Hagfish Insulin Research Articles

Structural Basis of the Aberrant Receptor Binding Properties of Hagfish and Lamprey Insulins

The insulin from the Atlantic hagfish (Myxine glutinosa) has been one of the most studied insulins from both a structural and a biological viewpoint; however, some aspects of its biology remain controversial, and there has been no satisfying structural explanation for its low biological potency. We have re-examined the receptor binding kinetics, as well as the metabolic and mitogenic properties, of this phylogenetically ancient insulin, as well as that from another extant representative of the ancient chordates, the river lamprey (Lampetra fluviatilis). Both insulins share unusual binding kinetics and biological properties with insulin analogues that have single mutations at residues that contribute to the hexamerization surface. We propose and demonstrate by reciprocal amino acid substitutions between hagfish and human insulins that the reduced biological activity of hagfish insulin results from unfavorable substitutions, namely, A10 (Ile to Arg), B4 (Glu to Gly), B13 (Glu to Asn), and B21 (Glu to Val). We likewise suggest that the altered biological activity of lamprey insulin may reflect substitutions at A10 (Ile to Lys), B4 (Glu to Thr), and B17 (Leu to Val). The substitution of Asp at residue B10 in hagfish insulin and of His at residue A8 in both hagfish and lamprey insulins may help compensate for unfavorable changes in other regions of the molecules. The data support the concept that the set of unusual properties of insulins bearing certain mutations in the hexamerization surface may reflect those of the insulins evolutionarily closer to the ancestral insulin gene product.

Electronic energy levels and hopping conductivity of an entire molecule of hagfish insulin

The quantum-chemical calculation of an entire molecule of hagfish insulin was done by the ENFC method in which the matrix elements were calculated at the ab initio level using a minimal basis set with simulation of the aqueous solution environment. The ac conductivity for hagfish insulin was also calculated at the ab initio level by random walk theory. All the results were compared with those of pig insulin. It is shown that the reduction of HOMOs and LOMOs localized on the active sites of hagfish insulin agrees with the decrease in the biological reactivity of the insulin. The analysis of primary hopping events showed that a different sequence could influence the biological activity of insulin through the distribution of the hopping centers and the quantities of the hopping frequencies. The curve of the frequency versus ac conductivity of hagfish insulin shows that the different amino acid sequences of proteins influence the hopping conductivity. However, the electronic properties of native proteins are dominated by the three-dimensional conformations. Finally, the electronic mechanism of trans-membrane signal transforms by insulin and its receptor, which had been proposed by Ye and Ladik, was clearly described.

Isolation and structure of lamprey ( Petromyzon marinus) insulin

Insulin has been purified to homogeneity from the caudal and cranial pancreas of the adult sea lamprey ( Petromyzon marinus). The final yield was 18.6 nmol/g (127.8 μg/g). The structures of both A- and B-chains have been determined using amino acid analyses, gasphase sequence analyses, and proteolytic mapping by fast atom bombardment-mass spectrometry. The sequence of the A-chain was found to be GIVEQCCHRKCSIYDMENYCN. The sequence of the B-chain, extended at the amino terminus, was determined to be SALTGAGGTHLCGSHLVEALYVVCGDRGFFYTPSKT. Lamprey insulin retains the common features of vertebrate insulins. Sea lamprey insulin has no more homology to hagfish ( Myxine glutinosa) insulin than it has to the teleost fish or to mammalian insulins.

Phylogenetical aspects on islet hormone families: A minireview with particular reference to insulin as a growth factor and to the phylogeny of PYY and NPY immunoreactive cells and nerves in the endocrine and exocrine pancreas

