding and internalization of PAP originate from this source in the cytoplasm.

A bacterially expressed single-chain Fv construct from the 2B4 T-cell receptor.

A single-chain Fv construct of the 2B4 T-cell receptor has been made and expressed in Escherichia coli as bacterial inclusion bodies. After solubilization in 6 M guanidine hydrochloride and formation of mixed disulfides with glutathione, the protein was refolded by diluting out the denaturant and allowing intramolecular disulfide bridges to form by disulfide exchange. Approximately 65-100 mg of refolded protein was obtained from 1 liter of bacterial culture, an appreciable fraction of which was monomeric in nondenaturing solvents. This protein bound to three monoclonal antibodies specific for allotypic or idiotypic determinants on the native 2B4 variable region but did not bind several other anti-T-cell-receptor monoclonal antibodies that lacked such specificity. These experiments show that T-cell-receptor variable regions, like the V regions of antibodies, can form a well-behaved single-chain Fv molecule and provide large amounts of recombinant single-chain Fv T-cell receptor that can be used to study T-cell function.

Retargeting of CTL by an efficiently refolded bispecific single-chain Fv dimer produced in bacteria.

A single-chain bispecific Fv dimer (bs(sFv)2) having specificity for mouse CD3 epsilon chain and human transferrin receptor was produced in bacterial inclusion bodies. To overcome difficulties associated with in vitro protein folding, we used a novel renaturation approach to obtain active bs(sFv)2. The protein was dissolved in the weak ionic detergent sodium lauroylsarcosine, and disulfides were formed by oxidation in air. After oxidation, the bs(sFv)2 exhibited very little covalent aggregation and migrated as a single species in nonreducing SDS-PAGE, suggesting that disulfides were correctly paired. The detergent was removed using an ion exchange resin and the protein fractionated by size exclusion chromatography. The recovered 65-kDa protein was monomeric in non-denaturing solvent, homogeneous by SDS-PAGE, and comprised 15 to 20% of material applied to the gel filtration column. This protein bound specifically to both mouse CD3 epsilon chain and human transferrin receptor with affinities indistinguishable from those of the parental Fabs or single-chain Fvs. The bs(sFv)2 specifically redirected mouse cytotoxic T cells to lyse target cells expressing human transferrin receptor at picomolar concentrations. Bacterially produced and detergent oxidized bs(sFv)2 molecules may therefore provide the abundant amounts of homogeneous active material required to redirect cytotoxic cells against tumors and other unwanted cells in animal models and in patients.

Changes in glycolytic network and mitochondrial design in creatine kinase-deficient muscles.

Skeletal muscles respond with high plasticity to pathobiological conditions or changes in physiological demand by remodeling cytoarchitectural and metabolic characteristics of individual myocytes. We have previously shown that muscles of mice without mitochondrial and/or cytosolic creatine kinases (ScCKmit(-/-) and/or M-CK(-/-)) partly compensate for the defect(s) by redirecting metabolic pathways and ultrastructural characteristics. Here, we show by semiquantitative Western blot analysis that the compensatory changes involve mutation- and fiber-type-specific coordinated regulation of divergent but functionally coupled groups of proteins. Fast-twitch gastrocnemius muscle of CK(--/--) mice display a two- to fourfold upregulation of mitochondrial cytochrome c oxidase, inorganic phosphate carrier, adenine nucleotide translocator, and voltage-dependent anion channel proteins. In parallel, cytosolic myoglobin is upregulated. Slow-twitch soleus muscle responds with changes in the glycolytic enzyme pattern, including a shift in lactate dehydrogenase isoenzyme composition. Adaptations in the network for oxidative adenosine triphosphate (ATP) production are already apparent at 17 days of age.

Alternative triggering molecules and single chain bispecific antibodies.

Although T cell receptors and Fc receptors are the best known cytotoxic triggering molecules, a number of adhesion molecules recently have been identified as alternative triggering molecules. We discuss how CD44, an adhesion molecule found on all leukocytes and many other cell types, becomes a triggering molecule on NK cells following stimulation with IL-2. We also describe a genetically engineered single chain bispecific antibody, produced in mammalian cells and in bacteria, that is capable of redirecting lysis in the nanogram per milliliter range.

Mammalian expression and secretion of functional single-chain Fv molecules.

Single-chain Fv (sFv) proteins are genetically engineered molecules that consist of the two variable domains of an antibody connected by a polypeptide linker; they contain the antigen binding function of the parental protein in a single 30-kDa polypeptide chain. sFvs are usually produced in bacteria where they are insoluble and therefore require extensive refolding in vitro. In this report we followed the processing of three antibody sFvs (145-2C11 directed against murine CD3 epsilon chain, OKT9 against the human transferrin receptor, and U7.6 against dinitrophenyl groups) by transfected mammalian (COS-7) cells to determine whether the mammalian protein folding machinery can produce and secrete active sFv with high efficiency. The sFvs contained an immunoglobulin light chain leader sequence, which directed them to the endoplasmic reticulum and allowed secretion into the medium. We found that the sFvs were secreted at different rates, with the rate-limiting step of secretion being their exit from the endoplasmic reticulum. We increased the secretion rate of one of the sFvs by introducing an asparagine-linked glycosylation site in FR1 of the heavy chain, and by using tunicamycin (an inhibitor of glycosylation) we found that glycosylated antibody sFvs were secreted faster than their nonglycosylated counterparts. All secreted sFvs specifically bound their antigens; where tested, at least 90% of the secreted sFv was functional. Therefore, mammalian cells can effectively fold and secrete sFv antibody and can provide a convenient system for testing and producing sFv proteins.

