Preparation and properties of fluorescent insulin derivatives

Synthesis and Identification of FITC-Insulin Conjugates Produced Using Human Insulin and Insulin Analogues for Biomedical Applications

Human insulin was fluorescently labelled with fluorescein isothiocyanate (FITC) and the conjugate species produced were identified using high performance liquid chromatography and electrospray mass spectroscopy. Mono-labelled FITC-insulin conjugate (A1 or B1) was successfully produced using human insulin at short reaction times (up to 5 h) however the product always contained some unlabelled native human insulin. As the reaction time was increased over 45 h, no unlabelled native human insulin was present and more di-labelled FITC-insulin conjugate (A1B1) was produced than mono-labelled conjugate with the appearance of tri-labelled conjugate (A1B1B29) after 20 h reaction time. The quantities switch from mono-labelled to di-labelled FITC-insulin conjugate between reaction times 9 and 20 h. In the presence of phenol or m-cresol, there appears to be a 10 % decrease in the amount of mono-labelled conjugate and an increase in di-labelled conjugate produced at lower reaction times. Clinically used insulin analogues present in commercially available preparations were successfully fluorescently labelled for future biomedical applications.

Site-specific fluorescein labeling of human insulin

Three fluorescein derivatives of human insulin (HI, 1) labeled at positions N(αA1) , N(αB1) and N(εB29) respectively, were synthesized using an N-trifluoroacetyl-based protecting group scheme. The Tfa protecting group introduced by reaction with ethyl trifluoroacetate was found to be stable in aqueous and organic media and efficiently removed under mild basic conditions.

Remote loading of diclofenac, insulin and fluorescein isothiocyanate labeled insulin into liposomes by pH and acetate gradient methods

Remote loading of the model drugs diclofenac, insulin and fluorescein isothiocyanate labeled insulin (FITC-insulin) into liposomes by formation of transmembrane gradients were examined. A trapping efficiency of almost 100% was obtained for liposomal diclofenac, by the calcium acetate gradient method, whereas liposomes prepared by the conventional reverse-phase evaporation vesicle method had 1-8% trapping efficiencies. Soybean-derived sterol was a better stabilizer of the dipalmitoylphosphatidylcholine bilayer membrane than cholesterol, as shown from trapping efficiencies and drug release. The pH gradient method resulted in a 5-50% of FITC-insulin liposomal trapping efficiency, while insulin could not be loaded by this method. Liposomes released calcein in response to insulin, showing insulin interacts with the liposomal membrane in the presence of a transmembrane gradient. The present work has demonstrated a remote loading method for weak acids such as diclofenac into liposomes by the acetate gradient method. From the result of remote loading of FITC-insulin into liposomes by the pH gradient method, this method may be available for the preparation of liposomal peptides.

Synthesis and characterization of insulin-fluorescein derivatives for bioanalytical applications

Human insulin was labeled with fluorescein isothiocyanate (FITC) and fully characterized to yield four distinct insulin-FITC species. High-performance liquid chromatography and electrospray mass spectrometry were used to determine the extent and location of fluorescein conjugation. By changing the reaction conditions (i.e., pH, time, and FITC/insulin ratio) the selectivity of the fluorescein conjugation was altered, and all conjugates could be separated. The isolated species of insulin-FITC were labeled at the following residues: A1(Gly), B1(Phe), A1(Gly)B1(Phe), and A1(Gly)B1(Phe)B29(Lys). All four insulin-FITC conjugates were then used to develop fluorescence polarization binding assays with monoclonal and polyclonal anti-insulin antibodies. The assay sensitivity differed between the conjugates depending on the site of modification (B1 > A1 > A1B1 > A1B1B29). Also, the type of antibody used had an important role in the binding of insulin-FITC conjugates. Finally, for the first time the biological activity of the four conjugates was demonstrated by an autophosphorylation assay. The positional substitution dramatically affected the biological activity, confirming insights into the residues responsible for the insulin binding region. The B1 conjugate was found to retain almost all biological activity while the A1 and A1B1 conjugates had approximately 10 times lower activity. The trisubstituted species (labeled at A1, B1, and B29) was determined to be least active.

