Part of this work was supported by the Swiss National Science Foundation (grant No .3200068286.02) and the Hans Wilsdorf Foundation.
Immunological descriptions support the role of direct contact between T‐lymphocytes and monocytes at the site of inflammatory lesions. The following will therefore focus on the role of direct contact between stimulated T‐lymphocytes and monocytes in the production of cytokines and cytokine inhibitors, as well as matrix metalloproteinases (MMPs) and tissue inhibitor of metalloproteinases (TIMP) by monocytes. The role of T‐lymphocytes and polymorphonuclear neutrophils (PMN), endothelial cells and fibroblastlike cells will also be discussed. Blocking the interaction between T‐lymphocytes and monocytes may provide a useful approach to therapeutic intervention.
Introduction
In many chronic inflammatory diseases, inflammation is characterised by the influx into the target tissue of immune cells such as dendritic cells, T‐ and B‐lymphocytes, granulocytes, and mononuclear phagocytes. This influx of inflammatory cells in the target tissue is associated with the proliferation of invading and resident cells, and frequently with destruction and remodelling of the extracellular matrix. Tissue destruction is ruled by proteases, mainly MMPs. The expression of these proteases and their inhibitors (TIMPs) is controlled by many factors including cytokines, contact with extracellular matrix components and direct cellcell contact 1, 2. In pathological conditions, the production of cytokines and MMPs by infiltrating and resident tissue cells escapes regulatory mechanisms. The activity of proinflammatory cytokines is counterbalanced by numerous mechanisms including cytokine inhibitors. It is generally acknowledged that imbalance between cytokines and their respective inhibitors is responsible for the persistence of chronic inflammatory conditions if not required for its initiation. There is now considerable evidence that cytokines, such as tumour necrosis factor (TNF)‐α and interleukin (IL)‐1β and ‐1α are involved in tissue destruction in many chronic inflammatory diseases affecting various organs.
Cytokines such as TNF‐α and IL‐1β play an essential role in inflammation and are mainly produced upon activation of monocytemacrophages. In immunoinflammatory diseases, in the absence of infectious agents, the factors triggering TNF‐α and IL‐1 production are still elusive. However, T‐cell cytokines such as IL‐4, ‐10, and ‐13 have predominantly antiinflammatory effects, and alone, interferon (IFN)‐γ, IL‐2, ‐15, or ‐17 display weak activation capacity in terms of IL‐1β and TNF‐α induction. This has prompted the author's group to hypothesise and to demonstrate that T‐cells exert a pathological effect through direct cellular contact with monocytemacrophages, inducing massive upregulation of IL‐1 and TNF‐α 3–6, such that the production of TNF‐α and IL‐1β induced in monocytes by membranes isolated from bloodderived stimulated T‐lymphocytes or stimulated T‐cells (HUT‐78) was equivalent to that induced by lipopolysaccharides 1. Besides triggering proinflammatory cytokine production, contactmediated activation of monocytes induced the production and/or shedding of cytokine inhibitors such as IL‐1receptor antagonist (IL‐1Ra), and soluble receptors of IL‐1 and TNF‐α 7–9.
T‐lymphocyte signalling of monocytemacrophages by direct cellcell contact
The relevance of T‐lymphocytes/monocyte interaction is illustrated by different chronic inflammatory diseases, affecting osteoarticular structures, lung parenchyma and the central nervous system, where T‐lymphocytes are likely to play a pivotal role 10–12. In the mid‐1980s, it was observed that the expression of membraneassociated IL‐1 (IL‐1α) in mouse monocytes was mediated by both soluble factors and direct contact with T‐cells 13. The importance of cellular contact was confirmed by experiments showing that IL‐1 was induced upon T‐cellmonocytes contact with both T‐helper cell (Th) type‐1 and ‐2 cells in the absence of lymphokine release 14. In human cells, direct contact with leptinstimulated T‐cells proved a potent stimulus of monocyte activation 3, 4; the production of IL‐1β by human monocytes also depending on the direct contact with anti‐CD3‐stimulated T‐cells 15. Following observations made by the author's group and others, several studies were carried out in human cells that have confirmed the importance of T‐cell contactmediated cytokine induction in monocytes. Studies have led to the concept that based on their effect on monocytes, T‐lymphocytes can be classified as cytokineactivated T‐lymphocytes or Tcell receptoractivated T‐lymphocytes 16.
