In humans, there is a decrease in pulmonary elasticity, loss of respiratory muscular strength, decreased ciliary beating and mucus velocity with age (Ho et al., 2001; Janssens and Krause, 2004). These changes occur from a combination of genetic predisposition, inflammageing and environmental exposure, which involve a wide range of molecular and cellular changes and impairment of cell-cell communications (Brandenberger and Muhlfeld, 2017).

It is becoming increasingly apparent that the microbiome is an important determinant of lung and gut homeostasis and the development of disease, particularly in older individuals (de Steenhuijsen Piters et al., 2016; Schenck et al., 2016; Man et al., 2017; Ragonnaud and Biragyn, 2021). It remains uncertain whether changes in the composition and diversity of microbiota represent a cause or consequence of pneumonia (de Steenhuijsen Piters et al., 2016). Older individuals with pneumonia exhibit increased abundance of species such as S. pneumoniae, Rothia and Lactobacilli, but decreased overall anaerobic bacterial diversity in the upper respiratory tract (URT) (de Steenhuijsen Piters et al., 2016). In a human experimental pneumococcal challenge model (EHPC), low density pneumococcal carriage was associated with a stable mucosal microbiome (baseline presence of Corynebacterium/Dolosigranulum species) and a less pro-inflammatory phenotype (de Steenhuijsen Piters et al., 2019). Whether different pneumococcal strains interact differently with the respiratory microbiome during colonization (Cremers et al., 2014) and how this affects older individuals, remains to be determined. The role of intestinal microbiota on lung susceptibility to pneumococcal infection also warrants further investigation in humans as murine studies suggest that Nod-stimulating microbiota in the gut induce GMCSF-dependent immunity, which influences alveolar macrophage function during pneumococcal infection (Schuijt et al., 2016; Brown et al., 2017).

An imbalance of cytokine expression in older individuals is referred to as “inflammageing”, where damage to the tissue, changes in composition of the microbiome and cellular and immune senescence, all contribute to this state of chronic inflammation (Franceschi and Campisi, 2014). The contributors to inflammageing may include microbial translocation, chronic infections, mitochondrial dysfunction and accumulation of DNA damage (Fulop et al., 2017; Ferrucci and Fabbri, 2018). This increased inflammation extends to the lungs, as healthy individuals >65yrs have elevated levels of IL-6, IL-8 and higher numbers of neutrophils in bronchoalveolar lavage (BAL) samples (Thompson et al., 1992; Meyer et al., 1996). The increased susceptibility to respiratory tract infections has also been linked to the heightened inflammatory status in older people. In a large prospective study of people aged 70-79yrs, being in the highest tertile for systemic IL-6 and TNF levels was associated with 1.6-1.7-fold increased risk for developing CAP (Yende et al., 2005). Inflammation, in particular IL-6 levels, at time of admission to hospital is also associated with CAP disease severity (Glynn et al., 1999; Antunes et al., 2002). Whether this represents more severe disease, or a pre-existing heightened inflammatory state is uncertain. For example, at admission, in a cohort of 22 patients with pneumonia, of which 19 had confirmed S. pneumoniae, patients <55yrs had increased levels of IL-6 compared to patients >68yrs (Bruunsgaard et al., 1999). However, at 7 days post admission, pro-inflammatory cytokines TNF and soluble TNF receptor I remained elevated in older adults, while they had returned to baseline in young adults. This increased inflammatory state with age may therefore contribute to dysregulation of immunomodulation of innate immune cells such as neutrophils, cytokines and chemokines (Williams et al., 2015) and ultimately, pneumococcal colonization, increasing the chance of IPD in older individuals.

S. pneumoniae binds to a variety of epithelial receptors including Keratin 10, laminin receptor and platelet-activating factor receptor (PAFr), the expression of which is altered in older individuals, thus potentially influencing pneumococcal adherence and susceptibility towards disease. For example, levels of Keratin 10, laminin receptor and PAFr are elevated in aged mice, human lung tissue and senescent A549 cells (Hinojosa et al., 2009; Shivshankar et al., 2011) which could contribute to altered outcomes after pneumococcal colonization. Pneumococcal micro-invasion of the epithelium in vitro is also associated with epithelial secretion of cytokines and chemokines and a transcriptomic enrichment of innate signaling pathways including Toll receptor cascades, NFκB and MAPK activation (Weight et al., 2019). In the EHPC model, an epithelial transcriptomic signature that was associated with bacterial clearance has been identified, indicating the involvement of epithelial activation in the control of pneumococcal colonization (Weight et al., 2019). Furthermore, in vitro epithelial cell models reveal that activation of p65, upregulation of the histone demethylase KDM6B and IL-11 secretion are associated with protection from epithelial damage following pneumococcal infection (Connor et al., 2020). We therefore speculate that in older adults, where pneumococcal micro-invasion and cellular senescence may be enhanced, dysregulation of these pathways and regulatory mechanisms could exacerbate invasive disease.

