Potential NALT

Potential of Nasopharynx-associated Lymphoid Tissuefor Vaccine Responses in the Airways

Per Brandtzaeg1,2

1Laboratory for Immunohistochemistry and Immunopathology, Centre for Immune Regulation, University of Oslo, Olso, Norway; and 2Departmentof Pathology, Oslo University Hospital, Rikshospitalet, Oslo, Norway

Nasopharynx-associated lymphoid tissue (NALT), constituting Wal-deyer’s ring in humans, is a unique inductive site for B-cell responsesand plasma cell generation. This makes the nasal route of vaccineadministration interesting for induction of mucosal and systemicantibodies. The unpaired nasopharyngeal tonsil (adenoids) and thepaired palatine tonsils are prominent NALT structures, functionallysimilar to the paired rodent NALT structures located dorsal to thecartilaginous soft palate. HumanNALT is more strategically located,however, because its elements are exposed to both airborne andalimentary antigens andhave antigen-retaining crypts. It also showssimilarities with lymph nodes and participates both in systemic- andsecretory-typemucosal immunity. Primary folliclesoccur at 16weeksof gestation, which is similar to Peyer’s patches but different fromrodent NALT whose organogenesis begins at birth. The formationof germinal centers reflecting B-cell activation does not take placeuntil shortly after birth, and terminal differentiation of plasma cellcan be seen about 2 weeks postnatally. Germinal centers arise in Tcell–dependent B-cell responses and are associated with somatichypermutation of Ig V-region genes. Downstream switching to vari-ous Ig isotypes also takes place, with or without concurrent expres-sion of the J-chain gene. The J chain is a crucial part of dimeric IgA andpentameric IgM, making these Ig polymers able to interact with theepithelial polymeric Ig receptor. This interaction is central in theformation of secretory IgA and secretory IgM. Accumulating evi-dence suggests a major role for NALT in antibody immunity of therespiratory tract and associated glands.

Keywords: mucosa-associated lymphoid tissue; B-cell homing; muco-

sal immunity; secretory IgA; IgD

The unpaired nasopharyngeal tonsil (also called the adenoids)and the paired palatine and lingual tonsils constitute the majorpart of Waldeyer’s ring, with the tubal tonsils and lateral pharyn-geal bands as less prominent components (1). These organs seemto be functionally comparable with the nasopharynx-associatedlymphoid tissue (NALT) in rodents, which is composed of twopaired lymphoepithelial structures besides the nasopharyngealduct, dorsal to the cartilaginous soft palate (2, 3). Of note,rodents do not have tonsils, and Waldeyer’s ring is more strate-gically situated than their NALT to generate mucosal immunitybecause its elements are exposed to both airborne and alimen-tary antigens. Human NALT may therefore play an importantimmune-inductive role as part of mucosa-associated lymphoidtissue (MALT). It also shows similarities with lymph nodes andmay, in addition, contribute as an effector site of local systemic-type and mucosal secretory-type of adaptive humoral immunity.

Tonsils contain four specialized tissue compartments con-tributing to immune functions (4, 5): (1) the reticular crypt

epithelium, (2) the extrafollicular area, (3) the mantle zones oflymphoid follicles, and (4) the follicular germinal centers (GCs).Primary follicles are present in human tonsils as early as 16 weeksof gestation (4), which is similar to Peyer’s patches of gut-associated lymphoid tissue (GALT) but different from rodentNALT, whose organogenesis begins at birth (3). Nevertheless,the formation of tonsillar GCs that reflects B-cell activation in-duced by exogenous antigens does not take place until shortlyafter birth; and terminal differentiation of effector B cells toextrafollicular plasma cells (PCs) can first be seen approximately2 weeks postnatally (4).

The GCs characteristically arise in T cell–dependent B-cellresponses and are associated with (1) clonal expansion of B cells;(2) somatic hypermutation in B-cell immunoglobulin (Ig) variableregion genes (Ig V-region genes); (3) positive selection of B cellsthat are able to receive antigen-specific signals by high affinity; (4)downstream switching of Ig heavy chain constant region genes (CH

genes) to various isotypes; (5) differentiation to memory and effec-tor B cells and PCs of the expressed Ig isotypes; and (6) concurrentinduction of the J-chain gene in a variable subset of these cells. Thisgene encodes a 15-kD peptide, the joining (J) chain, which is a cru-cial structural part of dimers and trimers of IgA (collectively calledpolymeric IgA [pIgA]) and pentameric IgM (6). Without the in-corporation of J chain, pIgs cannot bind to the epithelial pIg re-ceptor (pIgR), also called membrane secretory component (SC)(7). The pIg2pIgR interaction is a central step in the formationand export of secretory IgA (SIgA) and secretory IgM (SIgM)antibodies to mucosal surfaces (8).

This article discusses the putative importance of human NALTfor the immune function of the upper airways. Accumulating evi-dence suggests a role ofWaldeyer’s ring in humoral immunity of theregionalmucosae and associated glands; someof theNALT-derivedand activated B cells seem to seed preferentially these secretory ef-fector sites (9, 10). Therefore, administration of vaccines via thenasal route has an interesting potential of enhancing regional im-munity and does at the same time stimulate systemic immunity.

NAIVE B CELLS ENTER TONSILS VIA HIGHENDOTHELIAL VENULES

Migration of lymphoid cells is strictly regulated by the expressionof multiple adhesion molecules and receptors for chemoattrac-tants (chemokine receptors) that interact with correspondingligands on endothelial and stromal cells (11). The extravasationof mainly naive, or antigen-unexperienced, T and B cells intoinductive lymphoid compartments takes place through special-ized postcapillary so-called “high endothelial venules” (HEVs)and is regulated by similar molecular principles in the systemicand the mucosal immune system (Figure 1).

The naive lymphocytes express CD62L (L-selectin) that inter-acts with the adhesion molecule peripheral lymph node addressinexpressed by HEVs in both lymph nodes and tonsils (12). InGALT, however, peripheral lymph node addressin is replacedby a CD62L-binding adhesion molecule named mucosal addressincell adhesion molecule-1 (6, 9). The chemokines involved at thelevel of HEVs are secondary lymphoid tissue chemokine (CCL21)

(Received in original form November 4, 2010; accepted in final form March 18, 2011)

Correspondence and requests for reprints should be addressed to Per Brandtzaeg,

D.D.S., M.Sc., Ph.D., Department of Pathology, Rikshospitalet, P.O. Box 4950

Nydalen, N-0424 Oslo, Norway. E-mail: [email protected]

Am J Respir Crit Care Med Vol 183. pp 1595–1604, 2011

Originally Published in Press as DOI: 10.1164/rccm.201011-1783OC on March 18, 2011

Internet address: www.atsjournals.org

and Epstein-Barr virus (EBV)–induced molecule 1 ligand chemo-kine (CCL19). We have observed an important species differencein that both CCL21 and CCL19 are produced by stromal cells inhumans and become redistributed to the luminal face of HEVs,whereas CCL21 is actually produced by the HEV endothelium itselfonly in mice (12). Both these chemokines attract preferentiallyCCR71 naive T cells, and apparently less actively naive B cells (9).

Naive B cells are CXCR51 and extravasate, according tomouse experiments, mainly via modified HEVs presentingCXCL13 (in humans also called B-cell attracting chemokine-1).These HEV-like vessels occur juxtaposed to, or inside of, lym-phoid follicles (9). The B cells are next attracted to the mantlezones where CXCL13 is deposited on stromal or dendritic ele-ments, such as the tips of the follicular dendritic cells (FDCs)(13), an expression pattern that we have found both in humanGALT and tonsils (Figure 1). The origin of CXCL13 in GCsremains unclear but seems to be mainly macrophage-likedendritic cells (DCs) (14). Also, follicular B-helper T (TFH) cellsare attracted to the follicles by similar receptor2ligand interac-tions (Figure 1). The B cells are initially primed just outside thelymphoid follicle by interaction with cognate T cells and antigen-presenting cells, mainly interdigitating DCs (9); they then reenterthe follicle and end up as CCR71 GC cells after interactionswith FDCs and TFH cells (see below). Primed and selected B cellsare able to leave the follicle as memory and effector cells becauseof their CXCR5 down-regulation and temporal CCR7 up-regulation, allowing their attraction to the corresponding extra-follicular chemokines (9).

