|Home | About | Journals | Submit | Contact Us | Français|
Vaccination as a medical intervention has proven capable of greatly reducing the suffering from childhood infectious disease. However, newborns and infants in particular are age groups for whom adequate vaccine-mediated protection is still largely lacking. With the challenges that the neonatal immune system faces and the required highest level of stringency for safety, designing vaccines for early life in general and the newborn in particular poses great difficulty. Nevertheless, recent advances in our understanding of neonatal immunity and its responses to vaccines and adjuvants suggest that neonatal vaccination is a task fully within reach. Among the most promising developments in neonatal vaccination is the use of Listeria monocytogenes (Lm) as a delivery platform. In this review, we will outline key properties of Lm that make it such an ideal neonatal and early life vaccine vehicle, and also discuss potential constraints of Lm as a vaccine delivery platform.
It has been over 200 y since the breakthrough of vaccination as a life-saving tool.1 Yet despite great contribution by vaccines to the reduction of mortality and morbidity, the number of lives claimed by infectious diseases among newborns and infants under 6 mo of age still amount to more than 2 million every year.2 This is partly due to the fact that there are presently only 3 vaccines licensed for at-birth administration;3 for many other diseases that strike early, there still exists no vaccine.
Newborns have a higher risk of suffering from infection than any other age group.4 This is believed to be attributable to inherent limitations of the newborn’s immune system as compared with adults.5 However, a number of studies have now shown that given the appropriate stimulus, the neonate is indeed capable of generating an adult-like immune response,6-8 suggesting the possibility that newborns can respond to vaccination in a manner similar to their adult counterparts.4
Listeria monocytogenes (Lm) is a facultatively anaerobic, non-spore-forming, and gram-positive microbe. With a tolerance for low temperatures, high salt concentrations, and alkaline environments, Lm is able to survive and replicate in various niches, including food products–contamination of which is a common source of severe Listeria infection in humans.9Lm also possesses a host of devices that facilitate its pathogenesis, enabling it to effectively enter into and move within target cells, avoid autophagy, and spread cell-to-cell.9Lm has a relatively low infection rate in the general population (ranging from 0.1 to 11.3 per million people) despite its ubiquity.10 In addition, Lm displays non-uniformity in its age distribution, causing infection mostly in newborns or the elderly, with a disproportionately high mortality rate of 20–60% in those groups. Natural infection by Lm normally presents asymptomatically in the healthy host; clinical symptomatic infection by Lm on the other hand, known as “listeriosis,” may exist as a non-invasive form that develops as a febrile gastroenteritis, or an invasive form characterized by the systemic spread of Lm targeting mainly the placenta, liver and spleen, and the central nervous system.10 Clinical features of listeriosis in children most commonly include meningitis or septicemia, although a number of other complications may also occur.10
The serious impact that Lm has as a pathogen in newborns could preclude its consideration as a neonatal vaccine system. However, some of the very same aspects that make Lm such a formidable pathogen also make it a powerful vaccine delivery platform. Recent years have not only seen the establishment of Lm as a practical and versatile approach to vaccination in adults, but also the generation of very promising results as a neonatal vaccine vehicle. Following the success of increasing Lm’s safety profile while maintaining its potent immune stimulatory characteristics, we propose that Lm holds great potential as a vaccine vehicle for the human newborn.
The concept of Lm-based delivery of antigens first emerged at the University of Pennsylvania when Dr Yvonne Paterson demonstrated the ability of a recombinant Lm strain expressing and secreting the influenza virus nucleoprotein (NP) to efficiently access both major histocompatibility complex (MHC) class I and II pathways.11 Moreover, the Lm-NP vector induced an antigen-specific and T cell-dependent anti-tumor response that not only protected immunized mice against lethal challenge by NP-expressing tumors, but also caused the regression of established tumors.12,13 This suggested that Lm as a delivery platform effectively induces both prophylactic as well as therapeutic immunity.