Mucosal Barrier Function: The expression profiles of intercellular junctional proteins such as Claudins, ZO-1 and E-Cadherin, determine the permissiveness of the respiratory epithelium to the passage of microbes across the barrier. Changes in epithelial junctional expression during pneumococcal infection leads to structural reorganization of the barrier. For example, in vitro infection of human nasopharyngeal cells with S. pneumoniae is initially associated with decreased permeability to 4kDa FITC-dextran, indicating a strengthened barrier (Weight et al., 2019). However, over time (>8 hours post-infection, when bacterial replication and autolysis have also occurred) barrier integrity is altered, demonstrated through decreased expression of Occludin, ZO-1, Claudin-5 and VE-cadherin in the alveoli epithelium and endothelium in younger adults lung explants (Peter et al., 2017). Furthermore, downregulation of Claudin-7 and Claudin-10 in human and murine epithelial cells, led to increased pneumococcal transmigration across the epithelial barrier (Clarke et al., 2011).

The vitamin D receptor (VDR) regulates epithelial barrier function (Chen et al., 2018). In Salmonella infected VDR^-/- mice, Claudin 2 upregulation was associated with leaky intestinal barrier, increased pathology and upregulation of NFκB (Zhang et al., 2019), a critical regulator of inflammageing (Salminen et al., 2008) and tight junction protein expression (Ward et al., 2015). In the older human intestine, Claudin-2 upregulation has been detected, which was accompanied by decreased transepithelial electrical resistance and increased permeability (Man et al., 2015). Activation of NFκB following pneumococcal infection is widely reported (Malley et al., 2003; Weight et al., 2019), and Vitamin D deficiency is more severe in older generations (Hirani and Primatesta, 2005; Jolliffe et al., 2013), and there is evidence to suggest that supplementation of Vitamin D could be beneficial in boosting immunity and reducing acute respiratory infections (Martineau et al., 2017; Chambers et al., 2021). Whether regulation and junctional protein responses in the URT are altered with age in humans, and how this contributes to control of pneumococcal colonization, remains to be determined.

An important factor of epithelial innate immunity includes AMPs that neutralize toxins and eliminate pathogens (Hiemstra, 2001). Infection of human corneal epithelial cells with pneumococcus induced NFκB activation leading to the secretion of LL-37, β defensin -2, -3, -4, S100A7, S100A8, S100A9, Lipocalin and RNase 7 (Sharma et al., 2019). LL-37 plays a role in wound healing, can induce autophagy in a 1,25-dihydroxyvitamin D3 dependent manner, can activate the NLRP3 inflammasome in a model of P. aeruginosa and, is bactericidal against S. pneumoniae and Mycobacterium tuberculosis (Nijnik and Hancock, 2009; Yuk et al., 2009; McHugh et al., 2019; Sharma et al., 2019). Older individuals maintain similar levels of baseline production of cathelicidins and β defensin 2 in serum compared to younger adults (Castaneda-Delgado et al., 2013). However, in aged mice, CRAMP expression, the murine homolog of LL-37, was not upregulated following pneumococcal infection, compared to younger adults (Krone et al., 2013). This suggests a potential dysregulation of AMPs in older individuals and the implications for the control of S. pneumoniae at the mucosal surface warrants further investigation.

AMPs also induce the secretion of cytokines and chemokines like IL-6 and IL-8 from nasal epithelial cells, in an NFκB dependent manner (Pistolic et al., 2009). One might predict that given elevated levels of cytokines such as IL-6 and TNFα in older individuals (Yende et al., 2005; Man et al., 2015), epithelial cell responses may also differ in the response to pneumococcal carriage. IL-6 is a pleiotropic cytokine and so elevated baseline secretion in older individuals may either enhance or weaken barrier integrity upon pathogenic challenge. For example, IL-6 regulates the expression of tight junction proteins such as Claudin 2 and increases intestinal barrier permeability (Suzuki et al., 2011; Man et al., 2015). This could also occur in the respiratory setting, which may increase pneumococcal transmigration across the epithelial barrier. Alternatively, IL-6 is also known to confer epithelial repair and promote proliferation (Kuhn et al., 2014), which may inhibit pneumococcal adherence to the epithelium. For example, co-infection with Influenza A increases susceptibility to S. pneumoniae in both adult and older mice and in younger adults, characterized by increased bacterial burden in the URT (Mina et al., 2014; Jochems et al., 2018; Gou et al., 2019). In the murine study, IL-6 production was required to maintain barrier function and macrophage phagocytic function, which played a role in pneumococcal control and clearance (Gou et al., 2019). Although secreted by infected human nasopharyngeal cells in vitro (Weight et al., 2019), levels of epithelial IL-6 secretion in vivo have not been directly investigated in adults or older individuals.