B CELLS ARE ACTIVATED IN GCs TO BECOME MEMORYAND EFFECTOR CELLS

Primary follicles of secondary lymphoid organs, such as the ton-sils, consist mainly of recirculating naive B cells positive for sur-face IgD and IgM (sIgD1sIgM1), both coexpressed isotypes

exhibiting the same specificity for antigen. These CXCR51 lym-phocytes pass into the spaces of the network formed by theCXCL13-bearing FDCs (Figure 1). It is still unclear why bothsIgD and sIgM must be expressed to render B cells antigen-reactive. Likewise, the nature and origin of FDCs are obscure,but the existence of these cells and their accumulation in theprimary follicles depends on the presence of B cells. Thus, ani-mals depleted of B cells, or mice with severe combined immuno-deficiency do not have follicular aggregates of FDCs (15).

In contrast to lymph nodes, tonsils are not encapsulated andlack afferent lymph, but the reticular crypt epithelium containsmany DCs that can transport exogenous antigens to the extrafol-licular T-cell areas and to the B-cell follicles. Interdigitating DCsfunction as antigen-presenting cells in the extrafollicular primaryimmune responses and occur abundantly around HEVs, often inclusters with T cells that are mainly of the CD41 phenotype (4).These lymphocytes consist of both naive (CD45RA1) and memory(CD45R01) subsets, and some express the high-affinity IL-2 recep-tor (CD25) as a sign of recent activation (16). Altogether, the tonsilsseem able to mount both primary and secondary T-cell responses.

The GC reaction is initiated by stimulation of naive B cells im-mediately outside the lymphoid follicles through cognate help fromactivated CD41 T cells (Figure 1). These helper T cells have re-ceived processed foreign antigen from interdigitating DCs in thecontext of class II molecules of the major histocompatibility com-plex, in humans also called HLA class II, such as the classicalHLA-DR, -DQ and –DP molecules. The activated B cells canpick up antigen and further present it to the cognate T cells inan interaction that provides mutual support (Figure 1). Some ofthese B cells then colonize the follicles and act as “founder cells”for GCs (Figure 2).

Secondary follicles with GCs result from further stimulationof B cells, first via their interactions with FDCs on which nativeantigens are retained in the form of immune complexes. By thisactivation the founder B cells are induced to proliferate (Figure 2).

Figure 1. Main adhesion molecule– and

chemokine-regulated steps of T- and B-cell mi-gration to, and positioning within, mucosa-

associated lymphoid tissue (MALT). Naive T and

B cells enter both MALT and regional lymph

nodes via high endothelial venules (HEVs). Thechemokines involved (right panel) are secondary

lymphoid tissue chemokine (SLC, CCL21) and

Epstein-Barr virus–induced molecule 1 ligand

chemokine (ELC, CCL19), both produced bystromal cells and redistributed to the luminal

face of HEVs, as indicated, to attract preferen-

tially CCR71 naive T cells and, probably less ac-

tively, B cells. SLC may also be involved in theexit of lymphoid cells from MALT via draining

lymphatics as depicted. Naive B cells are

CXCR51 and, at least in mice, extravasatemainly via modified HEVs presenting CXCL13

(also called B-cell attracting chemokine [BCA]-

1 in humans) juxtaposed to, or inside of, lym-

phoid follicles; they are next attracted to themantle zone where BCA-1 is deposited on den-

dritic elements, such as the follicular-dendritic

cell (FDC) tips. The distribution of this chemokine, together with B and T cells, is shown by three-color immunofluorescence staining in the upper

left panel, and the receptor distribution is similarly shown in the lower left panel. Also, follicular B-helper T (TFH) cells (CXCR51) are attracted to the

follicle by similar interactions (e.g., white arrow, lower left panel ). B cells are primed just outside the lymphoid follicle by interaction with cognate

T cells in clusters with antigen-presenting cells, and the B cells may mutually act on the T cells, as indicated. The activated B cells then reenter the

follicle as “germinal center founder cells” to give rise to CCR71 B cells after interactions with FDCs and TFH cells (Figure 2). This temporal up-regulation of CCR7 allows the further activated B cells to leave the follicle as memory and effector cells by attraction to extrafollicular SLC and ELC.

Immunofluorescence pictures adapted from Carlsen and coworkers (13). HLA-II ¼ HLA class II molecules; TCR ¼ T-cell receptor.

1596 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 183 2011

A variety of adhesion molecules and other receptor proteins areinvolved in these interactions (16).

The GCs can be divided into various more or less well-definedmorphologic compartments (15). B-cell stimulation in the GCdark zone gives rise to exponential growth recognized by themonoclonal antibody Ki-67 (Figure 2) directed against a nuclearproliferation marker (16). The resulting centroblasts hypermutatetheir Ig V-region genes and give rise to centrocytes; these cells dieby apoptosis in the light zone unless they are rescued selectivelyby their ability to bind with high affinity to antigens present onFDCs, take up such antigens, and process and present them viaHLA class II molecules to TFH cells, which can be identified bytheir high level of CXCR5 expression (Figure 1).

The point in ontogeny when CD41 helper T cells becomeTFH cells, and whether these cells represent a separate T-celllineage, remains unsolved. Cell-surface markers and transcriptionfactors are currently under investigation (17–19). Tonsillar TFH

cells have been claimed to express CD57 in addition to CXCR5,but this is apparently not always the case. The best additionalmarkers seem to be inducible T cell costimulator (CD278),secretion of the immune activator IL-21, and expression ofthe costimulatory CD40 ligand (CD154). Cognate interactionbetween TFH cells and activated B cells, which express CD40,seems to be a very important event in the GC reaction (Figure2). The same is true for the induction of bcl-2 gene productsafter immune activation of centrocytes to prevent their apopto-sis, a rescuing event largely driven by IL-21 (19).

Although several costimulatory molecules are involved in thesynapse between TFH cells and HLA class II–expressing B cells(18, 19), GCs are not formed if the CD40 ligand–CD40 inter-action is experimentally blocked (20). Importantly, this costim-ulatory interaction also promotes switching of the Ig CH genesof B cells from Cm (IgM) to downstream isotypes, and differen-tiation to plasmablasts and PCs producing high-affinity anti-body. A prerequisite for the crucial consequences of cognateinteraction between B and TFH cells, in which activated GCB cells present processed antigen to the T cells (Figure 2), isthat the B cells express costimulatory B7 (CD80/CD86) mole-cules, which can bind to CD28 on the TFH cells. A recent ex-perimental study showed that CD28-dependent signaling isrequired for optimal TFH cell maturation and expansion (21).Classical memory B cells (sIgD2sIgM1) with strong B7 expres-sion are moreover found extrafollicularly in human tonsils

related to the crypt epithelium where they may likewise presentantigen to CD281 T cells (22).

VARYING ISOTYPE AND J-CHAIN EXPRESSIONBY TONSILLAR B CELLS

The tonsillar GC reaction normally generates a variable numberof intrafollicular Ig-producing plasmablasts or PCs dominated bythe IgG (55–72%) and IgA (13–18%) classes (4, 5). A large pro-portion of these two phenotypes shows concurrent J-chain ex-pression (26–43%), and even more so do those GC cells thatproduce IgM or IgD (49–82%) in healthy palatine tonsils of chil-dren (Figure 3A). Thus, the J chain–inducing capacity of theseorgans is much higher than that of peripheral lymph nodes butsimilar to that of mesenteric lymph nodes. The latter are func-tionally amplification organs for Peyer’s patches where the con-current induction of IgA and J chain is remarkably high (6, 9, 23).

The cytokine profiles and other microenvironmental factorsdetermining isotype differentiation and coexpression of J chainin B cells remain obscure (9, 23). According to the “decreasingpotential” hypothesis (24) it seems that clonal maturation inthe course of several proliferative GC cycles results in down-regulation of J-chain expression, thus promoting monomerproduction by the resulting IgA1 PCs rather than pIgA forsecretory immunity (9). This idea fits with the fact that theJ-chain expression was significantly suppressed both in intrafol-licular and extrafollicular IgA1 cells in tonsils of children aged2–12 years with recurrent tonsillitis (Figure 3A). These tonsilswere removed because of at least three annual attacks of acutetonsillitis during the past 2 years (25). Our studies were per-formed in 1980 and we were fortunate to obtain an age- andsex-matched unique biopsy control material from a group ofchildren undergoing surgery for inguinal hernias. None of themhad had any episode of acute tonsillitis, increased frequency ofcommon colds, or acute otitis media, and their serum levels ofimmunoglobulins were normal. Informed consent to performbiopsy was obtained from their parents (26).

