Hormonal control of inflammatory responses

Almost any stage of inflammatory and immunological responses is affected by hormone actions. This provides the basis for the suggestion that hormones act as modulators of the host reaction against trauma and infection. Specific hormone receptors are detected in the reactive structures in inflamed areas and binding of hormone molecules to such receptors results in the generation of signals that influence cell functions relevant for the development of inflammatory responses. Diversity of hormonal functions accounts for recognized pro- and anti-inflammatory effects exerted by these substances. Most hormone systems are capable of influencing inflammatory events. Insulin and glucocorticoids, however, exert direct regulatory effects at concentrations usually found in plasma. Insulin is endowed with facilitatory actions on vascular reactivity to inflammatory mediators and inflammatory cell functions. Increased concentrations of circulating glucocorticoids at the early stages of inflammation results in downregulation of inflammatory responses. Oestrogens markedly reduce the response to injury in a variety of experimental models. Glucagon and thyroid hormones exert indirect anti-inflammatory effects mediated by the activity of the adrenal cortex. Accordingly, inflammation is not only merely a local response, but a hormone-controlled process.


Introduction
An impressive challenge imposed by external factors to an organism in the animal kingdom, particularly in vertebrates, is the reaction to emergency conditions, such as trauma and infection. To deal efficiently with the corresponding demands, the organism activates a narrowly regulated set of events, characteristic of the inflammatory response. These include functional alterations of microvessels, migration of haematogenous cells to the injured area, phagocytosis by competent cells and local pain. Recruitment of components of the immune system leads to the amplification of the reaction.
Inflammation was considered a disease until Hunter in 1794, described it as a nonspecific response evoked by all manner of injury and both purposeful and beneficial in nature. This view was strongly supported by the results of Metchnikoff's investigations on the comparative pathology of inflammation dating from the end of the nineteenth century. The beneficial character of the response, however, was not easily assimilated at the time, inflammation being rather regarded as a passive and deleterious reaction, of no benefit to the host. Experiments in which pyogenic bacteria were used (C) 1993 Rapid Communications of Oxford ktd as the inflammatory stimulus, convincingly illustrate the defensive aspect of inflammation. In these experiments, recognized components of the early inflammatory reaction were blocked by systemic shock, local ischaemia, or administration of anticomplementary substances, and the effect of a modified host resistance on the final size of the developing bacterial lesion was compared to the lesion size in control animals. In all cases, the injection of bacteria into the dermis during periods of shock or ischaemia, or following the administration of anticomplementary substances resulted in maximal lesions substantially larger and more severe than the control lesions, am The results indicate that there is an early period in which the integrity of the response against injury is essential to prevent or limit the intensity of a future lesion, and stress the defensive role of inflammation.
A most significant aspect in earlier views on inflammation is the notion that the response is a pattern reaction developing regardless of the cause. Accordingly, any type of injury tends to evoke a similar response of the organism. It was from such notion that the idea of the response being mediated by endogenously mobilized active materialsthe inflammatory mediators--arose. Ideally, a potential inflammatory mediator should have the appropriate properties to bring about that part of the inflammatory response for which it is deemed to be responsible; it should be demonstrably present for a phase of the inflammatory response; and if it is possible to deplete the tissues of the mediator, this should also lead to a suppression of the inflammatory event for which it is responsible. 5'6 From numerous observations, generally it has been concluded that, in addition to chemical mediators, the inflammatory response is under the influence of substances of diverse origin which are capable of modulating its development. It is possible to identify local tissue factors, plasma factors, neurogenic factors, and hormones which 7<0 act as modulators of the inflammatory process. These substances provide the basis for the occurrence of regulatory mechanisms responsible for a fine-tune adjustment of the host response in accordance with the intensity of the damage inflicted. Were it only merely a local response, inflammation would hardly be so effective as a defensive reaction against injury.
Hormone studies from the early decades in this century, clearly indicate that many biological systems are subject to control by hormone actions. Binding of hormone molecules to specific receptors in target structures results in the generation of signals that influence cell functions. Specific ho.rmone receptors are detected in the reactive structures in inflamed areas and, in most instances, there is a striking similarity of the binding sites in those structures and those present in recognized target cells as far as the apparent aflqnity, specificity, and kinetics of the interaction are concerned. Accordingly, hormones can influence cell functions relevant for the development of inflammatory responses, including vascular reactivity to inflammatory mediators, oriented cell locomotion, particle ingestion by phagocytes, and microbicidal function. Granulation tissue formation, healing, and repair are also affected by the activity of endocrine glands.
Diversity of hormonal functions is likely to account for the recognized pro-and anti-inflammatory effects exerted by these substances.
Most hormone systems are capable of influencing the development of inflammatory reactions. 1 Insulin and glucocorticoids, however, appear to exert direct regulatory eects on the reactive structures in an inflamed area, at concentrations usually found in plasma.
The proinflammatory effect of insulin The insulin receptor: Insulin occurs throughout the entire vertebrate kingdom and is probably also present in many invertebrates. It serves the essential role of regulating the synthetic phase of energy metabolism by promoting glucose and amino acid uptake, glycogen and protein synthesis, and lipogenesis. The existence of a specific insulin receptor was postulated as early as 20 years ago.l-3 The receptor is a heterotetrameric structure consisting of two -subunits of M 135 kDa on the outside of the plasma membrane connected by disulphide bonds to/-subunits of M 95 kDa which are transmembrane proteins. Insulin binding to the -subunit induces conformational changes which are transduced to the/-subunit. This leads to the activation of a tyrosine kinase activity which is intrinsic to the cytoplasmic domain of the fl-subunit. Activation of the tyrosine kinase activity of the insulin receptor is an essential step in the transduction of an insulin signal across the plasma membrane of target cells. Signal transduction on the postkinase level possibly involves phosphorylation of substrate proteins at tyrosine residues, activation of serine kinases, and interaction with G-proteins, phospholipases and phosphatidylinositol kinases. 4 A large number of agents and altered physiological conditions are known to affect insulin binding to target cells. Addition of catecholamines, glucagon, or glucocorticoids to cells, as well as the diabetic state and obesity modify insulin receptor activity. s-18 Insulin and the endothelial cell: Insulin receptors are demonstrable on the surface of endothelial cells. 19 Regional dit:ferences, however, are observed in insulin receptor concentration, as revealed by insulin binding studies in primary cultures of endothelial cells. Both arterial and venous endothelial cells possess typical receptors for insulin on the basis of specifity of binding, curvilinear Scatchard plots, aflqnity profiles, pH dependence, and dissociation kinetics. Arterial cells, however, bind at least 2.5 times more insulin than do venous cells, whether studied at 4, 24, or 72 h after in vitro plating. 2 Scanning electron microscopy studies indicate that overt differences in surface morphology and surface area do not account for the differences in insulin binding between the two cell types. The cause of the marked difference in binding is not readily apparent. The observation that endothelial cells from several vascular locations can also manifest differences in a diverse group of functional properties, however, suggests that endothelial cells, from different vascular sites possess highly specialized and distinct surface characteristics and functional properties. 2 The characterization of insulin uptake by microvascular tissues is usually hampered by a lack of suitable tissue preparations and by the complexity involved in analysing binding studies in a tissue with more than one type of cell.