TLR1 levels are reduced on monocytes from older adults and TLR1/2 specific stimulation using Pam2SCK4 is associated with reduced responses in monocytes from older people (van Duin et al., 2007). Indeed, cytokine responses to pneumococcal or relevant ligands also appear decreased with age. Frail older individuals have increased baseline production of TNF by intermediate monocytes in particular, but show an impaired induction upon TLR1/2 or TLR4 stimulation (Hearps et al., 2012; Verschoor et al., 2014b). Upon heat-killed pneumococcal stimulation however, monocyte-derived macrophages in frail older people produce less TNF, IL-6, IL-1β and IL-8 and have a reduced capacity to kill S. pneumoniae. This is possibly related to defective PI3K-AKT signaling (Verschoor et al., 2014a), and/or insufficient activation of the NLRP3 inflammasome, as demonstrated in bone marrow derived macrophages from aged mice (Cho et al., 2018).

In younger adults, infiltration of classical monocytes into the nasal mucosa in the EHPC model coincides with initiation of pneumococcal clearance, while nasal myeloid cytokines correlate with clearance of colonization (Jochems et al., 2018). The impact of the respiratory monocyte/macrophage dysfunction described above on pneumococcal control in older people is not fully understood. However, altered monocyte subsets are an important contributor to a reduced ability to prevent pneumococcal infection in older people (Puchta et al., 2016). For example, reduced cytokine production to TLR1/2 agonists seem to be mediated by changes in CD14++ CD16+ intermediate and CD14+ CD16+ non-classical monocytes (Nyugen et al., 2010). In addition, alveolar macrophage numbers are higher in BAL samples from healthy older adults compared to younger adults (Thompson et al., 1992; Meyer et al., 1996), although how this affects innate immune responses to S. pneumoniae is unknown.

Murine studies have also identified age-related functional differences in monocyte/macrophage interactions with S. pneumoniae that may be relevant for disease pathogenesis. Puchta et al. demonstrated that TNF is a crucial mediator of the susceptibility to pneumococcal infections in inflammageing, as well as the alterations in monocyte subsets (Puchta et al., 2016). Increasing TNF with age led to premature egress of pro-inflammatory monocytes from bone marrow and increased levels of intermediate CD14++ CD16+ monocytes. Specific depletion of these monocytes or reduction in TNF levels enhanced immunity to pneumococcal infection and increased clearance in old mice (Puchta et al., 2016). Koppe found that following sensing of pneumococcal dsDNA by murine macrophages, STING (“stimulator of IFN genes”) binds to TBK1 (TANK-binding Kinase 1), leading to IRF3 (Interferon Regulatory Transcription Factor 3) activation and production of IFNβ, which assists pneumococcal clearance (Koppe et al., 2012). In aged mice, there is less STING/TBK1/IRF3 mRNA and protein expression compared to young mice infected with S. pneumoniae (Mitzel et al., 2014). This was associated with lower levels of IFNβ and higher bacterial burden in the lung, thought to be due to age-associated stress of the endoplasmic reticulum, resulting in increased autophagy-related protein 9a-STING complex formation (Mitzel et al., 2014), preventing STING complex formation with TBK1 (Saitoh et al., 2009).

There is a large body of evidence from both murine and human studies showing the importance of neutrophils in protecting against pneumococcal colonization of the nasopharynx (Lu et al., 2008; Weinberger et al., 2009; Jochems et al., 2018; Nikolaou et al., 2018). Impaired functional responses of neutrophils in older individuals may therefore also be detrimental during pneumococcal infection.