Fluoresceinthiocarbamyl-insulin: a potential analytical tool for the assay of disulfide bond reduction

We describe the synthesis of fluorescent derivatives of bovine pancreas insulin and its use as substrates of disulfide bond reduction in a spectrofluorometric assay. Amino groups of insulin were chemically modified with fluorescein isothiocyanate and proteins bearing one, two and three fluorescent groups were purified by ion-exchange chromatography. Upon incubation with dithiothreitol, di- and tri-fluoresceinthiocarbamyl-insulin evinced the highest and the lowest enhancement of fluorescence emission, whereas the mono-substituted protein had intermediate enhancement. Using di-fluoresceinthiocarbamyl-insulin, the reliability of this novel feature for the estimation of disulfide bond cleavage was assessed by (i) the separation of two fluorescent bands using sodium dodecyl sulfate-polyacrylamide gel electrophoresis, (ii) the linear response of the fluorescence signal within a range from 0.04 to 1 microM, and (iii) the correlation of the rate of fluorescence enhancement with concentrations of dithiothreitol ranging from 0.1 to 5 mM. Moreover, di-fluoresceinthiocarbamyl-insulin was a sensitive oxidant when the catalytic capacity of thioredoxin and protein disulfide isomerase was analyzed in the presence of dithiothreitol or glutathione, as reductants. On this basis, di-fluoresceinthiocarbamyl-insulin constitutes an analytical tool to test the capacity of biochemical preparations in the reduction of disulfide bonds.

Di-fluoresceinthiocarbamyl-insulin: a fluorescent substrate for the assay of protein disulfide oxidoreductase activity

We have developed a novel method for the continuous assay of protein disulfide oxidoreductase activity using as substrate bovine pancreas insulin in which both N-terminal amino groups are chemically modified with fluorescein isothiocyanate. The reduction of intercatenary disulfide bonds of di-fluoresceinthiocarbamyl-insulin with dithiothreitol initially lowers but subsequently enhances the emission intensity. In this biphasic kinetics, the rate of increase is sensitive enough for the estimation of Escherichia coli thioredoxin concentrations from 5 nM (0.06 microgram/ml) to 500 nM (6 micrograms/ml). Neither changes of pH over a range of 6.2 to 8.4 nor neutral salts (K+, Mg2+, and Ca2+) at concentrations lower than 100 mM affect this simple reaction system. Moreover, the fluorometric method is functional for measuring the reductive capacity of Brassica napus protein disulfide isomerase. Hence, a highly reproducible and accurate one-state assay for protein disulfide oxidoreductase activity not only greatly improves the sensitivity compared to the commonly used turbidimetric assay but also represents a reliable alternative to assays based on accessory enzymes or radiolabeled substrates.

Interaction of alpha-crystallin with spin-labeled peptides

alpha-Crystallin is a major protein of the vertebrate lens once thought to be highly specialized for conferring transparency. However, recent work has revealed a wide tissue distribution and a sequence homology to small heat shock proteins, suggesting a more general role for the protein. Like other molecular chaperons, alpha-crystallin is known to bind to unfolded proteins and suppress nonspecific aggregation in vitro. In the present work, spin-labeled derivatives of the insulin B chain and melittin were used to investigate the state of these proteins bound to alpha-crystallin. Insulin was selected since unfolding can be triggered by reduction of the interchain disulfide bonds, a treatment that does not affect alpha-crystallin. Upon reduction of insulin, the separated B chains aggregate. In the presence of alpha-crystallin, the B chains bind to alpha-crystallin and aggregation is suppressed. Melittin, a 26 amino acid peptide from bee venom, was selected for study since it is a random coil under physiological conditions, and its interaction with alpha-crystallin can be directly studied. EPR analysis of the spin-labeled peptides shows that the nitroxide side chains are immobilized in a polar environment on alpha-crystallin and that they are separated by 25 A or more in the complex, indicating that the bound proteins are not clustered. The bound B chains of insulin are not in a fully extended conformation, and melittin does not appear to bind to a hydrophobic surface in alpha-crystallin as an amphipathic helix, as it does to membranes and some other proteins.(ABSTRACT TRUNCATED AT 250 WORDS)