A common feature in the phylogeny of the four islet hormones (insulin, somatostatin, glucagon, PP) is that they do not seem to occur in the most primitive metazoan animals investigated so far, namely the coelenterates. However, already in the earliest protostomian invertebrates, such as flatworms and annelids, somatostatin and PP immunoreactive nerve fibres were found. In highly developed forms of protostomian invertebrates, such as insects, all the four islet hormones are represented as immunoreactive nerve cells and nerve fibres in the brain. In deuterostomian invertebrates a brain-gut-axis has evolved as regards somatostatin and PP, whereas insulin and glucagon now seem to occur exclusively as cells of open type in the gut mucosa. This brain-gut-axis for somatostatin and PP persists in all the vertebrates. The insulin cells, however, leave the gut mucosa already in the earliest forms of vertebrates and then appear only as cells in the islet parenchyma and in the mucosa of the bile duct (Agnatha) or in the pancreatic ducts (Gnathostomi). To some extent, glucagon islet cells evolve in a similar manner; here, however, cells immunoreactive with the precursor hormone, glicentin (enteroglucagon), persist in the gastrointestinal tract mucosa. A few PYY immunoreactive cells have been found in the pancreatic islet parenchyma of reptiles and mammals, often as disseminated cells in the acinar tissue. In the pancreas of these phyla NPY only occurs in neurons and nerve fibres. In pilot studies the effects of hagfish insulin as a growth factor have been compared with those of pig insulin on Swiss 3T3 mouse embryonic fibroblasts.

Identification and characterization of insulin receptors on foetal-mouse brain-cortical cells.

The occurrence of insulin receptors was investigated in freshly dissociated brain-cortical cells from mouse embryos. By analogy with classical insulin-binding cell types, binding of 125I-insulin to foetal brain-cortical cells was time- and pH-dependent, only partially reversible, and competed for by unlabelled insulin and closely related peptides. Desalanine-desasparagine-insulin, pig proinsulin, hagfish insulin and turkey insulin were respectively 2%, 4%, 2% and 200% as potent as bovine insulin in inhibiting 125I-insulin binding to brain-cortical cells, which corresponds to their relative biological potencies in classical insulin-target cells; no competition was observed with glucagon and nerve growth factor, even at high concentrations. Scatchard analysis of competitive-binding data resulted in curvilinear plots with a high-affinity binding of Ka = 3.6 X 10(8) M-1. Insulin binding to foetal brain-cortical cells differed, however, in two distinct aspects from that to classical insulin-binding cell types. Firstly, dilution of 125I-insulin-bound cells in the presence of unlabelled insulin did not accelerate dissociation of the labelled hormone. Secondly, exposure of brain-cortical cells to insulin before the binding assay enhanced insulin binding, suggesting up-regulation of insulin receptors in response to insulin. In conclusion, foetal-mouse brain-cortical cells bear specific binding sites for insulin. Their insulin receptor shows a marked specificity and affinity for insulin, but differs in at least two properties from most classical insulin receptors. These differences in hormone-receptor interaction could reflect structural differences between insulin receptors on embryonic and differentiated cells.

Effects of hagfish insulin in the Atlantic hagfish, Myxine glutinosa the in vivo metabolism of [ 14C]glucose and [ 14C]leucine and studies on starvation and glucose-loading

In hagfish, starved for 1 month at 4–6°, blood glucose decreased (1.9 to 0.8 m M) and serum insulin values diminished (2.2 to 1.1 n M). More than 90% of the glycogen in the liver and skeletal muscle was consumed, whereas protein and triglyceride contents were far more stable. The serum levels of amino nitrogen, triglycerides, and free fatty acids were all decreased after starvation. The results indicate that skeletal muscle glycogen was the prime source of energy. In starved hagfish, hagfish insulin (0.1 μg/g body weight) induced an approximate twofold stimulation of the synthesis of glycogen, protein, and neutral lipids from [ 14C]glucose after 33 hr at 4–6°. Most of the incorporation was detected in muscle glycogen. No insulin effects were seen in the liver. In analogous studies with [ 14C]leucine, hagfish insulin likewise stimulated the synthesis of glycogen and protein in muscle and protein synthesis in the liver. Despite the evidence of insulin-stimulated syntheses, the total glycogen and protein contents in muscle and liver were unaltered after 33 hr. Likewise, no insulin effects were seen on blood glucose, amino nitrogen, triglycerides, or free fatty acids. Only about 10% of the radioactive dose was incorporated into the muscle and liver, and the metabolic effects of insulin contributed to only about half of this fraction. Glucose-loading increased the serum insulin value from 0.7 to 1.9 n M. Pretreatment of hagfish with a mixture of glucose and amino acids for 3 days before and after injection of the above isotopes resulted in an increase of the serum insulin values. These endogenously elevated insulin levels were sufficient to stimulate the incorporation of the label into muscle glycogen and protein. The stimulations were similar to those obtained in experiments with exogenous insulin. It was concluded that the physiological role of insulin in skeletal muscle was similar to what has been observed in higher animals although the quantitative effects of insulin in the hagfish appeared smaller than in higher vertebrates.

