Nasal vaccination induces protective immunity without immunopathology

In the delicate structures of the gastrointestinal tract, the oropharynx, the respiratory tract, the salivary gland, the urogenital tract and the conjunctiva of the eye, the ‘kill everything foreign’ concept isn’t feasible. These structures are constantly bombarded by antigenic material and also have the important functions of, for example, absorbing nutrients and oxygen. Large inflammatory responses are inappropriate for gaseous exchange in the lung and lead to debilitating condition such as asthma and bronchiolitis.

Much of our knowledge on the mucosal associated lymphoid tissue (MALT) has been derived from studies on Peyer's patches in mice and humans. These structures contain 3 basic immune compartments: intraepithelial lymphocytes, diffuse lamina propria lymphocytes and organized B cell follicles. Overlying the B cell follicles, in the epithelium, are M cells whose function is to sample luminal contents and introduce them to lymphocytes in the pocket immediately beneath. Though it was originally thought that the majority of mucosal tissues contained these features significant differences exist. The most important is that MALT, though constitutive in the Peyer's patch and the appendix, is induced at other mucosal sites. The level of resident inflammatory cells also appears to depend on the species tested, age and antigenic load. Though the presence of M-like cells has been described for some locations outside the gut, their ability to transport antigens from the lumen to the lamina propria has not been demonstrated [1]. Data regarding the function, composition and structure of MALT should therefore be viewed in context and is likely to be site specific.

The mechanisms dictating lymphocyte migration patterns through MALT have been the subject of intense investigation. Originally a ‘common mucosal immune system’ was proposed, whereby lymphocytes activated by antigen in one mucosal tissue disseminated to and protected other mucosal sites. With the discovery of chemokines and their receptors we now appreciate that diape desis is far more complicated and dictated not only by selectin/integrin expression but also by epithelial derived chemokines [for an excellent review see 2]. It appears that there is even regional specialization within the previously unified gastrointestinal tract, which is likely to reflect changes in antigenic load and luminal contents. Regional specialization in the respiratory tract, though yet to be demonstrated, is reasonable to assume since the upper respiratory tract, especially the nose, is exposed to a high antigen load whereas the lower respiratory tract is essentially sterile. This is reflected by the presence of resident nasal associated lymphoid tissue (NALT) whereas few lymphoid cells surround the alveoli in a healthy individual. Naïve and memory lymphocytes may therefore continually traffic through the nasal tissue but are only attracted to the lower airways when absolutely required. Studies comparing recruitment of cells to different parts of the respiratory tract are limited [3]. Lymphocytes extracted from the nose preferentially home back to NALT and cervical and mesenteric lymph nodes but not to Peyer's patches thus indicating a fundamental difference in migratory properties of cells from this site [4]. More recently Csencsits et al. [5] demonstrated that adhesion molecule expression on high endothelial venules in NALT is similar to peripheral lymph nodes and distinct from the MAdCAM-1 dominated expression in the Peyer's patch. Even less is known about the signals mediating cell migration into the airways themselves. P-and to a lesser extent E-selectin expressing cells accumulate in the airways after intratracheal administration of sheep red blood cells [6]. Whether these are necessary for migration into the airway or simply reflect initial requirements for exiting the blood is not currently known.

Profound and damaging infiltration into the airways is common during viral infection and in asthma. Animal models of these conditions may therefore help to elucidate cell trafficking within the respiratory system. Lower respiratory tract infections are the third most common cause of death [7]. Respiratory syncytial virus (RSV) alone causes 70% of viral bronchiolitis cases. There is strong evidence that it is a T cell mediated immunopathological disease [8,9], which is supported by studies in established murine models. RSV infection of BALB/c mice results in the recruitment of a large inflammatory infiltrate containing CD8+ T cells and CD4+ T cells [10]. It appears to be the level, rather than the phenotype, of the inflammatory infiltrate that causes the clinical symptoms of disease. A reduction of T cells or TNF reduces pulmonary recruitment of inflammatory cells and the severity of illness without preventing virus clearance [11,12]. To prevent over exuberant immune responses we clearly need to investigate how T cells enter the lower airways during inflammation. This process is likely to involve a combination of selectins, integrins and chemokines [13]. A schematic representation of events during RSV infection is shown in Fig. 1.

