311Vet. Res. 37 (2006) 311–324© INRA, EDP Sciences, 2006DOI: 10.1051/vetres:2006003

Review article

The avian lung-associated immune system: a review

Sven REESEa, Grammatia DALAMANIa, Bernd KASPERSb*

a Institute for Animal Anatomy, Faculty of Veterinary Medicine, University of Munich, Germanyb Institute for Animal Physiology, Faculty of Veterinary Medicine, University of Munich,

Veterinärstr. 13, 80539 München, Germany

(Received 13 June 2005; accepted 21 November 2005)

Abstract – The lung is a major target organ for numerous viral and bacterial diseases of poultry. Tocontrol this constant threat birds have developed a highly organized lung-associated immunesystem. In this review the basic features of this system are described and their functional propertiesdiscussed. Most prominent in the avian lung is the bronchus-associated lymphoid tissue (BALT)which is located at the junctions between the primary bronchus and the caudal secondary bronchi.BALT nodules are absent in newly hatched birds, but gradually developed into the mature structuresfound from 6–8 weeks onwards. They are organized into distinct B and T cell areas, frequentlycomprise germinal centres and are covered by a characteristic follicle-associated epithelium. Theinterstitial tissue of the parabronchial walls harbours large numbers of tissue macrophages andlymphocytes which are scattered throughout tissue. A striking feature of the avian lung is the lownumber of macrophages on the respiratory surface under non-inflammatory conditions. Stimulationof the lung by live bacteria but not by a variety of bacterial products elicits a significant efflux ofactivated macrophages and, depending on the pathogen, of heterophils. In addition to the cellularcomponents humoral defence mechanisms are found on the lung surface including secretory IgA.The compartmentalisation of the immune system in the avian lung into BALT and non BALT-regions should be taken into account in studies on the host-pathogen interaction since thesestructures may have distinct functional properties during an immune response.

chicken / lung / BALT / avian respiratory phagocytes / IgA

Table of contents

1. Introduction ..................................................................................................................................... 3112. Anatomy of the avian respiratory system ....................................................................................... 3123. Bronchus-associated lymphoid tissue ............................................................................................. 3154. The interstitial immune system ....................................................................................................... 3175. The phagocyte system of the avian lung and air sacs ..................................................................... 3176. Particle uptake in the avian lung and air sacs ................................................................................. 3197. Secretory immunoglobulins – the IgA system ................................................................................ 3208. Concluding remarks ........................................................................................................................ 320

1. INTRODUCTION

Respiratory diseases play an importantrole in poultry and result in substantial eco-

nomic losses to the poultry industry [27,85]. In addition, several important poultrypathogens enter the host through the lungsurface and subsequently disseminate to

* Corresponding author: Kaspers@tiph.vetmed.uni-muenchen.de

Article published by EDP Sciences and available at http://www.edpsciences.org/vetres or http://dx.doi.org/10.1051/vetres:2006003

312 S. Reese et al.

their target organs in the body. Therefore,animal health significantly depends on thesuccessful control of pathogen invasion andpathogen replication at the bronchus-asso-ciated mucosal surface and in the lung tissue.This implicates that vaccination strategiesshould induce effector mechanisms whichhelp to prevent pathogen invasion in thefirst place. One way of achieving this goalis the vaccination through the mucosal sur-face itself, a technique frequently used bythe poultry industry. It allows for an eco-nomical, efficient and reliable way of vac-cination of large poultry flocks independentof the animal’s age. Today most of thesevaccines rely on live attenuated microorganisms but the development of noninfectious vaccines is highly desired. Therational development of novel vaccinesrequires a comprehensive knowledge of thestructure and function of the lung-associ-ated immune system in birds in order to tar-get vaccines appropriately and to provideefficient mucosal adjuvants. There is verylimited knowledge about the lung-associ-ated immune system in the chicken and inpoultry in general which might be a conse-quence of the unique and complex anatomyand function of the avian lung.

In birds, the body cavity is not divided bya diaphragm and ventilation is achievedthrough a unique way of gas transport whichrequires the action of the air sacs [23]. Asa consequence the avian lung is rigid, fixedat the thoracic walls and comprises a highlycomplex structure [62]. For a better under-standing of the lung-associated immunesystem in birds we will first introduce thebasic features of the avian lung and subse-quently discuss the structure and function ofthe bronchus-associated lymphoid tissue(BALT), the interstitial immune system andadditional immunological effector systemsof the lung.

2. ANATOMY OF THE AVIAN RESPIRATORY SYSTEM

The avian parabronchial lung differsmorphologically from the mammalian

bronchioalveolar lung in certain key aspects[18, 41, 44, 49]. While the arrangement ofthe bronchial airway system of the mamma-lian lung displays iterating, commonlydichotomous bifurcations which terminatein blind-ended air conduits [31, 48] in theavian lung, a highly intricate anastomoticsystem exists [6, 44, 90]. In complete con-trast to the tidally ventilated mammalianrespiratory system, where fresh inhaled airis mixed with residual stale air in the respi-ratory airways, the avian lung is a flow-through system [42]. Through the system ofair conduits, the primary bronchus, second-ary bronchi and tertiary bronchi (parabron-chi), the avian specific air sacs ventilate thelung continuously and unidirectionally likea pair of bellows [23, 43, 71]. Gas exchangetakes place in the parabronchial tissue man-tle, which is the fundamental functional partof the avian lung [40, 68].

The primary bronchus enters the lung atthe border of the anterior and medial thirdof the ventral lung surface and extends to thecaudal margin, where it opens into theabdominal air sac [18, 51]. Four groups ofsecondary bronchi (e.g. medioventral (Fig. 1),lateroventral (Fig. 1), mediodorsal (Figs. 1and 3a) and laterodorsal secondary bronchi)originate from the primary bronchus [36].The inspired air completely bypasses thecranial lying openings of the medioventralsecondary bronchi, a process which istermed inspiratory aerodynamic valving(IAV) [4, 5, 12, 47, 69, 86, 87]. In contrast,during the inspiratory phase as well as theexspiratory phase air flows in the medio-dorsal and lateroventral secondary bronchi.This flow pattern results in a continuousventilation of the parabronchial lung in acaudocranial direction (Fig. 2) [70]. A moredetailed discussion and illustrations of thefunctional morphology can be found inDuncker [19], McLelland [51], Maina [42],Scheid and Piiper [70].

