Mucosal vaccination: onward and upward

ABSTRACT

Introduction

The unique mucosal immune system allows the generation of robust protective immune responses at the front line of pathogen encounters. The needle-free delivery route and cold chain-free logistic requirements also provide additional advantages in ease and economy. However, the development of mucosal vaccines faces several challenges, and only a handful of mucosal vaccines are currently licensed. These vaccines are all in the form of live attenuated or inactivated whole organisms, whereas no subunit-based mucosal vaccine is available.

Areas covered

The selection of antigen, delivery vehicle, route and adjuvants for mucosal vaccination are highly important. This is particularly crucial for subunit vaccines, as they often fail to elicit strong immune responses. Emerging research is providing new insights into the biological and immunological uniqueness of mucosal tissues. However, many aspects of the mucosal immunology still await to be investigated.

Expert opinion

This article provides an overview of the current understanding of mucosal vaccination and discusses the remaining knowledge gaps. We emphasize that because of the potential benefits mucosal vaccines can bring from the biomedical, social and economic standpoints, the unmet goal to achieve mucosal vaccine success is worth the effort.

KEYWORDS:

• Adjuvant
• delivery system
• mucosal immunology
• mucosal vaccine
• subunit vaccine

1. The mucosal immune system

The mucosal surface consists of a large, exposed area of thin tissue lining the respiratory, digestive and genitourinary tracts. It not only facilitates important physiological functions such as nutrient absorption and gas exchange, but also serves a critical barrier role. The mucosa therefore possesses a robust immune system and can respond to the constant exposure to environmental substances including antigens and allergens from food and microorganisms. In higher mammals, this mucosal immune system (MIS) is the integration of an immunological network composed of tissues, lymphoid cells, mucous membrane-associated cells, myeloid cells, and effector molecules such as cytokines, chemokines, antimicrobial peptides, and secretory immunoglobulin A (SIgA).

The MIS is equipped with sophisticated innate and adaptive systems. The innate arm of MIS employs a range of mechanisms to provide first-line defense in the mucosa, which can be categorized into mechanical, chemical and cellular elements. The mechanical elements of the innate MIS include a physical barrier of tightly formed epithelium supplemented with a mucus coating and, as seen in the gut mucosa, peristalsis movements and enterocyte shedding [1]. Secondly, the chemical components include innate receptors, antimicrobial peptides and enzymes, specific protease inhibitors, and reactive oxygen species, which facilitate clearance of the pathogen through opsonization or cytokine-mediated immune responses. The cellular elements comprise of innate immune cells such as macrophages, dendritic cells (DCs), mast cells, and natural killer cells. These cell types either directly engulf pathogens via phagocytosis, or initiate signaling pathways that promote killing upon recognizing conserved structures on the pathogens [1]. These mechanisms work across the extra-epithelial, epithelial and subepithelial levels. In general, antimicrobial compounds work at the extra-epithelial level, whereas the mechanical barriers as well as the pattern recognition receptors (PRRs) that recognize highly conserved pathogen associated molecular patterns (PAMPs) function at the epithelial level. At the subepithelial level, the innate immune cells recognize pathogens and microbial molecules that have breached through the other two layers, and utilize a cascade of cytokines and chemokines to recruit effector cells as well as to trigger adaptive immune responses.

The adaptive MIS can be primarily divided into inductive and effector sites based on anatomical and functional properties (Figure 1). The mucosa-associated lymphoid tissue (MALT) is the principal inductive site and the largest lymphoid system in mammals, containing about 50% of the lymphocytes in the body [2]. MALT covers the surfaces of all mucosal tissues and can be divided by their anatomical location into gut-associated lymphoid tissue (GALT), nasopharynx-associated lymphoid tissue (NALT), and bronchus-associated lymphoid tissue (BALT), for example. Mucosal epithelia at different body locations have varied cellular organizations, but all contain similar compartmentation of follicles, interfollicular regions, subepithelial dome regions, and follicle-associated epithelium [2]. The linings in the oral cavity, pharynx, esophagus and urethra are formed by multi-layered squamous epithelia, in contrast to the single cell layer that covers the intestinal mucosa. In the respiratory and vaginal mucosa, the composition of the lining varies from pseudo-stratified tissue to simple epithelium. Due to the structural differences, these epithelia employ various antigen sampling mechanisms to control the cross-talk between the lumen and the lamina propria [3]. In stratified and pseudo-stratified epithelia, antigen presenting cells (APCs) can directly scout the antigens into the epithelium. For simple intestinal and respiratory epithelia of which the intercellular spaces are sealed by tight junctions, specialized M cells are required to capture and deliver the antigens. Through trans-epithelial transport, the antigens enter from the lumen to the underlying organized lymphoid tissue, where APCs reside [4,5]. Upon contact with the processed antigens from APCs, cognate naïve B cells and T cells undergo activation, clonal expansion, and differentiation into effector cells, which then migrate from the inductive site to effector sites present in all parts of the mucosa [6].

Figure 1. The mucosal adaptive immune system. Based on anatomical and functional properties, the mucosal adaptive immune system can be divided into inductive and effector sites. B cells and T cells undergo activation and clonal expansion in the inductive site. The differentiated effector cells then migrate through the peripheral lymphatic system to the effector sites present in all parts of the mucosa, where they perform specific functions upon activation, such as antibody production and cell-mediated immune responses. SIgA: secretory immunoglobulin A; pIgR: polymeric immunoglobulin receptor.