The development of Lm as a vaccine delivery platform against neoplastic disease has given rise to a notable repertoire of recombinant Lm strains capable of expressing and secreting a variety of tumor-associated antigens, a number of which have entered clinical-stage testing against a wide range of cancers; several of these Lm-based vectors have advanced to Phase II trials.14,15 One such strain expresses the E7 antigen of human papillomavirus (HPV) as a fusion protein with listeriolysin O (LLO), and is used against HPV-associated cancers such as cervical, head and neck, or anal cancer.16 Another Lm strain expresses human mesothelin and is being explored as an immunotherapeutic agent to treat pancreatic cancer.17 In preclinical studies using a murine prostate tumor model, it has furthermore been shown that the combination of Lm-based PSA (prostate specific antigen) immunotherapy and radiation therapy was more effective for the treatment of established tumors than either of the 2 treatment modalities used alone, with the combinatorial treatment demonstrating a multifold increase in PSA-specific cytotoxic T lymphocytes (CTLs) as well as a significant increase in CTL infiltration of tumor tissues.18 The immunotherapeutic efficacy of Lm-based vectors has been further improved when combined with anti-PD-1 antibodies.19 In a metastatic model of hepatic colorectal cancer, an AH1 peptide-expressing Lm strain elicited potent cytotoxic tumor-specific CD8+ T cell responses that could cure mice of hepatic metastasis and subsequently protect from tumor re-challenge.20 Recently, an Lm strain expressing and secreting a fusion of multiple peptides of the hepatocellular carcinoma (HCC)-related tumor-associated antigen was shown to slow the development of HCC in both prophylactic and therapeutic settings, demonstrating also the induction of strong cytotoxic T cell immunity against each epitope in the fusion peptide; this supports what we know regarding Lm’s ability to overcome self-tolerance to an endogenous antigen.21 While Lm is indeed capable of delivering a wide selection of proteins into the cells it infects, the versatility of Lm as a bacterial carrier system is further seen in its ability to also transport DNA into mammalian target cells, with the successful delivery of pro-drug converting enzymes by Lm into tumor cells having been reported.22
Lm-based immunotherapy has also been shown to severely reduce the suppressive activity of myeloid-derived suppressor cells as well as regulatory T cells in the tumor microenvironment (TME), thereby lessening the impairment of T cell function; this modulation of immunosuppression in the TME is an inherent property of all Lm-based vaccines.23 With a proven capability to target tumors both effectively and safely, as well as the surprising success observed when combined with various forms of cancer therapy (e.g., bacterial-based immunotherapy with chemotherapy), Lm-based immunotherapy has made significant contributions in the fight against cancer.
As a vaccine vehicle against infectious disease, immunization with recombinant Lm has been shown to protect against immunosuppressive strains of lymphocytic choriomeningitis virus (LCMV).24 An Lm strain expressing the human immunodeficiency virus (HIV) Gag protein was capable of inducing antigen-specific CD4+ T cells with a Th1 interferon-gamma (IFN-α)-producing phenotype, and immunized mice displayed accelerated clearance of viral challenge with a Gag-expressing vaccinia virus construct.25,26 Lm-based vaccination has also succeeded in inducing robust and functional HIV-specific cellular immune responses in mice regardless of underlying chronic infection status (e.g., helminth infections).27 Against the parasite Leishmania major, immunization of mice with recombinant Lm delivering an L. major antigen has been shown to generate in vivo CD4+ T cells with a Th1 phenotype that exerted protective antiparasitic function.28 Lastly, a recombinant Lm vaccine stably expressing the IglC protein of Francisella tularensis was found to induce protective immunity against lethal F. tularensis challenge in mice; this success was shown to be due to the powerful cellular CD4+ and CD8+ T cell immune responses against IglC.29 Moreover, mice immunized with this strain showed protection even against lethal intranasal challenge by F. tularensis LVS, i.e., mimicking the natural route of infection for humans (airborne).29
Cognizant of the safety concerns regarding neonatal vaccination, one of the first studies exploring Lm as a neonatal vaccine sought to evaluate the safety and immunogenicity of a hyperattenuated Lm strain expressing the HIV Gag protein, as well as the protection it could afford against challenge by a Gag-expressing recombinant of vaccinia virus.30 This Lm strain given to neonatal mice was found to be safe, and with the administration of a booster dose, also able to initiate a protective CD8+ cytolytic T cell response. The emergence of yet more promising results came several years later, when neonatal mice that were immunized at birth with only one dose of the attenuated strain ΔactA-Lm achieved protective immunity from severe Listeria infection.31 The same study not only demonstrated the induction of a T cell response in neonatal mice similar in kinetics to that displayed by their adult counterparts, but also identified the generation of robust and sustained Th1 CD4+ and CD8+ Lm-specific T cell memory responses, along with strong antigen-specific primary and memory CD8+ T cell responses.31