Most T‐cell types including T‐cell clones, freshly isolated T‐lymphocytes and T‐cell lines, such as HUT‐78 cells, induce IL‐1 and TNF‐α in monocytes 3–6. Various stimuli other than phytohaemagglutinin (PHA)/phorbol myristate acetate (PMA) induce T‐lymphocytes to activate monocytes by direct cellular contact including: 1) crosslinking of CD3 by immobilised anti‐CD3 monoclonal antibody with or without crosslinking of the costimulatory molecule CD28 9, 16–18; 2) antigen recognition on antigenspecific T‐cell clones 9; and 3) cytokines 19–21. Furthermore, depending on T‐cell type and T‐cell stimulus, direct cellcell contact with stimulated T‐lymphocytes can induce different patterns of products in monocytes. This suggests that multiple ligands and counterligands are involved in the contactmediated activation of monocytes, which are differentially induced in T‐cells depending on the stimulus. Imbalance in the production of proinflammatory versus antiinflammatory cytokines has also been observed, where Th1 cell clones preferentially induced IL‐1β rather than IL‐1Ra production 9, and cytokinestimulated T‐lymphocytes induced TNF‐α production while failing to produce IL‐10 19. It was also demonstrated that upon contact with stimulated T‐cells the balance between IL‐1β and IL‐1Ra production in monocytes is ruled by Ser/Thr phosphatase(s) 7 and that contactactivated human acute monocytic leukaemia cell line (THP)‐1 cells express membraneassociated protease(s) neutralising TNF‐α activity both by degrading the latter cytokine and by cleaving its receptors at the cell surface 8. Thus, the triggering of these intra and extracellular processes by direct contact with stimulated T‐lymphocytes may regulate the proinflammatory cytokines and their inhibitors, and the balance of their production in monocytes dictates in part the outcome of the inflammatory process.
Cell surface molecules involved in contactmediated monocyte activation
A crucial issue arising from these observations is the identity of the molecules on the T‐cell surface that are involved in contactmediated signalling of monocyte activation as well as their counterligands. It has been postulated that T‐cell membraneassociated TNF‐α was involved in monocyte activation. However, fixed, stimulated Th2 cells from a T‐cell line that did not express membraneassociated TNF induced both TNF and IL‐1 production in monocytes 22 demonstrating that TNF‐α may play a part but not a primary one. The author's group has shown that neither soluble TNF‐α receptors nor IL‐1Ra block T‐cell signalling of the monocytic cell line THP‐1. Moreover, neutralising antibodies to TNF‐α, IL‐1, IL‐2, IFN‐γ and granulocyte/macrophage colonystimulating factor all failed to affect monocyte activation by membranes from stimulated T‐cells 3, 4, 11.
In addition to membraneassociated cytokines, other surface molecules have been assessed as to their ability to activate monocytes upon contact with stimulated T‐cells, e.g. leukocyte function antigen (LFA)‐1/intercellular adhesion molecule (ICAM)‐1, CD2/LFA3, CD40/CD40L and lymphocyte activationantigen‐3 (LAG‐3). Thus, CD40/CD40L interaction was shown to be involved in the contact activation of both human and mouse monocytes by T‐lymphocytes stimulated for 6 h 23. However, when stimulated, T‐lymphocytes isolated from both CD40Lknockout and wildtype mice triggered monocyte activation, although to a lower extent 24. In the author's system, where stimulated human T‐lymphocytes proved to have a high capacity of inducing cytokines in monocytes, inhibition of contactinduced cytokine production was never observed, whether by blocking antibodies to CD40L or soluble CD40. Furthermore, HUT‐78 cells, which efficiently induce cytokine production in monocytes, do not express CD40L messenger ribonucleic acid in resting or activated conditions 25. Finally, THP‐1 cells that respond to contactmediated activation by membranes of stimulated T‐cells do not express CD40.