At the immune effector sites such as the lamina propria (LP), the surface epithelium consists of concentrated intraepithelial lymphocytes (IELs) and secretory immunoglobulin A (SIgA) that perform specific functions upon activation. IgA is the predominant antibody isotype in respiratory and digestive secretions; in contrast, IgG is the principal isotype in the bloodstream, lymph and interstitial fluids. While IgG mainly acts by opsonizing pathogens for phagocytosis and activating the complement system, IgA functions primarily as a neutralizing antibody that limits the access and entry of microorganisms and environmental antigens [7]. The secreted IgA molecules multimerize through the expression of the J chain, which facilitates binding between the secretory component (SC) and the multimerized IgA. The dimerized (and less commonly, the oligomerized) form of IgA is then recognized by the polymeric Ig receptor (pIgR) at the basolateral surface and subsequently transcytosed to the lumen. During the transcytosis process, SIgA can intercept invaded pathogens intracellularly and excrete them as non-virulent immune complexes to the luminal side. Upon arriving in the lumen, SIgA interacts with a large range of antigens and form different immune complexes that promote either protective (in response to antigens containing infectious agents) or homeostatic (in response to antigens associated with commensals) conditions. SIgA also possesses a natural resistance to proteolytic degradation, making it uniquely suited for functioning in mucosal secretions that are rich in proteases [8].

The production of antigen-specific SIgA starts with antigen presentation from APCs to naïve T cells, which then migrate to the B cell region of the MALT and induce class switching of IgM⁺ B cells to IgA⁺ B cells by the secretion of specific cytokines [9]. The antigen-primed mucosal B and T immunocytes then depart the inductive site and enter the circulation through the lymph, until they reside at selected mucosal sites and differentiate into memory or effector cells. The final destination of these matured cells is mainly the mucosa of origin, which is determined by the affinity between the site-specific integrins (‘homing receptors’) on their surface and the complementary mucosal tissue-specific receptors (‘addressins’) on the epithelial cells [10]. This cell migration is also directed by chemokines produced in the local microenvironment, which promote chemotaxis and regulate integrin expression [11]. After homing to the effector sites, IgA⁺ B cells develop into plasma cells that produce the monomeric form of IgA, which then go through the polymerization process described above and are released as SIgA in external secretions [12]. In addition, despite their anatomical separation, the functional connection between different MALT sites termed the ‘common mucosal immune system’ allows IgA secretion in mucosa distant from the site of antigen presentation and B cell activation [13].

2. Mucosal vaccination

The mucosa serves as the boundary between the internal body and the external environment and is the gateway for most infectious agents or antigens to enter the human body. Infectious agents can only disseminate further to cause a systemic disease upon successful invasion through the mucosa; therefore, mucosal vaccination can induce first-line immunity that prevents establishment and spread of the infection. The local immunity elicited by mucosal vaccination is especially crucial for protective efficacy against noninvasive infections at mucosal surfaces that are normally poorly permeable for serum antibody transduction or passive passage across the epithelia. Examples of such infections include enterotoxigenic Escherichia coli (ETEC), where vaccine-induced protection is conferred almost exclusively by locally produced SIgA antibodies [14]. Other examples include Helicobacter pylori gastrointestinal (GI) infection and chlamydial genital infection, in which the protective immunity is mediated predominantly by mucosal CD4⁺ T helper cells [15,16]. Upon mucosal vaccination, the strongest immune response occurs in the mucosal sites directly exposed to the antigen, followed by the surrounding tissues of antigen exposure. Furthermore, the notion of a ‘common mucosal immune system’ indicates that mucosal vaccination can elicit immune responses in multiple effector sites that are remote from the original vaccination site [17–19]. The migration of immune cells from inductive to effector mucosal tissues through the lymphatic system has been evidenced by the presence of antigen-specific mucosal IgA responses at extensive mucosal effector sites [20]. This can theoretically generate sterilizing immunity across epithelial surfaces, instead of just limiting the infection and protecting against the progression of disease symptoms [8].

Compared to injectable vaccines which generate predominantly serum IgG antibodies, mucosal immunization is able to induce both antigen-specific mucosal SIgA and systemic IgG antibodies, thus providing dual protection against pathogens [21]. Vaccine-induced mucosal immunity provides additional advantages over systemic immunity, as it may prevent invasion of the pathogen in the host as well as limit transmission to others. In this sense, mucosal vaccination has been regarded as an effective way to induce herd immunity for highly transmittable diseases, including the recent COVID-19 pandemic [22]. A growing chorus of experts have agreed that boosters administered through the mucosal route could reduce transmission or reinfections, and even potentially break the current reliance on the constant reformulation of vaccines targeting new variants of concern [23]. Furthermore, as preexisting systemic immunity generally does not interfere with the uptake of vaccine into mucosal inductive sites, a mucosal vaccine can be complementary to systemic immunization, or provide additional protection for infants with passive immunity acquired from maternal antibodies [24]. Antigens introduced by prior mucosal exposure can also prime the immune system to generate stronger responses to subsequent systemic immunization, and to induce high levels of both systemic and mucosal antibody responses [25,26]. This is evident in the case of systemic immunization of adults with pneumococcal polysaccharides who have often been primed by natural infections in early childhood. In these individuals, effective induction of serum IgG as well as IgA and IgG responses in nasal wash or saliva was observed after systemic vaccination [27]. The numerous advantages of mucosal vaccines compared with injected vaccines can be exemplified in influenza vaccination, in which several needle-based (subcutaneous or intramuscular) and needle-free (nasal) vaccines are available. It has been shown that injected influenza vaccines tend to reduce disease severity but often do not prevent disease onset. They also elicit poorer cross-reactive immunity compared to natural infections [28]. In contrast, nasal influenza vaccines can offer both systemic and mucosal immunity, and may provide better cross-protection [29–32]. Additional advantages of mucosal vaccination include the possibility of noninvasive and needle-free immunization. This mode of vaccination offers several benefits: it reduces physical and psychological discomfort, increases vaccine acceptance and safety, avoids the possibility of blood-borne infections due to needle re-use or accidents, and allows for more frequent boosting [33]. Needle-free delivery also reduces the need and costs associated with trained personnel for administration, and even offers the potential option of self-administration, which will be advantageous for mass immunization during pandemic situations [34,35]. These advantages are especially helpful in conquering barriers to vaccination in low-income or remote settings.