Nevertheless binding of insulin to microvessels isolated from adult bovine and neonatal porcine cerebral cortex is shown to be affected by pH, temperature, and other physiological conditions in much the same manner as the binding of this hormone to liver, adipose tissue, and muscle. Furthermore, chemical covalent cross-linking of insulin to its receptor in cerebral microvessels and the subsequent isolation of the hormone-receptor complex, indicate that the hormone is found associated with a polypeptide with a molecular weight which is indistinguishable from the -subunit of the liver insulin receptor. 21 The presence of endothelial surface receptors for insulin may be of physiological significance in several general ways: insulin could directly modulate endothelial function; surface receptors for insulin could play a role in the transfer of insulin between the circulation and the subendothelial tissues; insulin bound to the endothelial surface could be released into the circulation, the endothelium serving as an additional storage compartment that would influence blood levels of insulin, e Therefore, though the mechanism of glucose entry into vascular endothelial cells is thought to be independent of the action of insulin, 22 the hormone is likely to influence the activity and functions of the vascular endothelium. 23'24 Vascular responsiveness in insulin-deficient states: Diabetes mellitus, as a clinical or experimental entity, at:fects the microcirculation, as well as large arteries and occasionally veins. The ability of insulin to restore altered vascular responsiveness in experimental diabetes mellitus is an indication that the alterations are a consequence of the diabetic state. This is observed not only in preparations of large vessels from diabetic animals receiving insulin, 2s'26 but also in microvascular beds of insulin-treated diabetic animals.
Data obtained with tyramine, an indirectly acting sympathomimetic amine, provide evidence that no abnormal release of noradrenaline from storage sites occurs at the early stages of diabetes mellitus. Furthermore, they indicate that uptake mechanisms into the adrenergic nerve terminal are unlikely to be altered since displacement of noradrenaline requires tyramine uptake. Interaction of noradrenaline with specific receptors to produce sympathomimetic effects is also unaffected as shown by equivalent arteriolar constrictor responses induced by the catecolamine in normal and diabetic animals. In addition, equal concentrations of phentolamine are capable of blocking such responses in both groups of animals. Dilatation of arterioles and venules is observed in normal and diabetic animals with equivalent doses of acetylcholine, thereby suggesting that effector mechanisms for vasodilation are also preserved at the early stages of diabetes mellitus. Accordingly, release of endotheliumderived relaxing factor (EDRF) by acetycholine does not appear to be altered, nor does the relaxing mechanisms in vascular smooth muscle, as shown by the responses to papaverine, an endotheliumindependent vasodilator. 27 Therefore, if functional alterations of microvessels are observed at the early stages of the diabetic state, they are not necessarily related to any noticeable dysfunction of the autonomic nervous system. However, 2 months after the induction of streptozocin diabetes in rats, a nonspecific decrease in contractile responsiveness of the mesenteric vascular bed to nerve stimulation, and to intraluminal noradrenaline can be shown.
The diabetes-evoked depression of responses at this stage suggests the occurrence of functional changes in vascular reactivity to which an abnormality in sympathetic neuronal activity contributes. 28 Norepinephrine evokes a constrictor response of mesenteric microvessels in situ, the latency of which is analogous in normal and alloxan diabetic rats. Histamine and bradykinin are capable of antagonizing this response in normal but not in diabetic animals, unless the minimum effective doses are increased about 20-fold. In contrast, acetylcholine and papaverine are equally effective in both groups of animals. The altered response to histamine and bradykinin are not directly associated with hyperglycaemia since fasting renders the diabetic animals normoglycaemic and yet does not restore the reactivity of microvessels to these agents. Previous administration of insulin to diabetic animals corrects the impaired responses.
The functional changes observed in the responses to histamine and bradykinin, under these conditions, are unlikely to be associated with a defective response of the smooth muscle. First, because in extravascular smooth muscles obtained from either normal or diabetic animals equivalent responses to histamine and bradykinin are observed. Secondly, because concentration-effect curves, constructed from the response of isolated aortae to noradrenaline, are similar in normal and diabetic animals, provided the endothelium is removed. Differences are only observed in preparations in which the endothelium is left intact. Since histamine and bradykinin are potent permeability-increasing agents in most species, whereas acetylcholine and papaverine are devoid of such an action, histamine and bradykinin might antagonize the constrictor response of microvessels to noradrenaline through an action on lining endothelial cells resulting in interendothelial gaps and increased vascular permeability with transitory changes in the composition of extravascular t:luid. 29'30 Induced changes in local steady-state conditions, if the only from the osmotic point of view, can affect the reactivity of small vessels. 1 Accordingly, endothelial ceils may play a j. Garda-Leme and S. P. Farsky role when vasoactive substances, endowed with permeability-increasing properties, act on microvessels. That insulin is involved in such events is suggested by the restorative effect it exerts in diabetic animals and by the finding that a similar condition of impaired responses to histamine and bradykinin is produced in normal animals by the injection of 2-deoxyglucose, the acute effects of which are the result of intracellular glucopenia secondary to inhibition of glucose utilization. 32 The findings are consistent with the view that functional alterations of microvessels, characteristic of the diabetic state, involve the vascular endothelium and that, at least in part, they may be linked to continuing insulin deficiency.
Insulin and the vascular response to injury: The observations described above point to the in'volvement of insulin in events that are important for the development of inflammatory responses. It is, therefore, expected that many aspects of inflammation will be affected by the diabetic state. Edlung and co-workers in 1952 reported in a brief note that alloxan inhibits the dextran oedema in the rat. The et:fect, however, was attributed to alloxan itself and not to the diabetic state. 33 Subsequent studies revealed that regardless of what the immediate eects of alloxan may be, decreased inflammatory reactions in alloxan diabetic animals are obviously related to the diabetic state, since they can be obtained for several days following the injection of alloxan and can be reversed by insulin treatment. Alloxan diabetic rats fail to present the characteristic cutaneous oedema ('anaphylactoid' response) which follows the intravenous injection of dextran or egg white. The histamine content of the skin in these animals, however, is indistinguishable from that of controls. Pretreatment with insulin restores the ability of the animals to respond in a normal fashion to the injection of dextran or egg white. Furthermore, compound 48/80, a potent histamine releasing agent, is capable of evoking a shock-like picture of comparable intensity in alloxan treated and normal animals, thus indicating that the reduced inflammatory response rather reflects a reduced capacity of the animals to respond to inflammatory agents. 34 In addition, insulin considerably sensitizes rats to the dextran anaphylactoid reaction. Ineffective doses of dextran can initiate the reaction if given in association with crystalline insulin. s Light microscope investigation and electron microscopic studies revealed that microvessels of alloxan diabetic rats, challenged with histamine or serotonin, exhibit less labelling by intravenously injected colloidal carbon particles than do vessels of normal animals. Insulin corrects this condition and also potentiates leakage of carbon evoked by permeability-increasing factors in normal animals. The carbon particles label venules almost exclusively and abandon the lumen of leaking vessels through interendothelial junctions. 6 The vascular endothelium is highly permeable to water and water soluble molecules, but normally restricts the passage of macromolecules. Escape of plasma proteins in inflammation (or of colloidal carbon particles as in the experiments referred to above) apparently depend on a partial 'disconnection' of endothelial cells along the intercellular junctions. Several mechanisms were proposed in the attempt to explain how the molecules of inflammatory mediators succeed in creating intercellular gaps in microvessels through which leakage of macromolecules occurs. Contraction of endothelial cells is still a tenable explanation.  Insulin through an action on endothelial cells, might influence the response of these cells to inflammatory mediators, thus exerting a facilitatory effect on microvascular permeability. This view is supported by the following observations. The oedema which develops in an area after local injection of chemical irritants or application of physical stimuli is reduced when the animals are rendered diabetic by the administration of alloxan or streptozocin or by subtotal pancreatectomy. This inhibition is abolished by previous injection of insulin and is not associated with increased blood sugar concentrations. If diabetic animals are fasted for a few hours with water ad libitum, normal glucose levels are regained, but the animals still present reduced responses to the irritants. Accordingly, insulin deficiency, more than hyperglycaemia, is likely to be the cause of oedema inhibition in diabetic animals. 42'4 Diabetes mellitus does not affect the endogenous release of vasoactive agents, as referred to above. Furthermore, equivalent in vitro release of histamine from mast cells or of kinins from plasma by specific releasing agents is observed irrespective of whether the cells or plasma are obtained from normal or diabetic animals. In addition, diabetic animals exhibit decreased responses, in comparison with controls, to permeability factors such as histamine, bradykinin, or serotonin injected into the skin. Therefore, even if an irritant is capable of releasing permeability factors in diabetic animals, the ultimate effect of such factors on the microvasculature of the affected area is decreased, thus suggesting a modulatory role for insulin at this level. 42'43 Reduced responses to intracutaneous permeability factors are also observed in normal animals previously given 2-deoxyglucose to inhibit glucose utilization. Insulin does not improve the response of 2-deoxyglucose treated animals to the permeability-increasing agents. 2 The fact appears to indicate that whenever glucose utilization by the reacting structures is impaired, reduced micro-vascular permeability responses can be observed.