It was subsequently shown that Lm-immunized neonatal mice required only a single dose to reach maximal antigen-specific CD8+ T cell expansion, whereas adult mice required a booster dose; antigen-specific CD4+ T cell expansion on the other hand, required a boost to reach its peak, in both neonates and adults.32 With the ability to generate substantial protection and immune memory with only a single dose, Lm thus circumvents the requirement for boosting that subunit or inactivated vaccines often face.32 Specifically, a single administration of an attenuated Lm vaccine in mice during the first week of life was shown to induce protection from lethal challenge as early as 1 wk following immunization, lasting the lifetime of the mouse without the need for additional booster doses.33 In addition, mice immunized as neonates or adults displayed no difference in the functional avidity, sensitivity and T cell receptor Vβ(TCR-Vβ) repertoire of their antigen-specific T cells; receiving immunization as a neonate also did not compromise on protection or preclude booster responses with the same Lm vector at a later point in time.32 “Original antigenic sin” (a phenomenon where the specific immune response mounted against the initial antigen prevails even during subsequent infection by dissimilar variants of the antigen34) appears unlikely to impact the use of live Lm for neonatal vaccination, given that the introduction of new antigens in neonates via Lm vectors has been shown to allow for induction of protective cell-mediated immunity to the newly-introduced, while maintaining responsiveness to the original antigens.32,35 Furthermore, immune memory responses to specific multiple antigens in a live Lm-vectored vaccine were all found to last the lifetime of the mouse whether they were immunized as newborns or adults.33 Attenuated Lm has also been shown to clear from the host within 1 wk of administration in both neonates and adults, and would therefore be unlikely to alter the normal cytokine milieu for prolonged periods of time.33
Taken together, these data indicate that neonatal vaccination with Lm achieves at least similar—if not better—immune responses as compared with Lm immunization in adults, strengthening the notion of Lm being highly suitable as a neonatal vaccine delivery platform. The known pathogenicity of Lm, however, has meant that safety remains an issue of major concern.
Lm as a pathogen is well endowed with a range of virulence factors that enable it to cause disease. Many of these virulence factors have been targeted in ways that allow for attenuation while retaining immunogenicity. Understanding the functions of these virulence factors as well as their roles in pathogenesis is a central theme in the design of vaccines. Here, we briefly review the pathogenesis of Lm and current attenuation strategies (Table 1) aimed at providing optimal vaccine safety for this live vaccine vector.
Lm infection begins with oral ingestion of the bacteria.36,37 In the human gastrointestinal tract, Lm moves across the mucosal barrier by first adhering to the mucosal lining via the bacterial protein Ami;38 this is followed by entry into intestinal target cells either through phagocytosis or through the action of listerial internalin A (InlA).39 Transcytosis across the intestinal epithelium follows, after which Lm is released into the lamina propria by exocytosis and disseminates systemically.40
Lm primarily targets the liver, with the help of bacterial adhesin FbpA.41 FbpA binds to human fibronectin expressed on the surface of hepatocytes.41 Another listerial internalin, InlB, next binds to the host hepatocyte growth factor receptor (a tyrosine kinase receptor, Met) or complement component C1q receptor,42 in turn mediating efficient entry into hepatocytes (as well as fibroblasts and epithelioid cells).43 However, invasion efficiency is much higher in phagocytes, especially macrophages and monocytes.9 Macrophages and dendritic cells (DCs) represent the main carriers of the bacterium in infected tissue.44 The uptake of Lm by phagocytes occurs via the binding of scavenger receptors on host cells to lipoteichoic acid (a component of the listerial cell wall),45 or the binding of certain listerial cell surface components to host cell complement receptors, e.g., InlB.42 In order for Lm to avoid destruction inside the phagocyte, it has to escape from the phagocytic vacuole.46 Apart from several host factors involved in this process,46 the 3 bacterial factors most clearly responsible for enabling the escape of Lm into the cytosol are LLO and 2 phospholipase C enzymes, PlcA and PlcB. LLO is a secreted cholesterol-dependent cytolysin (CDC) toxin. In mice, LLO is essential for bacterial escape from primary vacuoles as well as secondary double-membrane vacuoles formed during cell-to-cell spread.47,48 In humans, the bacterial factors critical for phagosomal vacuolar escape are instead, the 2 phospholipases, PlcA and PlcB.9 PlcA is required for lysis of the phagosome formed during initial entry into the host cell,49 and PlcB is required for efficient escape from the secondary vacuole;50 the 2 phospholipases work synergistically with LLO in allowing Lm to escape into the cytoplasm.51,52 Once in the cytoplasm, Lm is able to replicate intracellularly by usurping nutrients provided by the host. Lm next induces polymerization of host actin filaments to move in the cytoplasm and spread from cell to cell;53,54 the only determinant that Lm requires for its actin-based motility is the bacterial surface protein ActA.55,56 The genes encoding Lm’s most prominent virulence-associated proteins (LLO, ActA, PlcA and PlcB) are located adjacently in a 9.6 kb virulence gene cluster,57 chiefly regulated by a pleiotropic virulence regulator, PrfA (a protein encoded by prfA).36,58