3. Current status of mucosal vaccine development

Due to the traditionally low efficiency of mucosal immunization, currently, there are only a few approved vaccines that are administered mucosally (Table 1). These vaccines are mostly live attenuated/inactivated whole organisms, and target mainly enteric pathogens [3]. Vaccines based on whole organisms carry their own ‘danger signals’ that provide adjuvant activities; however, they also contain the risk of side effects caused by the induction of inflammatory pathology. Furthermore, inactivated vaccines may lose their effectiveness due to denaturing of the antigens, while live attenuated viral vaccines can potentially regain their pathogenicity through spontaneous reversion and thus are not suitable for immunocompromised individuals [49]. Many recent efforts are focusing on subunit vaccines based on one or more recombinantly expressed or purified antigens (peptides, proteins, or polysaccharides). Subunit vaccines avoid the risks associated with pathogen contact, virulence reversion, and harmful inflammation. However, the major trade-off for the safety profile is the suboptimal stability and immunogenicity, which often need to be overcome by specialized delivery systems and/or potent adjuvants [49,50]. Subunit vaccines are also more susceptible to the intrinsic biological obstacles to the induction of effective mucosal immunity. As discussed above, antigens in vaccines are easily blocked by the multifaceted mechanical, cellular and chemical barriers. The physically and chemically harsh environment of mucosal surfaces is especially challenging for subunit mucosal vaccines. Taking orally administered subunit vaccines for example, degradation (e.g. by digestive enzymes such as pepsin), clearance (e.g. by ciliary movement and mucus flow) and biological barriers (e.g. by gastric acid, secretions and epithelial tight junctions) can all render the vaccination less effective [51]. In addition, the efficacy of mucosal subunit vaccines can be further reduced due to their short residence time in the mucosa and poor antigen absorption. Nonetheless, the advantage of vaccines containing only precise antigens lies not only in their safety but also in the flexibility of manufacturing and responsiveness to pandemic outbreaks. Therefore, continuous efforts are being made to improve existing and to discover new delivery systems for mucosal subunit vaccines. With the addition of mucoadhesives (e.g. chitosan, engineered grains such as MucoRice [52], starch-based Spherex, or other types of polymers such as carbopol, sodium alginate or cationic nanogel [53,54]), immunostimulants (e.g. adjuvants such as aluminum salts, innate receptor agonists), and permeation enhancers (e.g. bacterial toxins such as cholera toxin [CT], heat-labile enterotoxin [LT]), the above-mentioned challenges may be overcome [55–57]. Table 2 below summarizes selected advanced delivery strategies for mucosal subunit vaccines, including novel adjuvants, carrier systems, and formulations. Their advantages and disadvantages are also noted.

Table 1. Licensed mucosal vaccines to date.

Infection	Format	Trade Name	Administration	Formulation
Influenza	Live attenuated cold-adapted influenza virus reassortant strains	FluMist [36], Fluenz Tetra (Medimmune) [37]	Nasal	Spray
Poliomyelitis	Live attenuated polio vaccine (OPV)	Biopolio (bOPV), mOPV, tOPV, etc [38]	Oral	Liquid
Febrile acute respiratory diseases	Live-attenuated adenovirus type 4 and type 7	N/A (approved in new military recruits in the U.S [39].	Oral	Liquid
Rotavirus	Live attenuated monovalent human rotavirus	Rotarix (GlaxoSmithKline) [40]	Oral	Liquid
	Pentavalent live vaccine	Rotateq (Merk) [41]	Oral	Liquid
Recurrent urinary tract infections	Whole cell-inactivated	Uromume [42,43]	Oral (sublingual)	Spray
Typhoid	Ty21a Live attenuated vaccine	Vivotif (Crucell) [44,45]	Oral	Enteric coated capsule
Cholera	Live attenuated O1 serogroup (Inaba) – ctxA attenuation	Vaxchora [46]	Oral	Liquid
	Heat and formaldehyde-inactivated O1 serogroups (Inaba + Ogawa) + 0139	Euvichol, Shanchol [47]	Oral	Liquid
	Heat and formaldehyde-inactivated O1 serogroups (Inaba + Ogawa) + CTB	Dukoral [48]	Oral	Liquid

Table 2. Recently developed technologies and materials for mucosal delivery of primarily subunit, DNA and mRNA vaccines.

Strategy	Examples	Advantages	Considerations
Mucoadhesive/ mucopenetration	MucoRice [58], phospholipid-like amphiphilic polyphosphazene nanovesicle [59], chitosan [60,61], cationic nanogel [54] Polysaccharide alginate nanogel [62], LBL-Liposome [63], pluronic F127 [64], timethylchitosan-tripolyphosphate [65]	Enhanced antigen update in APCs that leads to improved antigen-specific immune responses Versatile and adaptable, allows modification of surface chemistry and carrier size for the case of synthetic polymers Stable, controlled release, and generally low toxicity	Antigen loading efficiency and inconsistent reproducibility in the case of synthetic polymers, nonspecific interactions, trade-off between adhesion and penetration means the antigen could be trapped in mucus, most relevant in oral delivery
Mucosal adjuvant	Synthetic surfactant SF10 [66], non-toxic enterotoxin derivatives [67,68], pathogen recognition pattern (PRR) ligands [69–71], Poly(I:C) [72], α-GalCer [73], nano-emulsion [74]	Increased antigen presentation, co-stimulation of immunomodulatory signals, prolonged immune response, reduced vaccine dosage and frequency	Toxicity, strategy for efficient co-administration of antigen and adjuvant,
Cell targeted vehicles	Epithelial cell targeted [75,76], dendritic cell targeted [65,77,78], M cell targeted [79].	Allows specific immunomodulation such as for Th1 (water-in-oil) or Th2 (oil-in-water)	Low efficacy by oral route
Recombinant viral vehicle	Adenoviral vector [80–84]	Safe, high loading capacity, self-assembly, intracellular delivery, intrinsic immunogenicity, versatile	Expensive and difficult to make, often requires additional adjuvant, effective release needs to be addressed
Recombinant bacterial vehicle	Lactic acid bacteria [85–89], recombinant Enterococcus faecalis [90], recombinant Listeria monocytogenes [91], recombinant salmonella typhimurium [92]	Natural adjuvant properties, safe, inexpensive production	Antigen presentation and stabilization, concerns about genetically modified status