The activity of vascular endothelial cells may, therefore, be metabolically regulated by insulin.
Activation of sensory neurons in unmyelinated fibres results in arteriolar dilatation and plasma exudation. The response is known as neurogenic inflammation and presumably involves the release of substance P and related substances. Neurogenic inflammation is also reduced in alloxan diabetic rats. 43 Furthermore, a marked decrease in plasma leakage evoked by substance P is observed in rats rendered diabetic by the administration of streptozocin. Determination of the content of substance P in sensory nerves and spinal ganglia, however, 44 shows that it is not altered by the diabetic state.
In contrast with the observations described above, altered transendothelial changes with diffusion of serum proteins to perivascular tissues, particularly at the capillary level of the microcirculation, may accompany the progression of the diabetic state. These changes, however, are not 'inflammatory' in their nature but rather reflect the occurrence of metabolic disorders responsible for the development of the diabetic microangiopathy. Recent evidence indicates that the endothelial cell, via specific receptors, is capable of responding to nonenzymatic glycosylation products originating from the interaction of simple sugars with amino groups. Nonenzymatic glycosylation is enhanced in patients with diabetes mellitus. When glycosylationmodified albumin is added to cultured endothelial cells, an increased permeability is observed. Accordingly, extravasation of serum proteins in diabetic vascular disease, unrelated to inflammatory stimuli, at least in part, appears to depend on the response of endothelial cells to glycosylation end products. 45 Insulin receptors in inflammatory cells: In white cells, the insulin binding can be studied under conditions that are most nearly physiological, employing normal ceils or cells obtained from patients with clinical disorders. Furthermore, cultured cells can also be used. Granulocytes, mononuclear leukocytes, granulocytic leukaemic white cells, cultured human lymphocytes--and human erythrocytes and cultured fibroblasts as well--all have specific insulin binding sites. These cells bind 12sI-insulin, and the bound hormone is displaced by nanogram quantities of unlabelled peptides.  When freshly isolated circulating leukocytes (monocytes) are incubated with 1251-insulin and examined by electron microscopic autoradiography, about 20% of the labelled material is internalized after 15 min at 37C. By 2h at the same temperature, approximately one-half of the 125I-insulin is internalized. The leukocytes bind and internalize insulin in a manner that mirrors that of major target tissues, such as the hepatocytes, sl Since cell surface receptors are internalized with the ligand, this provides a mechanism to initiate cell surface receptor loss. After removal of insulin from the incubation medium, the internalized receptors are rapidly reinserted back into the cell surface, s2 Accordingly, the concentration of available insulin receptors on the cell surface and, consequently, the sensitivity of the cell to insulin, are regulated by the ambient concentration of insulin. Cultured macrophages and macrophage-containing cell populations both possess specific receptors for insulin with properties indistinguishable from those of human leukocytes and other mammalian tissues, s3 In addition, platelets contain specific receptors for insulin, s4 Therefore it is entirely possible that a variety of functions of circulating leukocytes and platelets and of tissue macrophages are under the influence of insulin.
Studies on polymorphonuclear leukocytes obtained from diabetic patients greatly enhanced the knowledge of the role played by insulin in metabolic functions of inflammatory cells. The available evidence points towards glucose being freely permeable to the cell membrane. Comparative investigations show that the glycolysis of diabetic cells is decreased relative to controls. Glucose-6phosphate and fructose-6-phosphate accumulate in diabetes, suggesting a decreased activity of phospho-fructokinase. The increase in glucose-6phosphate concentration is probably responsible for the decrease in glucose utilization. The amount of glycogen and its synthesis is markedly reduced in diabetes. Insulin in vivo corrects the metabolic alterations observed in polymorphonuclear leuko- 55 cytes.
Abnormalities of leukocyte functions in diabetes mellitus: An enhanced susceptibility to infection is thought to occur in a poorly controlled diabetic state. In addition, critical evaluations of the topic show that infection is more serious and possibly more difficult to eradicate in the diabetic host. s6,s7 This might reflect an altered capacity of the diabetic patient to mount an inflammatory response, and the occurrence of immunoregulatory disorders.
Granulocyte dysfunction is a common finding in diabetes mellitus. Investigations in well-controlled diabetic subjects may fail to demonstrate any consistent defect that might predispose the patient to infection. However, in poorly controlled diabetic states, and in experimental diabetes mellitus, abnormalities in granulocyte chemotaxis, phagocytosis and microbicidal mechanisms are described.
Disturbances of the inflammatory cycle evoked by diabetes mellitus with reduction of leukocyte migration to the inflamed lesion are not uncommon. They have been clinically recognized, s8 For instance, the early granulocyte phase of the local cellular response in surgically abraded lesions is significantly delayed and diminished in poorly controlled diabetic patients, s9 Furthermore, the mean chemotactic index in diabetic patients is significantly less than in matching controls. The defect in chemotaxis is not correlated with plasma glucose levels, serum carbon dioxide, and blood urea nitrogen values.  The early local exudative cellular reaction in an inflammatory lesion is impaired in alloxan-induced diabetic rats due to a reduced migration of neutrophils to the inflamed area. Neutrophils, however, are capable of moving from reserve compartments (inflammatory leukocytosis) into blood in these animals, as much as in controls. Furthermore, an intrinsic cellular defect does not occur, because leukocytes obtained from diabetic animals are not devoid of chemotactic responsiveness when suspended in culture medium or normal serum. Suspended in the corresponding 'diabetic' serum, blockade of chemotaxis to lipopolysaccharide (LPS)-activated serum is observed. Hyperglycaemia alone, or hyperosmolality secondary to hyperglycaemia, the presence of ketone bodies, malnutrition, or a direct effect of alloxan do not explain the results. In addition, the capacity to generate chemotactic agents remains intact in serum of diabetic animals. The defect appears to be due to the presence of an inhibitory factor in plasma which is heat labile (56C), is destroyed by incubation with trypsin, and is retained after dialysis with 10 000 Mr retention dialysis tubing. Pretreatment of the animals with insulin results in the recovery of the chemotactic response either in vitro or ill vivo. 64 The inhibitory activity of chemotaxis is detected shortly after the onset of the diabetic state, being observed from the third day of alloxan administration. 6s The neutrophil migratory response in vitro to the chemotactic peptide N-formyl-methionyl-leucyl-phenyl-alanine (fMLP), and to leukotriene (LT) B 4 is not affected by the presence of diabetic rat serum. 66 Bacterial lipopolysaccharides have been demonstrated to activate both the classical and the alternative pathways of complement by a mechanism that does not require antibody to the LPS molecules. This results in the rapid appearance of chemotactic activity. Distinct subsets of chemotactic receptors mediate the response of neutrophils to fMLP, LTB 4 and complement-derived chemotaxins. Accordingly, the inhibitory activity of chemotaxis developing in diabetic rat serum appears to be restricted to some stage of the interactive process between neutrophils and complement-derived chemoattractants.