The selective and irreversible deletion of classical virulence factors is perhaps the most direct means of attenuating Lm.59,60 An Lm strain lacking prfA recently demonstrated a high level of safety in a Phase II clinical trial in patients with recurrent cervical cancer.61 Another Lm strain generated via the deletion of both actA and inlB exhibited diminished toxicity in vivo, primarily from impediment of the direct InlB-mediated infection of non-phagocytic cells as well as the reduction of ActA-mediated cell-to-cell spread from adjacent phagocytic cells; there was no compromise on Lm’s ability to infect phagocytic cells, which meant that immunogenicity was retained.62 In addition, actA deletion accompanied by deletion of plcB generated a strain that could be administered orally to adult volunteers without any adverse health sequelae.63
Auxotrophy has also been explored as a means of attenuation, where the Lm mutants generated require exogenous factors for in vivo and in vitro growth.59 The inactivation of 2 genes, dal and dat, has resulted in Lm strains that grow only when supplemented with d-alanine (a cell wall component in virtually all bacteria64,65).66 Consequently, these strains are unable to grow in the cytoplasm of eukaryotic host cells, Lm’s natural habitat during infection. Nevertheless, immunogenicity has remained as evidenced by the induction of T-lymphocyte responses and protective immunity against lethal challenge by wild-type Lm.66 A shuttle vector was designed containing a copy of the Lm dal gene, which could complement the growth of the Lm dal dat mutant both in vivo and in vitro.67 However, anticipated concerns over the recombination of plasmid with bacterial chromosomes led to the dalLm-containing plasmid being replaced with a plasmid containing an Lm actA promoter-regulated resolvase gene and the dal gene from Bacillus subtilis (dalBs) flanked by 2 res1 sites; this allowed the highly-regulated and transient expression of the dal gene upon exposure to the host cell cytosol, without the risk of reversion to a virulent microbe via recombination.68 Furthermore, the introduction of an irreversible deletion in the actA gene has resulted in a strain that is dal dat ΔactA attenuated and complementable by the dalBs-based antibiotic-free plasmid.69
Other approaches to attenuation have included deletions in aro genes, a family of genes belonging to the common branch of the biosynthesis pathway leading to aromatic compounds affecting oxidative respiration.70 Lm aro mutants displayed drastically reduced rates of cytosolic replication and cell-to-cell spread, yet retained immunogenicity.70 Mutation in the glcV gene of Lm was found to preclude the binding of certain listerial phages and produce profound attenuation by mechanisms that have yet to be elucidated, although it is speculated that glcV mutation leads to alterations in the listerial phage receptor and therefore interferes with normal host-pathogen interactions required for virulence.71 Such a strain, when administered orally to mice, induced robust and long-lasting protective immunity despite the near absence of vital organ infectivity.72 By removing genes required for nucleotide excision repair (uvrAB) and thus rendering Lm highly sensitive to photochemical inactivation, a Killed But Metabolically Active (KBMA) strain of Lm was designed that is unable to replicate yet still having sufficient metabolic activity for delivering antigens to the immune system.73 It was later on demonstrated that constitutive activation of prfA leads to the enhanced ability of KBMA Lm to induce protective cellular immunity.74 Deletion of the frvA gene resulted in an Lm strain incapable of iron homeostasis and strongly attenuated in virulence, yet retaining the ability for intracellular growth in antigen-presenting cells. Furthermore, immunization with the ΔfrvA mutant was found to offer complete protection from listerial infection.75 A self-destructing mutant of Lm achieved via the expression of a Listeria-specific phage lysin has also been described; however, safety concerns over the integration of the relevant Lm plasmid DNA into the host cell genome have prevented further development.76
Finally, forcing the expression of flagellin (flaA) or PrgJ (from the Type III secretion system of Salmonella typhimurium) by Lm in the host cell cytosol has been employed to specifically target the Caspase-1 activation pathway, bringing about the preferential clearance of bacteria via activation of the NLRC4 inflammasome.77 Although this strategy was indeed confirmed to bring about a high degree of attenuation in Lm,78 its impact on immunogenicity is not clear–with one group demonstrating the induction of protective immunity in mice against lethal challenge with Lm,77 and another showing instead, a poor induction of protective immunity.78 This conflicting data may be due to the different methods used in these 2 studies to increase flagellin expression, as well as the difference in bacterial species from which flagellin was taken.