Insulin binding to human cultured lymphocytes measured by flow cytometry using three ligands

The binding of insulin to cultured IM-9 human lymphocytes was studied by flow cytometry using FITC-insulin and biotinylated insulins coupled to streptavidin-phycoerythrin (N alpha beta 1-biotinylinsulin (B-insulin) and N alpha B1-(biotinyl-epsilon-aminocaproyl)insulin (NBC-insulin)). The reference methods were 125I-insulin binding and the insulin-antiinsulin antibody complexes for flow cytometry. There was a close correlation between 125I-insulin binding and increase in fluorescence for B-insulin, NBC-insulin, and insulin-anti-insulin antibody complexes, but not for FITC-insulin. NBC-insulin gave the largest increase in fluorescence (79 +/- 9 channels) and the the insulin-antiinsulin antibody complexes the smallest (34 +/- 2 channels) (P < 0.05). FITC-insulin and B-insulin gave similar results: 47 +/- 6 and 59 +/- 6 channels. The concentration reducing 125I-insulin binding by 50% was 1.1 x 10(-9) M for native insulin, 2.7 x 10(-9) M for B-insulin, 3.3 x 10(-9) M for NBC-insulin, and 6.6 x 10(-9) M for FITC-insulin (P < 0.05). Nonspecific binding was low for B-insulin and NBC-insulin but reached 75% for 10(-6) M FITC-insulin. These results suggest that B-insulin and NBC-insulin are suitable ligands for insulin binding studies using flow cytometry. This two-step procedure is easier than the insulin-antiinsulin antibody complex technique. Its poor affinity, specificity, and sensitivity make FITC-insulin less suitable.

Relationship between cell differentiation and binding of fluorescein isothiocyanate (FITC)-conjugated insulin of keratinocytes in culture

Basal-type keratinocytes, isolated from newborn rat skin and separated on Percoll density gradients, proliferate in low (0.1 mM) calcium medium and, after raising the calcium level to normal (1.96 mM), stratify. Cells in the low calcium culture do not have extensive cell-cell connections, as seen with fluorescein isothiocyanate (FITC)-conjugated insulin. Fluorescein isothiocyanate-conjugated concanavalin A and Griffonia simplicifolia isolectin B4, but not peanut agglutinin (PNA), fluorescently label these cells. In 3-day-old low calcium cultures, within 2 h after raising the calcium of the medium to the normal level, intense binding of PNA to cells appears and neighboring cells are connected through bundles of filaments that are fluorescently labeled by FITC-insulin. After 2 days in normal calcium medium, the cultures exhibit relatively smooth, straightlined, cell boundaries that are labeled by FITC-insulin and cell boundaries and intracellular granules that are stained by hematoxylin. One day later, similar cell boundaries are present, but they are not significantly decorated by FITC-insulin and, under phase contrast microscopy, are dark. Free FITC gives labeling patterns similar to those given by FITC-insulin, but the FITC labeling is blocked by mercaptoethanol and dithiothreitol in contrast to FITC-insulin binding. The present results suggest the insulin moiety is involved in the labeling by FITC-insulin and the labeling is chronologically related to the stage of cell differentiation.