The mechanisms underlying leukocyte accumulation in a tissue depend on the interaction between the cells and the vascular endothelium. The contact between white cells and the vessel wall, particularly in venules, may take the form of either rolling of leukocytes along the endothelium or their stationary adhesion. If adhesive forces are sufficient to fully counter the shearing forces exerted by the blood stream, then stationary leukocyte adhesion will result. In the case of lesser adhesive forces, leukocyte rolling may occur, provided the shearing forces remain below a critical level. 67 During inflammation leukocyte-endothelial interactions progress to a point where adherence and emigration of the cells occur. Defective leukocyte-endothelial interactions are observed in situ in diabetic states. Under basal conditions, the number of rolling cells in postcapillary venules of transilluminated tissues of diabetic rats is markedly reduced relative to controls. The finding is not dependent on the number of circulating leukocytes because total and differential counts in the peripheral blood are equivalent in both groups. If a noxious stimulus is applied to induce a local lesion, leukocytes accumulate in the connective tissue of normal animals in a pattern characteristic of the inflammatory reaction, whereas in diabetic rats only a few cells are found in an equivalent area of the perivascular tissue. Reversal of the defective leukocyte-endothelial interaction is attained by treatment of diabetic animals with insulin. Control rats injected intravenously with lyophilized plasma constituents, obtained after dialysis of diabetic rat plasma with 12 000 M retention dialysis tubing, behave as diabetic animals in that they exhibit a reduced number of leukocytes rolling along the venular endothelium. Heating of active samples for 1 h at 56C results in the complete loss of the inhibitory effect. 68 The findings indicate that substances present in plasma of diabetic subjects are capable of influencing leukocyte-endothelium interactions and chemotaxis, thereby reducing the number of cells which accumulate in an inflamed area. What is still incompletely defined is the precise nature and origin of these substances. Several families of adhesion receptors which play a role in leukocyte-endothelial interactions and, possibly, chemotaxis have been identified. These include: the integrins; the adhesion molecules of the immunoglobulin superfamily; and cell adhesion molecules with lectin-like domains. 69 Disturbances of the inflammatory cycle evoked by diabetes mellitus--and characterized by a reduced leukocyte migration to the inflamed area--might ultimately reflect an interference of plasma factors on the activity of such receptors.
Much effort has gone into studies of the phagocytic and microbicidal function in diabetes mellitus and, with a few exceptions, results reveal impairment of these functions. The percentage of neutrophils actively engaged in phagocytosis is considerably less in diabetic rats and the average number of organisms which have undergone phagocytosis by these cells is reduced relative to cells of normal control animals. However, the differential cell counts of the peripheral blood of the alloxan diabetic rats are equivalent to those of controls. 7 In addition, alloxan diabetic rats are more susceptible to experimental pneumococcal pneumonia than nondiabetic rats. The cumulative mortality in the diabetic group is significantly higher and more than ten times as many viable pneumococci are found in the pneumonic lesions of these animals as are present in the lesions of the nondiabetic controls. Serial histological studies revealed that phagocytosis is strikingly depressed in the alveolar exudates of the diabetic animals. 71 To separate ingestion from killing in 'diabetic' granulocytes, a technique was developed which distinguishes whether the altered efficiency of microbial killing is a consequence of an impairment in intracellular mechanisms or simply a consequence of a slower rate of microbial ingestion. From these studies it was concluded that diabetic subjects have a neutrophil phagocytic defect, an impaired intracellular killing, or a combined phagocytic and intracellular killing defect. 72 Since a clear-cut improvement in the microbicidal function of granulocytes is achieved following intensive diabetes management and reduction of fasting glucose concentrations, insulin may also participate in metabolic pathways regulating intracellular microbial killing. Glucose is freely permeable to the polymorphonuclear membrane. Most glucose, however, is metabolized via the glycolytic pathway and some enzyme reactions in this pathway are closed regulated by insulin availability. This suggests that the impairment in microbicidal functions in diabetes mellitus may have a metabolic basis. 73 Diabetes mellitus may also affect the mononuclear phagocytic system. Macrophages obtained from the peritoneal cavity of alloxan diabetic rats exhibit a reduced capacity to engulf opsonized sheep erythrocytes when compared with the activity of cells obtained from matching controls. Recovery of the impaired response is attained by pretreatment of the animals with insulin. 65 Immune responses: With the use of monoclonal antibodies to characterize lymphocyte subsets it was suggested that patients with type-I (insulindependent) diabetes mellitus may present immunoregulatory disorders already at the onset of the disease due to quantitative imbalance of lymphocyte subsets. In type-I diabetes of long standing the total T-cell population is reduced. Type-II (noninsulindependent) diabetic patients, however, show no abnormalities in T-lymphocyte subsets, v4 Suppressor T-cell activity is reportedly defective when lymphocytes from insulin-dependent diabetic patients are stimulated by concanavalin A. The abnormality may be related to a paucity of suppressor T-lymphocytes as evaluated by specific monoclonal antibodies. [75][76][77][78] Cell-mediated immune responses are significantly reduced in human insulin-dependent diabetes, 79 as well as the production of IL-2. 8 '81 Comparison of antibody responses in normal and diabetic animals has not always conclusively revealed a significant difference between both groups. 2 Diabetic and control rats sensitized by the subcutaneous injection of a solution of ovalbumin exhibit equivalent IgE antibody titres, as measured by homologous passive cutaneous anaphylaxis. However, a marked reduction in the number of leukocytes present in the bronchoalveolar lavage fluid after antigen challenge is observed in diabetic animals. Pretreatment with insulin restores the response to normal levels. The finding implies that despite equivalent antibody production, sensitized diabetic animals are unable to react normally to antigen challenge, and that the impaired response is likely to depend on the continuing deficiency of insulin associated with the diabetic state. 83 Conclusions: Overall, the experimental findings commented upon in the preceding sections agree with   j. Garda-Leme and S. P. Earsky the clinical evidence that disturbances of the inflammatory cycle are not uncommon in the diabetic state, rendering the diabetic patient more susceptible to infections. Under normal conditions the activity of the pancreatic islet B cells supplies optimal concentrations of circulating insulin which is then immediately available to target structures. Insulin receptors can be identified in the reacting structures of an inflamed area. It is plausible, therefore, to assume that the activity of those structures is under control by the ambient concentration of the hormone. Altered vascular responsiveness, defective leukocyte-endothelial interactions, and inflammatory cell dysfunctions resulting in chemotactic, phagocytic, and intracellular microbial killing defects are described in insulin-deficient states. In most instances, pretreatment with insulin corrects the alterations observed, thereby indicating that a continuing deficiency of the hormone is a major element in conditioning impaired responses to noxious stimuli. Evidence for the proinflammatory role of insulin is summarized in Table 1.