Optical diffraction analysis of crystalline inclusions in the rough endoplasmic reticulum of islet parenchymal cells of the hagfish, Myxine glutinosa.

Optical diffraction analysis was carried out on crystalline inclusions in the rough endoplasmic reticulum of the insulin and somatostatin cells in the islet organ of the hagfish. A striking difference in crystalline arrangement was observed between the inclusions of the insulin and somatostatin cells. The crystallographic arrangement of the inclusions observed in situ in the insulin cells differed from that previously found by means of X-ray diffraction analyses of hagfish insulin crystals formed in vitro.

Fast Fourier refinement of the structure of hagfish insulin at 1.9 Å resolution

Fast Fourier refinement of the structure of hagfish insulin at 1.9 Å resolution

1H nuclear magnetic resonance study of the histidine residues of insulin

Insulin has proved difficult to study by nuclear magnetic resonance spectroscopy because of its complex aggregation behaviour in solution and its insolubility between pH 4 and 7. Now for the first time it has been possible to assign the 1H nuclear magnetic resonances of the H-2 histidine protons of residues B5 and B10 of bovine 2 Zn insulin and Zn-free insulin, and the B5 and A8 residues of hagfish insulin. As expected, the addition of Zn to Zn-free insulin causes virtually no change in the chemical shift or the rate of H-D exchange of the H-2 proton of histidine B5, which is not involved in Zn binding in the 2 Zn insulin hexamer. The rate of H-D exchange of the H-2 proton of histidine B10 is decreased markedly on Zn binding at this residue, but the chemical shift of the resonance remains virtually constant owing to the balancing of an upfield ring current shift of the ordered histidine residues by a downfield shift due to electron withdrawal from the ring nitrogen by the Zn binding.

A specific antiserum against insulin from the Atlantic hagfish, Myxine glutinosa: Characterization of the antiserum, its use in a homologous radioimmunoassay, and immunofluorescent microscopy

The radioimmunoassay for hagfish insulin described employs a specific anti-hagfish-insulin serum raised in guinea pigs, 125I-labeled hagfish insulin, and purified hagfish insulin as standard. The apparent dissociation constant of the antiserum was 0.24 ± 0.05 n M. The cross-reactivity with ox insulin was on the order of magnitude of 0.01%. Equilibrium binding was reached within 24 hr at 4°. Insulin values, ranging from 50 pg to 5–10 ng, could be measured. The free and the bound label could be separated conveniently either with polyethylene glycol or a second precipitating antibody. Immunofluorescence microscopy showed that the antiserum reacted with hagfish islet cells similarly to what has earlier been reported using antiserum against mammalian insulins.

Insulin action on adipocytes. Evidence that the anti-lipolytic and lipogenic effects of insulin are mediated by the same receptor.

1. The dose-response relationships of insulin stimulation of lipogenesis and inhibition of lipolysis were studied simultaneously by using rat adipocytes to determine whether these different effects of insulin are mediated through the same or different sets of receptors. 2. The sensitivity (defined as the concentration of insulin required to produce a half-maximal effect) of the stimulated lipogenic response to insulin was not significantly different from the sensitivity of the anti-lipolytic response to insulin. The addition of different adrenaline and glucose concentrations did not alter the half-maximal concentration of insulin required to inhibit lipolysis. 3. The specificities of the lipogenic and antilipolytic responses were studied by using insulin analogues. The sensitivities of the lipogenic and anti-lipolytic responses were the same for five chemically modified insulins and hagfish insulin, which have potencies compared with bovine insulin of between 3 and 90%. 4. Starving rats for 48h significantly increased the sensitivities of both the antilipolytic and lipogenic responses to insulin, but the changes in the sensitivities of both lipogenesis and anti-lipolysis returned to that of fed rats. 5. We conclude that insulin stimulates lipogenesis and inhibits lipolysis over the same concentration range. These observations provide powerful evidence that the different effects of insulin are mediated through the same set of receptors.