Intracellular localization of glycosyl-phosphatidylinositol-anchored CD67 and FcRIII (CD16) in affected neutrophil granulocytes of patients with paroxysmal nocturnal hemoglobinuria.

Immunoelectron microscopical studies performed in healthy human neutrophils showed the presence of glycosyl-phosphatidylinositol (GPI)-linked CD67 in granules. The use of immunogold double-labeling of CD67 and lactoferrin (LF; as marker for specific granules) or CD67 and myeloperoxidase (MPO; as marker for azurophilic granules) showed that CD67 occurred only in the specific granules. Furthermore, flow cytometry showed that CD67 has a low level of expression on the plasma membrane of these cells. In paroxsymal nocturnal hemoglobinuria (PNH)-affected neutrophils, CD67 was not detected in any intracellular compartment by immunoelectron microscopy, and flow cytometry showed no CD67 on the plasma membrane. In earlier studies, FcRIII was found on the plasma membrane, in electron-lucent vesicles, and in the Golgi complex of healthy neutrophils, and in the Golgi complex of some of the PNH-affected neutrophils. Here we have studied FcRIII in PNH-affected cells of three other patients and found, by immunoelectron microscopy, that the receptor can not be detected in these cells. However, flow cytometry showed that FcRIII was not completely absent on the plasma membrane of the affected cells, but that the level of expression on these cells was low. Thus, PNH patients can differ from one another with respect to the occurrence of affected neutrophils that have a detectable level of FcRIII in the Golgi complex. In summary, these findings show not only that the expression of the two GPI-linked proteins, CD67 and FcRIII, is markedly lower on the plasma membrane, but also that neither occurred in any of the intracellular compartments of affected neutrophils of the PNH patients examined in this study.

Intracellular localization and de novo synthesis of FcRIII in human neutrophil granulocytes.

Immunoelectron microscopic studies in human neutrophils showed that FcRIII was present on the plasma membrane, in the Golgi complex, and in many small vesicles (120 to 180 nm). FcRIII was not found in specific or azurophilic granules as shown by immunogold double-labeling experiments, visualizing both FcRIII and either lactoferrin (a marker of specific granules) or myeloperoxidase (a marker for azurophilic granules). Because the occurrence of FcRIII in the Golgi complex suggested that biosynthesis of this receptor occurs in these cells, metabolic labeling experiments were performed. Immunoprecipitation of FcRIII from NP-40 lysates of cells labeled with 35S-methionine showed a diffuse 50- to 70-Kd band corresponding with the band noted after immunoprecipitation of FcRIII from surface iodinated cells. These findings show that de novo synthesis of FcRIII occurs in neutrophils and suggest that at least some of the small vesicles containing FcRIII derive from the Golgi complex and thus are involved in transport of newly synthesized FcRIII to the plasma membrane.

Redirection of T cell-mediated cytotoxicity by a recombinant single-chain Fv molecule.

We have produced two single-chain Fv (sFv) proteins by bacterial periplasmic secretion, one sFv with specificity for the hapten DNP, and the other for the human transferrin receptor. After solubilization and refolding, we recovered several mg of active sFv per liter of bacterial culture. Each sFv bound to cells bearing the appropriate Ag and could be used to direct targeted cellular cytotoxicity. Targeting relied on a universal bispecific antibody designed to cross-link CD3 on the cytotoxic T cell with a peptide fused to the sFv carboxyl-terminus. The universal bispecific antibody was used in combination with the Ag-specific sFv to redirect human cytotoxic T cells to kill a variety of target cells. Such an approach has a number of advantages that may make it useful for the immunotherapy of cancer and other diseases.

Redirection of cellular cytotoxicity. A two-step approach using recombinant single-chain Fv molecules.

In this article the authors discuss an indirect system for redirecting cellular cytotoxicity, which utilizes a "universal" bispecific antibody to redirect T-cells to kill cells targeted with single-chain Fv (sFv) fusion proteins that carry a peptide tag recognized by the bispecific antibody. This approach has a number of theoretical advantages in the immunotherapy of cancer.

Targeted cytokine production.

It has been well established that bispecific antibodies containing anti-T-cell receptor MAbs crosslinked to anti-tumor MAbs induce T cells to lyse tumor cells, as measured in a 51Cr-release assay. Such lysis requires direct attachment between target and cytotoxic cells and most probably involves the exocytosis of cytolytic substances into the cell:cell interface. In addition, targeted T cells mediate a second activity, the secretion into the medium of factors that can block the growth of bound tumor cells and unbound bystander cells. In order to test how targeted effector cells mediate anti-tumor effects in vivo, we are currently developing a totally syngeneic murine system in which murine T cells are targeted against mouse mammary tumors. The system allows us to treat both primary tumors and tumor transplants, using a mammary-tumor-virus antigen as the entity that is specifically recognized on the tumor cells.