Glucocorticoids and the feedback mechanism regulating the progression of inflammatory responses The glucocorticoid receptor: The molecular mechanisms by which glucocorticoid effects are produced have many common features with those of other classes of steroids. They readily penetrate the cell membrane. In the cell the hormone binds to receptor proteins and exerts its effects through receptor-mediated genomic actions. From numerous observations, generally it has been concluded that the uncomplexed glucocorticoid receptor is a heteromer composed of a single steroid and DNA-binding subunit, and two 90 kDa heat shock proteins (hsp90). Such proteins are abundant cytoplasmic materials. The DNA-binding domain is a highly conserved sequence of about 65 amino acids required for interactions with the glucocorticold response elements (GREs), that correspond to short sequences of DNA to which the receptor binds to alter transcription. The carboxy-terminal amino acids comprise the steroid binding domain that, in the absence of the ligand, appears to maintain the receptor in a repressed state. Binding of the hormone to its receptor results in the formation of a complex that undergoes activation with acquisition of an enhanced affinity for DNA. Activation is thought to involve a conformational change of the hormone-receptor complex with dissociation of the hsp90 subunits. Translocation of the activated complex into the cell nucleus forms the nuclear bound complex, primary effects of glucocorticoids being exerted at the gene level and resulting in transcriptional induction or repression. In humans, the glucocorticoid receptor is found in nearly all cell types. The molecular basis for target cell specificity cannot, therefore, be due to cell-specific expression of the receptor, but is more likely controlled at the level of the target gene itself.  Glucocorticoid administration and the sequence of inflammatory events: Glucocorticoids have long been used for the management of inflammatory diseases and deleterious immune responses. Steroids employed in glucocorticoid therapy are usually analogues of cortisol, the major active glucocorticoid in humans. Therefore, their administration under experimental conditions greatly helps the understanding of glucocorticoid action.
The microcirculation: Glucocorticoids are known to play a role in the maintenance of the functional integrity of the microcirculation. Corticosterone depresses the permeability effects of inflammatory mediators in normal animals, while the enhanced permeability response in adrenalectomized animals is decreased by the steroid to levels somewhat lower than in controls. 91  Maximum inhibitory effects are obtained a few hours after drug administration. 92 Passive cutaneous anaphylaxis evoked in mice is blocked by glucocorticoids. A latency period is required for the expression of the inhibitory elCfect. 92 In addition, vascular leakage in an inflammatory reaction induced by immune complexes is counteracted by the steroids. 93 In exteriorized tissues, local exposure of the microcirculation to a glucocorticoid blocks microvascular leakage evoked by histamine or LTB 4. The effect, observed after a latency period, apparently is exerted at the vascular endothelial cell level. 94'95 Glucocorticoid receptors are demonstrable in cultured endothelial cells and constitute a population of high-affinity sites in these cells, 96 where the steroids exert several effects including changes in morphology and protein synthesis. 9v Accumulation of leukocytes in inflammatory infiltrates: Leukocytes always adhere to the vascular endothelium before penetrating the vessel wall. Two categories of substances formed or released in an inflamed area enhance leukocyte adherence to endothelial cells: chemotactic products and cytokines. The effect of chemoattractants (complementderived factors, fMLP, LTB4) is primarily directed towards the leukocyte, since pretreatment of the endothelial cell with such substances usually fails to enhance adherence. In addition, cytokines such as tumour necrosis factor (TNF) and granulocyte-microbicidal activity are dependent on the conmacrophage colony-stimulating factor (GM-CSF) centrations used. In vitro, high concentrations of the also enhance the ability of neutrophils to adhere to steroids are required to block phagocytic functions the endothelium.  The stimulation is usually and microbicidal capability of polymorphonuclear transient. Endothelial cells invitroarestimulated by leukocytes. Moderate to severe impairment of TNF and interleukins (IL) to become more neutrophil phagocytosis occurs when cortisone is adhesive to neutrophils.  As referred to above, administered daily to human subjects for 7-9 specific adhesion glycoproteins expressed on the days. In addition, lysosomal membrane stabilization surface of leukocytes and endothelial cells play a as measured by the retarded release of acid relevant role in the adhesive phenomenon, phosphatase, and impairment in aerobic lactate Chemotactic factors acting on leukocytes and formation and glucose utilization are observed with cytokines acting on both the leukocyte and the the use of cortisone. Hydrocortisone impairs the endothelial cell induce the expression of adhesion bacterial capacity of human polymorphonuclear molecules on their surface favouring cell-to-cell leukocytes, and decreases 02 consumption, the interactions. IL-1 is produced by nearly all cell types production of H202, glucose utilization, and the in response to antigens, toxins, and inflammatory extracellular release of granular enzymes by the processes. It is capable of inducing synthesis of cells. 112'13 other cytokines (lymphokines) arid shares many of In normal mice, hydrocortisone does not alter the its properties with TNF. 4'5 Reduction of serum number of macrophages already present in the IL-1 activity is observed after treatment with peritoneal cavity, but the transit of mononuclear glucocorticoids. 16 Glucocorticoids decrease the phagocytes from the circulation into the peritoneal production of IL-1 by macrophages, lv Treatment cavity is arrested. However, during an inflamof human lung fragments in vitro with glucocortimatory response in the peritoneal cavity, hydrocoids results in a dose-dependent inhibition of IL-1 cortisone is able to suppress the increase in production. Nonglucocorticoid steroids have no number of peritoneal macrophages. This effect effect. 8 Glucocorticoids strongly inhibit TNF appears to be due to a diminished influx of production. The inhibiting effect may be demon-mononuclear phagocytes from the peritoneal blood. strated in vitro in the presence of cortisol Hydrocortisone, however, was found to have no concentrations corresponding to normal free effect on phagocytosis by mouse peritoneal cortisol levels in vivo. is Since locomotion of macrophages. 2'113 leukocytes in response to a chemical gradient of chemotactic factors is thought to remain unaffected Leukocyte kinetics: Because glucocorticoids are capable following exposure of the cells to glucocorti-of affecting the kinetics of circulating leukocytes, coids, 19 the steroids, at least in part, appear to their adminstration causes marked changes in blood inhibit leukocyte accumulation in an inflamed area leukocyte counts. Administration of glucocortiby blocking the production of mediators, particu-coids leads to a decrease in the number of larly cytokines, involved in the recruitment process, lymphocytes, monocytes, eosinophils, and baso-The steroids might also inhibit the expression of phils in the blood, whereas the number of adhesive molecules on the surface of inflammatory neutrophils increases. and endothelial cells. There are considerable differences in suscept-Both hydrocortisone and betamethasone have a ibility to glucocorticoids among various species. suppressive effect on neutrophil mobilization into Although the basis of these differences is not clearly the peritoneal cavity of mice and rats iniected with understood, animal species have been divided into living microorganisms, endotoxins, or bacterial steroid-sensitive (hamster, mouse, rat, rabbit) and extracts. Betamethasone also decreases by more steroid-resistant (monkey, guinea-pig, humans) than 50% the number of neutrophils which are groups. The differentiation is usually based on the mobilized into the pleural cavity of rats injected relative ease of producing lymphoid depletion after with kaolin. Several anti-inflammatory steroids are a given regimen of systemic glucocorticoids. In capable of reducing leukocyte emigration into the addition, there is a heterogeneity in the response of pouch fluid when tested by suspending them lymphoid cells to glucocorticoids even within the directly in the irritant solution employed to produce same species. 14 the response. Labelled leukocytes previously In steroid-sensitive species lymphocytopenia incubated with hydrocortisone and transferred to results from a significant lympholytic effect of the untreated recipients fail to accumulate at the glucocorticoids. Apparently, the lympholytic effect inflamed site. 1'111 is a receptor-mediated event. 114 In resistant species, the decrease in number of circulating lymphocytes Phagocytosis and microbicidalactivity: In many instances, involves redistribution of the cells from the the effects of glucocorticoids on phagocytosis and circulation into other body compartments. Even suprapharmacological concentrations of glucocortilymphocytes. The cell-mediated response is amcoids are not capable of inducing lymphocyte lysis plified by yIFN, that enhances the processing of in humans. 