Structure and biological activity of hagfish insulin

An isomorphously phased electron density map of hagfish ( Myxine glutinosa) insulin has been calculated at a resolution of 3·1 Å spacing. The molecule crystallises with one molecule per asymmetric unit but is organised as a symmetric dimer lying on a 2-fold crystal axis. The structure of the hagfish insulin monomer is much more similar to that of pig insulin molecule 2 than molecule 1 of the dimer that constitutes one third of the 2 Zn insulin hexamer. There are different conformations however at the N and C termini of the B-chain. At the C terminus the two final residues on hagfish insulin partially obscure the A1 glycine residue, which in pig insulin is exposed. This structural difference has been shown, however, not to be responsible for the reduced activity of the hagfish insulin.

The Insulin Receptor and Insulin of the Atlantic Hagfish: Extraordinary Conservation of Binding Specificity and Negative Cooperativity in the Most Primitive Vertebrate

The North Atlantic hagfish, a cyclostome that is representative of the most primitive vertebrates still alive, diverged from the other vertebrates about 500 million years ago. Hagfish insulin, which differs from porcine insulin in 18 (38%) of its amino acids, had a potency of 5-10% that of porcine insulin in stimulating glucose oxidation and deoxyglucose transport in rat adipocytes and was 5-10% as potent as porcine insulin in binding to insulin receptors on rat adipocytes and human (IM-9) lymphocytes. Like all other naturally occurring insulins, hagfish insulin accelerated the dissociation of 125I-porcine insulin from insulin receptors and the degree of the acceleration was related to its occupancy of the receptor. The insulin receptor of the hagfish erythrocyte showed the time, temperature, and pH dependence of binding and the negative cooperativity that are characteristic of all other insulin receptors. That the negative cooperativity is fully conserved in such an ancient insulin and receptor suggests that it is an important functional feature of this hormone-receptor system. The hagfish receptor showed the same absolute affinity and rank order of preference for insulins and insulin analogues (chicken > pork > proinsulin > guinea pig > desoctapeptide) found with other receptors of less primitive vertebrates, which supports the conclusion that the receptor for insulin is functionally better conserved evolutionary than the hormone. However, uniquely, hagfish insulin was more potent in binding to hagfish receptors than to mammalianreceptors; with all other species of insulins studied, the affinity of the hormone for homologous receptor was the same as for receptors of heterologous species.

Evidence for a biological role of somatostatin in the Pacific hagfish, Eptatretus stouti

Immunoreactive somatostatin was found in acetic acid extracts of the intestine (10–11 pg/mg wet wt) and the islet-organ (approximately 10 ng/mg wet wt) of the Pacific hagfish, Eptatretus stouti. Hagfish islet-organs were incubated in vitro in a modified Krebs-Ringer bicarbonate medium, and the concentration of insulin released into the medium was measured with a specific radioimmunoassay for hagfish insulin. The amount of insulin released by islet-organs incubated in medium containing 20 mg of glucose/dl did not differ significantly from that released by tissues incubated in medium containing 120 mg of glucose/dl. Release of insulin decreased markedly when tissues were incubated in medium containing 120 mg of glucose/dl and 100 ng of cyclic somatostatin/ml. Subsequent incubation of the somatostatintreated islet-organs in medium without somatostatin resulted in recovery of insulin secretion to the expected rate. The results of this study provide evidence that somatostatin may function in the regulation of insulin secretion in the cyclostomes.

PHYLOGENY OF INSULIN:Some Evolutionary Aspects of Insulin Production with Particular Regard to the Biosynthesis of Insulin in Myxine glutinosa

