Key points are not available for this paper at this time.
Time for primary review 32 days. Tumour necrosis factor alpha (TNFα) is a pleiotropic cytokine that has many proinflammatory actions with negative inotropic effects. It has been implicated in the pathogenesis of many non-infectious disorders, from rheumatoid disease 1, to multiple sclerosis 2. This cytokine also affects the heart 3 where it is produced by immune cells and the myocardium in some diseases. Raised serum TNFα is seen in patients with cardiomyopathy, myocardial infarction, and chronic heart failure 4, thus implicating TNFα in disease pathogenesis. Raised serum levels of this cytokine have also been identified in transplant patients following episodes of acute cellular rejection. There is a correlation between TNFα expression and rejection grade 5 suggesting that the cytokine is a candidate marker of rejection. In this review, we consider the basic biology of TNFα in relation to certain cardiac diseases. There are interrelationships between some of the conditions described, but these have been arbitrarily divided into three main groups: allograft rejection, coronary artery disease, and heart failure. Tumor necrosis factor alpha, cachectin (TNFα), was originally identified for its potent toxicity against tumour cells, hence its name 6. Today, TNFα is recognised as a pleiotropic cytokine functioning within a complex and tightly regulated cytokine network. It activates multiple transduction pathways, inducing or suppressing a wide variety of genes, including those encoding the production of cytokines, adhesion molecules, and inducible nitric oxide synthase (iNOS). TNFα has many proinflammatory actions: orchestrating the inflammatory response through activation of proinflammatory cytokine genes, such as IL-1 and IL-6, as well as its own production 7. This section will summarise our current understanding of the biology and function of TNFα with particular emphasis on the heart, and review recent work on the regulation of its production. The TNFα molecule (a 157 amino acid polypeptide; Mr 17 356) exists as both a membrane bound and a secreted molecule, both bioactive 8. TNFα acts at the cellular level via both type I (p55) and type II (p75) receptors. Both receptors have been localized in the human myocardium 9. Activation by TNFα of different transduction mechanisms may operate through separate TNFα receptors. Both receptors can be up or downregulated under different pathophysiological conditions 9. The activated macrophage is the main source of TNFα, containing both cell associated and membrane bound TNFα 10. Analysis of the kinetics of TNFα secretion, demonstrated that bioactivity appeared after 2 h of macrophage culture stimulation, reaching a maximum by 4–8 h and disappearing by 12 h. The cytokine is rapidly synthesised and released on demand, and not stored in the cytoplasm. Other cells releasing TNFα include lymphocytes, fibroblasts, neutrophils, smooth muscle and mast cells. The ability of adult mammalian myocardial cells to release TNFα after endotoxin stimulation has been shown 11. The TNFα gene is one of the earliest genes to be transcribed after T lymphocyte activation. Transcription of mRNA does not require de novo protein synthesis, as do other cytokines, such as IL-2 12. There is a high baseline transcription of TNFα mRNA by comparison to other cytokines such as TNFβ 13. Both positive and negative regulatory gene elements affect TNFα gene transcription 14, as well as other cytokines, such as GM-CSF and IL-4. These can differentially regulate transcription, through acting on upstream promoter elements 15. Production of the TNFα gene product is regulated at the post-transcriptional level. The TNFα mRNA has a very short half-life (30 min) by comparison to the mRNA half-life of TNFβ (5.5 h) 13. A conserved consensus sequence (UUAUUUAU) in the 3′ untranslated region of the mRNA destabilizes the mRNA and shortens its half-life 16,17. Various degradative pathways can degrade mRNA 18, for example, via poly A tail removal 19. These regulatory mechanisms limit the use of in situ hybridisation techniques in detecting mRNA transcripts. Messenger RNA’s sensitivity to tissue fixation, can further result in false negative results. TNFα production is also subjected to translational control. Translational derepression, occurs following lipopolysaccharide (LPS) induction and the translational rates of TNFα mRNA are accelerated 18,20. TNFα mRNA expression is not constitutive in adult myocardium, but induced 10,21. However, constitutive expression has been documented in other normal tissues, such as the spleen, liver, and kidney 21,22. Within the normal population, there is considerable variation in the amount of TNFα produced. Such variation is also seen when peripheral blood lymphocytes 23 and mononuclear cell populations 24 are stimulated with endotoxin 25,26. Polymorphisms within microsatellite regions flanking the TNF locus have been identified 27,28. Five microsatellite regions are described: TNF a, b, c, d and TNF-e. High levels of in vitro TNFα production from mononuclear cell populations are associated with the TNFa2 allele and low levels with TNFa6 allele, respectively. In addition, there are polymorphisms within the TNFα promoter region as follows: a point mutation of G to A in position -308 (termed TNF1 and TNF2, respectively) and in position -238 predisposes to higher TNF production. Studies using reporter gene assays have demonstrated an increase in the level of TNF gene transcription when the TNF2 allele is inserted 29–30. Two additional G/A mutations at position -376 and -163 are described 31. An insertional ‘C’ polymorphism is also