114'115 Both Tand B-cells are depleted antigen by macrophages. 14 from the circulation, but there is a relative greater Suppression of antibody production by glucoeffect on the T-cell as compared to other corticoids can be observed in some species, 14 but lymphocyte subpopulations. 1s'6 A transient in humans, where blockade of immunological monocytopenia with a return to normal counts reactions by glucocorticoids has been described, it within approximately 48 h, and eosinopenia occur appears to have no accompanying effect on antibody in sensitive and resistant species following glucosynthesis. 4'6 Patients on glucocorticoid therapy corticoidadministration. The responseisapparently exhibit a nearly normal antibody response to dependent on redistribution of the cells. 1 Steroidantigenic challenge. 7 Furthermore, there is no induced neutrophilic leukocytosis reaches a peak convincing evidence at present to suggest that 4-6 h after drug administration. The response is glucocorticoids directly affect the complement thought to occur mainly by an enhanced mobilizasystem. 8 Glucocorticoid therapy does not sigtion of cells from the marginated granulocyte nificantly suppress the immediate wheal-and-flare pool. 12 Neutrophilia induced by glucocorticoids response to an antigen skin test in atopic patients, resembles peripheral blood changes produced by thus suggesting that histamine release is not altered epinephrine and exercise. In'these circumstances, by the steroids. 18 the time to achieve peak responses is short, the Monocytes and macrophages are among the most neutrophil alkaline phosphatase content of cells is sensitive cells to the actions of glucocorticoids, low which may signify re-entry or shifting of older whereas the lymphocyte is relatively resistant. mature cells from the marginated pool, and the Inhibition of recruitment of macrophages to an number of immature cells in the peripheral blood affected area is achieved by the use of glucocortiremains low. Following administration of glucocoids. Though production of macrophage micorticoids the time to peak neutrophilia is also gration inhibitory factor (MIF) by sensitized relatively short; the alkaline phosphatase content lymphocytes is not affected, the effects of MIF on that is inversely related to neutrophil age is macrophages are blocked by glucocorticoids. 1<119 normally reduced in released cells; and immature Antigen uptake by macrophages is not influenced neutrophils in the peripheral blood are practically by the steroids. 9 However, the facilitatory effect absent. These findings corroborate the suggestion of ylFN on the processing and display of antigens that development of neutrophilia following glucois inhibited. This suggests that activation of corticoid administration is primarily due, at least in macrophages is regulated by opposing actions of humans, to mobilization of the marginated granulcytokines and glucocorticoids. In addition, glucoocyte pool. 2 corticoids decrease the production of IL-1 as referred to above. Two distinct forms of IL-1, The immune system: In many instances, undesirable termed IL-10 and IL-lfl, have been purified and consequences of an antigenic challenge can be cloned. Production of both forms of IL-1 is controlled by the use of glucocorticoids. The suppressed by the steroids in a dose-dependent control is chiefly the result of glucocorticoid effects fashion and this is accompanied by equivalent on cytokine production or cytokine actions, reductions in mRNA expression for IL-1.12 An antigen processed by the macrophage is Accordingly, glucocorticoids through inhibition of presented to resting T-cells together with a major IL-1 production are capable of influencing histocompatibility antigen. Under the influence of activation of resting T-cells. Furthermore, gluco-IL-1, secreted by the macrophage, the resting T-cell corticoids reduce the generation of IL-2 and inhibit becomes activated and synthesizes a variety of expression of IL-2 receptors on T-lymphoproteins, of which some are released and others cytes.
2q23 Therefore, T-cell proliferation is effectbecome integral components of the cell membrane, ively blocked by the steroids, since IL-2 directs Gamma-interferon (2-IFN), IL-2,-3 and-4, as well clonal expansion of specific T-lymphocytes. A as the B-cell differentiating factor (BCDF) are direct consequence of these effects is the attenuation released. Receptors for IL-2 and other cytokines are of cell-mediated immune responses. examples of membrane-associated proteins generated during antigenic challenge. The antigen can Endogenous glucocorticoids and the control of injqammatory also activate B-cells through surface antibodies, responses: Increased blood concentrations of gluco-B-cells under the influence of 2IFN, IL-1, -2 and corticoids are observed after injection of en--4, and BCDF become antibody-secreting plasma dotoxins. This increase results from an enhanced cells. Activated T-cells undergo clonal expansion release of steroids by the adrenal gland rather than under the influence of IL-2, that also activates a from decreased metabolism of the normally secreted particular group of cells to become cytotoxic adrenal steroids. 24 Hypophysectomy abolishes the adrenal cortical responses to endotoxins, thereby implying that endotoxins do not act directly upon the adrenal cortex. Since blockade of the hypothalamic release of corticotropin-releasing hormone (CRH) also abolishes the increase of plasma glucocorticoids resulting from the injection of endotoxins, the substances appear to cause an increase of plasma corticosteroids by an action on the hypothalamus-pituitary-adrenal axis at the level of the nervous system. During the early stages of experimentally induced inflammatory responses, blood levels of glucocorticoids are consistently and markedly increased.  Evidence has been obtained that the development of acute inflammatory responses is regulated by a feedback mechanism which is induced comparatively early. In carrageenan or dextran lesions of the rat's paw, a factor has been harvested in perfusates of lesions 2-3 h in age, though not less than 1 h old. The effects of the factor are demonstrable by a decrease of exudation in other animals given the perfusate intravenously prior to injection of an irritant into the paw. Inhibition of oedema formation is abolished by prior adrenalectomy or electrolysis lesions in the hypothalamic median eminence of the test animals. The perfusate also elevates the serum level of corticosterone, concomitantly depressing the content of ascorbic acid of the adrenal. On the other hand, the perfusate is effective in rats whose adrenal medulla has been ablated, suggesting that adrenal catecholamines play no significant role in the inhibitory effect. Accordingly, the factor harvested in perfusates from inflamed lesions seems to owe its effects to the release of glucocorticoids (corticosterone) via the hypothalamus-pituitary-adrenal axis. In addition, when both hind paws of the rat are injected simultaneously with an irritant, responses of the same magnitude and parallel time-course developments are observed. If the injections, however, are made at an interval of 2.5 h, the paw that is first injected exhibits a normal response, whereas the response in the other paw is reduced to about 50%. This attenuation is not observed in adrenalectomized animals. These data have been interpreted to indicate that glucocorticoids secreted in larger amounts during the early stages of an inflammatory reaction, apparently govern the development of such reaction. The increased secretion, induced via the hypothalamus-pituitary-adrenal axis under the influence of factors produced in the inflamed area would thus represent the last event of a feedback mechanism that regulates the progression of inflammatory responses. 26 More recent investigations support the occurrence of a regulatory mechanism governing the development of inflammation. First, a relationship between the temporal development of carrageenan-induced inflammatory lesions and pulsed release in plasma of increased amounts of corticosterone in rats was reported. I28 Secondly, it was shown that susceptibility of inbred Lewis female rats to develop arthritis in response to streptococcal cell wall peptidoglycan polysaccharide (SCW) was related to a defective hypothalamus-pituitary-adrenal axis responsiveness to inflammatory mediators. 129 Thirdly, it was observed that resistance of histocompatible Fisher rats to SCW arthritis is regulated by an intact hypothalamus-pituitary-adrenal axis-immune system feedback loop. 29 IL-1 provokes the secretion of adrenocorticotropic hormone (ACTH) and corticosterone in mice and rats. 13 Immunoneutralization of CRH blocks the stimulatory effect of IL-1 on glucocorticoid secretion, pointing to an activation of CRH secretion in response to the cytokine. TM Similarly, IL-6 stimulates the secretion of ACTH in conscious, freely moving rats. The effect can be abolished by a previous injection of antiserum against CRH, thereby indicating that one site of action of IL-6 is at the hypothalamic level. Accordingly, a mutual regulatory control mechanism occurs between the reacting structures in an inflamed area and the activity of the hypothalamus-pituitary-adrenal axis. 32 The regulatory mechanism apparently involves the participation of cytokines generated at the inflamed site.