The wall of the primary bronchus and theinitial parts of the secondary bronchi arecovered by a mucosa, which is longitudinallyfolded and ciliated [52]. The epithelium of

The avian lung-associated immune system 313

the thin walled, poorly vascularized air sacsconsists of squamous and cuboidal cells,with only a few ciliated columnar and non-ciliated columnar cells [14, 51]. Theentrances to the secondary bronchi with the

exception of the medioventral bronchial open-ings show a modified mucosa with a non-cil-iated flat zone as demonstrated by scanningelectron microscopy [52] (Fig. 3b). Thiszone was found to be the primary location

Figure 1. Medial view of the right lung of a chicken showing the main bronchi (modified after King[35]). Dots mark the localization of the BALT nodules in the primary bronchus. Legend: PB primarybronchus; 1: medioventral secondary bronchi; 2: mediodorsal secondary bronchi; 3: lateroventralsecondary bronchi; A: ostium to the clavicular air sac; B: ostium to the cranial thoracis air sac; C:ostium to the caudal thoracis air sac; D: ostium to the abdominal air sac. (Reprinted from Form andFunction in Birds, Vol. 4, A.S. King and J. McLelland, p. 226, 1989, with permission from Elsevier.)(A color version of this figure is available at www.edpsciences.org.)

Figure 2. Air flow pattern in the avian respiratory tract during inspiration (a) and expiration (b).1: Clavicular air sac; 2: cranial thoracis air sac; 3: caudal thoracic air sac; 4: abdominal air sac (mod.after König H.E., Liebich H.G., Anatomie und Propädeutik des Geflügels, Stuttgart, New York,2001, p. 253, reproduced by permission of Schattauer, Stuttgart, New York).

314 S. Reese et al.

The avian lung-associated immune system 315

of the BALT in the avian lung [21, 84](Fig. 1).

A large number of parabronchi originatefrom the internal surfaces of the secondarybronchi connecting the mediodorsal and lat-eroventral secondary bronchi with themedioventral bronchi [17, 18, 51]. The wallof the parabronchi and most parts of the sec-ondary bronchi are formed by the gasexchange tissue of the avian lung [18, 40].The inhaled air flows through the parabron-chial lumen and then centrifugally into theexchange tissue through the atria, theinfundibula, and the network of air capillar-ies (Figs. 5a and 5b). The air capillaries areclosely surrounded by a network of bloodcapillaries, which together constitute themost efficient gas exchanger unit amongair-breathing vertebrates [41, 43].

The blood-gas barrier essentially con-sists of an endothelium, a single basal lam-ina, and a very thin squamous epithelial celllayer [45]. The epithelial lining cells areextremely thin, their cytoplasm being soattenuated that the thickness is often notgreater than that of the unit membrane [1].The blood-gas barrier in the avian lung isapproximately 56–67% thinner than that ofa mammal of the same body mass while therespiratory surface area is approximately15% greater [49]. While large surface areaand thin tissue barrier enhance respiratoryefficiency, these structural features predis-

pose birds to pulmonary injury from envi-ronmental toxicants and invasion by path-ogenic micro organisms [42].

3. BRONCHUS-ASSOCIATED LYMPHOID TISSUE

Organized lymphoid structures in thebronchial mucosa of the chicken lung werefirst described by Bienenstock et al. in 1973[8]. These lymphoid aggregates were foundto be highly similar to Peyer’s patches andother gut-associated lymphoid tissues (GALT)and were designated BALT. In the originalreport, finger-like mucosal projections con-taining lymphoid nodules with germinalcenters (Fig. 3c) were described. They werefound far more frequently in the chickenthan in any other species examined. How-ever, these early observations have not beenfollowed up for nearly 20 years and onlyrecently new interest in respiratory diseaseshas led to a more detailed analysis of thelung–associated immune system in birds [84].

BALT structures of the chicken and tur-key are located at the junctions of primaryand secondary bronchi [21, 84] (Fig. 1) aswell as at the ostia to the air sacs [54] andare found regularly in birds raised underconventional conditions [20, 21]. In thechicken they are confined to the openingsof the most caudal secondary bronchi, the

Figure 3. (a) Scanning electron micrograph of the primary bronchus of a chicken (longitudinal sec-tion) showing the openings to the mediodorsal secondary bronchi (*). (b) Scanning electron micro-graph showing the non-ciliated FAE around the opening to a mediodorsal secondary bronchus.(c) Transverse section (hematoxylin and eosin staining) of this nonciliated zone showing a BALTnodule with germinal centers (*).Figure 4. BALT nodules in the primary bronchus of a 13 week old chicken. (a) Immunohistochem-ical demonstration of CD3+ T-lymphocytes; (b) immunohistochemical demonstration of BU1+

B-lymphocytes.Figure 5. (a) Longitudinal section (methylene blue staining) of the parabronchus wall of a chicken.1 – Atrium; 2 – Interatrial septum; 3 – Infundibulum; 4 – air capillaries; 5 – interparabronchial sep-tum. (b) Scanning electron micrograph of a parabronchus of a chicken cut in longitudinal section.(c) Horizontal section of the atria demonstrating macrophages in the interatrial septa by immuno-histochemistry (stained with mab Kul01). Figure 6. Lymphocytes in the parabronchial wall of a 13-week-old chicken. (a) Immunohistochem-ical demonstration of CD3+ T-lymphocytes; (b) immunohistochemical demonstration of BU1+

B-lymphocytes.

316 S. Reese et al.

mediodorsal, lateroventral and laterodorsalsecondary bronchi [21], while a wider dis-tribution pattern including the regions oflongitudinal foldings of the mucosa hasbeen observed in turkey [20]. Mature BALTstructures consist of lymphocyte aggregateswhich are covered by a distinct layer of epi-thelial cells harbouring considerable num-bers of lymphocytes [22], the lymphoepi-thelium or follicle-associated epithelium(FAE) [7, 20, 21]. The FAE is made up ofciliated and non-ciliated cells, the relativenumbers of which differ significantly at dif-ferent stages of BALT development [20,21]. Four epithelial cell types have beendescribed in the FAE [20, 21] includingcells which have irregular microvilli ontheir luminal surface and close contact tolymphocytes and therefore display somefeatures of microfold (M) cells. However,particle uptake by these cells has not beenobserved and characteristic organelles suchas apical vesicles are absent [21]. It has beensuggested that these cells may not beinvolved in antigen uptake and processing,but participate in tissue repair [20]. Moredetailed studies are required to understandthe role of the non ciliated epithelium in thedevelopment and function of BALT nod-ules in birds.

