Rapid development of affinity matured monoclonal antibodies using RIMMS.

Affinity matured murine monoclonal antibody producing cell lines can now be rapidly generated using a novel repetitive, multiple site immunization strategy designated RIMMS. RIMMS capitalizes on rapid hypermutation and affinity maturation events which occur in B cell populations localized within secondary lymphatic tissue early in response to antigenic challenges. A murine myeloma cell line, P3XBcl-2-13, stably transfected with Bcl-2, enhances the outgrowth of hybridomas following somatic fusion with immune lymphocytes isolated from pooled peripheral lymph nodes (PLN) 8-14 days after the initial immunization. Immunizations somatic fusion, screening and isolation of affinity matured IgG secreting monoclonal antibody cell lines occur within a one month time period. By using RIMMS, we have been able to expedite the isolation of affinity matured monoclonal antibodies to numerous antigens, including a drug hapten.

Preparation and characterization of interleukin-2-gelonin conjugates made using different cross-linking reagents.

Conjugates of IL-2 with the ribosome-inactivating protein gelonin were prepared using heterobifunctional reagents to link the proteins via disulfide, acid-labile, and noncleavable linkers. In each case, one protein was modified using 2-iminothiolane. The sulfhydryl groups so introduced were then reacted either with 2-nitro-5-dithiobenzoate groups or with iodoacetamido groups which had been introduced into the second protein. In the case of the acid-labile linkage, a reagent which forms a labile bond upon reaction with amino groups, 4-(iodoacetamido)-1-cyclohexene-1,2-dicarboxylic acid anhydride (its synthesis is described in this paper) was used to modify the toxin. The conjugates were separated from nonconjugated proteins by gel filtration on Sephadex G100 (SF). Each was analyzed with respect to its ribosome-inactivating activity, its ability to bind to the IL-2 receptor, and its in vitro cytotoxicity. The ribosome-inactivating activity of gelonin was unaffected by modification with 2-iminothiolane and was retained in conjugates prepared using this reagent. Modification of the toxin with 4-(iodoacetamido)-1-cyclohexene-1,2-dicarboxylic acid anhydride to form the acid-labile link drastically reduced the activity of the toxin. However, the activity of the toxin was recovered following acid treatment to release the native protein. Conjugates containing each type of linkage exhibited both specific binding and selective cytotoxicity toward cells expressing the IL-2 receptor. The most potent of these toxins, that containing the disulfide linkage, exhibited a cytotoxicity which was 2 orders of magnitude greater than that of unconjugated gelonin.

Cytotoxicity of gelonin and its conjugates with antibodies is determined by the extent of their endocytosis.

Conjugates of the single-chain ribosome-inactivating protein gelonin with ligands that bind to cell surface molecules vary greatly in their cytotoxicity. Conjugates that are not endocytosed after binding to cells exhibit low cytotoxicity similar to that of free gelonin, while conjugates that are endocytosed demonstrate enhanced cytotoxicity relative to free gelonin. However, the number of internalized gelonin molecules needed to intoxicate cells to the same degree has been found to be similar for all conjugates and for free gelonin. The intracellular concentration of gelonin has to be between 2,000-10,000 molecules/cells to achieve a surviving fraction of 0.37. Our studies revealed the presence of three distinct categories of cell surface molecules, those that are efficient in mediating endocytosis of immunotoxins, those that are only moderately efficient, and those that seem not to cause internalization of bound immunotoxins.

Activation of human cytolytic cells through CD2/T11. Comparison of the requirements for the induction and direction of lysis of tumor targets by T cells and NK cells.