Thus, whereas downstreamCH gene switching in GCs of healthytonsils gives rise to a relatively high fraction of extrafollicularIgA1J-chain1 PCs (50% of total IgA1 PCs), most extrafollicu-lar IgG1 PCs show little or no J-chain expression (Figure 3A).The J chain can only bind to the heavy chains of IgA and IgM,so in IgG1 and IgD1 PCs it has no apparent function and isbroken down in the cytoplasm (9). Notably, the phenotypic Ig

Figure 2. Immune events taking place in the

dark and light zones of germinal center in ton-

sillar secondary lymphoid follicle. The “germinalcenter founder cells” receive their initial stimu-

lation through cognate interaction with acti-

vated CD41 helper T cells just outside the

lymphoid follicle before they enter it to becomecentroblasts (Figure 1). Further details are given

in the text. CD40 ligand (CD40L)/CD40 ¼ co-

stimulatory molecules; FDC ¼ follicular dendriticcell; GCDC ¼ germinal center dendritic cell;

HLA-II ¼ HLA class II molecule; IC ¼ immune

complex; sIg ¼ surface immunoglobulin; TCR ¼T-cell receptor; TFH ¼ follicular B-helper T cell.

Pulmonary Perspective 1597

class distribution was similar in healthy and diseased tonsils(26), so the only striking tonsillitis-associated change in the localB-cell system was suppressed J-chain expression in IgA1 PCs(Figure 3A). There was a similar but nonsignificant trend inhypertrophic adenoids removed because of obstructive symp-toms in a group of age-matched children (26).

The fact that tonsillar IgA1 PCs are mainly of the IgA1 iso-type (at least .95%), along with the presence of IgD1 PCs,supports the notion that tonsillar B-cell differentiation takesplace mainly with classical downstream CH gene switching to-gether with some “nonclassical” switching to IgD (5, 6, 10, 23).A regionalized microbial impact on B-cell differentiation is sug-gested by the unique sIgD1IgM–CD381 centroblast subset identifiedin the dark zone of tonsillar GCs (27). This subset shows

deletion of the Sm and Cm gene segments, therefore selectivelygiving rise to IgD-producing PCs by the so-called “nonclassicalswitching.” Compartmentalized B-cell dispersion explains therelatively high frequency of IgD-producing PCs normally occur-ring in this region (10), and particularly the striking PC replace-ment with this class that often is seen in IgA deficiency (23, 28).Most strains of Hemophilus influenzae and Moraxella catarrha-lis, which are frequent colonizers of the nasopharynx, expressouter-membrane IgD-binding factors that can activate sIgD1

B cells by crosslinking sIgD of the B-cell receptor (29). In thismanner it seems likely that sIgD1 tonsillar centrocytes are stim-ulated to proliferate and differentiate polyclonally, therebydriving V-gene hypermutation and Sm/Cm gene deletion (30).

Such regional microbial influence on B-cell differentiation issupported by our observation that Sm/Cm deletion is morefrequently detected in diseased than in clinically healthy ton-sils and adenoids (10); and extrafollicular IgD-producing PCsare, as mentioned previously, relatively numerous in recurrenttonsillitis and adenoid hyperplasia (4, 5). However, there arelarge individual variations, and the IgD class percentage reportedin some publications (up to 20–25%) is much too high (30–32).We found the mean proportion of IgD-producing PCs to bewell below 5% of all isotypes in the extrafollicular compartment(4).

It is interesting that sIgD1IgM2 B cells generated by non-classical switching seem to express predominantly V-gene rep-ertoires that may allow considerable crossreactivity, includingautoimmunity, but understanding the biologic significance of thisobservation requires further studies (33). Although numerousantimicrobial and other IgD antibody activities have been mea-sured both in mouse and human serum, the protective or patho-genic role of circulating IgD has only recently been explored (34).

Because IgD does not activate the classical complement path-way, it is likely that such antibodies can block defense functionsof IgG and IgM within the mucosae and reduce the immune-exclusion efficiency of SIgA and SIgM antibodies in the upperairways in the face of bacterial infections that drive local IgDproduction (28). The ability of IgD to bind to monocytes andmacrophages and basofils (its cross-linking thereby inducing re-lease of proinflammatory cytokines, such as IL-1, IL-6, IL-8, andtumor necrosis factor-a) may add to the pathogenic potential ofIgD antibodies (32, 35). Of note, selective IgA-deficient patientsshowing substantial replacement with only IgM- and IgG-producingPCs in their nasal mucosa have less clinical problems in theirairways than those with abundant mucosal IgD-producing PCs(36).

DISSEMINATION OF ACTIVATED TONSILLAR B CELLSTO SECRETORY TISSUE SITES

Because the J chain is a key peptide in the formation of pIgA andpentameric IgM that can bind to the pIgR expressed basolater-ally on secretory epithelia (8), tonsillar B-cell differentiationexhibits features compatible with precursor generation for thesecretory immune system. It may be envisioned that only a frac-tion of the pIgA-expressing plasmablasts that exit from GCs willterminate their differentiation as extrafollicular PCs (5, 6); in-stead, many of them may home to regional secretory effectorsites for terminal differentiation to pIgA-producing PCs there(Figure 3B). The adenoids possess, in addition, a local secretoryimmune system because patches of the crypt epithelium expresspIgR/SC (4, 5). This is not the case in the palatine tonsils whereonly passive paracellular transfer of IgG and IgA takes placethrough the reticular crypt epithelium. Because this epitheliumis of squamous type in the palatine tonsils, it is additionallyprotected by its production of the antimicrobial peptide called

Figure 3. Putative B-cell developmental stages in germinal center of

palatine tonsils and adenoids. The pathways from surface Ig (sIg)–expressing B lymphocytes to terminal plasma-cell differentiation may

include coexpression of cytoplasmic J chain (J) and any of the four Ig

classes depicted, but the J chain can only combine with cytoplasmic IgAand IgM to form polymers (IgA1J, dimeric IgA; IgM1J, pentameric

IgM). (A) Plasmablasts and plasma cells in germinal center (GC) and

extrafollicular compartment near crypts show cytoplasmic coexpres-

sion of J chain in the proportions (%) indicated. Tissue samples werefrom children with clinically healthy or diseased (recurrent tonsillitis)

palatine tonsils. An extrafollicular area immunostained for IgA and J

chain is illustrated to the right (see color key) to show double expression

(IgA 1 J chain, yellow fluorescence; arrow). Data adapted from Korsrudand Brandtzaeg (25, 26). (B) The J chain–expressing B cells terminate to

some extent their differentiation locally in the extrafollicular compart-

ment (small inclined arrows at right) but may preferentially show a po-tential for homing to distant secretory effector sites, particularly in the

region of the upper respiratory and digestive tracts. Effector B cells with

little or no J-chain expression become extrafollicular plasma cells (or

may join the systemic immune compartment; not indicated).

1598 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 183 2011

calprotectin or L1 protein (4, 5). Therefore, topical surface pro-tection, including antibody-mediated immunity in the crypts,differs between the two lymphoid organs.

Nasal and bronchial mucosae, and salivary and lacrimal glands,contain an IgA11 and IgD1 PC distribution in their stroma similarto that of the extrafollicular tonsillar compartment (9). This factsupports the notion that such regional secretory tissues are seededby GC-derived B-cell blasts fromWaldeyer’s ring (5, 9, 10, 23). Itsimmune-inductive function would hence be similar to that of ro-dent NALT (37). Regionalization of MALT also concerns the gutbecause the intestinal lamina propria receives most of its activatedB cells from Peyer’s patches and other GALT structures, such asthe numerous solitary or isolated lymphoid follicles (6, 9, 23). Apossible minor blast contribution to the airways from bronchus-associated lymphoid tissue remains possible. Although isolatedlymphoid follicles do not regularly exist in the normal adult hu-man lung (23, 38), such bronchus-associated lymphoid tissue struc-tures were found in all of nine large bronchial autopsy specimensof children aged 2–15 years who died from traumatic causes andhad no airway disease (39). Likewise, in large autopsy specimensof nasal mucosa occasional scattered isolated follicles were seen in38% of children below 2 years of age (40). These follicles weremost likely induced postnatally like rodent NALT (3).