Light and transmission electron micros-copy revealed a similar cellular composi-tion of the lymphoid nodules in chickensand turkeys. In 6 to 8-week-old birds ger-minal centers are found in most nodules andoccasionally plasma cells are seen under theFAE. Macrophages, cells with dendritic cellmorphology and heterophils are distributedin considerable numbers throughout theBALT nodules [20, 21]. Due to the availa-bility of an extended set of leukocyte spe-cific monoclonal antibodies in the chickensystem, immuno-histological studies on thecomposition of BALT structures were car-ried out (Figs. 4a and 4b). This work con-firmed the presence of large numbers of Bcells, with IgM+ cells in extend of IgG+ cellsand these again in extend of IgA+ lym-phocytes [22]. In cases where germinalcenters are not present in the BALT struc-

ture, B-cells are confined to the edge of thelymphoid tissue, whereas lymphoid nod-ules are composed of aggregates of T-cellsin the centre surrounded by B cells [22, 32].

The development of the BALT dependson the age and on environmental stimula-tion [32]. In day old chickens and turkey noor very few infiltrating lymphocytes areseen in the primary bronchus [20–22].Within the first week after hatch CD45+ leu-cocytes start to migrate into the primarybronchus [33] and small lymphocytic infil-trates are found consistently at the openingsof the secondary bronchi [20]. Within thenext 3–4 weeks lymphoid nodules developat these locations and IgM-, IgG- and IgA-producing cells can be detected [20, 22].CD4+ and CD8+ T cells are found at thesame time localized in distinctive patterns.In the absence of germinal centers CD4+

cells form single clusters while CD8+ lym-phocytes are distributed throughout thenodules. Germinal centers can first bedetected in 2 to 4-week-old birds [20, 22,33]. They are composed of blast-like cellswhich stain positive for either of IgG, IgAor IgM and with the pan B cell markers HIS-C1 or CB3 [22, 33]. In addition, IgM andIgG bearing cells with long cytoplasmicprocesses resembling follicular dendriticcells are present in the germinal centers [22,33]. In these BALT structures CD4+ T cellsform large parafollicular caps around thegerminal centres while CD8+ cells are scat-tered throughout the tissue [22]. Thenumber of IgG-, IgA- or IgM-producingcells continues to increase during the fol-lowing weeks and BALT nodules can befound surrounding the entire openings tothe secondary bronchi in 18 weeks old birds[20]. In contrast to the situation in healthyhumans and similar to the rabbit and rat[83], BALT is a constitutive structure in theavian lung of specific pathogen free (SPF)and conventionally reared birds [20–22, 33].

BALT development was shown to fol-low the same time course in SPF and con-ventional chickens [22] but seems to beinfluenced by environmental stimuli. Using

The avian lung-associated immune system 317

a Bordetella avium infection model in tur-key Van Alstine and Arp showed that thenumber of BALT nodules significantlyincreased in infected birds [84]. From theselimited experimental data it can be con-cluded that SPF containment still providessufficient stimulation by inhaled antigensfor the development of the BALT. Whetheran antigenic stimulus is essential for theinduction of BALT formation in chicken isunknown. Studies by Bang and Bang [3]demonstrating lymphocytic infiltrates inthe trachea and the Harderian gland of germfree chickens indicate that BALT structuresmight also be induced without microbialstimulation. However, the presence ofinflammatory substances such as endotox-ins in the inhaled air can not be excluded andmay play a role in BALT formation. Itis therefore conceivable that antigenicstimulation is required for the induction ofBALT formation and for the developmentinto fully mature structures in the first fewweeks after hatch.

The demonstration of a highly developedand constitutively present BALT structurein the chicken and turkey stimulated someinvestigators to suggest that these mucosa-associated lymphoid structures may func-tionally compensate for the lack of lymphnodes in birds [22]. However, it remains tobe shown if these structures specificallyrespond to antigenic stimulation by the gen-eration of memory B and T lymphocytes.

4. THE INTERSTITIAL IMMUNE SYSTEM

In addition to the highly organisedBALT structures a diffuse infiltration ofleucocytes into the interstitial lung tissuehas been described. Using a pan-leukocyte(HIS-C7) marker Jeurissen et al. identifiedleucocytes located in the tissue surroundingthe parabronchi as early as day five afterhatch [33]. The majority of these cellsstained positive with an antibody specificfor monocytes, macrophages and interdig-

itating cells suggesting that these cells areantigen presenting cells. Furthermore, IgM+

but not IgG+ or IgA+ lymphocytes werefound at that time point. In older birds,Bu-1 positive B lymphocytes and CD3+

T cells are found throughout the parenchyma(Fig. 6). The majority of these T lym-phocytes were described as CD8+ α/β TCRexpressing cells but few CD4+ and γ/δT cells were also found. Macrophages seemto be abundantly present throughout theinterstitial tissue of the parabronchial wall(Fig. 5c). These cells were initially identi-fied by their expression of MHC class IImolecules and their reactivity with mono-clonal antibodies CVI-ChNL-68.1- and74.2 [34]. This observation is in good agree-ment with data obtained from standard his-tological and TEM studies as discussedbelow.

5. THE PHAGOCYTE SYSTEM OF THE AVIAN LUNG AND AIR SACS

In the mammalian lung alveolar macro-phages constitute an important defence lineat the gas-blood barrier [9, 10] that protectsthe extensive surface area which is only sep-arated by a thin, easily assailable tissue bar-rier [55]. Relatively little is known about thecorresponding cell type in the avian lung-air sac system. In the literature these cellshave been designated as “avian respiratorymacrophages or phagocytes – ARP” [26,76] or “free avian respiratory macrophages– FARM” [42, 77]. Alveolar macrophagescan easily be collected by pulmonary lavagein mammals, however this procedure yieldsonly few FARM in birds [25, 46, 55, 77, 81].From these studies it was estimated thatalveolar macrophages in mice or rats were20 times more frequent compared to FARMin the lung of the much larger chicken [46,79]. Some earlier studies even reported ona complete failure to obtain free avian res-piratory macrophages through lung lavage[37, 39, 65]. In this context it should benoted that FARM obtained through lunglavage are not only derived from the surface

318 S. Reese et al.

of the lung but also from the air sacs aselegantly shown in domestic fowls andducks [55]. Differences in the samplingtechnique as well as in the chicken strainsand immune status of the birds may accountfor the variable results obtained within [76]and between studies.