We studied the mechanisms whereby human T cells and NK cells are activated and directed to lyse tumor targets through the CD2 (T11/E-rosette) Ag. Using two cloned NK lines, we showed that these cells, as had previously been shown for T cells, could be directed to lyse an "NK-resistant" tumor target in the presence of antibody heterodimers. These heterodimers consisted of a (mAb) to CD2 (anti-T11(2) or anti-T11(3] linked to a mAb recognizing the tumor cell (J5, anti-CALLA). However, distinct differences between NK cells and T cells were observed with regard to the requirements for such directed lysis: first, only one epitope of CD2 on NK cells (either T11(2) or T11(3] needed to be recognized by the antibody heterodimer in order for directed lysis to occur, whereas for T cells both T11(2) and T11(3) epitopes had to be recognized. Second, in confirmation of previous data with monomeric anti-T11(2) or anti-T11(3) antibody, heterodimers constructed with these reagents enhanced conjugate formation between NK cells and tumor targets, whereas no such enhancement was seen with T cells. All types of heterodimer directed lysis were dependent on the adhesion molecule LFA-1, as an anti-LFA-1 antibody-blocked lysis. Third, whereas in T cells lysis mediated through CD2 appeared to be regulated by CD3 but not vice versa, all types of lysis by NK cells appeared to be regulated through CD2. Finally we showed that F(ab')2 fragments of the anti-T11(2) and anti-T11(3) antibodies could activate NK cells, but were unable to activate T cells either as cloned cytolytic lines, or in populations of PBL. The implications of our findings with regard to the role of CD2 in the activation of cytolytic cells is discussed.

Immunotoxins containing single-chain ribosome-inactivating proteins.

We have summarized what is currently known about the distribution, biological role, and the mechanism of action of the single chain ribosome-inactivating proteins and described the purification of one of them, gelonin, as an example. ITs have been made with several of these proteins and, depending upon the antibody used for conjugation, these immunoconjugates can show specific in vitro cytotoxicity which is similar to that shown by equivalent ITs prepared with ricin A chain. The most potent of these conjugates have shown antitumor efficacy in a variety of animal tumor models, including both syngeneic rodent tumors and xenografts in nude or immunosuppressed mice. An important point needs to be addressed, however, before concluding that ITs containing single chain toxins will be clinically useful. A major problem with this approach is that it is likely that both the antibody and the toxin components of these conjugates will be immunogenic. Both antitoxin and antixenogenic immunoglobulin responses have been shown to occur in animals after infusion of IT, although it has not yet been clearly demonstrated that such antibody responses adversely effect the pharmacokinetics or the efficacy of immunoconjugates. Thus, preliminary enthusiasm over the efficacy of these new reagents must be tempered with the knowledge that their use in the clinic may be limited by the host immune responses or other as yet undefined factors. The fact that there are many immunologically distinct single chain ribosome-inactivating proteins does suggest one way of evading the antitoxin response, by a sequential treatment with a panel of immunoconjugates, each containing a different single chain toxin.

An immunotoxin composed of a monoclonal antitransferrin receptor antibody linked by a disulfide bond to the ribosome-inactivating protein gelonin: potent in vitro and in vivo effects against human tumors.

An immunoconjugate was prepared containing a disulfide linker between a murine monoclonal antibody (5E9), which recognized the human transferrin receptor, and the ribosome-inactivating protein gelonin. This immunoconjugate was found to consist of two major species, 5E9-gelonin2 and 5E9-gelonin1, and a minor species of 5E9-gelonin3 and less than 10% of either free antibody or gelonin. 5E9-gelonin was extremely toxic in vitro to human tumor cell lines expressing the 5E9 antigen, including a Burkitt's lymphoma, an adult T-cell acute lymphocytic leukemia, an acute myelogenous leukemia, a promyelocytic leukemia, and a cervical carcinoma line. A 24-hour exposure to 10(-9) M immunoconjugate killed 90-99.9% of tumor cells, depending on the cell line. A 5E9-negative murine leukemia was not sensitive to this conjugate. Pharmacokinetic analysis of the disappearance of this immunoconjugate from the murine circulation revealed that it had a biphasic clearance, with an initial rapid phase with a half-life (t1/2) of 3 hours and a later, slower phase with a t1/2 of about 1 day. Analysis of blood samples by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis revealed that a substantial degree of disulfide-linker breakdown occurred in vivo and that the 5E9-gelonin2 species was cleared more rapidly than the 5E9-gelonin1. With use of the same clonogenic assays used to measure in vitro toxicity, biologically active immunoconjugate could be detected in murine plasma for up to 24 hours after iv administration, but the concentration of immunoconjugate by this measure was considerably less than that predicted by SDS-gel electrophoresis. The ability to deliver immunoconjugate to tumor cells in vivo was studied with use of the Burkitt's lymphoma Namalwa as a xenograft in nude mice. It was possible to deliver substantial amounts of immunoconjugate to Namalwa cells in xenografted ascites with direct ip inoculation; lower but significant amounts of immunoconjugate could be delivered to this xenograft after systemic iv administration, provided the tumor burden was low. The 5E9-gelonin conjugate, when administered iv at the time of ip tumor inoculation, prolonged survival of nude mice bearing Namalwa or other human tumors as ascites xenografts and delayed or prevented the growth of subcutaneous nodules of Namalwa in an antigen-specific fashion after a single iv injection. Direct intratumoral administration also inhibited the growth of visible subcutaneous nodules of Namalwa. This immunoconjugate may be useful in the treatment of human cancer.