Another study showed that in cocultures of living cells stimulated with IL‐15, Th1 unlike Th2 clones induce IL‐1β production in monocytes 21. In the latter system, blockade of the CD40‐CD40L interaction results in inhibition of IL‐1β production while IL‐1Ra induction is unaffected. This differential effect indicates the selective relevance of CD40‐CD40L engagement upon monocyte activation by Th1 clones. However, the levels of CD40L expression did not differ in Th1 and Th2 cell clones, implying that additional, unidentified molecule(s) preferentially expressed by Th1 cells are involved due to their capacity to induce IL‐1β. Therefore, CD40/CD40L may be a cofactor in the contactmediated activation of monocytes by stimulated T‐lymphocytes. From the knowledge gained to date, CD40‐CD40L may be important in terms of quality of stimulation, but it is not crucial. In the author's system LAG‐3 did not induce the production of IL‐1β and TNF‐α. Others found soluble CD23 induced cytokine production on monocytes 26, 27. LFA‐1 (CD11a/CD18) and CD69 contribute to the activation of human monocytic cells by stimulated T‐cells 4, 28. This was substantiated by a study showing that IL‐15 induced synovial T‐cells from rheumatoid arthritis patients to activate the production of TNF‐α by monocytes. This effect was inhibited by antibodies to CD69, LFA‐1 and ICAM‐1 20.
Together these studies suggest that some known surface molecules are involved in T‐cell signalling of monocytes. However, inhibitors (e.g. antibodies) of these molecules fail to abolish monocyte activation altogether, suggesting that the factor(s) required for T‐cell signalling of human monocytes by direct contact remain(s) to be identified. Furthermore, hierarchy and the sequence of events during this crosstalk need to be established.
Identification of a plasma inhibitor of T‐cell contactmediated activation of monocytemacrophages
The inhibition of T‐cell signalling of monocytes may be important in that it would maintain a low level of monocyte activation within the bloodstream. After several steps of purification, apolipoprotein (apo) A‐I was identified as being an inhibitor of contactmediated activation of monocytes. Apo A‐I appears to bind preferentially to stimulated T‐lymphocytes in this context 29. These results were further confirmed by using recombinant apo A‐I 30. Apo A‐I is a “negative acutephase protein” and the principal protein of highdensity lipoproteins (HDL). Variations of apo A‐I concentration were observed in several inflammatory diseases including rheumatoid arthritis 31, and antibodies to apo A‐I have been described in serum of patients with systemic lupus erythematosus 32. Low levels of apo A‐I in patients with chronic inflammatory diseases may be a link between infection and chronic inflammation 33.
The identification of HDL‐associated apo A‐I ligand(s) on stimulated T‐cells may lead to the elucidation of the mechanisms and molecules involved in T‐cell signalling of monocytes. HDL‐associated apo A‐I has been shown to bind specifically to a number of cellsurface molecules 34 including HDL‐binding protein, scavenger receptor B1, HB2, cubillin, adenosine triphosphatebinding cassette A1 transporter, and a 95 kD protein at the surface of human foetal hepatocytes. All these proteins have a high molecular weight (≥80 kD) and to date have not been identified on T‐cells. Thus, there is no easy answer as to the identity of the apo A‐I ligand at the surface of stimulated T‐cells, although a specific HDL‐binding site on human T‐lymphocytes has been reported but not identified 35, 36. Furthermore, none of the identified apo A‐I ligands display characteristics compatible with the preliminary characterisation of the surfaceactivating factor on stimulated T‐cells, particularly its molecular weight of ∼40 kD.