As there is an increasing number of studies describing the disconnection between current systemically-delivered vaccines and mucosal immune protection, mucosal delivery of existing or new vaccines has become a focus for many diseases. For example, a H1N1 influenza challenge performed in consenting human volunteers showed that preexisting mucosal IgA provided higher protection against severe disease than systemic IgG, suggesting a vaccine strategy that can induce mucosal IgA may provide better protection against influenza [93]. In light of this, several groups have attempted the development of mucosal vaccines against influenza virus using different delivery systems. As examples, Dong et. al. proposed using hemagglutinin (HA) embedded in graphene oxide polyethyleneimine (GP) nanoparticles as an un-adjuvanted intranasal vaccine and showed that the vaccine induced protection against homo- and hetero-subtypic influenza infection in mice [94]. Another nasal vaccine strategy developed by Varma et. al. demonstrated that the hydrogel plus cGAMP formulation increased mucoadhesion and retention of a HA antigen that has been computationally optimized, and resulted in significantly higher IgG and IgA levels following a prime-boost schedule [95]. In regard to the recent pandemic, data have shown that current mRNA vaccination against SARS-CoV-2 does not induce sufficient mucosal immunity in the human lower respiratory tract that can prevent infection by higher transmittable strains such as the Omicron BA.1.1 variant. This is partly due to the fact that current vaccines, while being robust in inducing circulating humoral and cellular immunity, fail to provoke notable lung tissue-residing Spike-specific memory B and T cells [23,96,97]. In their recent research, Tang et. al. demonstrated that the combination of systemic mRNA plus mucosal adenovirus type-5 vectored S protein (Ad5-S) vaccination provoked strong cellular immunity in the respiratory tract, as well as neutralizing mucosal IgA against Omicron BA.1.1 [23]. Similarly, Pitcovski et. al. reported that the oral subunit SARS-CoV-2 booster vaccine following subcutaneous priming elicited serum IgG and mucosal IgA levels that were comparable to those induced by three subcutaneous doses, as well as a T cell response as evidenced by IFN-γ and IL-2 secretion [98]. This supports the use of a mucosal booster strategy in systemically vaccinated individuals as a promising approach to achieve more robust immunity and protection from re-infection by future variants [99]. In addition, the combination of parenteral and mucosal vaccines can be highly effective in combatting pathogens that infect through the mucosal route and then spread systemically, such as poliovirus and hepatitis A [100]. Recent research has shown that mucosal exposure to microbes or vaccines can induce airway memory macrophages, which possess properties to enhance innate protection against both the intended target as well as heterologous pathogens in the lungs [101–103].

4. Special considerations for mucosal vaccines

4.1. The multiple administration routes

Although in some cases such as poliovirus, systemic IgG is sufficient to protect infections occurring at the mucosal sites, in general, local immune responses are desirable for protection against diseases that mainly occur via the corresponding mucosal routes. As the mucosa covers a large area of the body surface, it provides several versatile routes for vaccination. As discussed previously, the epithelial lining differs dramatically across different mucosal surfaces. Therefore, when considering the routes for mucosal immunization, the mechanism of antigen transport across epithelial barriers is a key consideration. Research that delves into the specific molecular recognition systems and the nonspecific adhesion mechanisms, as well as the interactions between microorganisms with these transepithelial transport pathways, will greatly facilitate more effective targeting of the mucosal immune system [104].

Oral administration of vaccines has been of interest for years, and all but one of the currently licensed mucosal vaccines are given through this route (Table 1). Compared to other mucosal vaccination routes, oral delivery provides easier access and is associated with less discomfort, which will assist in large-scale vaccine roll-outs and higher compliance. Oral vaccination also has a safer profile compared to nasal administration as it avoids the potential risk of antigen migration to the brain through olfactory epithelium and bulbs [105,106]. In the last decades, extensive research has improved the immunogenicity of orally delivered antigens; however, the harsh GI environment and the presence of oral tolerance mechanisms remain major challenges. Alternatively, vaccine delivery through the sublingual and buccal mucosa may be advantageous. Both sites consist of non-keratinized epithelium and thinner cell layers, which offer an anatomically and histologically suitable structure and cell composition for vaccine delivery [107]. Furthermore, sublingual vaccination can generate mucosal immune responses in the respiratory, digestive and reproductive tracts, whereas immunity induced by oral immunization is mostly restricted to the digestive tract [108]. Traditionally these routes have been used for the delivery of low molecular weight drugs to the bloodstream, such as sublingual immunotherapy (SLIT) for counteracting allergic hypersensitivity [107]. However, the concept of these former applications, which takes advantage of immune tolerance, is vastly different from mucosal vaccination which aims at inducing an immune reaction.