Among the array of attenuation strategies that exist for Lm-based vectors, perhaps the method that would best achieve safety for use in the neonate is the irreversible deletion of Lm virulence genes. The major genes that have been shown to be important for the virulence of this pathogen include actA, plcA, plcB, hly, prfA, inlA and inlB.59 Of these, the deletion of actA, which is responsible for cell-to-cell spread of Lm, causes at least a thousand-fold attenuation while still retaining immunogenicity.79 The deletion of hly or prfA completely eliminates the ability of Lm to grow intracellularly, reducing immunogenicity to a level that hinders its use as a vaccine platform.59 The presence of other deletions in an actA mutant strain (e.g., Lm dal dat ΔactA; Lm ΔactA/ΔInlB) has been demonstrated to further enhance the overall attenuation to Lm. With actA deletion having already been proven safe yet effective in newborn mice, coupled with the additional credibility ascribed to it by its utility in multiple Lm vectors (Table 1), such a strategy, when employed with efforts to further reduce the risk for reversion to wild-type Lm (e.g., dal dat deletion), currently represents the most favorable option for Lm-based neonatal vaccination.
Proof-of-concept for safety of neonatal immunization using live-attenuated vaccines has been established with the BCG and oral polio vaccines.3 Moreover, while it impossible to predict at this stage if current strategies of Lm attenuation would definitely be safe for the human newborn, the currently available data summarized above strongly suggest this to be the case.30,31,33
Vaccine adjuvants are agents that serve to activate the innate immune system, directing the quantity and quality of the adaptive immune response following an antigen-specific stimulus.80 The inclusion of adjuvants has been key to the efficacy of most vaccines given early in life, which are often subunit vaccines that lack inherent adjuvant activity required for the generation of a favorable immune response.81 The increasing attention that Lm has been receiving as a vaccine vector is in part due to its adjuvant activity, i.e., its ability to elicit a strong innate immune response. The potent and predictable immune modulatory activity of Lm allows for the induction of robust Th1-type cell-mediated immunity by Lm-based vaccines. The capacity to function as a powerful adjuvant relates to the specific signaling pathways that Lm activates.
Lm infection of antigen presenting cells (APCs) results in the activation of at least 3 distinct bacterial recognition pathways82: 1) a TLR/MyD88-dependent pathway that induces the expression of inflammatory as well as suppressive/regulatory cytokines (such as TNF-α, IL-12 and IL-10), autophagy and production of reactive oxygen species (ROS);83-85 2) a STING/IRF3-dependent pathway that leads to expression of interferon (IFN)-β and co-regulated genes;86 and 3) an AIM-2/Caspase-1-dependent inflammasome pathway that results in proteolytic activation and secretion of IL-1β and IL-18, in addition to pyroptotic cell death.87 Importantly, activation of STING/IRF3- or AIM-2/Caspase-1-dependent pathways require that Lm be alive and escape into the cytosol.88,89
Activation of the 3 pathways mentioned above by Lm initiates the innate immune response in monocytes and macrophages.36,82 The release of IL-6 by infected cells leads to recruitment of neutrophils to the site of infection, which in turn destroy extracellular bacteria, digest apoptotic cells and secrete chemokines that recruit monocytes/macrophages.59 Infected macrophages also produce IL-12, inducing the synthesis of IFN-γ by NK cells and bystander CD8+ T cells; IFN-γ subsequently activates macrophages to become listericidal through the production of ROS and reactive nitrogen species (RNS).90,91 Lm is also known for its ability to induce type I IFNs (e.g., IFN-α and IFN-β), which although typically associated with anti-viral immune responses and essential for the clearance of many intracellular pathogens, have also been suggested to be detrimental to the host during the immune response to Lm.92,93 Lm infection also activates autophagy, where the formation of a double membrane vacuole around cytosolic Lm leads to subsequent degradation of the bacterium via the lysosomal pathway.94 In DCs, Lm induces the release of IL-2; IL-6; IL-12; and TNF-α, as well as the subsequent upregulation of other proteins (e.g., CD40, PD-L1) that promote the maturation and activation of high-affinity T cells.59,95