Direct contact of T‐lymphocytes with human lung tissue macrophages and alveolar macrophages
Human alveolar macrophages (AM) and lung tissue macrophages (LTM) have a distinct localisation in the cellular environment. Their response to direct contact with activated T‐lymphocytes was studied in terms of the production of interstitial collagenase (MMP‐1), 92 kD gelatinase (MMP‐9), and TIMP‐1, one of the counterregulatory tissue inhibitors of metalloproteinases 37. Either AM obtained by bronchoalveolar lavage or LTM obtained by mincing and digestion of lung tissue were exposed for 48 h to plasma membranes of T‐lymphocytes previously activated with PMA and PHA for 24 h. In LTM exclusively, but not in AM, membranes of activated T‐cells strongly induced the production of MMP‐1, MMP‐9, and TIMP‐1 whereas membranes from unstimulated T‐cells failed to induce the release of MMPs (table 1⇓). Both populations of mononuclear phagocytes spontaneously released only small amounts of MMPs and TIMP‐1. Similar results were obtained when MMP and TIMP‐1 expression were analysed at pretranslational and biosynthetic levels, respectively. Blockade experiments with cytokine antagonists revealed the involvement of T‐cell membraneassociated IL‐1 and TNF‐α on MMP production by LTM upon contact with T‐cells. These data suggest that the ability of lung macrophages to produce MMPs after direct contact with activated T‐cells is related to the difference in phenotype of mononuclear phagocytes and cell localisation. In addition, these observations indicate that cellcell contact represents an important biological mechanism in potentiating the inflammatory response of mononuclear phagocytes in the lung. Recently, HDL‐apo A‐I was observed to interfere with the contact between T‐lymphocytes and lung macrophages (fig. 1⇓).
Direct contact between T‐lymphocytes and other target cells
Cellular contact with T‐lymphocytes induces an imbalance between MMP‐1 and TIMP‐1 production by dermal fibroblasts and monocytic cells 17 as well as prostaglandin E2 production by dermal fibroblasts 38. The T‐cell surface molecules involved are mainly IL‐1 and TNF‐α and not the ligand and counterligand involved in T‐cell/monocyte interactions. Plasma membranes or fixed, stimulated T‐cells markedly inhibited the synthesis of type‐I and ‐III collagen in fibroblasts, whether treated with transforming growth factor‐β or not. This inhibition of extracellular matrix production mediated by T‐cell contact was partially due to cumulative effects of T‐cell membraneassociated IFN‐γ, TNF‐α and IL‐1β 39. Thus, direct contact with stimulated T‐cells favours extracellular matrix catabolism by enhancing MMP production while diminishing collagen synthesis and decreasing the repair process.
The author's group has demonstrated that membranes of stimulated T‐lymphocytes induce the expression of ICAM‐1, vascular cell adhesion molecule‐1, and E‐selectin on microvascular endothelial cells from human brain (HB‐MVEC). In addition to cell adhesion molecules, contactmediated activation of HB‐MVEC induced the production of IL‐6 and IL‐8. Cell contactinduced expression of cell adhesion molecules and the production of IL‐6 and IL‐8 were inhibited by TNF‐α inhibitors demonstrating that membraneassociated TNF‐α was largely responsible for the activation of endothelial cells 40.
At an early stage of inflammation T‐lymphocytes may also be present simultaneously with PMN in the tissue. PMN produce reactive oxygen radicals, a large variety of proteolytic enzymes and various cytokines, especially IL‐8 and IL‐1Ra. PMN oxidative metabolism leads to the production of highly reactive oxygen species and contributes to the elimination of pathogenic microorganisms, but it could be harmful to host tissue at the site of inflammation, by causing membrane lipid peroxidation leading to both cell and tissue damage. In previous studies, cellular contact with activated T‐cells was shown to stimulate and prime neutrophil oxygendependent respiratory bursts 41 and this activity on PMN correlated with that on THP‐1 cells, suggesting that similar factors at the surface of stimulated T‐cells were involved in the activation of neutrophils and monocytes 5. The author's laboratory recently confirmed this and demonstrated that HDL inhibited T‐cell contactinduced respiratory bursts in neutrophils 42. This strongly suggests that analogous molecules at the surface of T‐cells are involved in the activation of both monocytes and neutrophils. However, lipidfree apo A‐I, but not HDL, proved to inhibit immunoglobulin G‐induced superoxide production by PMN 43, hinting at a specific inhibitory activity of HDL to the contactmediated activation of PMN.