The distribution pattern and the origin ofFARM in the avian lung were addressed byseveral studies using light and transmissionelectron microscopy. Interestingly, FARMwere never found on the surface of the aircapillaries which represent the functionalequivalent to the mammalian alveoli butwere regularly present on the surfaces of theatria and the infundibulae [42, 46]. Thus,macrophages seem to be located at strategiccheck points where fresh air is distributedinto the gas exchange areas and where par-ticles can be trapped and removed. In supportof this hypothesis, clusters of macrophageswere identified in interatrial septa (Fig. 5c)and the loose connective tissue at the floorof the atria at the entrance to the infundibulaeand the gas-exchange tissue proper [37, 46].

Phagocytic cells are also present in theair sacs. Studies using fiberoptic endoscopyidentified heterophils as the most abundantcell population followed by macrophagesand small numbers of lymphocytes [14].Additionally, scattered solitary phagocyteswere identified in the connective tissue ofair sac walls [37].

Some controversy exists as to whetherthese tissue macrophages can be mobilisedunder non inflammatory conditions tomigrate onto the lung surface. WhereasKlika et al. [37] did not find evidence for themigration of FARM from the interstitial tis-sue to the surface of the conducting airwaysother investigators reported the opposite[25, 77]. Nganpiep and Maina [55] com-pared the number of macrophages obtainedby first and subsequent lung lavages andfound a clear increase in cell numbers. Theauthors concluded that these cells emanatedthrough efflux from the subepithelial com-partment or even from the vascular systemitself.

Inflammatory stimuli applied to the res-piratory system efficiently elicit macrophagesand heterophils to the respiratory surface.This was clearly shown by the high numbersof macrophages recovered from the lungand air sacs of turkeys after inoculation ofincomplete Freund’s adjuvant (IFA) intothe abdominal air sacs. These macrophagesshowed rapid adhesion to glass surfaces,phagocytosis of zymosan particles and kill-ing of Escherichia coli by in vitro assays[25]. Likewise, intratracheal instillation ofheat-killed Propionibacterium acnes [79]as well as avirulent Salmonella typhimu-rium [82] induces a rapid and high levelinflux of avian respiratory macrophages,whereas the intravenous inoculation of Pro-pionibacterium acnes resulted in a weakinduction of respiratory macrophages [79].No mobilisation of phagocytes was foundafter intravenous application of E. coli LPS,Saccharomyces cervisiae glucan, IFA orsubcutaneous application of LPS-G-IFA[79], compounds which are known to effi-ciently induce the migration of phagocytesto the mammalian respiratory system [76].Comparable results were obtained in a Pas-teurella multocida model, where intratra-cheal, but not oral administration of anapathogenic vaccine strain efficiently acti-vated the nonspecific cellular defense sys-tem of the respiratory tract [28]. Migrationof cells to the lungs and air sacs was evidentwithin 2 h and showed peak activity at 8 hafter stimulation [56]. Importantly, thisresponse was only seen after administrationof a relatively high number of stimulatinglive bacteria (> 108) and vanished within 3–4 days [78]. ARP elicited in this wayshowed increased phagocytic capacities incomparison to those obtained from controlchicken [80, 81]. Similar results were foundafter inoculation of Pasteurella multocidainto the caudal thoracic air sac in turkeys.The air sac reacted rapidly and intenselywith exudation of heterophils (94%) and asmaller amount of macrophages (6%) [24].

There is a very limited knowledge aboutadhesions mechanisms of micro organismsto avian respiratory macrophages. In one

The avian lung-associated immune system 319

investigation it was demonstrated that Sero-group A strains of Pasteurella multocidaadhere to turkey air sac macrophages. Theadhesion of the bacteria was mediated bycapsular hyaluronic acid, which was recog-nized by a specific cell surface glycoprotein[63]. Additional experiments demonstratedthat capsular hyaluronic acid bound to anisoform of the CD44 molecule expressed oncultured turkey peripheral blood monocytes[64]. This corresponds to the observationthat mammalian alveolar macrophagesexpress the hyaluronan receptor CD44 [15].In another study where chickens wereexperimentally inoculated with avian fim-briated Escherichia coli it was shown thatvariations in the expression of the F1 fim-brial phase influenced the potential of avianrespiratory macrophages for phagocytosisof these bacteria. E. coli isolates whichexpressed F1 fimbriae were rapidly killedby the macrophages, whereas in the absenceof F1 fimbriae they were resistant [61].

A more extended discussion of the non-specific cellular defence system of the avianrespiratory tract can be found in a recentreview by Toth [76].

6. PARTICLE UPTAKE IN THE AVIAN LUNG AND AIR SACS

Inhaled aerosol particles are eliminatedfrom the respiratory system by several inde-pendent mechanisms including aerody-namic filtration, mucociliary clearance andphagocytosis. A detailed study by Hayterand Besch [29] showed that the depositionof particles critically depends on their size.Large particles (3.7–7 µm in diameter) areremoved in the nasal cavities and the prox-imal trachea, while smaller particles aredeposited throughout the respiratory sys-tem. Midsize particles (1.1 µm) are trappedprimarily in the lung and cranial air sacswhile smaller particles (0.091 µm) passthrough the entire lung and are finallytrapped in the abdominal air sacs. Removalof small inert particles (non-toxic iron oxideaerosol, particle diameter 0.18 µm) from the

lung was first investigated by Stearns et al.[74] in a duck model. It was shown thatthese particles were not only phagocytosedby FARM but also removed by the epithe-lial cells of the atria and the initial portionsof the infundibula. Moreover particles werefound in interstitial macrophages. Theseobservations were subsequently confirmedin chickens, pigeons and ducks by Mainaand Cowley [46] as well as Nganpiep andMaina [55] who observed phagocytosis ofputative foreign particular material and ofred blood cells by the epithelial lining of theatrial muscles, the atrial floor and theinfundibulae. From this work the pictureemerges that the epithelium and the inter-stitial macrophages of the atrial andinfundibular area play an important role inthe removal of particles from the air on theirway to the thin, extensive and highly vul-nerable tissue of the gas exchange area [55].With surface phagocytic cells at the atrialand infundibular levels [42, 46], subepithe-lial phagocytic cells and interstitial macro-phages [32, 39, 72] and resident pulmonaryintravascular macrophages [46], birdsappear to possess a highly diversified pul-monary defence armoury [42]. Unfortu-nately, the pathway of antigen uptake andprocessing in the chicken lung has not beenstudied to date. It is therefore unclear, ifphagocytosed material is presented to thelocal immune system in the BALT andparenchyma or transported to other second-ary lymphoid organs such as the spleen.