The adaptive immune response against Lm infection serves 2 main functions: the specific lysis of infected cells and the rapid secretion of IFN-γ in response to innate production of IL-12 and IL-18.96,97 IFN-γ is a crucial contributor to the cell-mediated immune response via macrophage activation; the increasing of antigen presentation via the MHC class I and II pathways; and the inhibition of Th2 cell expansion.91 During infection, Lm secretes a limited number of proteins into the cytosol of the host cell, which when rapidly degraded by the proteasome, generate peptide fragments that enter the MHC class I antigen processing pathway.59,98 Lm infection also generates a robust MHC class II-restricted CD4+ T cell response and drives CD4+ T cells toward a Th1 phenotype;25,59,99 occurring simultaneously is the expansion of CD8+ T cell responses.59,100 A strong and lasting Th1 cell-mediated response dominated by CD8+ T cells and made optimal by CD4+ T cells is what drives protective immunity to Lm infection101,102 and ensures complete and final clearance of the microbe, with humoral responses playing only a minimal role.91
The particularly high risk of severe outcome from Lm infection in the newborn suggests that possible differences exist between the immune signaling pathways activated by Lm in neonates vs. the adult.103,104 The importance of the MyD88-dependent pathway for host resistance is clearly seen from the extreme vulnerability displayed by MyD88-deficient mice to Lm infection.105,106 Even though the ability for innate recognition of pathogens via the TLR/MyD88 pathway early in life does not appear to be different compared with adults,107 there are stark contrasts between the downstream effector responses generated in the human newborn as compared with the young adult103,104 (discussed in the following section). The TLR/MyD88-dependent response of human neonatal monocytes specifically to Lm infection has yet to be explored.104
While IRF3-dependent production of type I IFN via the STING/IRF3-dependent pathway in human newborns is reduced compared with adults,103 the production of IFN-β in humans in response to Lm appears not IRF3-dependent but instead, p38 MAPK-dependent.108 The role of the STING/IRF-3 pathway in human neonatal Lm infection is thus still unclear.104
Finally, although the developmental patterns of the various inflammasome pathways in humans in response to Lm have not been unravelled, it is known that activation of the inflammasome is one way through which aluminum hydroxide (alum, the most common vaccine adjuvant) exerts its function.109 Given that innate immune responses induced by alum decrease over the first 2 y of life,110 age-dependent differences in at least some inflammasome activities are likely to exist.104 This has, however, not been investigated in relation to Lm.104
Splenocytes from infected neonates of a murine neonatal listeriosis model have been shown to exhibit much lower expression levels of Th1-supporting cytokines (IL-12p70 and IFN-γ), even when presented with Th1-driving stimuli.104,111 IL-10 is also produced at elevated levels by neonatal mice upon infection with Lm.112 In addition, CD71+ erythroid cells in neonatal mice and human cord blood have been shown to exert immunosuppressive effects; in mice, they appear to be crucial in rendering the neonate more susceptible to Lm infection.113 In humans, the response to Lm has not been studied as a function of age.104 As a proxy, it is known that the generation of proinflammatory cytokines (e.g., TNF-αand IL-1β) by TLR activation in the neonate can differ according to stimuli, and reaches adult levels of production between 1–2 y of age.114,115 There is also a progressive decline in the production of IL-10, IL-6 and IL-23 during this timeframe, from levels initially higher than those found in adults.114,115 Type I IFN production induced by TLR agonists reaches adult levels within only a few weeks of life, despite being significantly lower at birth.114,115 Th1-supporting innate cytokines (IL-12p70 and IFN-γ) likewise eventually reach adult levels of production.114,115 In addition, neonates display significantly reduced TLR-mediated production of cytokines that induce the production of ROS or RNS.116,117
Neonatal CD4+ T cell responses in mice appear to have a Th2 bias at birth.118 Although murine primary neonatal CD4+ Th1 cells and Th2 cells develop in tandem, only Th1 cells were found to undergo apoptosis when re-exposed to antigen.119 Additionally, the production of less IL-12p70 and more IL-10 by neonatal innate cells upon stimulation (as compared with their adult counterparts) would likely lead to suboptimal activation of neonatal CD4+ Th1 and CD8+ T cells.103,104,114,115 However, in spite of the established differences between the human newborn and adult adaptive immune responses,4 given the appropriate stimuli, the human newborn is very much capable of displaying strong and protective Th1-type responses even prior to birth.104,120
While the limitations of the immune system in early life have been speculated to be due to a functional impairment of neonatal T cells, studies in recent years have suggested that functional alterations in neonatal APCs also play a part. APCs are known to be key players during the innate immune response and serve as the link to the adaptive immune response. In the context of vaccination, the effectiveness of the role APCs play depends broadly on 2 factors: the delivery of antigens to APCs and the ability of APCs to subsequently process and present the antigens.