Dichotomy of the mucosal immune system between the gutand the airways has also been suggested by B-cell homing studiesin the rat (41, 42). Furthermore, activated human tonsillar B cellswere found to migrate to the lung, but not to the gut mucosa,when transferred to mice with severe combined immunodefi-ciency (43). Notably, direct immunization of the palatine tonsils,and for nasopharyngeal tonsils particularly vaccination by thenasal route, gave rise to local B-cell responses but the inducedcirculating specific B cells did not enter the small intestinal mu-cosa (44). Moreover, in infants dying of sudden infant deathsyndrome, the palatine GCs were shown to be overactivated asrevealed by an increased number of IgG1 and IgA1 plasmablastsand PCs (45). These activated B cells apparently seeded regionalsecretory effector sites, such as the parotid glands, in excessivenumbers to become local PCs (46). It has also been documentedin several human studies that nasal immunization induces specificIgA antibodies in nasopharyngeal secretions, in addition to en-hancing systemic immunity (6, 9, 47, 48).

Attempts have beenmade to track directly the dissemination ofB-cell blasts fromWaldeyer’s ring by means of molecular markers.Because the human herpesvirus EBV preferentially establishespersistent infection of tonsillar memory and effector B cells, theirmigration to other organs has been mapped by DNA analysis forlatent EBV infection (49). On the basis of this approach, B-celltrafficking was suggested to take place from Waldeyer’s ring toperipheral blood and to a lesser extent into systemic lymphoidorgans, such as mesenteric lymph nodes and the spleen (49). Us-ing another DNA marker, namely a deletion of the IgM CH gene,we confirmed and extended these results by showing that acti-vated tonsillar B cells undergoing so-called “nonclassical switch-ing” to IgD1J-chain1 plasmablasts by means of this deletion,preferentially home through cervical lymph nodes to the upperairways and associated glands (9, 10, 23).

The extravasation of activated memory and effector B andT cells into effector tissues takes place through the local microvas-cular endothelium and is controlled in a site-specific manner. Thisprocess is much better defined for the intestinal lamina propriathan for other secretory tissues (9, 11, 23). Thus, the B-cell homingdichotomy between the gut and the upper aerodigestive tractalluded to previously clearly has a molecular basis in terms ofadhesion molecules and chemokines (Figure 4). Endothelial mu-cosal addressin cell adhesion molecule-1 is expressed by the mi-crovasculature throughout the gut lamina propria, where it binds

GALT-derived memory and effector T- and B-cell blasts withhigh surface levels of the integrin a4b7 (9, 11, 23).

Homing to the small intestinal lamina propria is, in addition, de-termined by the chemokine receptor CCR9, which interacts with itsligand TECK (CCL25) produced preferentially by the epitheliumin that part of the gut. Conversely, homing to the colonic laminapropria is fine-tuned by CCR10 interacting with MEC (CCL25).The latter molecular pair is apparently also important for homingof tonsillar B-cell blasts to the upper respiratory tract (10, 50).However, these cells show poor gut-homing properties becauseof low levels of surface a4b7 but express, instead, CD62L (Figure4). This adhesion molecule enables such NALT-derived cells toenter not only regional secretory tissues related to the airways butalso peripheral lymph nodes and to some extent the uterine cervixmucosa (9, 10). Integration of mucosal immunity induced in Wal-deyer’s ring with systemic immunity is further enhanced by theexpression of CCR7 on tonsillar B-cell blasts (9, 10).

EFFECT OF ADENOTONSILLECTOMY ONREGIONAL IMMUNITY

In the United States, approximately 400,000 surgeries annuallycan be ascribed to removal of palatine tonsils or adenoids. Theindication for such operations has shifted from infection to upperairways obstruction. The frequency of tonsillectomy has thereforedeclined significantly and progressively since the 1970s. This de-velopment can probably also be ascribed to the increasing aware-ness of the immune function of Waldeyer’s ring. The informationreviewed in the previous sections provides strong support for thenotion that this lymphoid tissue functions as inductive NALT inhumans and supplies secretory effector sites of the aerodigestivetract with activated pIgA precursor cells (Figure 4). To evaluateclinically this notion, it is important to study the effect of adeno-tonsillectomy on the regional SIgA levels. The pioneering reportby Ogra (51) showed that combined tonsillectomy and adenoi-dectomy in children reduced the level of IgA antibody to polio-virus threefold to fourfold in their nasopharyngeal secretions anddelayed or abrogated their local SIgA response to subsequentlive oral poliovaccine. Notably, however, this result could par-tially have reflected an abolished local secretory immune functionbecause of removal of the adenoids (5).

Jeschke and Stroder (52) performed tonsillectomy in childrenand found that their serum Ig and salivary IgA levels decreased forup to 3 years. Furthermore, although D’Amelio and coworkers (53)observed no salivary IgA reduction (but decreased serum IgA) inpreviously tonsillectomized adults (16–24 yr old), Cantani and cow-orkers (54) found in children that salivary IgA and serum IgA (andless so IgG and IgM) were significantly reduced 4 months aftercombined adenotonsillectomy. Subsequent studies in tonsillectom-ized children showed, however, elevated salivary Ig levels after3–4 years (55), whereas no effect was found in tonsillectomizedyoung adults after 6 months except for a slight reduction of totalIgM and salivary IgG antibodies to Streptococcus mutans and EBV(56).

Altogether, there is a need for more extensive immunologicstudies focusing collectively on the adenoids and palatine tonsils.Considerable redundancy of inductive lymphoid tissue in Wal-deyer’s ring (such as the lingual tonsils) might mask a potentiallyunwanted immunologic effect of adenotonsillectomy. This pos-sibility is supported by studies that have reported reduced SIgAlevels in saliva from children with pharyngitis involving recurrenttonsillitis (57) or adenoid hyperplasia (58), perhaps reflectingdecreased global pIgA induction in Waldeyer’s ring caused byinflammation-induced down-regulation of J-chain expression asseen in recurrent tonsillitis (Figure 3A). It should also be notedthat the cervical lymph nodes apparently function as reserve

Pulmonary Perspective 1599

immune-inductive organs draining the nasopharyngeal and oro-pharyngeal region.

DISEASE- AND AGE-RELATED CHANGES OF TONSILLARIMMUNE FUNCTION

Recurrent tonsillitis, and to a lesser extent adenoid hyperplasia, isassociated with decreased J-chain expression by tonsillar plasma-blasts and PCs (Figure 3A); their putative contribution to theregional SIgA system is thereby compromised (Figure 3B). More-over, recurrent tonsillitis seems, in an irreversible manner, to speedup the age-related involution of the tonsils as immunologic organsin terms of reduced B-cell differentiation to PCs (4).

The underlying immunoregulatory alterationsmight be relatedto increased shedding of antigen-transportingmembrane (M) cellsin the reticular crypt epithelium, thereby also influencing the reg-ulated balance between expansion of early (J-chain positive) andmature (J-chain negative) effector B-cell clones (4, 9). The mi-croenvironmental conditions necessary for tonsillar proliferationof early B-cell clones might depend on the presence of M cellsand particular subsets of helper and regulatory T cells. Increasedshedding of M cells changes the way in which foreign material ispresented to the lymphoid tissue, and relatively more direct pas-sage through the reticular crypt epithelium could conceivablyresult in a different level or mode of antigen processing and pre-sentation that would preferentially favor the expansion of mature(i.e., recycled through GCs), rather than early, B-cell clonesaccording to the “decreased potential” hypothesis (24).

Even healthy tonsils are to a large extent involved in the gen-eration of mature memory and effector B-cell clones as evidencedby the predominance of extrafollicular J chain–negative IgG1 PCs(Figure 3A). In older children and adults with recurrent tonsillitis,expansion of both early and mature memory clones is apparentlydecreased. This development is indicated by reduced numbers ofIg-producing PCs in the intrafollicular and extrafollicular tonsillarcompartments (4). In addition to possible changes in helper andregulatory T cells, an underlying mechanism could be a decreasedregulated translocation of antigens into the tonsils. The number ofactive M cells is largely reduced in recurrent tonsillitis; even directpassage of foreign material may be hampered as the reticularepithelium becomes covered by a stratified and partly keratinizedlining. Similar changes occur in clinically normal tonsils above 25years of age (4).

Altogether, inflammatory conditions continuing after the age ofabout 10 years seem to accelerate the aging process of tonsils.