Compared to oral vaccination, the nasal route presents several advantages such as the lack of acidity and secreted enzymes, and a smaller mucosal surface area that does not disperse vaccines. The nasal mucosa has been recognized as a highly immune-competent area therefore only requires a smaller dose of antigen to elicit protective responses. Nasal vaccination has been shown to induce an immune response in the bloodstream, upper respiratory tract, and other effector sites including the lungs and the genitals [8]. However, the natural defense barriers such as nostril hair, cilia, mucus, and the keratinized stratified squamous epithelium can reduce the delivery efficacy of nasally administered vaccine antigens. Both oral and nasal immunization can target the Waldeyer’s ring, which is a large aggregate of mucosal lymphoid tissue consisting of tonsils and adenoids in the nasopharynx, and has drawn interest as a unique inductive site for B cell responses and plasma cell generation [109].

Other less explored mucosal immunization routes include the ocular, vaginal and rectal routes. The eye mucosa shares some common immunological features with the other mucosa. For example, the conjunctiva-associated lymphoid tissue (CALT) contains CD4⁺ and CD8⁺ T cells, mast cells, DCs and Langerhans cells [110,111]. Studies have provided evidence that the eye mucosa contains functional M cells with the capacity to uptake luminal antigens [112], and can be an effective and safe alternative immunization route against human papillomavirus (HPV) and influenza virus [113–115]. The vaginal mucosa has several different features compared to the oral and nasal mucosa discussed above. The male and female genital tracts possess only a few inductive mucosal sites therefore producing a lower humoral and cellular immune response upon infection [116]. However, several studies have demonstrated that vaginal vaccination can induce adequate immune responses consisting of both humoral and T cell-mediated immunity, especially in the context of sexually transmitted diseases [117–120]. Similarly, rectal administration offers unique advantages compared to the more conventional routes, such as avoidance of gastric degradation that challenges oral administration [121].

When choosing an immunization route for mucosal vaccination, one should be aware that the common mucosal immune system might be more restricted than previously thought, as apparent compartmentalization exists (Figure 2). For example, while oral immunization can induce substantial antibody responses in salivary glands, the small intestine, ascending colon and mammary glands, it is relatively less effective at eliciting an IgA response in the distal segments of the large intestines or female genital tract [122–124]. On the other hand, nasal immunization results in antibody responses in the upper airway mucosa, local secretions (including saliva and nasal secretions), and cervicovaginal mucosa, but less immune response in the gut [125]. Eye drop vaccination resulted in antigen-specific IgA in mucosal samples including tears, saliva, nasal washes and vaginal wash [110,115], whereas vaccination via intravaginal route or rectal route usually results in an immune response limited to the genital tract or colon and rectum, respectively [8]. Hence, mucosal compartmentalization should be taken into consideration for the choice of vaccination route in order to achieve effective immune responses at the desired sites.

Figure 2. Compartmentalization of the common mucosal immune system. Although different inductive sites can share the location of effector sites where IgA-secreting cells are seeded, the phenomenon of compartmentalization in the common mucosal immune system is evident. Different mucosal vaccination routes can produce an immune response in specific MALT areas, therefore exploitation of compartmentalization within the common mucosal immune system can direct the immune response to a particular site used by the pathogen for invasion, to effectively counter the development of infection.

Furthermore, it has been shown that the route of immunization not only determines tissue tropism of effector/memory cells, but also the tolerogenicity of the specific antigen. Stray et. al. demonstrated that intra-uterine but not intranasal inoculation of ultraviolet light (UV)-inactivated Chlamydia trachomatis (Ct) induced regulatory T cells that subsequently exacerbated susceptibility to new Ct infections in mice [126]. Although this tolerance could be reverted by complexing UV-Ct with a novel charge-switching synthetic adjuvant particle (cSAP) that targets the immunogenic CD11b⁺CD103⁻ DCs in uterine mucosa. Interestingly, only mucosally delivered cSAP-UV-Ct induced effector T cells that facilitated the rapid seeding of resident memory T cells (T_RM) in uterine mucosa that was not seen in subcutaneous injection of the same preparation [126].

4.2. Intrinsic difficulties in inducing mucosal immunity

As the mucosal tissues are constantly exposed to foreign substances including those harmless to the host (e.g. food substances and commensal bacteria), tolerance is the default response to mucosal antigens instead of immune activation. This immunotolerance is essential in the maintenance of mucosal homeostasis and avoidance of unnecessary inflammation, but also limits mucosal immunity [127]. Due to the relatively low levels of absorption and high chances of proteolytic degradation of the antigens at mucosal surfaces, larger amounts of vaccine antigens may be required to induce a sufficient immune response. However, prolonged and repeated mucosal exposure to antigens can in turn possibly induce mucosal tolerance [128]. Several regulatory mechanisms are in place to maintain the control of mucosal immune reactivity. However, precise knowledge regarding the ontogeny of different mucosal regulatory cells is still lacking, impeding the development of appropriate delivery systems and immunomodulating agents. It is known that the type of mucosal immune response is dependent on the nature of the antigen, the type of antigen-presenting cell and the local microenvironment [8]. Protein antigens generally induce a regulatory and/or inhibitory response that leads to tolerance, whereas natural infection or microorganism components such as toll-like receptor (TLR) ligands can induce potent cellular and humoral responses that overcome mucosal tolerance. Effective mucosal vaccination therefore should also involve the induction of a local innate inflammatory response, which can recruit additional unconditioned DCs that produce polarizing cytokines to induce an effector T and B cell response. This should manifest as a strong Th1, Th2 and/or Th17 cellular response along with the induction of an antigen-specific IgA response (Figure 3).