As described above, smaller particles aretransported through the lung into the airsacs. The thin air sac walls are primarilycovered by squamous and cuboidal epithe-lial cells. In contrast to the epithelium of theparabronchial atria particle uptake by the airsac epithelium has not been described. Inaddition, a mucociliary escalator systemwas only found in the most proximal part ofthe air sacs right next to bronchial openings[14]. Thus, particle clearance is largelyaccomplished by phagocytic cells and is sig-nificantly slower than in the lung [25, 53].

320 S. Reese et al.

7. SECRETORYIMMUNOGLOBULINS – THE IgA SYSTEM

Adaptive humoral immune responseson mucosal surfaces are primarily mediatedby secretory IgA [60]. In the chicken anIgA like immunoglobulin was originallydescribed by Orlans and Rose [58] and waslater shown to be present as a polymericimmunoglobulin in serum and in severalexternal secretions [88]. Cloning of thechicken Cα gene provided structural dataallowing the definite designation of thisgene as the avian homologue for the mam-malian α H chain [50]. Even though theexistence of a secretory component inchicken IgA was proposed for many years[59], final proof has been provided onlyrecently through the cloning of the chickenpolymeric immunoglobulin receptor [89].IgA can be recovered from intestinal fluid,bile, tears and tracheal as well as lungwashings [75]. However, the highest con-centration of IgA is found in the bile fluid[66].

The presence of IgA+ B cells in thechicken respiratory tract was demonstratedby immunohistology. IgA+ B cells andplasma cells were found beneath the tra-cheal epithelium in 6-weeks-old birds, butnot in younger individuals [33]. In the lung,IgA+ cells were already present in 2-week-old birds within the epithelium of bothBALT and non-BALT areas and a dramaticincrease in cell numbers was observed withage [22]. Using haemolytic plaque assaysLawrence et al. [38] demonstrated that thesecells actively secret IgA. Since IgA can beobtained by lung lavage at readily detecta-ble concentrations [13] it seems likely thatIgA+ cells in the tissue are the prime sourceof secretory IgA in the chicken lung. How-ever, it is still unknown if the poly-Ig recep-tor is expressed on the lung epithelium anddistribution of IgA secreted in the upper res-piratory [2, 57] tract into the lung can notbe excluded so far.

8. CONCLUDING REMARKS

The lung-associated immune system haslargely been neglected in avian immunol-ogy research despite its importance in thecontrol of respiratory diseases. The mor-phological studies discussed above clearlyshow that an organized mucosa-associatedlymphoid tissue, the BALT, is constitu-tively present in normal chicken and turkeylungs. These structures show features char-acteristic for mucosal inductive sites [11]including T cell zones intervening betweenB cell follicles, the presence of antigen-presenting cells and a follicle-associatedepithelium in contact with inhaled particles[22]. However, functional studies on anti-gen sampling properties, the transfer andprocessing of antigen and the stimulation ofnaïve T and B lymphocytes has not beenpublished to date. Information on mucosaleffector sites in the avian lung is even morelimited. The presence of intraepithelial andinterstitial lymphocytes has been described[33]. More recently, the isolation and func-tional characterization of CD8+ effectorT cells expressing IFN-γ from lung tissueof influenza virus infected birds wasdescribed [73]. Unfortunately, this workdid not specify whether these cells wereobtained from the mucosal inducer oreffector sides of the lung tissue. Similarly,a recent study in an infectious bronchitisvirus model in chickens applied micro-array analysis to investigate the hostresponse in the lung but did not discrimi-nate between BALT structures and theinterstitial immune system as the source ofRNA [16]. With the chicken genomesequence available [30], future work onhost-pathogen interaction will increas-ingly apply this approach or more targetedgene expression analysis [67] to the lungtissue. Such work should take the distinctanatomical structures of the lung-associ-ated immune system into account whichmay very well show strikingly differentphenotypic properties and functional activ-ities.

The avian lung-associated immune system 321

ACKNOWLEDGEMENTS

This work was supported by the “BayerischesStaatsministerium für Umwelt, Gesundheit undVerbraucherschutz” and the BMBF programFUGATO.

REFERENCES

[1] Abdalla M.A., Maina J.N., King A.S., KingD.Z., Henry J., Morphometrics of the avianlung. 1. The domestic fowl (Gallus gallus var-iant domesticus), Respir. Physiol. 47 (1982)267–278.

[2] Avakian A.P., Ley D.H., Protective immuneresponse to Mycoplasma gallisepticum dem-onstrated in respiratory-tract washings fromM. gallisepticum-infected chickens, AvianDis. 37 (1993) 697.

[3] Bang B.G., Bang F.B., Localized lymphoidtissues and plasma cells in paraocular andparanasal organ systems in chickens, Am. J.Pathol. 53 (1968) 735–751.

[4] Banzett R.B., Butler J.P., Nations C.S., BarnasG.M., Lehr J.L., Jones J.H., Inspiratory aero-dynamic valving in goose lungs depends ongas density and velocity, Respir. Physiol. 70(1987) 287–300.

[5] Banzett R.B., Nations C.S., Wang N., FredbergJ.J., Butler J.P., Pressure profiles show fea-tures essential to aerodynamic valving ingeese, Respir. Physiol. 84 (1991) 295–309.

[6] Bellairs R., Osmond M., The atlas of chickdevelopment, Academic Press, London, 1998.