described at position +70 32 resulting in an 8‘C’ rather than 7‘C’ repeated sequence. Administration of large amounts of TNFα to the circulation causes shock due to a decrease in peripheral vascular resistance and direct cardiac effects 33. Mann has argued that the net effect of TNFα on cardiac function will depend on the amount and duration of TNFα expression. Short term expression of TNFα within the heart may be an adaptive response to ‘stress’, whereas long term expression may be maladaptive by producing cardiac decompensation 34. Excessive TNFα levels can produce left ventricular dysfunction 35, cardiomyopathy, and the clinical manifestation of heart failure 36,37. TNFα can affect heart failure, in part, by stimulating myocyte hypertrophy, through the generation of reactive oxygen intermediates in cardiac myocytes 38, and also by inducing ventricular remodeling, through stimulating extracellular matrix protein production and increased turnover of matrix 35,39. TNFα will cause cardiomyocyte loss, through necrosis, or apoptosis, as demonstrated in in vitro models. It can induce apoptosis directly, via the TNF receptor, or indirectly, through stimulation of nitric oxide (NO) production 40. Evidence now suggests that cardiotropin-1, produced by cardiomyocytes, is able to inhibit cytokine-induced cardiomyocyte apoptosis in vitro 40. The effects of TNFα on myocardial contractility and left ventricular dysfunction, were first demonstrated in animal models 41,42. TNFα can depress myocardial function through two major pathways. The NO-dependent pathway 35,36,43, and the sphingomyelinase (sphingosine-dependent) pathway 44. Activation of the NO-dependent pathway can induce negative inotropic effects on isolated cardiac myocytes, causing immediate cell contraction, through stimulation of iNOS production 45. The resultant increase in NO will act as an important intracellular signalling molecule, that mediates the negative inotropic effects. Even though TNF has negative inotropic effects on isolated myocytes, its net effect on the whole heart is an increase in end diastolic pressure, resulting in increased ventricular volume. The induction of iNOS and its effects on cardiac function in allograft rejection, have been reviewed in detail 46. Activation of the sphingomyelinase pathway will result in breakdown of the phospholipid, sphingomyelin, to its metabolites ceramide and sphingosine 47. Both can act as second messengers in the signalling pathways. Edmunds and Woodward showed that the early increase in coronary perfusion pressure following TNFα treatment of rat perfused hearts, was due to early coronary vasoconstriction, mediated by sphingosine and thromboxane A2 44. TNFα, at low non-toxic concentrations, depresses cultured myocyte contractile performance independently of NO, through blocking α- and β-adrenoceptor-stimulated increase in contractility 35,48. Several other studies have analysed the effects of TNFα on myocardial calcium handling as one mechanism of TNFα-induced contractile dysfunction 49. TNFα-induced disruption of calcium handling may lead to dysfunctional excitation–contraction coupling causing systolic and/or diastolic dysfunction 37. The observations of Schreiner suggest that the effects of TNFα on cardiac contractile cell function, are reversible, both in animal models, and in clinical manifestations of heart failure, including allograft rejection 35. The actions of TNFα on left ventricular dysfunction were shown to be partially reversible in animal models following treatment with TNFα anatgonists 50. The effects of TNFα are listed in Table 1. Summary of the effects of TNFα on the cardiovascular system CAD: coronary artery disease; CHF: congestive heart failure. Summary of the effects of TNFα on the cardiovascular system CAD: coronary artery disease; CHF: congestive heart failure. Acute cardiac allograft rejection is an immune-mediated response, hallmarked by cellular infiltration and myocyte damage in the transplanted heart. The infiltrate consists largely of T lymphocytes and macrophages. In more severe forms, polymorphonuclear cells and eosinophils are also recruited 51. Both the CD4+ T lymphocytes and macrophages play a major role in directing the rejection response, through elaboration of initiator cytokines, such as TNFα, and the induction of effector molecules such as NO 52. Moderate to severe rejection episodes may initially be asymptomatic in some cases, but are often followed by breathlessness, pyrexia, and raised intracardiac pressures (end-diastolic). Lesser grades of rejection have no clinical signs or symptoms. In the early 1970s, endomyocardial biopsy (EMB) was introduced, and is now the ‘gold’ standard for monitoring cardiac transplants. Rejection is graded histologically from mild to severe, depending on the extent and type of cellular infiltrate and degree of myocyte damage, according to the Working Formulation of the International Heart and Lung Transplant Study Group (ISHLT) 51. The immunosuppressive regimen following transplantation is altered according to the grade of rejection and clinical symptoms. However, due to limitations of the biopsy procedure (such as sampling error), the risk of over-immunosuppression or under-immunosuppression remains a problem, particularly for the management of mild to focal moderate rejection (ISHLT grades 1 and 2, respectively) 53–55. Hence there is a pressing need to identify a marker that can predict rejection. An alternative non-invasive method of diagnosis, such as serum analysis, would be of great benefit. TNFα has many proinflammatory functions and hence has been implicated in the initiation and orchestration of the rejection response. The