Figure 3. Tolerance versus immunity at mucosal sites. The nature of the antigen as well as the type of antigen-presenting cell determines the type of immune response induced at the mucosa. In general, protein antigens induce regulatory and/or inhibitory response which by default leads to tolerance. Tregs negatively regulate Th1 and Th2 cells by producing cytokines such as IL-10 and TGF-β, thereby maintaining homeostasis. On the other hand, adjuvanted antigen or natural infection can induce strong cellular and humoral immune responses and prevent mucosal tolerance. Adjuvants support the induction of effectors by triggering innate immunity, which recruits unconditioned DCs from other sites to initiate an antigen-specific immune response or acquired immunity.

While temporarily breaking tolerance is the goal for most mucosal vaccines, exploiting mucosal tolerance may be useful for some applications. Ilan et. al. induced tolerance to adenoviral proteins in experimental animals by oral ingestion. This resulted in markedly reduced anti-adenoviral humoral and cellular immunity, which was presented as increased TGF-β, IL-4 and IL-10 and decreased IFN-γ production by lymphocytes upon in vitro exposure to adenoviral antigens [129]. In the context of gene therapy using adenoviruses as delivery vectors, this intentionally induced tolerance to adenovirus can extend viral survival thus allowing for longer expression of the desired gene product encoded in a recombinant adenoviral vector. In addition, mucosally-induced tolerance may also provide therapeutic benefits in the treatment of preexisting autoimmunity and allergies [130,131].

4.3. Unanswered questions in mucosal immunology

The adaptive arm of mucosal immunity is relatively complex and less understood than its innate counterpart. A more complete picture of adaptive mucosal immunity is required to inform the design of effective vaccines that can launch a rapid adaptive immune response to combat fast replicating pathogens such as respiratory viruses. Compared to circulating IgA, mucosal IgA antibodies exhibit complexity in many forms. Serum IgA has a monomeric structure, whereas mucosal SIgA exists in dimeric to polymeric forms. These polymeric SIgA can consist of either IgA1 or IgA2 in humans, and the subclass proportions vary in different tissues. SIgAs also play a wide variety of roles, including nonspecific immune exclusion, pathogen-specific neutralization, and regulating homeostasis with commensal bacteria [132]. More research is required to harness this multifaceted immunoglobulin isotype for potential prophylactic or therapeutic applications. As for the cellular immune component, the importance of T cells in mucosal immunity has long been acknowledged but strategies to elicit appropriate T helper responses are still lacking. A well-studied aspect of different T cell types is that they target pathogens of different nature. Th1 cells secret IFN-γ and promote cell-mediated immunity against intracellular bacteria and viruses, whereas cytotoxic T lymphocytes (CTLs) eliminate virus-infected cells; Th2 cells secrete IL-4, IL-5 and IL-6 to assist B cells in producing neutralizing antibodies that can effectively recognize extracellular pathogens and toxins; finally, Th17 cells produce IL-17 and IL-21 that are important for defense against extracellular bacteria [133,134]. Therefore, adjuvants that can specifically promote a certain type of T cell response will enhance protection against particular diseases.

APCs such as DCs play an important role in influencing T cell differentiation. Although studies have reported that CX3CR1⁺ DCs promote Th1/Th17 cell differentiation, mucosal DCs preferentially induce Th2 cell differentiation and induce IgA-secreting B cells, indicating that Th1 responses may be difficult to induce by the interactions between the mucosal vaccine and mucosal DCs [135]. Therefore, adjuvants that can successfully recruit ‘unconditioned’ DCs from other sites could in theory facilitate the induction of protective Th1 and Th2-type responses at the mucosal sites. Compared to Th1 and Th2 responses, the role and scope of Th17 cells and their IL-17 cytokine production upon infection or vaccine-induced immunity still await more understanding. Reports on the involvement of IL-17 in vaccine-induced protection and IgA secretion suggest that IL-17 may be a novel indicator of mucosal immunity [136,137]. Cytokines produced by Th17 cells have been shown to induce inflammatory responses, increase neutrophil recruitment, modulate neutrophil homeostasis, and thus play an important role in controlling many bacterial and fungal pathogens [138,139]. Progress has been made to understand the effect of Th17 responses in facilitating protective immunity produced by vaccination [140–142]. To this end, the mechanism of adjuvant-induced Th17 immunity has also been investigated in a nanoemulsion adjuvant [143]. However, data regarding mucosal immunity remains limited.

Memory T cells play an important role in durable protection and immune surveillance. The most recently identified subset are the tissue-resident memory T (T_RM) cells, which reside locally at mucosal tissues and provides frontline defense against different pathogens [144]. CD8⁺ T_RM primarily target neoplastic cells and intracellular pathogens through cytotoxic lysis [145–147]. On the other hand, CD4⁺ T_RM cells are active in the protection against various respiratory and sexually transmitted infections via the provision of helper functions [148–151]. Intranasal vaccination studies have demonstrated that successful induction of CD4⁺ T_RM in animals conferred protection against pneumococcal colonization, influenza challenge and SARS-CoV infection [152–154]. This indicates that knowledge obtained from T_RM studies can inform mucosal vaccine strategies that are able to induce long-lasting responses. However, many biological features such as the differentiation, maintenance, and plasticity of T_RM cells remain unclear, partly due to the difficulties in accessing T cells as blood sampling can only capture circulating T cells but not resident T cells [155].