Inflammatory cell functions are also markedly influenced by the level of circulating glucocorticoids. The number of leukocytes rolling along the venular endothelium of the microcirculation network is markedly increased in adrenalectomized rats relative to sham-operated and normal controls. Blood leukocyte counts, however, are equivalent in the three groups of animals. No noticeable haemodynamic changes occur to explain the increase in number of rolling leukocytes in adrenalectomized rats. The rolling movement that represents the primary leukocyte-endothelial interaction, is presumably the result of two forces, one an adhesive force towards the wall of the vessel and the other the shearing force of the flowing blood. In a given venule with approximately constant blood flow velocity the shear force also remains constant. 13 Adhesion changes affecting the primary leukocyte-endothelial interaction are likely, therefore, to prevail following the ablation of the adrenal glands, resulting in an increased number of rolling leukocytes. That the effect is under control by glucocorticoids is indicated by the fact that methyrapone-treated animals also exhibit an enhanced number of cells rolling along the venular endothelium. Methyrapone, through inhibition of the 11-/3-hydroxylation reactions in the adrenal cortex, markedly reduces the secretion of cortisol, corticosterone and aldosterone, while the secretion of 11-deoxycorticosteroids is relatively unimpaired. Accordingly, tle administration of methyrapone results in a compensatory increase in the secretion of ACTH and in enhanced secretion of 11deoxycortisol, a relatively inert steroid, and 11-deoxycorticosterone, a mineralocorticoid.
Therefore, methyrapone blocks the production of glucocorticoids but does not typically cause a deficiency of mineralocorticoids.
Under the influence of local inflammatory stimuli, cells emerge into the perivascular tissue. The number of cells, chiefly neutrophils, present in a standard area in the tissue is greater in adrenalectomized than in control animals. Endogenous glucocorticoids, therefore, appear to control leukocyte migration to an inflamed area through an interference with leukocyte-endothelium interactions. 6s Adhesion glycoproteins of the integrin family (f12 integrins), collectively termed CD11/ CD18 complex, 134'13s are only expressed in haematopoietic precursor cells and in white blood cells. The complex is composed of ofl glycoprotein heterodimers with distinct o chains (CD11a, CD 1 lb, CD 1 lc), and a common fl subunit (CD18).
Administration of a monoclonal antibody anti-CD18 to adrenalectomized rats completely blocks the previously enhanced migration of leukocytes from the microcirculation to the perivascular inflamed tissue. The finding suggests that glucocorticoids may govern the expression or the activity of the CD11/CD18 complex on the surface of leukocytes, thereby limiting the migration of these cells to inflamed areas. Glucocorticoids vs insulin: A possible interrelationship may exist between endogenous glucocorticoids and insulin in controlling the development of acute inflammation. 127 Insulin appears to facilitate the action of vasoactive substances on vascular endothelial cells, thus contributing to the formation of interendothelial gaps and to increased vascular permeability to plasma proteins. The activation of the hypothalamus-pituitary-adrenal axis by factors formed in inflamed tissue results in increased levels of circulating glucocorticoids which, in turn, are relevant for the maintenance of the integrity of the microcirculation. The immediate opening of the endothelial intercellular junctions at the venular side of the microvasculature produced by inflammatory stimuli is a transient event, even in the continuous presence of permeability-increasing agents. 137 It is plausible, therefore, to suggest that increased concentrations of circulating glucocorticoids, as observed during the early stages of the inflammatory response, might antagonize the facilitatory effect of insulin upon endothelial cells of the reacting vessels, thereby limiting the development of the response.
Lipocortins: Lipocortin has been described as a glycoprotein whose synthesis or secretion is stimulated by glucocorticoids through receptormediated actions and which specifically inhibits phospholipase A 2 in vitro and in vivo. 8 It is now known that this substance belongs to a much larger class of proteins which share the property of binding membrane phospholipids in a calciumdependent manner. Due to inhibition of the phospholipase A 2 activity, a controlling enzyme in the production of eicosanoids and PAF, lipocortins have been suggested to act as mediators of the anti-inflammatory effect of glucocorticoids. More recently, the relevance of the inhibitory action of lipocortins on phospholipase A 2 activity has been questioned. 19'4 Lipocortins are thought to act extracellularly and must, therefore, be secreted but there is no direct evidence that they are, in fact, secreted proteins and it is not known whether they act as inhibitors of intracellular phospholipase. Purified lipocortins block the stimulated release of arachidonic acid from membrane phospholipids, but not basal release. It is not clear whether any of the recombinant or otherwise structurally characterized lipocortins are actually induced by glucocorticoids particularly in a way that correlates with anti-inflammatory activity. However, the possibility has been raised that certain proteins can be secreted by pathways other than those involving the trans-Golgi network. Following the recent discovery of putative receptors on phagocytes for lipocortins (lipocortin-1), and the observation that these receptors influence the inflammatory actions of the cells by mechanisms independent of inhibition of phospholipase A 2 activity, a novel interpretation for the role of lipocortins on inflammation has been proposed. Activation of the hypothalamus-pituitary-adrenal axis by potential mediators, including interleukins and TNF, during the early stages of the inflammatory response would result in an enhanced secretion of glucocorticoids, as referred to above. This in turn, would lead to the production of lipocortins at discrete sites throughout the body. Saturation of lipocortinbinding sites in phagocytes would then reduce the migratory and proinflammatory activity of these cells with a consequent downregulation of the host response. 4 Apparently, even physiological levels of lipocortin 1 in the absence of inflammation are regulated by glucocorticoids. 142 The anti-inqammatory action of glucocorticoids: molecular mechanisms: As for most biological et:fects of steroids, the anti-inflammatory activity of glucocorticoids involves receptor occupancy and induction of gene expression. The molecular mechanisms underlying this activity were established by two lines of evidence. One derives from experiments in-which Occurrence of specific receptors in endothelial and inflammatory cells. (e) Inhibition of recruitment of mononuclear phagocytes to injured areas; (f) Interference with leukocyte kinetics; (g) Interference with the activity of the immune system. inhibition of RNA and protein synthesis were used to study the effects of dexamethasone. The existence of a time lag for manifestation of the antiinflammatory effect of the steroid was interpreted to represent time required to bring about production of specific proteins through gene expression. The effect is suppressed by actinomycin D or cycloheximide, thereby indicating that RNA and protein synthesis are essential for the expression of the anti-inflammatory activity of the glucocorticoid. 143 The other line of evidence is based on the effects of RU38486, a synthetic glucocorticoid receptor antagonist. The substance exhibits a high affinity for the corresponding receptor and displays antiglucocorticoid activity in vivo and in vitro. 144 Treatment with RU38486 blocks, in a dosedependent manner, the anti-inflammatory eect of dexamethasone under various experimental conditions, confirming that this action, like most steroid actions, occurs by receptor-mediated proces- 145,146 ses. The mechanism through which oestrogens influence acute inflammation is not clear. A protecting activity on microvessels which is independent of a systemic action is suggested, since the effect is observed by direct application of the hormones to the target tissue, ls2,s3 This has been thought to be accomplished through an influence on acid mucopolysaccharides of the perivascular ground substance or through the activation of the adenylate cyclase system that regulates leukocytic lysosomal enzyme release. 154-1s6 The mechanism by which oestradiol, oestrone and oestriol confer protection in adjuvant arthritis is also unknown. Since lymphocytes are involved in the pathogenesis of this disease, the well-known depleting effects of oestrogens on lymphoid tissues might account in part for the protective action of these hormones.