However, recurrent tonsillitis may significantly change the tonsillarimmunocompetence even before that age (Figure 3A). The ob-served alterations may be irreversible because they were revealedin periods when the patients had been without inflammatory

Figure 4. Putative scheme for compartmentalized muco-

sal B-cell homing from inductive (top) to effector (bottom)

sites in humans. Depicted are more or less preferred path-

ways (graded arrows) presumably followed by mucosal Bcells of any isotype activated in nasopharynx-associated

lymphoid tissue (NALT) represented by Waldeyer’s lym-

phoid ring (including palatine tonsils and adenoids), andbronchus-associated lymphoid tissue (BALT), versus gut-

associated lymphoid tissue (GALT) represented by Peyer’s

patches, appendix, and colonic-rectal isolated lymphoid

follicles. The principal homing receptor profiles of the re-spective B-cell populations, and compartmentalized adhe-

sion and chemokine cues directing their extravasation at

different effector sites, are indicated (pink and blue panels).

The gland-associated distribution of plasma cells (green),after terminal differentiation of extravasated mucosal B

cells, is schematically depicted at the bottom.

Figure 5. Presumed immunobiology underlying induction of T (violet)

and B (green) cells after nasal vaccine administration, and local gener-ation of secretory IgA (SIgA) antibodies by export via the polymeric Ig

receptor (pIgR). 1 ¼ delivery device for nasal vaccine administration

(nasal spray, drops, or OptiMist); 2 ¼ adjuvanted uptake of vaccine

antigen through nasal mucosa; 3 ¼ immune-induction in adenoidsand palatine tonsils (human NALT); 4 ¼ antigen targeting and migration

of mucosal dendritic cells (DCs); 5 ¼ immune induction and amplifica-

tion in regional (cervical) lymph nodes by antigen-loaded DCs and mac-rophages (Mfs); 6 ¼ compartmentalized homing and extravasation of

NALT-induced T and B cells to secretory effector sites in airways, gut, and

uterine cervix; and 7 ¼ local production and pIgR-mediated external

transport of dimeric IgA to generate SIgA. As discussed in the text,NALT-derived B cells preferentially extravasate at regional effector sites

and in systemic lymphoid organs, while showing only poor homing

capacity for the small intestinal lamina propria (Figure 4).

1600 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 183 2011

symptoms for at least 4 weeks (4, 25). Altered immune functionmay contribute to the recurrence of tonsillitis, and a vicious cyclemay then develop. It is noteworthy, however, that considerableimmunologic activity persists even in diseased palatine tonsils andadenoids of children, so the described functional changes cannotby themselves justify surgical removal of these lymphoid organs.Therefore, a conservative attitude toward adenotonsillectomy seemsto be immunologically justifiable, especially at an early age.

VACCINATION AGAINST PATHOGENS BYTHE NASAL ROUTE

Thepotential advantage of nasal immunization is illustrated by theprotection achieved against influenza. At present, parenteral vac-cination is recommended in vulnerable subjects, but this approachinduces little or no cross-protection. Thus, there is a continuingneed for interpandemic manufacturing of the actual vaccines;when a genomic drift occurs in a virus, the vaccine strain has tobe replaced, and it usually takes at least 6 months before a newvaccine is available. Conversely, many studies in experimental ani-mals and humans have demonstrated that nasal vaccination givesrise to cross-protection against drifted strains (48). With the avail-able live attenuated influenza vaccine for intranasal administra-tion (FluMist; MedImmune, Gaithersburg, MD), good protectionwas achieved despite the fact that the epidemic strain was notpart of the vaccine (59). Cross-clade immunity against experi-mentally applied HIV in mice has been reported after nasalDNA prime followed by nasal peptide boost; the vaccine con-tained epitopes of clade B, but high and long-lasting serum anti-body titers against the neutralizing gp41 ELDKWAS epitopesfrom clades A, B, C, and D were observed (60).

In nasal mucosa of unvaccinated adult subjects, antibody-producing PCs with specificity for influenza virus are present(61), but it remains unknown whether they are induced by pre-vious (subclinical) infection or reflect cross-reactivity of the mu-cosal IgA system. Notably, parenteral immunization with aninactivated trivalent virus vaccine did not result in an increaseof influenza-specific PCs in nasal mucosa (62), although an IgAresponse was elicited in tonsils and saliva (63, 64).

A dense population of putative antigen-presenting cells withamacrophage orDC phenotype exists in and below the normal sur-face epithelium of human nasal mucosa (65). The previously men-tioned results suggest that to stimulate a regional immune response,a vaccine should be targeted both against these cells, which maymigrate to the cervical lymph nodes, and against the crypts withM cells characteristic of NALT structures, probably the adenoidsin particular (Figure 5). Such local B-cell induction apparentlyimprints the necessary homing properties of the primed cells toextravasate efficiently in airway mucosa and associated glandsand give rise to secretory immunity at these regional effectorsites (10, 66). The NALT-derived plasmablasts may to some ex-tent reach the uterine cervix mucosa but do not express sufficient

gut-homing properties (Figure 4) to enter consistently the intes-tinal lamina propria, particularly not in the small bowel (9, 10).

The amount of antigen reaching the lymphoid tissue of Wal-deyer’s ring and cervical lymph nodes after parenteral immuni-zation is clearly insufficient to induce a nasal immune response,although some SIgA antibodies may occur in nasopharyngealsecretions. This most likely reflects local production in theadenoids where, as pointed out previously, epithelial expressionof pIgR/SC exists, in contrast to the palatine tonsils (4, 5). Notunexpectedly, there is some communication in terms of memoryand effector B-cell distribution between the mucosal and thesystemic immune systems, particularly so in cervical lymph nodesand Waldeyer’s lymphoid ring because of shared homing mole-cules, as discussed in a previous section (10).

The efficacy and effectiveness of the trivalent, cold-adapted liveattenuated nasal spray influenza vaccine (CAIV-T; MedImmune incooperationwithWyeth,Madison,NJ) havebeen documented bothin healthy children and adults (67). Although nasal vaccination alsoefficiently induces systemic immunity, a combination of intranasaland parenteral immunization may be preferable for optimal pro-tection when an inactivated influenza vaccine is used (68). Alter-natively, the effect of subunit vaccines applied topically can beenhanced by incorporation into liposomes or with the addition ofa nontoxic mucosal adjuvant (Eurocine; Eurocine Vaccines AB,Karolinska Institutet, Solna, Sweden). Adjuvantation of nasalvaccines with mucoadhesive polymers, such as chitosan deriva-tives, has been promising in mouse experiments (69), particu-larly when combined with antigen-loaded nanoparticles (70).

The initial optimism regarding Escherichia coli heat-labileenterotoxin as an adjuvant in humans (71) vanished with theoccurrence of Bell’s palsy after nasal application of an adju-vanted inactivated influenza vaccine (Nasalflu; Berna BiotechAG, Basel, Switzerland) (72). This problem apparently reflectsthe possibility for toxins to enter the central nervous systemfrom the olfactory bulb or cause irritation and swelling of nervesgoing through bony canals to the brain. Other adjuvants, such asthe hydrophobic outer-membrane protein preparations (proteo-somes) from Neisseria meningitidis, may be an efficient and safealternative (73–75). In mice, the uptake of proteosomes can beenhanced by incorporation of the Toll-like receptor 2 ligandPorB (76), but it remains to be shown whether Toll-like recep-tors are expressed on M cells or other parts of the follicle-associated epithelium in human NALT. Finally, virus-derivedparticles may function without adjuvants as demonstrated fora trivalent inactivated whole-cell influenza vaccine (77), prob-ably because of enhanced targeting to M cells and DCs. How-ever, a more recent clinical trial reported superior efficacy ofthe live attenuated influenza vaccine in children 12–59 monthsof age, both for antigenically well-matched and drifted viruses(78).

An inactivated whole-virus monovalent influenza vaccinewas recently tested in Oslo, Norway, with different devices for

TABLE 1. VACCINATED SUBJECTS (% OF N ¼ 15–19) WITH HEMAGGLUTINATION INHIBITIONSERUM ANTIBODY TITER $ 40 BEFORE AND AFTER TWO DOSES (3 WK) OR FOUR DOSES (6 WK)OF INACTIVATED WHOLE-VIRUS INFLUENZA VACCINE, AND SHOWING A SIGNIFICANT NASAL IgARESPONSE AFTER FOUR DOSES

Vaccine

Delivery

Serum:

Preimmunization

Serum: After

Two Doses

Serum: After

Four Doses

Nasal Fluid

IgA Response

Nasal spray (OptiMist) 32% 89% 100% P ¼ 0.0003

Nasal spray (conventional) 32% 79% 94% P ¼ 0.0006

Nasal drops 28% 94% 100% P ¼ 0.0006

Oral spray 29% 76% 87% No response*

Based on data from References 79 and 80.