4.4. The human mucosal immune system is difficult to model

Although many principles of mucosal immunology are effective in animal models, the same phenomena often fail to be established in humans. This can be attributed to the vast differences in anatomy, physiology and immunity between different species, making it difficult to extrapolate results from animal models to humans [3]. For example, the mouse vaginal mucosa is a keratinized tissue, whereas human vaginal mucosa has a non-keratinized surface and therefore tends to be more permeable and less resistant to damage [156]. The important innate immune receptor TLR exists as TLR1-TLR10 in humans; in contrast, mouse cells express a different set of TLRs, namely TLR1-TLR9 and TLR11-TLR13 [157]. Humans have two IgA subclasses which are evenly distributed in mucosal tissues, whereas mice only express a single IgA isotype and do not have any functional FcαRI homologue [158]. Furthermore, differences in the size, number, distribution and composition of mucosal lymphoid components such as Peyer’s patches in different species can also have a fundamental effect on the production of immune responses [159].

Furthermore, as opposed to highly controlled settings in animal models, variability in human subjects, for example, in the commensal flora, nutritional status and past immunological experience, have all been found to affect mucosal vaccine efficacy [8]. This might partially explain the different efficacies of several mucosal vaccines seen in low- and middle-income countries versus high-income countries. The ‘tropical barrier,’ where responses to oral vaccines in low- and middle-income countries have been found to be lower than those in high-income countries, presents a challenge that requires additional intervention such as probiotic supplements for controlling and/or optimizing commensal microflora [160,161].

The implications of the mucosal flora in mucosal immunization have drawn attention since increasing information on the commensal microbiome has become available. Interactions between different microbes within the mucosal microenvironment affect the level of nutrients, toxins, metabolic end-products and anti-microbial compounds. Furthermore, as the microbiota outnumber our own cells by 10-fold, this constant contact with microbes stimulates a large quantity of intestinal lymphoid cells and IgA (3–5 g per day) at any given time of our lives [162,163]. This dynamic and complex ecosystem can affect the efficacy of mucosal vaccines, as the vaccine antigen essentially needs to compete for the attention of the mucosal immune system with existing numerous microbial antigens in the gut [164]. For example, an antigen delivered by a vector composed of a commensal bacterium may be subject to the endogenous regulation imparted to the vector. On the other hand, a system using an attenuated pathogen vector may induce a stronger immune response to the vaccine antigen, but this initial response to the neoantigen may be overwhelmed by the response to the more immunodominant antigens of the vector [128].

4.5. Technical constraints in assessing mucosal immune responses

Most vaccine trials rely on the measurement of serum antibody levels as a correlation of efficacy and protection. However, this usually does not reflect mucosal secretory antibody levels. Mucosal immunization is noninvasive, however, ironically the sampling methods for mucosal immune responses often require more invasive procedures. Serum antibodies and blood cells are readily sampled, while secretory antibodies and mucosal effector cells are relatively difficult to obtain. The local variability of mucosal tissues and secretions add further challenges to the accurate quantification of comprehensive mucosal immune responses [24]. Several methods have been developed in research laboratories for measuring mucosal vaccine-specific antibodies or T cell responses [24]; however, standardized and validated methods for use in clinical settings remain limited. The recently developed nasal and bronchial fluid sampling techniques provide a noninvasive technique, which has been increasingly used in various studies [165,166]. Establishing robust correlates of vaccine-induced adaptive immunity is a priority in the current research scene [117,167]. For example, salivary IgA has been shown to highly correlate with intestinal IgA responses in an ETEC challenge study [168]. The assessment of cellular mucosal immune responses is often restricted by access to tissues, and sampling from mucosal sites can be invasive and unpleasant. An alternative is the measurement of vaccine-specific antibody-secreting cells (ASCs) that carry mucosal homing markers (e.g. integrin α₄/β₇or CCR9) in the circulation, which are considered a better representation of local immune responses [169,170]. These effector cell populations can be characterized by flow cytometry or ELISPOT assays [171], or their culture supernatant can be analyzed by ELISA or multiplex assays [172].

4.6. Regulatory hurdles that restrict mucosal vaccine development

Effective mucosal vaccination requires the use of carriers and adjuvants; however the safety of exposing the sensitive mucosal tissues to these compounds remains a major concern. A well-known example are the enterotoxin adjuvants cholera toxin (CT) and heat-labile toxin (LT). CT has been successfully used in mice to promote antigen delivery via endocytosis and transportation by M cells, which eventually leads to the stimulation of DCs [173]. LT exhibits a similar immunostimulatory effect as CT, and has been extensively studied for its adjuvant activity in mice [174]. Despite these promising attributes as a vaccine adjuvant, toxicity of CT and LT has limited their applications [175]. In human volunteers, as little as 5 micrograms of purified CT could induce serious diarrhea [176], and just 2.5 micrograms of LT was sufficient to cause fluid secretion [177]. Significant efforts have been made to dissociate these molecules’ enterotoxicity from adjuvanticity. For example, several mutants created by alternating the active site and the protease site have been shown to have lower toxicity while retaining adjuvant activities [178,179].

The risks of vaccine-induced toxicity are even more pronounced for nasal administration due to the possibility of the vaccine passing into the brain through olfactory nerves. A well-known example is Nasal flu, an experimental inhaled influenza vaccine which has been associated with Bell’s palsy (a condition that causes temporary facial paralysis) [106]. This unfortunate side effect was speculated to result from the retrograde transportation of the LT adjuvant to the brain through olfactory nerves and led to the termination of the vaccine’s development [106]. In nasal mucosa, the universally expressed ganglioside receptors offer a direct portal to the central nervous system (CNS), thus permit enterotoxins such as CT and LT to be transported into the olfactory bulb [175]. Modification of LT and CT, such as the use of the CTB subunit alone, can significantly reduce the adverse neurological effects [175,180]. Due to the risk of live viruses entering CNS from the olfactory neuroepithelium [181–185], there is a need to evaluate the neurotropism of live-attenuated viruses for nasal vaccine development. Advanced strategies such as the neurotropic viral vector delivery system developed by Jerusalmi et al. have demonstrated successful delivery of a protein without any viral RNA detected in the olfactory bulb [186]. Similarly, Kong et al. reported that the cationic nanogel-based nasal vaccine effectively delivers the pneumococcal antigen PspA to DCs without CNS accumulation in mice and monkeys [187,188].