Preceding phylogenetic studies on the occurrence of insulin have shown--e.g. by bioassays and immunocytochemical procedures--that insulin producing B-cells are present in all vertebrates and even in several invertebrates, both protostomian and deuterostomian. The most original B-cells are obviously endocrine cells of open type, situated in the mucosa of the alimentary tract. Moreover, the results of these studies show that insulin is not only a polypeptide hormone of considerable age but also that the insulin molecule seems to have been kept surprisingly stable during evolution. Best known of all non-mammalian insulins is that from the hagfish, Myxine glutinosa. It is probably the most original insulin of all in the vertebrate series. Both the amino-acid sequence and the three-dimensional structure of the dimer of hagfish insulin differ only little from those of pig insulin. The biosynthesis occurs via proinsulin and is also in most respects similar to mammalian insulin biosynthesis. There are, however, some differences. Although it readily crystallizes as tetragonal bipyramids, hagfish insulin does not form hexamers. In a test system, with isolated rat fat cells, its binding affinity is 23% and its potency 5% of that of pig insulin, a discrepancy indicating a "partial antagonism" on the receptors. Although the conversion of proinsulin to insulin seems to occur in the secretion granules, they contain no crystalline cores. Since a strictly tryptic-like enzyme was found to destroy hagfish insulin rapidly, the enzyme converting proinsulin to insulin must--in addition to a carboxy-peptidase-B-like activity--have a different specificity in Myxine.

ZINC AND MANGANESE CONTENTS OF MICRO-DISSECTED PANCREATIC ISLETS OF SOME RODENTS

By a flameless atomic absorption spectrophotometry procedure, using a graphite furnace, it was possible to assay the contents of zinc and managanese in micro-dissected pancreatic islets of several rodents. Interest was focused upon the islets of guinea-pigs, due to the fact that guinea-pig insulin lacks a histidine residue in the B10 position of the molecule which normally binds zinc (or other heavy metals) in the hexamer formation, probably involved in the storage of insulin. Both the zinc and manganese contents were too low in the guinea-pig islet parenchyma to be reasonably involved in the storage of insulin in the beta-granules. Instead, it was suggested that guinea-pig insulin, like hagfish insulin, might crystallize without access to zinc or other heavy metals. Low zinc and manganese contents were also observed in newborn and diabetic guinea-pigs. The islet zinc content was high in the Wistar rat, the Chinese hamster, and the spiny mouse. No significant amounts of manganese were found in any of these kinds of islet parenchyma.

Degradation, receptor binding affinity, and potency of insulin from the Atlantic hagfish (Myxine glutinosa) determined in isolated rat fat cells.

Insulin from the Atlantic hagfish, Myxine glutinosa, a primitive vertebrate, was studied with respect to degradation, receptor binding, and stimulation of glucose transport and metabolism in isolated rat adipocytes. The degradation was studied in a concentrated suspension with about 100mul of cells/ml of suspension. 125I-labeled hagfish insulin and 125I-labeled pig insulin were degraded at the same rate when present in concentrations of 0.3nM. Native hagfish insulin inhibited the rate of degradation of 125I-labeled pig insulin half-maximally at a concentration of 12+/-2 nM (S.D., n=6) as compared to 130+/-32 nM (S.D.,n=6) for pig insulin. Native hagfish insulin in a concentration of 130 nM was biologically inactivated at a rate several times slower than pig insulin in the same concentration. The results indicate that the maximal velocity (Vmax) of degradation of hagfish insulin as well as the concentration causing half-maximal velocity (Km) are about 10 times lower for hagfish insulin than for pig insulin. The receptor binding and the biological effects of hagfish insulin were studied in dilute cell suspensions where the degradation of hormone in the medium was negligible. The receptor binding affinity of hagfish insulin was 23+/-7 per cent (S.D., n=10) of that of pig insulin. Hagfish insulin was able to elicit the same maximal stimulation of both 3-o-methylglucose exchange and lipogenesis from glucose as pig insulin. However, the potency of hagfish insulin with respect to activation of lipogenesis was only 4.6+/-0.6 per cent (S.D., n=15) of that of pig insulin. Hagfish insulin thus constitutes the first described insulin which exhibits a discrepancy between relative binding affinity and relative potency. This discrepancy was not due to the methionine residue (B31) at the COOH-terminal end of the B chain of hagfish insulin, since removal of this residue caused no marked change in the binding affinity or the potency. The results indicate that the receptor occupancy must be 5 times higher with hagfish insulin than with pig insulin to cause a particular degree of activation of lipogenesis. Hagfish insulin might therefore be characterized as a "partial antagonist" on the receptors. However, it was not possible to demonstrate antagonistic properties of hagfish insulin on the cells. The effect of hagfish insulin plus pig insulin in submaximally stimulating concentrations was additive. Furthermore, the decay of activation of adipocytes after incubation with hagfish insulin followed the same time course as the decay of activation after incubation with pig insulin in a concentration of equal potency. These phenomena are in agreement with the concept that adipocytes possess a large excess of receptors which can mediate the effect of insulin on lipogenesis from glucose.