rejection response is initiated by activation of CD4+ T helper cells by alloantigen, either through direct stimulation by donor antigen presenting cells, or indirectly by recipient antigen presenting cells. Activated T helper cells will release initiator cytokines such as IL-1β, IL-2, and interferon γ, which in turn activate macrophages to release TNFα 56. TNFα participates in initiating the response through upregulation of MHC molecule expression required for specific T cell activation and increased cellular infiltration through endothelial cell activation and adhesion molecule expression 57. TNFα will further maintain the inflammatory response within the rejection infiltrate through upregulation of adhesion molecules, increased vascular permeability, and activation of inflammatory cells 57 (Table 1). Early suggestions that TNFα may play a role in the rejection response, came from two sources. Firstly, observations were made that cardiac 58, renal 59,60, and liver 61, transplant recipients, had raised serum levels of TNFα following rejection episodes, although no causal relationship was established. The second source of evidence came from animal models of acute cardiac allograft rejection. Anti-TNF antibody therapy prolonged cardiac allograft survival in the rat 62,63, although this was not demonstrated for allogenic skin in Rhesus monkeys 64. In addition, localization of the protein and mRNA transcripts within the rejection infiltrate of both cardiac 5,65, and renal 66,67 allografts, provided supporting evidence for the role of TNFα in rejection. In cardiac allografts, the presence of protein TNFα correlated with higher grades of rejection 5. When EMB were examined for both TNF mRNA transcript and protein product expression, using combined in situ hybridisation and immunohistochemistry 65, approximately half of the cases positive for TNF mRNA transcript were also positive for protein product. Although there was no correlation between expression and grade of rejection, there was a trend to an increased number of positive cells in samples taken early after transplantation. Analysis of serum TNFα levels have shown that in most cases with elevated levels, the amount of TNFα in samples taken within the first 30 days post-transplantation did not relate to grade of rejection 65. In serum levels were increased in patients with moderate to severe rejection for samples taken in the first post-transplantation showed that for three patients with allograft rejection, elevated levels with rejection episodes However, in other studies there was no between serum TNFα levels and grade of cellular rejection. that when samples were taken after the 30 days of to stimulating effects of or therapy on TNFα levels there was no relationship between elevated levels of TNFα and grade of rejection. the in situ expression of TNFα is in the the expression of TNFα in situ both mRNA and was with serum TNFα levels 65. TNFα levels did not with the number of TNFα positive within the biopsy TNFα is a acting potent serum will not in situ levels within a the in studies to TNFα mRNA and from a number of limitations and is rejection is sampling is the major of the EMB at samples are and three samples are taken variation in serum TNFα levels have to studies on the of TNFα production the relationship between TNFα gene polymorphisms and TNFα production in heart transplant stimulation of whole blood samples demonstrated a between the microsatellite allele with TNF production When TNFα gene polymorphisms are in relation to those of immunosuppressive cytokine that the expression of TNFα certain of and promoter gene polymorphisms are associated with acute cardiac rejection. In patients with high levels of rejection, more patients were as high with patients of low rejection levels In addition, we have identified a of heart transplant acute cellular rejection that did not with immunosuppressive therapy and to TNFα of these cases has identified a increase in the of the TNF2 by comparison to the whole of heart transplant than of acute had the TNF2 allele, by comparison to in the whole transplant Although these studies into a of many different for TNFα and its but showed a of higher expression of TNFα early after and suggest that its release may not be specific to rejection. of TNFα following therapy and have been well cytokine including TNFα of cytokines this may not be a of rejection induction of TNFα mRNA transcripts can be by immunosuppressive such as and term may cytokine of the but not release of TNFα following may this in between may to in the of cytokine The studies described in this review have not analysed the relationship between TNFα expression and grade of rejection at biopsy when are including therapy therapy 61, and The of TNF protein in cardiac in the of or clinical evidence of rejection to be in from these When such are our own studies have shown that TNFα expression can be of the biopsy cases with positive for the in situ protein expression of TNFα of will often have a higher grade of rejection on the biopsy This is for mild grades of rejection (ISHLT grades and of samples need to be analysed to these studies further that expression of TNFα is not in the the of such levels as an alternative method of TNFα Although the studies described in this review, do not increased TNFα expression is the or of rejection, the studies that stimulation of the inflammatory response alloantigen, or will result in large amounts of TNFα secreted by with a to produce higher levels of A large inflammatory may within the of The inflammatory response is initiated by the presence of T cells specific to low The elaboration of inflammatory cytokines, such as