[7] Bienenstock J., Befus D., Gut- and bronchus-associated lymphoid tissue, Am. J. Anat. 170(1984) 437–445.

[8] Bienenstock J., Johnston N., Perey D.Y.,Bronchial lymphoid tissue. I. Morphologiccharacteristics, Lab. Invest. 28 (1973) 686–692.

[9] Bowden D.H., Macrophages, dust, and pul-monary diseases, Exp. Lung Res. 12 (1987)89–107.

[10] Brain J.D., Macrophages in the respiratorytract, in: Fishman A.P., Fisher A.B. (Eds.),Handbook of Physiology, Vol. 1. Circulationand Nonrespiratory Functions, AmericanPhysiological Society, Bethesda, Maryland,1985, pp. 447–471.

[11] Brandtzaeg P., Pabst R., Let’s go mucosal:communications on slippery ground, TrendsImmunol. 25 (2004) 570–577.

[12] Brown R.E., Kovacs C.E., Butler J.P., WangN., Lehr J., Banzett R.B., The avian lung: isthere an aerodynamic expiratory valve? J.Exp. Biol. 198 (1995) 2349–2357.

[13] Cihak J., Hoffmann-Fezer G., Ziegler-HeitbrockH.W.L., Stein H., Kaspers B., Chen C.H.,Cooper M.D., Lösch U., T cells expressing theVbeta1 T-cell receptor are required for IgAproduction in the chicken, Proc. Natl. Acad.Sci. USA 88 (1991) 10951–10955.

[14] Crespo R., Yamashiro S., Hunter D.B., Devel-opment of the thoracic air sacs of turkeys withage and rearing concitions, Avian Dis. 42(1998) 35–44.

[15] Culty M., O’Mara T.E., Underhill C.B.,Yeager H., Swartz R.P., Hyaluronan receptor(CD44) expression and function in humanperipheral blood monocytes and alveolarmacrophages, J. Leukoc. Biol. 56 (1994) 605–611.

[16] Dar A., Munir S., Vishwanathan S., ManujaA., Griebel P., Tikoo S., Townsend H., PotterA., Kapur V., Babiuk L.A., Transcriptionalanalysis of avian embryonic tissues followinginfection with avian infectious bronchitisvirus, Virus Res. 110 (2005) 41–55.

[17] Duncker H.R., The lung air sac system ofbirds. A contribution to the functional anat-omy of the respiratory apparatus, Ergeb. Anat.Entwicklungsgesch. 45 (1971) 7–171.

[18] Duncker H.R., Structure of the avian respira-tory tract, Respir. Physiol. 22 (1974) 1–19.

[19] Duncker H.R., Vertebrate lungs: structure,topography and mechanics. A comparativeperspective of the progressive integration ofrespiratory system, locomotor apparatus andontogenetic development, Respir. Physiol.Neurobiol. 144 (2004) 111–124.

[20] Fagerland J.A., Arp L.H., A morphologicstudy of bronchus-associated lymphoid tissuein turkeys, Am. J. Anat. 189 (1990) 24–34.

[21] Fagerland J.A., Arp L.H., Structure and devel-opment of bronchus-associated lymphoid tis-sue in conventionally reared broiler chickens,Avian Dis. 37 (1993) 10–18.

[22] Fagerland J.A., Arp L.H., Distribution andquantitation of plasma cells, T lymphocytesubsets, and B lymphocytes in bronchus-asso-ciated lymphoid tissue of chickens: age-related differences, Reg. Immunol. 5 (1993)28–36.

[23] Fedde M.R., Relationship of structure andfunction of the avian respiratory system to dis-ease susceptibility, Poult. Sci. 77 (1998)1130–1138.

322 S. Reese et al.

[24] Ficken M.D., Barnes H.J., Acute airsacculitisin turkeys inoculated with Pasteurella multo-cida, Vet. Pathol. 26 (1989) 231–237.

[25] Ficken M.D., Edwards J.F., Lay J.C., Induc-tion, collection, and partial characterization ofinduced respiratory macrophages of the tur-key, Avian Dis. 30 (1986) 766–771.

[26] Fulton R.M., Reed W.M., DeNicola D.B.,Light microscopic and ultrastructural charac-terization of cells recovered by respiratory-tract lavage of 2- and 6-week-old chickens,Avian Dis. 34 (1990) 87–98.

[27] Glisson J.R., Bacterial respiratory disease ofpoultry, Poult. Sci. 77 (1998) 1139–1142.

[28] Hassanin H.H., Toth T.E., Eldimerdash M.M.,Siegel P.B., Stimulation of avian respiratoryphagocytes by Pasteurella multocida: effectsof the route of exposure, bacterial dosage andstrain, and the age of chickens, Vet. Micro-biol. 46 (1995) 401–413.

[29] Hayter R.B., Besch E.L., Airborne particledeposition in the respiratory tract of chickens,Poult. Sci. 53 (1974) 1507–1511.

[30] Hillier L.W., et al., Consortium I.C.G.S.,Sequence and comparative analysis of thechicken genome provide unique perspectiveson vertebrate evolution, Nature 432 (2004)695–716.

[31] Horsfield K., Pulmonary airways and bloodvessels considered as confluent trees, in: CrytalR.D., West J.B., Weibel E.R., Barnes P.J.(Eds.), The lung: Scientific foundations, Lip-pincott-Raven, Philadelphia, 1997, pp. 1073–1079.

[32] Jeurissen S.H., Vervelde L., Janse E.M.,Structure and function of lymphoid tissues ofthe chicken, Poult. Sci. Rev. 5 (1994) 183–207.

[33] Jeurissen S.H., Janse E.M., Koch G., De BoerG.F., Postnatal development of mucosa-asso-ciated lymphoid tissues in chickens, Cell Tis-sue Res. 258 (1989) 119–124.

[34] Jeurissen S.H., Janse E.M., Kok G.L., De BoerG.F., Distribution and function of non-lym-phoid cells positive for monoclonal antibodyCVI-ChNL-68.2 in healthy chickens andthose infected with Marek’s disease virus,Vet. Immunol. Immunopathol. 22 (1989)123–133.

[35] King A.S., Structural and functional aspectsof the avian lungs and air sacs, in: Felts W.J.L.,Harrison R.J. (Eds.), International review ofgeneral and experimental zoology, AcademicPress, New York, 1966, pp. 171–267.