The amino acid sequence of the insulin from a primitive vertebrate, the atlantic hagfish (Myxine glutinosa).

Insulin has been isolated and purified from the islet organs of the cyclostome, Myxine glutinosa, by means of acid-ethanol extraction, fractional precipitation, and gel filtration. The complete amino acid sequence of the hormone has been determined by Edman degradation of the S-carboxymethylated or performic acid-oxidized A and B chains, and of various tryptic peptides derived from the chains. The 52-residue hagfish insulin has many structural features in common with other vertebrate insulins including the locations of the 6 half-cystine residues, the NH2-terminal 7 residues and the COOH-terminal 6 residues of the A chain, and several shorter sequences in the B chain that are known to comprise the dimer interface in porcine insulin crystals. Of the 24 residues which are invariant among the other known insulins, 23 are identical in hagfish int 16 of these sites it contains residues not previously observed in vertebrate insulins. The B chain also contains an additional COOH-terminal residue of methionine, making it 1 residue longer than the usual 30-residue mammalian B chains. Several features of the tertiary and quaternary structure of hagfish insulin, including the probable absence of a metal ion-stabilized hexameric form, are discussed on the basis of these findings. The results suggest that the conformation of the insulin molecule has been well conserved throughout the entire evolution of the vertebrates.

Low resolution crystal structure of hagfish insulin

Insulin from the Atlantic hagfish, Myxine glutinosa, crystallizes in space group P4 12 12 with a monomer in the asymmetric unit. The application of the Rossmann & Blow (1962) rotation function, utilizing the known 2-zinc pig insulin crystal structure, has established the existence of an insulin dimer containing a crystallographic 2-fold axis. The position of the hagfish insulin molecule in the unit cell has been determined and a set of calculated phases derived. These are compared to phases found from isomorphous replacement studies. A 6 Å resolution electron density map has been calculated which shows the A and B chains are folded in a similar way to pig insulin and that the monomers are similarly organized into dimers.

Insulin in Invertebrates and Cyclostomes

It seems increasingly clear that insulin is a hormone that does not occur exclusively in vertebrates. Several independent reports now exist giving evidence of insulin production in the digestive tissues of both deuterostomian and protostomian invertebrates. Cells with some light-microscopical and ultrastructural characteristics of vertebrate B-cells have also been observed. Recently, evidence has been obtained that insulin can act as a hypoglycemic hormone, promoting glycogen synthesis, also in a protostomian invertebrate, the gastropod mollusc, Strophocheilus oblongus . The endocrine pancreas of the cyclostomes occupies a key position in the comparative endocrinology of the islet parenchyma and in the evolution of insulin production. It may represent an evolutionary link between the presumably gut-connected dispersed insulin-producing cells of deuterostomian invertebrates and the pancreatic islets of gnathostomian vertebrates. This hypothesis was supported by the fact that cells with light-microscopical and ultrastructural similarities to the islet B-cells were observed in the bile duct mucosa of the hagfish, Myxine glutinosa . However, immunofluorescent studies with antisera against human insulin and C-peptide did not show any immunoreactive material outside the B-cells of the endocrine pancreas. Particular attention was paid to elucidate the biological significance of the large cystic cavities that are so typical for the cyclostomian islet parenchyma. The working hypothesis that they may contain stored insulin, proinsulin (or even “proto-proinsulin”) was not supported by immunofluorescence, autoradiographic, or ultrastructural investigations, nor by proinsulin assays. It is possible that hagfish islet B-cells contain zinc, despite the fact that the amino acid residue in B10-position is aspartic acid instead of histidine. The biosynthesis of hagfish insulin shows a pattern similar to that in gnathostomes. Its rate is related to the ambient temperature and at 11 C the conversion of proinsulin to insulin lasts several days.