Fig. 1.

Events mediating cell recruitment to the airway during RSV infection. (1) RSV infects the respiratory epithelium of the lower lung causing an increase in molecules involved in cell adhesion and the release of proinflammatory cytokines (e.g. TNF) and chemokines (e.g. RANTES); (2) Released TNF up-regulates adhesion molecule expression by high endothelial cells causing naïve T cells to migrate out of the blood and into the lung interstitium or mediastinal lymph node; (3) Naïve T cells are then activated by antigen presented on APCs, which increases the expression of chemokine receptors (e.g. CCR5); (4) Cells become intimately attached to airway epithelial cells via CCR5/RANTES and ICAM-1/LFA-1. Migration into the airways occurs along a chemokine concentration gradient; Cells in the lumen are retained there by similar interactions.

Due to the debilitating and long-term effects of RSV infection, vaccine development is a priority. Attempts have been hampered however, by the failed formalin-inactivated RSV vaccine trials of the 1960s where immunized children developed more severe disease than unvaccinated controls. A fundamental problem therefore exists. Efficient virus clearance requires T cells but T cells cause disease if allowed to accumulate unchecked. Clearly, an effective vaccine needs to limit viral replication but not fill the airways with inflammatory cells. The answer may lie in the step-wise reduction of viral burden as it migrates down the respiratory tract from NALT to larynx to BALT and finally the alveoli.

An article by Matsuoka et al. [14] in this issue, clearly shows the benefit of inducing immunity in both NALT and the lower lung compared to the lung alone. This paper also indicates that regional specialization may exist within the respiratory tract. Enteric or intradermal vaccination of mice with vaccinia virus expressing the fusion protein of RSV (rVV-F) display reduced RSV replication in the lung but not in the nasal tissue. This lack of protection in the upper respiratory tract by vaccination at sites distant to the nose has previously been demonstrated using vaccinia constructs in the cotton rat model [15] and also using the fusion protein together with mutated cholera toxin [16]. Together these studies suggest that vaccination in the skin or intestine induces RSV specific cells capable of extravasating into the lung tissue but not into the nose. More importantly, mice vaccinated intranasally are protected at both respiratory sites [14,17]. The fact that immunization at distant sites only protects the lung may imply, firstly, that the recruitment of lymphocytes to, or their retention in, the nasal tissue differs from the lung, secondly, that it may be nothing to do with homing properties but rather the level of resident memory cells at each site and finally, that differences in the respiratory tract simply reflect the large vascular bed in the lower airways making it more likely for cells to exit the blood. It should be noted however, that other studies have demonstrated protection in the upper and lower respiratory tract even when vaccines are administered at distant sites [18]. Protection may therefore also depend on vaccine formulation. To unravel these mysteries integrin and chemokine receptor expression needs to be compared in the nose and lung during infection of vaccinated mice.

It is well established that intranasal RSV challenge causes immunopathology in mice previously sensitized with rVV-F [19]. The phenotype and extent of disease can alter depending on the initial vaccination site [20]. In this issue Matsuoka et al. [14] show that dermal administration leads to more immunopathology after intranasal RSV challenge compared to intranasal vaccination. This difference may be explained by a requirement for initial control of RSV in NALT thereby reducing the pathogen burden in the lower airways in mice vaccinated by the intranasal route. In support of this theory Chen et al. [21] show that immunopathology after RSV challenge can only be avoided when the vaccine is delivered to the nasal cavity. A similar beneficial effect of intranasal RSV fusion protein vaccination has previously been demonstrated when coadministered with cholera toxin [22]. Therefore nasal vaccination induces protection without the pathology and can even protect distant mucosal sites from a variety of infectious diseases [23–25].

In our quest for an effective RSV vaccine it is clear that we need more knowledge on what causes cell extravasation and retention in different parts of the respiratory system. Knowledge on how this alters with age and exposure to other pathogens [26,27], will be important for the suitable delivery of future vaccines. Protection of the lung alone is not enough to prevent immunopathology, which seems to require a step-wise reduction in viral titre; the NALT could fulfil this role by reducing the pathogen burden to a level that only induces minimal inflammation in the lower lung. For this to occur most effectively it is clear that intranasal vaccination will be required.