TNFα, is by into the and into macrophages. TNFα affects and by in cultured It acid and is associated with increased levels of TNFα a role in This is by of signalling by both and the of the in cultured TNFα to the of through its direct on endothelial function, stimulation of and as well as and stimulation of adhesion TNFα the risk of by with the by and suppressing the protein pathway in endothelial cells (Table 1). TNFα has been demonstrated in of but is from normal TNFα expression with the of the suggesting it may play a role in disease showed by and immunohistochemistry that TNFα was in with or It was in normal coronary and in in a muscle cells, macrophages and cells were for this cells do not TNFα whereas of smooth muscle cells this TNFα expression in smooth muscle be by was by human smooth muscle cells produce TNFα when with low TNFα mRNA is synthesised by smooth muscle cells and by macrophages The cytokine was as a of in reactive where there was and muscle cells from to both IL-1 and TNFα by GM-CSF may act in an within a for TNFα may induce expression of TNFα mRNA TNFα gene polymorphisms in relation to coronary artery disease and The were in cases and in the populations in and However, the TNFα -308 A allele was more in than in and of this allele were more than polymorphisms are to to risk but TNFα -308 polymorphism may be to The TNFα -308 allele was associated with a of myocardial in both but the of a this be TNFα is not released in patients coronary The of TNFα myocardial an acute response to myocardial necrosis rather than A of taken at the of coronary and were The were and with and after had increased expression of TNFα and with the primary There was a trend for a number of T cells and increased expression of IL-1 in An to and may a of to In this there were no in the amounts of or when primary tissue with studies on coronary suggest there is a of TNFα and Both cytokines regulate endothelial and smooth muscle of TNFα in after cardiac transplantation both the and number of coronary artery This was associated with and a of in the A of have cytokine levels following A of two of one with an of than and those of an of than coronary artery were The and did not with to number of circulation or duration of There were no in the In the there were higher levels of IL-2, TNFα, and a higher maximum cytokine response to circulation for IL-2, IL-2 receptor, and left ventricular dysfunction is associated with a higher degree of proinflammatory cytokine release coronary artery Such a response is associated with and a higher of This examined and samples taken at in h h after TNFα release This was under which have to the higher level of cytokine release with Other TNFα release circulation are Other studies have cytokines such as TNFα, IL-6, and cytokines, such as in patients The cytokines were by coronary and artery of TNFα and were higher in coronary than blood after TNFα and levels were higher in than blood within 1 h of the myocardium was a major source of TNFα and in patients of TNFα and by has effects on but the do not cytokine release from other other may have an blood and be an important source of such The release of endotoxin and seen cytokine The release TNFα and it is that the does damage TNFα and levels were higher in than blood after The were to rather than release cytokines in the early of the levels of myocardial TNFα and after and showed that it induced an increase in myocardial techniques to that myocardial TNFα was to by the myocytes levels of TNFα and its receptors have been demonstrated in congestive heart failure TNFα has effects on cardiac function, to heart failure in TNFα may thus have a Mann has argued that as TNFα is produced following of cardiac it may act as a gene in the heart. whole mammalian hearts, TNFα mRNA can be induced within 30 of a with endotoxin TNFα mRNA levels to baseline levels after removal of the TNFα stimulation of cultured cardiac myocytes in increased shock protein expression, which was by a antibody against TNFα high of TNFα, within the pathophysiological in levels of expression 34. for NO has been demonstrated in was from patients with end primary cardiac transplantation or endomyocardial was in myocytes in cardiomyopathy, in and may to the low contractility and the in these may have an inflammatory and this will to the pathogenesis and cytokine production and iNOS induction The increased levels of iNOS in with an source of TNFα, not in heart disease or normal There is no increase in in IL-2 or TNFα in mild to moderate heart failure In and cardiomyopathy, TNFα was increased in up to of patients in some TNFα may be raised in than of patients with acute myocardial with large by and has been associated with and with other major complex is associated with and raised of TNFα in and are associated within a major were to identify between and the of TNFα levels associated with congestive heart failure The for increased TNFα production in patients with remains but to the associated with the blood In vitro on perfused rat suggest that the resultant may activate a protein which in turn TNFα production The of this production by suggests a mechanism for cardiac TNFα is a potent cytokine which acts not on coronary but also on myocardial Although TNFα be as a marker of allograft rejection, it be as a of or TNFα may act in with other cytokines the rejection response, and may to the of in The of TNFα with heart failure is although further is to causal by to cytokines have not up to This may be are not at the a cytokine to have an effect when it is of an is The of with a to high TNFα is and may require its in relation to the cytokine in to of high risk
May Azzawi (Wed,) studied this question.