Eur Respir J 1993, 6, 120-129 REVIEW Regulation of antigen-presenting cell function(s) in lung and airway tissues P.G. Holt Regulation of antigen-presenting cell function(s) in lung and airway tissues. Correspondence: P.G. Holt P.G. Holt. Division of Cell Biology ABSTRACT: A variety of cell populations present in respiratory tract tissues The Western Australian Research can express the function-associated molecules on their surface which are re­ Institute for Child Health quired for presentation of antigen to T-lymphocyte receptors. However, the GPO Box 0184 Perth, Western Australia 6001. potential role of individual cell populations in regulation of local T -lymphocyte· dependent immune reactions in the lung and airways depends on a variety of Keywords: Antigen presentation additional factors, including their precise localisation, migration characteris­ airway epithelium tics, expression ofT-cell "eo-stimulatory" signals, responsiveness to inflamma­ class li MHC (I a) antigen expression tory (in particular cytokine) stimuli, host immune status, and the nature of the T-lymphocyte activation antigen challenge. Recent evidence (reviewed below) suggests that tJ1e induc­ tion of primary immunity (viz. "sensitisation") to inhaled antigens is normally Received: April 30, 1992 controlled by specialised populations of Dendritic Cells, which perform a sur­ Accepted: September 8, 1992 veillance role within the epithelia of the upper and lower respiratory tract; in the pre-sensitised host, a variety of other cell populations (both bone-marrow derived and mesenchymal) may participate in re-stimulation of "memory" T­ lymphocytes. Eur Respir J .. 1993, 6, 120-129. Aberrant immunoinflammatory responses to exog­ control of T-cell activation is normally operative to enous antigens are important aetiological factors in a protect the lung against unnecessary and/or exagger­ large proportion of the common non-malignant respi­ ated responses to foreign antigens. ratory diseases. The inflammatory processes which produce the characteristic pathology in these diseases, are controlled directly or indirectly by the secreted Identification of APCs in respiratory tract tissues products ofT-cells, and result ultimately from a chain in situ: immunostaining of cells expressing class ll of cellular reactions initiated by the "presentation" of major histocompatibility complex (MHC) inhaled foreign antigen to the T-cell receptor. This immune region associated (la) antigen presentation event may occur locally, giving rise to im­ mediate activation of cytokine producing memory T­ Surface expression of Ia is a necessary prerequisite cells in the airway mucosa [1], and lumen [2), as has for "presentation" of antigen to the T-cell receptor, been observed in chronic asthmatics, or systemically, and, in general terms, the concentration of la on the particularly in the regional lymph nodes draining the cell surface correlates with the potency of APCs in respiratory tract, where initial "priming" of the naive T-cell activation l4J. Thus, potential APCs may be immune system occurs following inhalation of antigens identified in tissue sections by immunoenzymatic stain­ not previously encountered [3). ing employing monoclonal antibodies against la. The immunological sequelae of these initial T -cell A substantial literature is available from studies which activation events are highly variable and can range have applied this technology to lung tissue. from T-cell anergy or tolerance, to the manifestations Cells expressing la are abundant in lung tissues of of hypersensitivity and associated tissue pathology. all speci.es examined, and are found both within epi­ The eventual outcome of individual encounters with thelial and submucosal sites throughout the conduct­ inhaled antigens depends to a significant extent on the ing airways, and within the connective tissues functional capacity of the antigen-presenting cells sur- rounding blood vessels and alveolar septal walls (APCs), which initially sequester the antigens. Recent in the lung parenchyma. The precise identity of information on the nature of APC populations in dif­ immuno-stained cells in lung tissue sections has been ferent tissue microenvironments in the respiratory tract, understandably controversial, in view of the difficul­ reviewed below, suggests that a high level of local ties associated with controlling the plane of section; REGULATION OF APC FUNCTION IN AIRWAYS 121 particularly with small biopsy samples from human co-stimulator(s) by DCs is constitutive [8, 9), whereas lung. Interpretation of the human literature is further­ macrophages and B-cells produce Httle, unless they are complicated by the paucity of data on normal control pre-stimulated with natural "adjuvants", in particular