Biliary excretion of FITC metabolites after administration of FITC labeled asialo orosomucoid as a measure of lysosomal proteolysis

An isolated perfused rat liver system was used to study the hepatic uptake and degradation of asialo orosomucoid (asialo alpha 1-acid glycoprotein). To this aim we coupled the fluorochrome FITC to the asialoglycoprotein. The covalent attachment of FITC to the glycoprotein did not affect its perfusate disappearance. The disappearance rate was characterized by a t1/2 approximately equal to 6.1 min, the clearance being 11.2 ml/min. The internalized ligand was probably extensively degraded in the lysosomes as demonstrated by the appearance of low molecular weight fluorescent compounds in the bile, having a higher fluorescence yield than the native conjugate. Lysosomal degradation of ASOR-FITC was shown to be the rate limiting step in FITC excretion into the bile. Treatment of a perfused liver with varying doses of the protease inhibitor leupeptin did not influence the perfusate disappearance rate of the protein. However, leupeptin inhibited the biliary output of FITC metabolites in a dose dependent fashion, half maximal inhibition occurring at 210 nM (in the perfusion medium), corresponding with a dose of 0.05 mg leupeptin per 10 g liver. It is concluded that the rate of lysosomal degradation of proteins in vivo can be determined by measuring the biliary excretion of fluorescent material originating from fluorescent probes covalently coupled to the particular protein.

Renal biopsies performed on diabetics

Metabolic disorders and immunological factors are discussed in connection with the pathogenesis of diabetic microangiopathy. Renal biopsies were obtained from 22 diabetics (8 women aged 18 to 53, 14 men aged 15 to 52). 7 of the 22 patients had been suffering from diabetes for 2 weeks to 3 years, 10 for 7 to 25 years, 2 showed a pathological glucose-tolerance test, i.e., they had been "latent" diabetics, and 3 patients, had been so-called "potential" subjects of diabetes due to hereditary traits or delivery of big babies. They were examined by light miroscopy as well as by immunofluorescence microscopy. A number of cases were chosen for the differentiation and counting of glomerular cells (n=8) as well as for electron microscopic (n=7) and polarizing-microscopic (n=6) examinations. Histologically, focal proliferations of mesangial cells as well as an increase in mesangial substance in the glomeruli was found in all cases, although in a varying degree of intensity. These results were confirmed by both the glomuerular cell count and electron-microscopic examination. Immunofluorescence microscopy made it possible to detect frequently both IgA (9/17) and IgG (9/17), usually in either linear or mesangial arrangements whereas it was less frequently possible to detect IgM (1/17) and albumin (1/8) and impossible to detect beta1C in the glomerulus. Labeled insulin was detected five times in the glomerulus. Polarizing-microscopic measurements made in order to discover possible submicroscopic variations in the structure of GBM showed deviations in the average values of anisotropic indices from the controls in the group of long-term diabetics only. The pathogenesis of diabetic microangiopathy may be described as an inflow of immunoglobulins and serum proteins into the mesangium because of an alteration of the capillary endothelium, the mesangial cell being thus caused to overfunction, proliferate and produce an excess of mesangial matrix. In prolonged diabetes the mesangial cell, so far as its own metabolism is concerned, will finally be affected to the point where its power of synthesis is modified in the sense of an excess and/or faulty composition of GBM (glomerular basement membrane).

Acetoacetylation of insulin

Insulin was treated with diketen at pH6.9. The reaction mixture was resolved into four components by DEAE-Sephadex chromatography. The first component was unchanged insulin. The second and third components were shown by end-group analysis to be substituted on phenylalanine B-1 and glycine A-1 respectively. The fourth component was disubstituted on both phenylalanine B-1 and glycine A-1. The in-amino group of lysine B-29 was not involved in the reaction at low reagent concentrations. The purity of these derivatives was checked by their electrophoretic behaviour and by measurement of the rate of their reaction with trinitrobenzenesulphonic acid. The hormonal activity of the derivatives was determined. The effect of the modifications on the hormonal activity and the tertiary structure of insulin is discussed.