[36] King A.S., Apparatus respiratorius, in:Baumel J.J. (Ed.), Handbook of avian anat-omy: Nomina anatomica avium, Nuttal Orni-thological Club, Cambridge, Massachusetts,1993, pp. 257–300.

[37] Klika E., Scheuermann D.W., De Groodt-Lasseel M.H., Bazantova I., Switka A., Pul-monary macrophages in birds (barn owl, Tytotyto alba), domestic fowl (Gallus gallus f.domestica), quail (Coturnix coturnix), andpigeons (Columbia livia), Anat. Rec. 246(1996) 87–97.

[38] Lawrence E.C., Arnaud-Battandier F., KoskiI.R., Dooley N.J., Muchmore A.V., BlaeseR.M., Tissue distribution of immunoglobulin-secreting cells in normal and IgA-deficientchickens, J. Immunol. 123 (1979) 1767–1771.

[39] Lorz C., Lopez J., Incidence of air pollution inthe pulmonary surfactant system of the pigeon(Columba livia), Anat. Rec. 249 (1997) 206–212.

[40] Maina J.N., Scanning electron microscopestudy of the spatial organization of the air andblood conducting components of the avianlung (Gallus gallus variant domesticus), Anat.Rec. 222 (1988) 145–153.

[41] Maina J.N., Comparative respiratory mor-phology: themes and principles in the designand construction of the gas exchangers, Anat.Rec. 261 (2000) 25–44.

[42] Maina J.N., Some recent advances on thestudy and understanding of the functionaldesign of the avian lung: morphological andmorphometric perspectives, Biol. Rev. Camb.Philos. Soc. 77 (2002) 97–152.

[43] Maina J.N., Fundamental structural aspectsand features in the bioengineering of the gasexchangers: comparative perspectives, Adv.Anat. Embryol. Cell Biol. 163 (2002) III-XII,1–108.

[44] Maina J.N., A systematic study of the devel-opment of the airway (bronchial) system ofthe avian lung from days 3 to 26 of embryo-genesis: a transmission electron microscopicstudy on the domestic fowl, Gallus gallus var-iant domesticus, Tissue Cell 35 (2003) 375–391.

[45] Maina J.N., Morphogenesis of the laminated,tripartite cytoarchitectural design of theblood-gas barrier of the avian lung: a system-atic electron microscopic study on the domes-tic fowl, Gallus gallus variant domesticus,Tissue Cell 36 (2004) 129–139.

[46] Maina J.N., Cowley H.M., Ultrastructuralcharacterization of the pulmonary cellulardefenses in the lung of a bird, the rock dove,

The avian lung-associated immune system 323

Columba livia, Proc. R. Soc. Lond. B Biol.Sci. 265 (1998) 1567–1572.

[47] Maina J.N., Africa M., Inspiratory aerody-namic valving in the avian lung: functionalmorphology of the extrapulmonary primarybronchus, J. Exp. Biol. 203 (2000) 2865–2876.

[48] Maina J.N., van Gils P., Morphometric char-acterization of the airway and vascular sys-tems of the lung of the domestic pig, Susscrofa: comparison of the airway, arterial andvenous systems, Comp. Biochem. Physiol. AMol. Integr. Physiol. 130 (2001) 781–798.

[49] Maina J.N., King A.S., Settle G., An allomet-ric study of pulmonary morphometric param-eters in birds, with mammalian comparisons,Philos. Trans. R. Soc. Lond. B Biol. Sci. 326(1989) 1–57.

[50] Mansikka A., Chicken IgA H chains. Impli-cations concerning the evolution of H chaingenes, J. Immunol. 149 (1992) 855–861.

[51] McLelland J., Anatomy of the lungs and airsacs, in: King A.S., McLelland J. (Eds.), Formand function in birds, Academic Press, Lon-don, 1989, pp. 221–280.

[52] McLelland J., MacFarlane C.J., Solitary gran-ular endocrine cells and neuroepithelial bod-ies in the lungs of the Ringed Turtle Dove(Streptopelia risoria), J. Anat. 147 (1986) 83–93.

[53] Mensah G.A., Brain J.D., Deposition andclearance of inhaled aerosol in the respiratorytract of chickens, J. Appl. Physiol. 53 (1982)1423–1428.

[54] Myers R.K., Arp L.H., Pulmonary clearanceand lesions of lung and air sac in passivelyimmunized and unimmunized turkeys follow-ing exposure to aerosolized Escherichia coli,Avian Dis. 31 (1987) 622–628.

[55] Nganpiep L.N., Maina J.N., Composite cellu-lar defence stratagem in the avian respiratorysystem: functional morphology of the free(surface) macrophages and specialized pul-monary epithelia, J. Anat. 200 (2002) 499–516.

[56] Ochs D.L., Toth T.E., Pyle R.H., Siegel P.B.,Cellular defense of the avian respiratory sys-tem: effects of Pasteurella multocida on res-piratory burst activity of avian respiratorytract phagocytes, Am. J. Vet. Res. 49 (1988)2081–2084.

[57] Ohshima K., Hiramatsu K., Distribution of T-cell subsets and immunoglobulin-containingcells in nasal-associated lymphoid tissue(NALT) of chicken, Histol. Histopathol. 15(2000) 713–720.

[58] Orlans E., Rose M.E., An IgA-like immu-noglobulin in the fowl, Immunochemistry 9(1972) 833–838.

[59] Peppard J.V., Rose M.E., Hesketh P., A func-tional homologue of mammalian secretorycomponent exists in chickens, Eur. J. Immu-nol. 13 (1983) 566–570.

[60] Phalipon A., Cardona A., Kraehenbuhl J.P.,Edelmann L., Sansonetti P.J., Corthesy B.,Secretory component: A new role in secretoryIgA-Mediated exclusion in vivo, Immunity 17(2002) 107–115.

[61] Pourbakhsh S.A., Boulianne M., Martineau-Doize B., Fairbrother J.M., Virulence mecha-nisms of avian fimbriated Escherichia coli inexperimentally inoculated chickens, Vet.Microbiol. 58 (1997) 195–213.

[62] Powell F.L., Respiration, in: Whittow G.C.(Ed.), Sturkie’s Avian physiology, AcademicPress, London, 2000, pp. 233–264.