Lm is an excellent candidate for a neonatal antigen delivery vehicle given its ability to deliver antigen (in the form of DNA, or expressed and secreted as proteins) directly and efficiently into APCs.121,122 This capacity has led to Lm being among the most commonly employed bacterial vectors for efficacious antigen delivery.123 In most Lm vaccine strains, the antigen of interest is expressed from an episomal origin, delivered in the form of a multicopy plasmid and expressed under the control of an Lm promoter.123 The retention of such plasmids by Lm in vivo is achieved by prfA complementation from the plasmid in a prfA-negative mutant background.11,124,125 In the absence of prfA complementation, phagolysosome escape is not possible, thereby leading to loss of intracellular growth and antigen presentation.62 The expression system used in Lm-vectored transgene expression is derived from 2 different plasmids, with one portion containing the p15 origin of replication (providing a low copy number in Escherichia coli), and the other with its original copy control gene deleted (resulting in an upregulation of plasmid copy number).126,127 Besides plasmid-based strategies, chromosomal integration techniques have also been utilized to generate recombinant vaccine strains that express antigen from a chromosomal origin, e.g., a phage-based system that integrates genes at specific locations in the bacterial genome with the help of a site-specific integrase,67 or homologous recombination using allelic exchange.69 Between the 2, however, plasmid-based strategies might offer a greater advantage than chromosomal integration, in that the former allows for higher expression levels than a single copy chromosomal gene.123 Furthermore, the development of antibiotic marker-free plasmid expression systems for Lm has made this bacterial vector highly suitable for use in humans.67,69
The intracellular lifestyle of Lm involves phagosomal and cytoplasmic phases, thus giving it access to both MHC class I and II pathways as well as the consequent activation of CD8+ T cells and CD4+ T cells. The LLO and ActA virulence factors of Lm have been found to contain PEST-like sequences, which have been associated with ubiquitin-mediated protein degradation.128,129 In studies involving Lm as an anti-tumor vaccine, the expression and secretion of antigens fused to LLO or ActA enhanced anti-tumor efficacies and were highly effective at inducing tumor regression with complete regression of established tumors.128 In addition, the fusion of antigens with a truncated non-hemolytic fragment of LLO has been shown to significantly enhance the antigen-specific T cell-mediated immune response.124,125 LLO is also capable of inducing various cytokines such as IL-12, IL-18 and IFN-γ, which are believed to be critically important for the expression of nonspecific resistance and the generation of acquired immunity in an infected host.130,131
DCs are perhaps the most important APCs of the immune system. This is significant because as previously mentioned, DCs are one of the phagocytic cells that Lm can infect.121 More importantly, associated with Lm-induced DC maturation is the increased efficiency of antigen processing as well as a slower turnover rate of surface-expressed MHC-peptide complexes, which leads to more effective antigen presentation to CD8+ or CD4+ T cells.132 Although Lm is known to induce death in other cell types, the ability of human DCs to resist death and undergo maturation by the upregulation of costimulatory signals despite Lm infection is indeed an important component in their role as effective APCs for listerial immunity.120
Antigen processing capabilities are important for APC maturation as well as the generation of fully-activated effector T cells.133,134 In mice, neonatal DCs compared with their adult counterparts exhibit lower efficiency in the classical pathways leading to antigen processing and presentation;135 this has been shown to result in lower antigen-specific CD8+ T cells.136,137 Cross-presentation refers to the ability to process and present exogenous antigens via the MHC class I pathway;138 and neonatal DCs have been shown to be deficient in MHC class I cross-presentation of soluble antigen as well.139 However, such an age-dependent deficit in antigen presentation can be overcome if the antigen is delivered into (or produced within) the cytoplasm of neonatal DCs.7,137 In light of the neonate’s capacity for effective antigen processing and presentation, as well as the ability to generate adult-like Th1 responses when immunized in the presence of strong adjuvants,6-8 the impaired capacity of neonates to mount optimal immune responses appears less likely to be due to reduced antigen presentation, and more likely a result of lower or altered responses to stimuli leading to activation of APCs, such as adjuvants.140,141 Adjuvants can enhance antigen presentation to and antigen uptake by APCs (e.g., alum), and/or directly induce innate immune responses (e.g., the TLR4 agonist, monophosphoryl lipid A).79,142 The fact that Lm possesses both kinds of adjuvant characteristics is significant, and likely provided the basis for its success as a vaccine delivery vehicle. In other words, in addition to being an ideal vehicle for antigen delivery, Lm also possesses the adjuvant properties that appear capable and required for overcoming the apparent defects reported in neonatal antigen processing and presentation.