Radio-immunoassay for plasma alpha human atrial natriuretic peptide: a comparison of direct and pre-extracted methods.

Plasma atrial natriuretic peptide (ANP) concentrations were measured by both direct radio-immunoassay and with pre-extraction of the peptide from plasma using C18 reverse phase columns. Peptide concentrations were measured in normal subjects (including a group of eight volunteers who received an intravenous infusion of 0.9% NaCl solution), patients with renal failure (including a group with end-stage disease undergoing renal dialysis) and patients with a spectrum of cardiac dysfunction. The overall correlation of results from direct and extracted assay methods was good. However, absolute values from extracted assays were significantly lower than from parallel direct assays. This discrepancy was due to interference from platelets and from another, as yet unidentified, plasma component demonstrated by gel filtration experiments. Extraction of the peptide from plasma by C18 columns largely eliminated these sources of interference and was particularly important for accurate measurement of peptide concentrations within the normal range. Plasma peptide concentrations were elevated in cardiac and renal failure, fell with renal dialysis and rose in normal subjects challenged with an intravenous isotonic fluid load. These findings suggest that ANP participates in the regulation of body fluid volumes and arterial pressure.

Plasma atrial natriuretic peptide concentrations during exercise in sodium replete and deplete normal man.

To explore the effects of exercise on plasma atrial natriuretic peptide (ANP) concentrations, eight normotensive volunteers performed maximal treadmill exercise in sodium replete and deplete states. Baseline immunoreactive plasma ANP concentrations were significantly lower during sodium depletion. During exercise plasma ANP rose in all subjects on both occasions. Plasma peptide responses were attenuated by sodium depletion with peak exercise levels only double baseline values, in contrast to the threefold increase in ANP concentrations observed when subjects were sodium replete. Plasma renin and aldosterone concentrations also rose with exercise. In contrast to changes in plasma ANP, the responses of both were enhanced by sodium depletion.

Immunocytochemistry of renin in renal tumours.

We used a panel of two polyclonal antisera and two monoclonal antibodies to human renin to assess the tissue distribution of immunoreactive renin in a range of tumours and normal human tissues. The only tissue showing positive staining for renin was kidney and all four antisera stained the myoepithelioid cells in the renal cortex. In the survey of tumours we found immunoreactive renin only in renal tumours, namely, renal cell carcinoma, and nephroblastoma (Wilms' tumour). The renin-positive cells were sparse and distributed mainly along the course of the tumour blood vessels. They stained positively with all four antibodies and, in pairs of serial sections, we showed that the same cell reacted with two different antisera. This suggests that renal cell carcinoma and nephroblastoma have within them cells which contain renin.

Renin and inactive renin (prorenin) in the plasma of patients with malignant renal tumours.

Plasma active and inactive renin (porenin) concentrations were measured in 26 adults with renal carcinoma and 8 children with nephroblastoma. In patients with carcinoma, plasma total renin concentration was slightly raised in 8 patients and markedly in one, in whom the increase was due to a more than 10-fold excess of inactive renin. Total renin concentration was raised in 7 out of 8 cases of nephroblastoma, the increase being due to excess inactive renin. It is likely that this represented oversecretion of a renin precursor, as inactive renin from the plasma of a patient with renal carcinoma showed similar biochemical characteristics to pure human prorenin.

Plasma atrial natriuretic peptide: responses to modest and severe sodium restriction.