Conclusion
The administration of biologics and biological inhibitors is common practice in medicine, and the use of cytokine antagonists, whether antagonists to ligand binding or to the receptor, has been more successful than the administration of cytokines with inhibitory properties. Of interest, in the human therapeutic arsenal, it is substances acting at the ligandreceptor level that are the most specific. As far as the persistence of inflammation is concerned, the contact between stimulated cells, either migrating from the blood or resident, is most likely to be a major cause of chronic destructive or fibrotic manifestations. According to a body of evidence, contact between stimulated T‐cells and monocytes or mesenchymal cells, which regulate the production of IL‐1, TNF, MMP and eicosanoidderived metabolites, can be inhibited by numerous antibodies to either ligands or counterligands, but other plasma components such as HDL apo A‐I can also inhibit cellcell interaction either in the blood stream or by diffusing in the inflamed tissue. This mechanism of control affords a new link and approach between cytokines (i.e. IL‐1 and TNF production), lipid metabolism and acutephase proteins. Contact between T‐lymphocytes and fibroblasts can also affect collagen synthesis, and the response of the fibroblasts may be strongly dependent on membraneassociated cytokines (i.e. IL‐1, TNF), which does not apply to the contact between T‐lymphocytes and monocytes.
A major goal in the future will probably be to gain more insight into the agents controlling the regulation of ligands and counterligands expressed during cellcell contact, which clearly varies from one cell to another (table 2⇓).
A new approach in therapeutic intervention in chronic inflammatory diseases
One of the hallmarks of chronic inflammatory diseases affecting many organs (e.g. lung, joints, kidney) is the persistent contact between stimulated immune cells migrating from the blood stream (i.e. T‐lymphocytes and monocytemacrophages) to the organspecific resident cells (i.e. fibroblastlike cells in the lung or the synovium). This contact leads to the production of large amounts of proinflammatory cytokines (i.e. IL‐1 and TNF), which induce the production of proteases that contribute to the destruction of the matrix (i.e. collagen, proteoglycans). Simultaneously, the new synthesis of matrix components is either decreased (lack of repair process) or, on the contrary, abnormally increased (fibrosis), entailing the loss of normal functions. Previous successful therapeutic advances have consisted of blocking the action of IL‐1 and/or TNF.
The author's new approach is based on impeding the production of both interleukin‐1 and tumour necrosis factor during the contact between T‐lymphocytes and monocytes. This will hopefully be achieved after elucidating the cellsurface molecules involved in this cellcell contact and investigating the natural soluble molecule(s) that can prevent this contact. Indeed, it was found that lipoprotein such as highdensity lipoprotein/apolipoprotein A‐I is a potent inhibitor of this cellcell contact. This is of particular interest in that highdensity lipoprotein/apolipoprotein A‐I is considered to be a beneficial component of lipid metabolism and apolipoprotein A‐I a negative acutephase protein, which unfortunately decreases during inflammation and therefore leads to the lack of inhibition during cellcell contact. This new concept bridges the gap between inflammation, lipid metabolism, acutephase proteins and other diseases such as arteriosclerosis (fig. 2⇓).
Acknowledgments
In preparing this article the author would like to thank D. Burger, C. Chizzolini, S. FerrariLacraz, N. Hyka, P. RouxLombard, R. Chicheportiche, M‐T. Kaufmann and L. Gruaz. Thanks are also due to C. Edwards and T. Kohno (Amgen, Thousand Oaks, CA, USA) for the sequencing of apolipoprotein A‐I.
- © ERS Journals Ltd