However, the finding that the hormones diminish the incidence of arthritis when administered in the latent period either starting on the day of inoculation, or 5-10 days thereafter, associated with the fact that the immunological reaction leading to arthritis is quickly initiated, suggest a direct hormonal effect on the capacity of involved tissues to react to the inflammatory stimulus in this disorder. TM Oestrogens are reported to influence antibody formation, and to suppress cell mediated immunity. In addition, the hormones inhibit lymphocyte responses to nonspecific mitogens.
Effects of oestradiol and related steroids on leukocyte function and metabolism in vitro have been investigated. Human blood leukocytes, incubated with oestradiol for I h before the addition of live Staphylococci, ingest the same number of bacteria as do control leukocytes. The normal increase in oxygen consumption after phagocytosis, however, is reduced in the presence of oestradiol.
Cells incubated with oestradiol have diminished granule lysis after phagocytosis. s7 Oestrogens, apparently influence the microbicidal actvity of the myeloperoxidase system in polymorphonuclear leukocytes, ls8 Accordingly, changes in the hormonal environment involving oestrogens may modify leukocyte responses to inflammatory stimuli. Further information, however, is required for a clear interpretation of the way these hormones may act.
Indirectly acting hormones Glucagon" Any stimulus of sucient intensity to constitute a stress and cause a nonspecific increase in sympathetic efferent activity is likely to produce a rise in plasma glucagon concentration. s9 Plasma glucagon levels are elevated in patients with bacterial infections, in burns, after endotoxin administration, or following injection of bacterial pyrogens, ls9-61 Glucagon interferes with the development of inflammatory responses. Given daily by the subcutaneous route, after a local reaction to Freund's adjuvant has already developed in the rat, glucagon produces a decrease in the response, comparable to that attained with daily effective doses of indomethacin. Administered subcutaneously 30 min beforehand, it reduces oedema formation in the 4-h interval which follows the injection of chemical irritants into the 194 Mediators of Inflammation. Vol 2.1993 rat's paw, and decreases the local exudation of Evans blue previously given intravenously. Apparently this is an indirect effect since it is not observed in adrenalectomized animals, but persists after the removal of the adrenal medulla. 1.2 It is improbable that metabolic changes may play a role in this circumstance, because increased insulin secretion can quickly compensate for and oppose the effects of glucagon.
The investigation of possible effects of glucagon on inflammatory cells has received limited attention despite the fact that hormones that influence carbohydrate metabolism are intimately involved in the host defence. During the early febrile phases of bacterial and viral illnesses, fasting plasma concentrations of glucose, insulin, glucagon, and even growth hormone tend to be increased. If carbohydrate tolerance is measured by means of an intravenous glucose load at the time, glucose disappearance rates show changes approaching those characteristic of maturity onset diabetes. Furthermore, concurrent hypercortisolism, documented in clinical studies by increased urinary excretion of free cortisol during fever, has been shown to influence glucose tolerance deleteriously not only through enhancement of gluconeogenesis, but also through augmentation of glucagon secretion. .3'1.4 This interplay of hormone actions is an integral part of the systemic response of the host against invading microorganisms and may serve a useful purpose in terms of adaptative metabolic changes, cell kinetic alterations, and energy supply during infection. The increased activity of pancreatic islets may reflect, therefore, the need for homeostatic adjustments in this condition.
Human mononuclear leukocytes specifically bind glucagon. The molecular basis for the actions of this hormone in target tissues is the activation of the adenylate cyclase system in plasma membranes following the interaction with specific macromolecular receptors. Elevation of intracellular concentrations of cAMP appears to mediate the various physiological actions of glucagon. Its ei:fects on mononuclear leukocytes, however, are much less pronounced than in liver plasma membranes. To date, there is no indication of a possible involvement of glucagon in the metabolism of peripheral leukocytes. Therefore, the significance of its binding to mononuclear leukocytes remains uncertain. *s Thyroid hormones: Conflicting results have been obtained from investigations on the influence of thyroid gland on inflammation. This might reflect the multiplicity of effects of thyroid hormones on metabolic pathways, enzyme activities, and on the response of target tissues to other hormones. In addition, the age of the animals at which the hormonal disorder is produced and the duration of the prevailing alteration might be factors in the variation of the results observed. Nevertheless, there is evidence to show that altered inflammatory responses occur in hyperthyroid states.
In an attempt to identify the mechanisms through which thyroid gland activity can influence the development of inflammatory reactions, the capacity to respond to noxious stimuli was tested in rats when thyroid defects induced by hormone administration of thyroidectomy, were fully established. Whereas animals kept in a condition of sustained excess of circulating thyroid hormones present consistently depressed inflammatory responses, hypothyroid rats respond in a normal fashion. 166'167 Decreased reactions to intracutaneously injected histamine and serotonin; inhibited swelling response to carrageendn injected into the paws; and depressed primary and secondary reactions to Freund's adjuvant only occur in the group of hormone-treated animals. In addition, enlargement of the adrenal glands; reduced content of adrenal ascorbic acid; and decreased number of circulating eosinophils, which characterize a circumstance of adrenal cortical hyperactivity, are observed only in this group. A spontaneous reversal of the previously inhibited acute inflammatory response to carrageenan occurs 3-4 days after interruption of hormone administration to the animals and this is coincidental with the return to normal of the formerly enlarged adrenal glands.
Similarly, specific inhibition of adrenal cortical steroid biosynthesis in hormone-treated animals with aminoglutethimide restores the previously depressed response to carrageenan without interference with the increased levels of serum thyroxine. 167 The findings suggest that the inhibitory eects of thyroid hormones on inflammatory responses are likely to be indirect.
There are considerable interactions between the activities of the thyroid gland, the adrenal cortex and pancreatic islet B-cells. Hypertrophy of the adrenals and reduced content of adrenal ascorbic acid are observed following thyroid hormone administration,  whereas thyroidectomy or injection of antithyroid drugs alone may result in atrophy of the adrenal gland. 171'172 The influence of thyroid hormones on the activity of the adrenal gland is either direct or indirect through changes in the release of ACTH. [173][174][175] Pancreatic B-cell dysfunction has been suggested by observations of decreased insulin secretion in isolated islets from animals made thyrotoxic, and by either decreased plasma insulin responses to an intravenous glucose challenge or lack of increased insulin levels in the face of hyperglycaemia. [176][177][178] Such interactions are a reasonable explanation for the effects of thyroid hormones on inflammation.

Concluding remarks
Almost any stage of inflammatory and immunological responses are affected by hormone actions. Proand anti-inflammatory effects are recognized and this forms the basis for the suggestion that hormones exert an integrative function on inflammation. The effect of a hormone ultimately depends on its interaction with a receptor, and a major determinant of this step is the circulating level of the hormone. Specific hormone receptors are demonstrable in the reacting structures of an inflamed area. Endocrine disorders alter host defences against trauma and infection. Inflammation, therefore, is a hormone-controlled process. Consequently, management of inflammatory diseases should take into account the general condition of the patient in addition to the clinical evaluation of local signs and symptoms.