tissue. bacterial cell wall antigens conttlining lipopolysaccha­ At least seven major cell types have been positively ride [8]. Mesenchymal cells, such as epithelia and identified as expressing la in healthy or diseased lungs: fibroblasts, are believed to be generally incapable of B-lymphocytes, activated T-lymphocytes, fibroblasts, delivering an effective co-stimul.ator signal [ 10], dendritic cells (DCs), and macrophages - the latter although this issue is controversial. can be further subdivided (only in humans) into resi­ Host immune status is also an important determinant dent pulmonary alveolar macrophages (PAMs), infil­ in selecdon of appropriate APCs. It is now recognized trating monocytes present in either extraceilular fluids that naive (i.e. non-immune) and memory-T-cells have on the alveolar surface or within solid ti ssues, and ma­ different requirements for anligen-specific activation. ture tissue macrophages. With the exception of la+T­ In particular, with respect to soluble protein antigens, Iymphoblasts, all of Lllese latter cells have been monocytes/macrophages and B-cells funcrion very invoked as potential regulators of antigen specific T­ poorly in presentation of inductive signals to naive T­ cell responses to inhaled antigens, and the salient as­ cells. whereas they effecrively stimulate responses in pects of rhe supporting evidence for these claims is memory T-cells [9, 11 , 12, 13]. DCs appear unique reviewed below. in their capacity to present soluble protein antigens to non-immune T-cells [9J and hence prime/sensitize the naive host. Factors determining the contribution of potential The situation with respect to the induction of pri­ APC populations to individual immune responses mary immunity to particulate antigens is less cle~u·, as in the lung DCs are actively endocytic. but appear non-phagocytic The cell populations detailed above ctul be broadly in vitro L9). However. at the ultrastrucrural level, c lassified as "professional" APCs (B-cells, mature DCs exhibit cytoplasmic inclusions character­ macrophages and DCs) or "opportunistic" APCs (air­ istic of secondary lysozomes, prompting suggestions of wHy epithelial cell s . Type ll pneumocytes and phagocytic activity al an earlier stage of their life fibroblasts). The potential for each cell population to cycle in vivo [14]. participate in an immune induction event in the respi­ Based on the above, our current understanding of the ratory tract is determined by two sets of interacting role of the various candidate APCs is outlined below. factors: hos t· immune status, and the physicochemi­ cal properties of the inhaled antigen. APC populations in normal lung It is axiomatic that large insoluble particulate anti­ gens will gain only limited access to cell populations Dendritic cells below epithelial surfaces, and hence U1e transmission of antigenic signals to the T-cells system from such The presence of DCs in human parenchyma] lung sources requires the participation of actively phagocytic tissue was initially associated with underlying inflam­ cells, which function at the level of L11 e airway matory and fibrotic disease states [15. 16]. However, lumenal suli'ace. Moreover. the relevam APC requires they were subsequently identified as normal residents the additional capacity to migrate selectively to T-cell of the uirway epithelium and alveolar septa in all spe­ zones in regional lymph nodes, in order to induce a cies [17-20]. primary immune response in a naive host. The first evidence of a role for these cells in anti­ In contrast, low molecular weight soluble antigens, gen presentation to T-lymphocytes, was provided by may be expected to be readily translocated to intra­ a study on rat lung from our laboratory [211. These epithelial or submucosal. microenvironments via intrn­ experiments employed collagenase digestion to extract ceUular (pinocytic) or imercellular pathways [5], and mononuclear cells from lung parenchyma) tissue, prior hence gain access to all the potential APC populations to assessment of APC activity in various cell fractions listed above. This class of antigen, typified by Dcr of the digest. The phenotype of Lh e principal APC in p I allergen from the house dust mite. which leaches lhese digests was non-adherent, non-phagocytic, sur­ rapidly into epithelial lining nuids after impaction on face immunoglobulin (slg-), la• and of ultra-low den­ the airway surface [6], appears particularly effective in sity on percoiJ gradients, consistent wilh their identity inducing immunologically-mediated disease. as DCs L21]. These results were confirmed and ex­ A further factor of importance, relates to the capac­ tended to include an identical population in the airway ity of individual antigens to directly "stimulate" aspects epithelium