5. Expert opinion

Traditional vaccination strategies have been largely focused on generating strong systemic antibodies and cellular immunity that help control the worsening of infectious diseases. Increasing research in mucosal immunology and vaccinology has prompted considerations for adding mucosal immunity to the current vaccination paradigm. The main attraction of using the mucosal route for vaccination is that mucosal vaccination has the potential to induce immunity that can either prevent the attachment and colonization of the infectious agent at the mucosal epithelium (for noninvasive bacteria), or from penetrating and replicating in the mucosa (for invasive bacteria and viruses), and/or block microbial toxins from entering and affecting epithelial and other cells [8]. However, mucosal vaccines face a different set of hurdles compared to their injectable counterparts. The unique biological architecture and immunological features of the mucosal immune system require different vaccination strategies as opposed to systemic immunization. The underlying mechanisms for the generation of effective vaccine-induced immune responses in mucosal tissues are complex, and can be different from one mucosal surface to another. Advancement in mucosal vaccine development can only be achieved with a better understanding of the molecular and cellular innate and acquired immune mechanisms and regulation of adaptive mucosal protection. The optimization of mucosal vaccine formulations requires further understanding of mucosal APC subsets and their functions, crosstalk between mucosal compartments, the induction and maintenance of residence memory effector cells, as well as the mechanisms and development of IgA B cells and Th17 responses.

Recently there have been significant advances in the development of mucosal vaccine delivery systems and mucosal adjuvants, although their application in humans still awaits to be established. This is partly due to the fact that experimental methodology in this relatively young discipline is limited. Methods that facilitate the monitoring of mucosal immune responses in humans have only been partly developed, and standardized protocols for generating reproducible and comparable results are not widely available. Existing assays are primarily for measuring secretory antibody responses, whereas practical methods for evaluating mucosal cellular responses are still lacking. The development of accurate and reproducible measurement of mucosal immune responses is key for the selection and approval of vaccine candidates. This is even more critical in the neonatal, pediatric and aged populations, as they are the at-risk populations for infections that need effectively and safely induced protection. Furthermore, mucosal vaccines that can generate robust and persistent mucosal immunological memory in infants and young children will make a huge difference in reducing health burdens associated with infectious diseases.

Compared to the advancement in parenteral vaccination, mucosal vaccines have significantly lagged in development. This is evidenced by the fact that the most recently licensed mucosal vaccine was approved more than 10 years ago. In addition, all of the currently available mucosal vaccines were built upon the older live attenuated/inactivated whole-organism approach and utilize the more traditional oral route. The recent COVID-19 outbreak and the increasing emergence of new pathogens present both a challenge and an opportunity to update knowledge and approaches in current vaccinology. Mucosal vaccination offers an alternative to improve or complement traditional parenteral approaches. The ease of administration and relative cost-effectiveness can also reduce logistical and economic burden, while increasing public acceptance and adherence. Furthermore, emerging studies have provided a rationale for the effectiveness of mucosal immunization in combatting epithelial cancers, modulating autoimmunity and dampening allergic responses. The combination of parenteral and mucosal immunization also offers promise in providing superior protection compared to current regimens. In this sense, the potential advantages of mucosal vaccination largely outweigh the above-mentioned obstacles, therefore continuous efforts in exploring innovative mucosal vaccination strategies should be encouraged.

Article highlights

• Successful mucosal vaccination can lead to the induction of both protective mucosal immunity and sterilizing systemic immunity. This double-layered guard not only can protect the vaccinee but may also reduce the potential of interpersonal transmission.

• Recent scientific advancement in the molecular and cellular understanding of the mucosal immune system opens new doors for the development of novel mucosal vaccines. New adjuvants, delivery systems, and formulations have propelled the progression of several mucosal vaccines and showed promise in pre-clinical and clinical studies.

• Mucosal vaccine development still faces several significant hurdles. These include the intrinsic difficulties in inducing mucosal immunity, the incomplete understanding of the mucosal immune system, the lack of reliable experimental tools and the heightened regulatory restrictions resulting from past incidents.

• Apart from the ability to induce both local and systemic immune responses, mucosal vaccination also offers additional benefits in safety and versatility. Therefore, the outstanding obstacles are worth constant efforts, as they are largely outweighed by the proven advantages.

Declaration of interests

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

Catherine JY Tsai: conceptualization; writing – original draft; writing – review and editing. Jacelyn MS Loh: creation of figures (figures created with BioRender.com); writing – review and editing. Kohtaro Fujihashi: Funding acquisition; supervision; writing – review and editing. Hiroshi Kiyono: Funding acquisition; writing – review and editing.

Additional information

Funding

The authors involved in this review are supported by the IMSUT International joint research program [K22-3062] (to C. J.-Y.T.), [K23-3044] (to J.M.S.L.); Agency for Medical Research and Development (AMED)-SCARDA [223fa627003h0002] (to KF and HK); Japan Society for the Promotion of Science (JSPS) Postdoctoral Fellowship P20801 (to C. J.-Y.T.); JSPS Grant-in-Aid for Scientific Research Postdoctoral Fellowship (22F20801 to K.F.); JSPS Grant-in-Aid for Scientific Research B (20H03856 to K.F.); JSPS Grant-in-Aid for Scientific Research Challenging Research Pioneering (20K20495 to K.F.); University of California San Diego (UCSD) Digestive Diseases Research Center (DK120515 to H.K.); and a 3 M donation (to H.K.). Sponsors had no control over the interpretation, writing, or publication of this work.