* No increase of IgA antibody titer in nasal fluid, but a slight increase in whole saliva.

Pulmonary Perspective 1601

intranasal or intraoral spray application, and exhibited promisingresults for induction of antibodies both in serum and nasal secre-tions (Table 1). Mild side effects were deemed to be acceptable(79, 80). Most importantly, the same intranasal vaccine alsoinduced cellular immunity in addition to the desirable two-tiered antibody response, mucosal and systemic. A serum hem-agglutination inhibition titer of 40 or higher, which is considereda protective level, was obtained in most volunteers after twovaccine doses given 1 week apart; and an additional memoryeffect was revealed after three or four doses in that 100% ofthe individuals had acquired protective antibody titers. This re-sult was best achieved by vaccine application in nasal dropsor with a special breath-activated device (OptiMist; OptinoseAS, Oslo, Norway), whereas oral spray only induced serumantibodies.

CONCLUSIONS

The enormous innate drive of the mucosal immune system doesnot only enhance diversity but also memory (48). However, withregard to NALT, the situation in humans and mice seems to besignificantly different. First, NALT of rodents is an organizedbell-shaped structure in the floor of the nasal cavity, and itsorganogenesis is different from that of GALT and human pal-atine and nasopharyngeal tonsils (3, 81). Both these types oftonsils have deep and branched antigen-retaining crypts,whereas rodent NALT has a smooth surface, like GALT inall mammalian species (23). This disparity probably explainsthat GCs develop shortly after birth in tonsils as in the heavilymicrobe-exposed GALT structures, whereas rodent NALTrequires an infection or another danger signal, such as choleratoxin, to drive GC formation (82, 83).

The relative lack of innate stimulation is therefore probably thereason that a nonreplicating rotavirus vaccine adjuvanted witha mutant E. coli toxin did not induce a substantial memory re-sponse in murine NALT (84). By contrast, human NALT with itsantigen- and microbe-retaining crypts is liable to polyclonal stim-ulation for enhanced development of B-cell diversity and mem-ory. In addition, the crypt epithelium is activated by microbialproducts to secrete the B-cell activating factor of the tumor ne-crosis family (BAFF) and the cytokine thymic stromal lympho-poietin (TSLP). Class switch and broad reactivity of local B cells ispromoted by thymic stromal lymphopoietin when it activatesBAFF-producing DCs (85). Together, these features of Waldeyer’sring constitute an intriguing basis for the current interest in exploit-ing the nasal route (Figure 5) for vaccine administration to combata variety of diseases (86). The adjuventation and delivery systemsof inactivated nasal vaccines should be as tissue-compatible aspossible. Many approaches are explored and several have beentested in phase I clinical trials (87). Side effects have to be carefullymonitored, but nasal vaccine administration seems to be much lessrisky than pulmonary delivery by aerosol technology (88).

Disclosure Statement: The author does not have a financial relationship with acommercial entity that has an interest in the subject of this manuscript.

Acknowledgment: Hege Eliassen and Erik K. Hagen provided excellent assistancewith the manuscript and figures.

References

1. Dolen WK, Spofford B, Selner JC. The hidden tonsils of Waldeyer’s

ring. Ann Allergy 1990;65:244–250.

2. Kuper CF, Koornstra PJ, Hameleers DMH, Biewenga J, Spit BJ,

Duijvestijn AM, van Breda Vriesman PJC, Sminia T. The role

of nasopharyngeal lymphoid tissue. Immunol Today 1992;13:219–224.

3. Fukuyama S, Hiroi T, Yokota Y, Rennert PD, Yanagita M, Kinoshita N,

Terawaki S, Shikina T, Yamamoto M, Kurono Y, et al. Initiation of

NALT organogenesis is independent of the IL-7R, LTßR, and NIK

signaling pathways but requires the Id2 gene and CD3-CD41CD451

cells. Immunity 2002;17:31–40.

4. Brandtzaeg P. Immune functions and immunopathology of palatine and

nasopharyngeal tonsils. In: Bernstein JM, Ogra PL, editors. Immu-

nology of the ear. New York: Raven Press; 1987. pp. 63–106.

5. Brandtzaeg P. Immunological functions of adenoids. In: van Cauwenberge P,

Wang D-Y, Ingels K, Bachert C, editors. The nose. The Hague: Kugler

Publications; 1998. pp. 233–246.

6. Brandtzaeg P, Farstad IN, Johansen F-E, Morton HC, Norderhaug IN,

Yamanaka T. The B-cell system of humanmucosae and exocrine glands.

Immunol Rev 1999;171:45–87.

7. Brandtzaeg P, Kiyono H, Pabst R, Russell MW. Terminology: nomenclature

of mucosa-associated lymphoid tissue. Mucosal Immunol 2008;1:31–37.

8. Brandtzaeg P, Prydz H. Direct evidence for an integrated function of J

chain and secretory component in epithelial transport of immuno-

globulins. Nature 1984;311:71–73.

9. Brandtzaeg P, Johansen F-E. Mucosal B cells: phenotypic characteristics,

transcriptional regulation, and homing properties. Immunol Rev 2005;

206:32–63.

10. Johansen F-E, Baekkevold ES, Carlsen HS, Farstad IN, Soler D, Brandtzaeg

P. Regional induction of adhesion molecules and chemokine receptors

explains disparate homing of human B cells to systemic and mucosal

effector sites: dispersion from tonsils. Blood 2005;106:593–600.

11. Kunkel EJ, Butcher EC. Chemokines and the tissue-specific migration

of lymphocytes. Immunity 2002;16:1–4.

12. Carlsen HS, Haraldsen G, Brandtzaeg P, Baekkevold ES. Disparate

lymphoid chemokine expression in mice and men: no evidence of

CCL21 synthesis by human high endothelial venules. Blood 2005;106:

444–446.

13. Carlsen HS, Baekkevold ES, Johansen F-E, Haraldsen G, Brandtzaeg P.

B cell attracting chemokine 1 (CXCL13) and its receptor CXCR5 are

expressed in normal and aberrant gut associated lymphoid tissue. Gut

2002;51:364–371.

14. Carlsen HS, Baekkevold ES, Morton HC, Haraldsen G, Brandtzaeg P.

Monocyte-like and mature macrophages produce CXCL13 (B cell-

attracting chemokine 1) in inflammatory lesions with lymphoid neo-

genesis. Blood 2004;104:3021–3027.

15. MacLennan ICM. Germinal centers. Annu Rev Immunol 1994;12:117–139.

16. Brandtzaeg P, Halstensen TS. Immunology and immunopathology of

tonsils. Adv Otorhinolaryngol 1992;47:64–75.

17. Rasheed AU, Rahn HP, Sallusto F, Lipp M, Muller G. Follicular B helper

T cell activity is confined to CXCR5hiICOShi CD4 T cells and is in-

dependent of CD57 expression. Eur J Immunol 2006;36:1892–1903.

18. Linterman MA, Vinuesa CG. T follicular helper cells during immunity

and tolerance. Prog Mol Biol Transl Sci 2010;92:207–248.

19. Yu D, Vinuesa CG. The elusive identity of T follicular helper cells.

Trends Immunol 2010;31:377–383.

20. Kelsoe G. The germinal center reaction. Immunol Today 1995;16:324–326.

21. Platt AM, Gibson VB, Patakas A, Benson RA, Nadler SG, Brewer JM,

McInnes IB, Garside P. Abatacept limits breach of self-tolerance in

a murine model of arthritis via effects on the generation of T follicular

helper cells. J Immunol 2010;185:1558–1567.

22. Liu Y-J, Barthelemy C, de Bouteiller O, Arpin C, Durand I, Bachereau

J. Memory B cells from human tonsils colonize mucosal epithelium

and directly present antigen to T cells by rapid up-regulation of B7–1

and B7–2. Immunity 1995;2:239–248.

23. Brandtzaeg P. Function of mucosa-associated lymphoid tissue in anti-

body formation. Immunol Invest 2010;39:303–355.

24. Ahmed R, Gray D. Immunological memory and protective immunity:

understanding their relation. Science 1996;272:54–60.