In the nose and lungs of healthy older people, neutrophils are highly abundant (Thompson et al., 1992; Meyer et al., 1996; Reiné et al., 2019). Excessive neutrophil recruitment to the lung following infection can mediate tissue damage and may exacerbate inflammation in older people (Menter et al., 2014). In frail older individuals, neutrophils have an immature profile with increased levels of intracellular reactive oxygen species (ROS) and cell surface expression of proinflammatory markers like CD11b (Verschoor et al., 2015). They have an impaired capacity to produce neutrophil extracellular traps (Hazeldine et al., 2014) and altered chemotaxis responses to respiratory infection, which leads to prolonged production of proteinase and a pro-inflammatory milieu (Sapey et al., 2014; Sapey et al., 2017). Neutrophils in older individuals may also be impaired in their opsonophagocytotic ability to several bacterial pathogens including S. pneumoniae and S. aureus (Wenisch et al., 2000; Butcher et al., 2001; Simell et al., 2011). Interestingly, vitamin E supplementation in aged mice prevented neutrophil migration and mortality during pneumococcal pneumonia (Ghanem et al., 2015), suggesting that regulating neutrophil activity in older individuals could be beneficial to the host.

Unconventional, innate-like T cells called MAIT cells, play important roles in the defense against bacterial and viral infections. Recently, studies with the EHPC model demonstrated that blood and nasal MAIT cells were associated with protection from pneumococcal colonization (Jochems et al., 2019b). They can be activated by conserved bacterial ligands derived from vitamin B (riboflavin) biosynthesis or indirectly via cytokines (Godfrey et al., 2019; Toubal et al., 2019). MAIT cells recognize precursors of the riboflavin synthesis pathway after presentation via MHC class I–related protein 1 (MR-1). This pathway is highly conserved in the pneumococcal genome, and MAIT cells can respond to pneumococcal isolates in both MR-1 dependent and independent manners (Kurioka et al., 2018). MAIT cells are depleted from blood in old age, with levels progressively dropping from 3-5% of T cells in young adulthood to below 1% in older adults (Novak et al., 2014; Walker et al., 2014; Loh et al., 2020). Remaining MAIT cells in this population show clonal expansion, similar to conventional CD8 T cells, and increased basal inflammation, although they retain potent anti-microbial function (Loh et al., 2020). There appears to be a substantial depletion of total CD8+ T cells from the nasal mucosa in older adults (Reiné et al., 2019) and the ratio of CD4/CD8 T cells increases in the lung in middle age (Meyer et al., 1996). Therefore, as MAIT cells are predominantly found within the CD8+ T cell compartment, we postulate that there is loss of MAIT cells at the mucosal surface in older individuals which negatively impacts on the control of pneumococcal colonization. This warrants further investigation.

In murine models, a Th17-mediated recruitment of monocytes and neutrophils leads to clearance of S. pneumoniae colonization (Lu et al., 2008; Zhang et al., 2009). Some human studies also suggest a role for Th17 cells in protection against colonization showing increased ratios of pneumococcal-specific Th17/Tregs with increasing age as carriage rates decrease (Mubarak et al., 2016) and a SNP in the IL17A gene associated with ~2-fold increased risk of pneumococcal colonization in children in the first year of life (Vuononvirta et al., 2015). However, a protective role for Th17 cells in humans has not been fully substantiated. We have shown acquisition of pneumococcal antigen-specific tonsillar Th1 T cells but not Th17 cells with age (Pido-Lopez et al., 2011). In the EHPC model, Th17 cells were found in the lung after colonization, which associated with increased bacterial killing in macrophages (Wright et al., 2013), but were not identified in the nasopharynx. Higher Th17 responses were found in children from Bangladesh with high carriage rates, compared to children from Sweden with lower carriage rates (Lundgren et al., 2012). In HIV⁺ individuals in Malawi where carriage rates are high, there was no evidence of a Th17 protective phenotype (Glennie et al., 2013). Furthermore, colonization by S. pneumoniae has been associated with decreased Th17/Treg ratios in children, possibly mediated by TGF-β induction leading to regulatory responses (Zhang et al., 2011; Neill et al., 2014; Jiang et al., 2015). Stimulation of PBMCs from individuals of different ages with three different pneumococcal proteins revealed a non-significant decrease in responses of older adults, although polyfunctionality (co-production of IFNγ by same donor) was decreased (Schmid et al., 2011). In older adults, total Th17 numbers in blood are decreased and total Treg numbers increased (van der Geest et al., 2014). Together, these observations highlight that a role of Th17 cells in conferring protection against colonization in humans remains unclear. Further investigations of mucosal Th17 and Tregs responses in older people using longitudinal studies and the EHPC model are required.