Plasma atrial natriuretic peptide (ANP), active renin and aldosterone concentrations were measured during modest and severe sodium restriction in 16 normal volunteers. Eight volunteers took, in two separate 4-day periods, diets containing 200 and 15 mmol sodium/day (plus frusemide, 40 mg, on day 1 of low sodium intake), in random order. With sodium restriction plasma ANP concentrations fell in parallel with cumulative sodium balance. Changes in plasma renin and aldosterone were reciprocal to those in plasma ANP. To assess the threshold and reproducibility of the effects of modest sodium restriction on plasma ANP concentrations, a further group of eight volunteers took identical constant sodium diets (80 mmol/day in two cases and 120 mmol/day in the remainder) for 4 days on two successive weeks. On both occasions similar individual and mean falls in plasma ANP values accompanied the small net sodium loss occurring on these regimes. Plasma active renin concentration rose significantly during both diet phases. These data suggest plasma ANP concentrations alter in response to minor changes in sodium status, and are consistent with a role for plasma ANP in physiological regulation of body fluid volumes.

Plasma alpha natriuretic peptide in cardiac impairment.

Regional plasma alpha human atrial natriuretic peptide concentrations were measured, and their relation to intracardiac pressures assessed, in an unselected series of 45 patients undergoing diagnostic cardiac catheterisation. Arteriovenous gradients in plasma concentrations of alpha human atrial natriuretic peptide were consistent with its cardiac secretion and its clearance by the liver and kidneys. Plasma concentrations of the peptide in the pulmonary artery, aorta, and superior vena cava correlated closely with the mean right atrial and pulmonary arterial pressures, and similar, though weaker, positive relations were seen with the left ventricular end diastolic and pulmonary artery wedge pressures. Concentrations of both atrial natriuretic peptide and renin showed significant inverse relations with serum sodium concentrations. Plasma concentrations of alpha human atrial natriuretic peptide are an additional objective indicator of the severity of haemodynamic compromise in patients with cardiac impairment.

Binding of human renin to plasma proteins and comparison with plasma active and inactive renin.

Inactive renin was partially purified from normal male plasma. It showed no enzymatic activity at pH 7.4, and its molecular weight by gel filtration was 53 000 compared with 45 000 for partially purified plasma active renin and 41 000 for pure renal renin. After exposure to pH 3.0 the inactive renin became enzymatically active but its molecular weight did not change. The acid-activation could be reversed when the renin was re-adjusted to pH 7.4 and warmed. Pure renal renin labelled with I125 was added to normal male plasma. It had a molecular weight of 40 000 by gel filtration. When the mixture was acidified and neutralized, some I125 label appeared in a high-molecular-weight peak, as might occur if the renin was associated with a binding protein. If the mixture of plasma and labelled renin was treated with guanidine hydrochloride, most of the label appeared in the high-molecular-weight peak. However, after both acid treatment and treatment with guanidine hydrochloride the 'high-molecular-weight' peak appeared in the void volume of the column (Mr 108 000) rather than in the position of inactive renin (Mr 53 000). Also, the I125 label in the peak was neither immunologically nor electrophoretically similar to renin. It may represent denatured renin bound to plasma proteins, e.g. alpha-2 macroglobulins. We conclude that 'inactive renin' is a prorenin-like material rather than a protein-bound form of active renin.

Evidence from peptide mapping for a human renin zymogen.

Data obtained from peptide mapping of the active and inactive forms of human renin show that there are extensive regions of common sequence in the two forms of the enzyme, and are consistent with the hypothesis that inactive renin is a renin zymogen.

The purification of a high-molecular-weight, enzymatically inactive renin precursor from human kidney.

A high-molecular-weight enzymatically inactive form of renin has been purified to homogeneity from human kidney. It has an Mr of 48 000 as determined by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulphate and an Mr of 51 000 by gel filtration on Sephadex G100. It is activated by treatment with trypsin and reversibly activated by exposure to acid. We conclude that this material represents a human prorenin.

Purification of human renin by affinity chromatography using a new peptide inhibitor of renin, H.77 (D-His-Pro-Phe-His-LeuR-Leu-Val-Tyr).

A new affinity column for renin was prepared by coupling the isosteric peptide inhibitor of renin, H.77 (D-His-Pro-Phe-His-LeuR-Leu-Val-Tyr, where R is a reduced isosteric bond, -CH2-NH-), to activated 6-aminohexanoic acid-Sepharose 4B. Chromatography of a crude extract of human kidney cortex on this material resulted in a 5500-fold purification of renin in 76% yield. The purified enzyme (specific activity 871 units/mg) was free of non-specific acid-proteinase activity and was stable at pH 6.8 and -20 degrees C over a period of several weeks.