[63] Pruimboom I.M., Rimler R.B., AckermannM.R., Brogden K.A., Capsular hyaluronicacid-mediated adhesion of Pasteurella multo-cida to turkey air sac macrophages, Avian Dis.40 (1996) 887–893.

[64] Pruimboom I.M., Rimler R.B., AckermannM.R., Enhanced adhesion of Pasteurella mul-tocida to cultured turkey peripheral bloodmonocytes, Infect. Immun. 67 (1999) 1292–1296.

[65]Qureshi M.A., Marsh J.A., Dietert R.R., SungJ., Bolnet C.N., Petite J.N., Profiles of chickenmacrophage effector functions, Poult. Sci. 73(1994) 1027–1034.

[66] Rose M.E., Orlans E., Payne A.W., HeskethP., The origin of IgA in chicken bile: its rapidactive transport from blood, Eur. J. Immunol.11 (1981) 561–564.

[67] Rothwell L., Young J.R., Zoorob R., WhittakerC.A., Hesketh P., Archer A., Smith A.L., KaiserP., Cloning and characterization of chickenIL-10 and its role in the immune response toEimeria maxima, J. Immunol. 173 (2004)2675–2682.

[68] Scheid P., Mechanisms of gas exchange inbird lungs, Rev. Physiol. Biochem. Pharma-col. 86 (1979) 137–186.

[69] Scheid P., Piiper J., Aerodynamic valving inthe avian lung, Acta Anaesthesiol. Scand.Suppl. 90 (1989) 28–31.

[70] Scheid P., Piiper J., Respiratory mechanicsand air flow in birds, in: King A.S., McLellandJ. (Eds.), Form and function in birds, Aca-demic Press, London, 1989, pp. 369–392.

324 S. Reese et al.

[71] Scheid P., Slama H., Piiper J., Mechanisms ofunidirectional flow in parabronchi of avianlungs: measurements in duck lung prepara-tions, Respir. Physiol. 14 (1972) 83–95.

[72] Scheuermann D.W., Klika E., Lasseel D.G.,Bazantova I., Switka A., An electron micro-scopic study of the parabronchial epitheliumin the mature lung of four bird species, Anat.Rec. 249 (1997) 213–225.

[73] Seo S.H., Peiris M., Webster R.G., Protectivecross-reactive cellular immunity to lethal A/Goose/Guangdong/1/96-Like H5N1 influen-zavirus is correlated with the proportion ofpulmonary CD8+ T cells expressing gammainterferon, J. Virol. 76 (2002) 4886–4890.

[74] Stearns R.C., Barnas G.M., Walski M., BrainJ.D., Deposition and phagocytosis of inhaledparticles in the gas exchange region of theduck, Anas platyrhynchos, Respir. Physiol. 67(1987) 23–36.

[75] Tizard I., The avian antibody response, Sem.Avian Exotic Pet. Med. 11 (2002) 2–14.

[76] Toth T.E., Nonspecific cellular defense of theavian respiratory system: a review, Dev.Comp. Immunol. 24 (2000) 121–139.

[77] Toth T.E., Siegel P.B., Cellular defense of theavian respiratory tract: paucity of free-resid-ing macrophages in the normal chicken,Avian Dis. 30 (1986) 67–75.

[78] Toth T.E., Siegel P.B., Cellular defense of theavian respiratory system: dose-response rela-tionship and duration of response in intratra-cheal stimulation of avian respiratory phago-cytes by a Pasteurella multocida bacterin,Avian Dis. 37 (1993) 756–762.

[79] Toth T.E., Siegel P., Veit H., Cellular defenseof the avian respiratory system. Influx ofphagocytes: elicitation versus activation,Avian Dis. 31 (1987) 861–867.

[80] Toth T.E., Veit H., Gross W.B., Siegel P.B.,Cellular defense of the avian respiratory sys-tem: protection against Escherichia coli air-sacculitis by Pasteurella multocida-activatedrespiratory phagocytes, Avian Dis. 32 (1988)681–687.

[81] Toth T.E., Pyle R.H., Caceci T., Siegel P.B.,Ochs D., Cellular defense of the avian respi-ratory system: influx and nonopsonic phago-cytosis by respiratory phagocytes activated byPasteurella multocida, Infect. Immun. 56(1988) 1171–1179.

[82] Toth T.E., Curtiss R. 3rd, Veit H., Pyle R.H.,Siegel P.B., Reaction of the avian respiratorysystem to intratracheally administered aviru-lent Salmonella typhimurium, Avian Dis. 36(1992) 24–29.

[83] Tschernig T., Pabst R., Bronchus-associatedlymphoid tissue (BALT) is not present in thenormal adult lung but in different diseases,Pathobiology 68 (2000) 1–8.

[84] Van Alstine W.G., Arp L.H., Histologic eval-uation of lung and bronchus-associated lym-phoid tissue in young turkeys infected withBordetella avium, Am. J. Vet. Res. 49 (1988)835–839.

[85] Villegas P., Viral diseases of the respiratorysystem, Poult. Sci. 77 (1998) 1143–1145.

[86] Wang N., Banzett R.B., Butler J.P., FredbergJ.J., Bird lung models show that convectiveinertia effects inspiratory aerodynamic valv-ing, Respir. Physiol. 73 (1988) 111–124.

[87] Wang N., Banzett R.B., Nations C.S., JenkinsF.A. Jr., An aerodynamic valve in the avianprimary bronchus, J. Exp. Zool. 262 (1992)441–445.

[88] Watanabe H., Kobayashi K., Isayama Y.,Peculiar secretory IgA system identified inchicken. II. Identification and distribution offree secretory component and immunoglobu-lin of IgA, IgM, and IgG in chicken externalsecretions, J. Immunol. 115 (1975) 998–1001.

[89] Wieland W.H., Orzaez D., Lammers A.,Parmentier H.K., Verstegen M.W.A., SchotsA., A functional polymeric immunoglobulinreceptor in chicken (Gallus gallus) indicatesancient role of secretory IgA in mucosalimmunity, Biochem. J. 380 (2004) 669–676.

[90] Woodward J.D., Maina J.N., A 3D digitalreconstruction of the components of the gasexchange tissue of the lung of the muscovyduck, Cairina moschata, J. Anat. 206 (2005)477–492.