25. Korsrud FR, Brandtzaeg P. Influence of tonsillar disease on the expression

of J-chain by immunoglobulin-producing cells in human palatine and

nasopharyngeal tonsils. Scand J Immunol 1981;13:281–287.

26. Korsrud FR, Brandtzaeg P. Immunohistochemical evaluation of J-chain

expression by intra- and extra-follicular immunoglobulin-producing

human tonsillar cells. Scand J Immunol 1981;13:271–280.

27. Liu Y-J, Arpin C. Germinal center development. Immunol Rev 1997;156:

111–126.

28. Brandtzaeg P, Jahnsen FL, Farstad IN, Haraldsen G.Mucosal immunology

of the upper airways: an overview. Ann N Y Acad Sci 1997;830:1–18.

29. Janson H, Heden L-O, Grubb A, Ruan M, Forsgren A. Protein D, an

immunoglobulin D-binding protein of Haemophilus influenzae:

1602 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 183 2011

cloning, nucleotide sequence, and expression in Escherichia coli. In-

fect Immun 1991;59:119–125.

30. Liu Y-J, Arpin C, de Bouteiller O, Guret C, Banchereau J, Martinez

Valdez H, Lebecque S. Sequential triggering of apoptosis, somatic

mutation and isotype switch during germinal center development.

Semin Immunol 1996;8:169–177.

31. Arpin C, de Bouteiller O, Razanajaona D, Fugier-Vivier I, Briere F,

Banchereau J, Lebecque S, Liu YJ. The normal counterpart of IgD

myeloma cells in germinal center displays extensively mutated IgVH

gene, Cm-Cd switch, and l light chain expression. J Exp Med 1998;187:

1169–1178.

32. Chen K, Xu W, Wilson M, He B, Miller NW, Bengten E, Edholm ES,

Santini PA, Rath P, Chiu A, et al. Immunoglobulin D enhances im-

mune surveillance by activating antimicrobial, proinflammatory

and B cell-stimulating programs in basophils. Nat Immunol 2009;10:

889–898.

33. Zheng N-Y, Wilson K, Wang X, Boston A, Kolar G, Jackson SM, Liu

YJ, Pascual V, Capra JD, Wilson PC. Human immunoglobulin se-

lection associated with class switch and possible tolerogenic origins

for Cd class-switched B cells. J Clin Invest 2004;113:1188–1201.

34. Preud’homme JL, Petit I, Barra A, Morel F, Lecron JC, Lelievre E.

Structural and functional properties of membrane and secreted IgD.

Mol Immunol 2000;37:871–887.

35. Drenth JP, Goertz J, Daha MR, van der Meer JW. Immunoglobulin D

enhances the release of tumor necrosis factor-a, and interleukin-1b as

well as interleukin-1 receptor antagonist from human mononuclear cells.

Immunology 1996;88:355–362.

36. Brandtzaeg P, Karlsson G, Hansson G, Petruson B, Bjorkander J,

Hanson LA. The clinical condition of IgA-deficient patients is related

to the proportion of IgD- and IgM-producing cells in their nasal

mucosa. Clin Exp Immunol 1987;67:626–636.

37. Yanagita M, Hiroi T, Kitagaki N, Hamada S, Ito HO, Shimauchi H,

Murakami S, Okada H, Kiyono H. Nasopharyngeal-associated lym-

phoreticular tissue (NALT) immunity: fimbriae-specific Th1 and

Th2 cell-regulated IgA responses for the inhibition of bacterial at-

tachment to epithelial cells and subsequent inflammatory cytokine

production. J Immunol 1999;162:3559–3565.

38. Pabst R, Tschernig T. Bronchus-associated lymphoid tissue: an entry site

for antigens for successful mucosal vaccinations? Am J Respir Cell

Mol Biol 2010;43:137–141.

39. Heier I, Malmstrom K, Sajantila A, Lohi J, Makela M, Jahnsen FL.

Characterisation of bronchus-associated lymphoid tissue and antigen-

presenting cells in central airway mucosa of children. Thorax 2011;66:

151–156.

40. Debertin AS, Tschernig T, Tonjes H, Kleemann WJ, Troger HD, Pabst

R. Nasal-associated lymphoid tissue (NALT): frequency and locali-

zation in young children. Clin Exp Immunol 2003;134:503–507.

41. Van der Brugge-Gamelkoorn GJ, Claassen E, Sminia T. Anti-TPN-

forming cells in bronchus-associated lymphoid tissue (BALT) and

paratracheal lymph node (PTLN) of the rat after intratracheal priming

and boosting with TNP-KLH. Immunology 1986;57:405–409.

42. Sminia T, van der Brugge-Gamelkoorn GJ, Jeurissen SH. Structure and

function of bronchus-associated lymphoid tissue (BALT). Crit Rev

Immunol 1989;9:119–150.

43. Nadal D, Albini B, Chen C, Schlapfer E, Bernstein JM, Ogra PL. Distri-

bution and engraftment patterns of human tonsillar mononuclear cells

and immunoglobulin secreting cells in mice with severe combined im-

munodeficiency: role of the Epstein-Barr virus. Int Arch Allergy Appl

Immunol 1991;95:341–351.

44. Quiding-Jarbrink M, Granstrom G, Nordstrom I, Holmgren J, Czerkinsky

C. Induction of compartmentalized B-cell responses in human tonsils.

Infect Immun 1995;63:853–857.

45. Stoltenberg L, Vege A, Saugstad OD, Rognum TO. Changes in the

concentration and distribution of immunoglobulin-producing cells in

SIDS palatine tonsils. Pediatr Allergy Immunol 1995;6:48–55.

46. Thrane P, Rognum TO, Brandtzaeg P. Increased immune response in up-

per respiratory and digestive tracts in SIDS. Lancet 1990;335:229–230.

47. Brandtzaeg P. Immune functions of human nasal mucosa and tonsils in

health and disease. In: Bienenstock J, editor. Immunology of the lung

and upper respiratory tract. New York: McGraw; 1984. pp. 28–95.

48. Brandtzaeg P. Induction of secretory immunity and memory at mucosal

surfaces. Vaccine 2007;25:5467–5484.

49. Laichalk LL, Hochberg D, Babcock GJ, Freeman RB, Thorley-Lawson

DA. The dispersal of mucosal memory B cells: evidence from per-

sistent EBV infection. Immunity 2002;16:745–754.

50. Kunkel EJ, Kim CH, Lazarus NH, Vierra MA, Soler D, Bowman EP,

Butcher EC. CCR10 expression is a common feature of circulating

and mucosal epithelial tissue IgA Ab-secreting cells. J Clin Invest

2003;111:1001–1010.

51. Ogra PL. Effect of tonsillectomy and adenoidectomy on nasopharyngeal

antibody response to poliovirus. N Engl J Med 1971;284:59–64.

52. Jeschke R, Stroder J. Verlaufsbeobachtung klinischer und immunolo-

gischer Parameter, insbesondere des Speichel-IgA, bei tonsillekto-

mierten Kindern. Klin Padiatr 1980;192:51–60.

53. D’Amelio R, Palmisano L, Le Moli S, Seminara R, Aiuti F. Serum and

salivary IgA levels in normal subjects: comparison between tonsil-

lectomized and non-tonsillectomized subjects. Int Arch Allergy Appl

Immunol 1982;68:256–259.

54. Cantani A, Bellioni P, Salvinelli F, Businco L. Serum immunoglobulins

and secretory IgA deficiency in tonsillectomized children. Ann Al-

lergy 1986;57:413–416.

55. Lenander-Lumikari M, Tenovuo J, Puhakka HJ, Malvaranta T, Ruuskanen

O, Meurman O, Meurman P, Vilja P. Salivary antimicrobial proteins

and mutans streptococci in tonsillectomized children. Pediatr Dent

1992;14:86–91.

56. Kirstila V, Tenovuo J, Ruuskanen O, Suonpaa J, Meurman O, Vilja P.

Longitudinal analysis of human salivary immunoglobulins, non-

immune antimicrobial agents, and microflora after tonsillectomy. Clin

Immunol Immunopathol 1996;80:110–115.

57. Ostergaard PA. IgA levels and carrier rate of pathogenic bacteria in 27

children previously tonsillectomized. Acta Pathol Microbiol Scand

[C] 1977;85:178–186.

58. Hess M, Kugler J, Haake D, Lamprecht J. Reduced concentration of

secretory IgA indicates changes of local immunity in children with

adenoid hyperplasia and secretory otitis media. ORL J Otorhinolar-

yngol Relat Spec 1991;53:339–341.