There are presently several limitations in considering Lm as a vector for neonatal vaccination. First and foremost, as a live vaccine for the neonate, the utmost level of safety has to be guaranteed if Lm were to be administered early in life. However, currently available data on the safety of Lm vectors is sufficiently convincing to suggest that further exploration of this approach is warranted. Second, Lm lacks the ability to carry out the full range of post-translational modifications observed in mammalian cells, and proteins greater than 60 kDa are not always properly folded.143 In addition, the need for secretion of protein antigens expressed by bacteria places restraints on the choice of antigens.143 Nevertheless, existing data on the processing and delivery of antigen via live Lm suggest that these theoretical limitations do not hinder Lm’s efficacy. Third and perhaps also the most important limitation of Lm as a vaccine vehicle is its poor induction of antibody responses, given that antibodies are known to be key components for many vaccines achieving protective immunity.141,144 High-affinity and diverse antibody repertoires have been shown to be inducible in neonatal mice.145,146 The mechanisms of infection and intracellular lifestyle of Lm typically result in poor induction of humoral immune responses by Lm.147 However, considering that diminished T cell help as well as immature antigen presentation have been implicated as factors contributing to the limited vaccine B cell responses in infants,144 Lm as a neonatal vaccine vehicle might well be a possible solution to such a shortcoming of the newborn–rather than a shortcoming itself–given its powerful induction of T cell responses and antigen presentation capabilities. In fact, we have shown that Lm induces antibody responses in neonatal mice surpassing those of their adult counterparts.148
Lm offers great potential against a number of infectious diseases that strike very early in life. With its major strength being the immediate induction of highly effective cell-mediated protective immunity, Lm should benefit vaccines against microbes for which the role of cell-mediated immunity in protection from infection has been clearly and most consistently identified; namely, microbes with an intracellular life cycle. These include viruses, several parasites (e.g., Leishmania, malaria), and certain bacteria (e.g., Burkholderia, Mycobacteria, Yersinia). Importantly, against many of these pathogens, Lm has, in fact, already been shown to effectively induce cell-mediated immunity in animal models.24,25,28,149,150Lm-vectored vaccines targeting microbes with an intracellular lifestyle would fill a vacant niche, given that there are currently no licensed vaccines against them (with the exception of BCG, whose effectiveness has been contentious). Furthermore, Lm-vectored vaccines also stand to address the important medical need of inducing immunity near birth, especially against microbes that are known to infect early in life (e.g., Mycobacteria, Leishmania, Burkholderia, respiratory syncytial virus, malaria). Yet given the limitations of Lm-based vaccine delivery, most notably the relatively poor induction of humoral immunity by Lm, vaccines that protect mostly on the basis of high titers of specific antibodies (e.g., those directed against toxins such as diphtheria or tetanus, or against pathogens such as hepatitis, influenza virus, or B. pertussis) may not be well served if delivered via an Lm-based vector. For many of those pathogens, however, existing vaccines already provide at least reasonable protection in most circumstances.
The burden of infectious disease leading to suffering and death early in life is immense, underscoring the urgency for an effective approach to neonatal immunization. The particular advantages that Lm has to offer as a neonatal vaccine vehicle are substantial. Attenuated strains of Lm safe for neonates have now been identified; more importantly, these attenuated strains are very efficient at inducing robust Th1-type immunity in neonates. Furthermore, with Lm’s ability to confer lasting and possibly lifelong immunity when given during the neonatal period, Lm-based vaccines not only stand to benefit the infant with protection throughout early life, but also offer a favorable solution to the logistical issues often faced in the need for administration of booster doses later in life. The merits of Lm as a neonatal vaccine platform are clear and significant, and as such, worthy of further exploration.
A.W. is an employee and shareholder of Advaxis Inc.