59. Belshe RB, Gruber WC, Mendelman PM, Cho I, Reisinger K, Block SL,

Wittes J, Iacuzio D, Piedra P, Treanor J, et al. Efficacy of vaccination

with live attenuated, cold-adapted, trivalent, intranasal influenza virus

vaccine against a variant (A/Sydney) not contained in the vaccine. J

Pediatr 2000;136:168–175.

60. Devito C, Zuber B, Schroder U, Benthin R, Okuda K, Broliden K,

Wahren B, Hinkula J. Intranasal HIV-1-gp160-DNA/gp41 peptide

prime-boost immunization regimen in mice results in long-term HIV-1

neutralizing humoral mucosal and systemic immunity. J Immunol 2004;

173:7078–7089.

61. Brokstad KA, Cox RJ, Eriksson JC, Olofsson J, Jonsson R, Davidsson

A. High prevalence of influenza specific antibody secreting cells in

nasal mucosa. Scand J Immunol 2001;54:243–247.

62. Brokstad KA, Eriksson JC, Cox RJ, Tynning T, Olofsson J, Jonsson R,

Davidsson A. Parenteral vaccination against influenza does not induce

a local antigen-specific immune response in the nasal mucosa. J Infect

Dis 2002;185:878–884.

63. Brokstad KA, Cox RJ, Olofsson J, Jonsson R, Haaheim LR. Parenteral

influenza vaccination induces a rapid systemic and local immune re-

sponse. J Infect Dis 1995;171:198–203.

64. El-Madhun AS, Cox RJ, Søreide A, Olofsson J, Haaheim LR. Systemic

and mucosal immune responses in young children and adults after

parenteral influenza vaccination. J Infect Dis 1998;178:933–939.

65. Jahnsen FL, Gran E, Haye R, Brandtzaeg P. Human nasal mucosa

contains antigen-presenting cells of strikingly different functional

phenotypes. Am J Respir Cell Mol Biol 2004;30:31–37.

66. Quiding-Jarbrink M, Nordstrom I, Granstrom G, Kilander A, Jertborn

M, Butcher EC, Lazarovits AI, Holmgren J, Czerkinsky C. Differ-

ential expression of tissue-specific adhesion molecules on human

circulating antibody-forming cells after systemic, enteric, and nasal

immunizations. A molecular basis for the compartmentalization of

effector B cell responses. J Clin Invest 1997;99:1281–1286.

67. Glezen WP. Influenza vaccination for healthy children. Curr Opin Infect

Dis 2002;15:283–287.

68. Keitel WA, Cate TR, Nino D, Huggins LL, Six HR, Quarles JM, Couch

RB. Immunization against influenza: comparison of various topical

and parenteral regimens containing inactivated and/or live attenuated

vaccines in healthy adults. J Infect Dis 2001;183:329–332.

Pulmonary Perspective 1603

69. Hagenaars N, Mania M, de Jong P, Que I, Nieuwland R, Slutter B,

Glansbeek H, Heldens J, van den Bosch H, Lowik C, et al. Role of

trimethylated chitosan (TMC) in nasal residence time, local distri-

bution and toxicity of an intranasal influenza vaccine. J Control

Release 2010;144:17–24.

70. Slutter B, Bal S, Keijzer C, Mallants R, Hagenaars N, Que I, Kaijzel E, van

EdenW, Augustijns P, Lowik C, et al. Nasal vaccination withN-trimethyl

chitosan and PLGA based nanoparticles: nanoparticle characteristics

determine quality and strength of the antibody response in mice

against the encapsulated antigen. Vaccine 2010;28:6282–6291.

71. de Bernardi di Valserra M, Zanasi A, Ragusa S, Gluck R, Herzog C. An

open-label comparison of the immunogenicity and tolerability of in-

tranasal and intramuscular formulations of virosomal influenza vac-

cine in healthy adults. Clin Ther 2002;24:100–111.

72. Mutsch M, Zhou W, Rhodes P, Bopp M, Chen RT, Linder T, et al. Use

of the inactivated intranasal influenza vaccine and the risk of Bell’s

palsy in Switzerland. N Engl J Med 2004;350:896–903.

73. Berstad AK, Andersen SR, Dalseg R, Dromtorp S, Holst J, Namork E,

Wedege E, Haneberg B. Inactivated meningococci and pertussis

bacteria are immunogenic and act as mucosal adjuvants for a nasal

inactivated influenza virus vaccine. Vaccine 2000;18:1910–1919.

74. Plante M, Jones T, Allard F, Torossian K, Gauthier J, St-Felix N, White

GL, Lowell GH, Burt DS. Nasal immunization with subunit proteo-

some influenza vaccines induces serum HAI, mucosal IgA and pro-

tection against influenza challenge. Vaccine 2001;20:218–225.

75. Treanor J, Nolan C, O’Brien D, Burt D, Lowell G, Linden J, Fries L.

Intranasal administration of a proteosome-influenza vaccine is well-

tolerated and induces serum and nasal secretion influenza antibodies

in healthy human subjects. Vaccine 2006;24:254–262.

76. Chabot SM, Chernin TS, Shawi M, Wagner J, Farrant S, Burt DS,

Cyr S, Neutra MR. TLR2 activation by proteosomes promotes up-

take of particulate vaccines at mucosal surfaces. Vaccine 2007;25:

5348–5358.

77. Greenbaum E, Furst A, Kiderman A, Stewart B, Levy R, Schlesinger M,

Morag A, Zakay-Rones Z. Mucosal [SIgA] and serum [IgG] immu-

nologic responses in the community after a single intra-nasal immu-

nization with a new inactivated trivalent influenza vaccine. Vaccine

2002;20:1232–1239.

78. Belshe RB, Edwards KM, Vesikari T, Black SV, Walker RE, Hultquist

M, Kemble G, Connor EM; CAIV-T Comparative Efficacy Study

Group. Live attenuated versus inactivated influenza vaccine in infants

and young children. N Engl J Med 2007;356:685–696.

79. Bakke H, Samdal HH, Holst J, Oftung F, Haugen IL, Kristoffersen AC,

Haugan A, Janakova L, Korsvold GE, Krogh G, et al. Oral spray

immunization may be an alternative to intranasal vaccine delivery to

induce systemic antibodies but not nasal mucosal or cellular immu-

nity. Scand J Immunol 2006;63:223–231.

80. Bakke H, Haneberg B. The development of mucosal vaccines (in

Norwegian; see http://www.tidsskriftet.no/). Tidsskr Nor Laegeforen

2006;126:2818–2821.

81. Kunisawa J, Fukuyama S, Kiyono H. Mucosa-associated lymphoid tis-

sues in the aerodigestive tract: their shared and divergent traits and

their importance to the orchestration of the mucosal immune system.

Curr Mol Med 2005;5:557–572.

82. Henriksson G, Helgeland L, Midtvedt T, Stierna P, Brandtzaeg P. Im-

mune response to Mycoplasma pulmonis in nasal mucosa is modu-

lated by the normal microbiota. Am J Respir Cell Mol Biol 2004;31:

657–662.

83. Shikina T, Hiroi T, Iwatani K, Jang MH, Fukuyama S, Tamura M, Kubo

T, Ishikawa H, Kiyono H. IgA class switch occurs in the organized

nasopharynx- and gut-associated lymphoid tissue, but not in the

diffuse lamina propria of airways and gut. J Immunol 2004;172:6259–

6264.

84. Ogier A, Franco MA, Charpilienne A, Cohen J, Pothier P, Kohli E.

Distribution and phenotype of murine rotavirus-specific B cells in-

duced by intranasal immunization with 2/6 virus-like particles. Eur J

Immunol 2005;35:2122–2130.

85. Xu W, He B, Chiu A, Chadburn A, Shan M, Buldys M, Ding A, Knowles

DM, Santini PA, Cerutti A. Epithelial cells trigger frontline immu-

noglobulin class switching through a pathway regulated by the in-

hibitor SLPI. Nat Immunol 2007;8:294–303.

86. Vajdy M, Singh M. Intranasal delivery of vaccines against HIV. Expert

Opin Drug Deliv 2006;3:247–259.

87. Jabbal-Gill I. Nasal vaccine innovation. J Drug Target 2010;18:771–786.

88. Lu D, Hickey AJ. Pulmonary vaccine delivery. Expert Rev Vaccines

2007;6:213–226.

1604 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 183 2011

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.