A new study led by researchers from Griffith University, Gold Coast campus, Queensland -Australia has found that activities such as nose picking or plucking nose hairs can lead increase risk for Alzheimer’s disease and dementia.
The study findings were published in the peer reviewed journal: Scientific Reports (Nature) https://www.nature.com/articles/s41598-022-06749-9
In the current study, we showed that (1) C. pneumoniae rapidly infected both the olfactory and trigeminal nerves in mice, (2) C. pneumoniae entered the CNS via nerves within 24–72 h after intranasal inoculation and without concurrent blood infection, (3) injury to the nasal epithelium exacerbated peripheral nerve infection, but reduced brain infection, (4) C. pneumoniae inclusions in the olfactory nerve and bulb were associated with accumulations of Aβ, (5) the glial cells populating the olfactory/trigeminal nerves and brain supported C. pneumoniae replication, and (6) C. pneumoniae infection leads to differential regulation of Alzheimer’s disease related genes.
Thus, C. pneumoniae can very rapidly spread from the periphery to the CNS via the nerves extending between the nasal cavity and the brain, without blood infection. To our knowledge, this study is the first report of Aβ deposition in response to C. pneumoniae infection of the primary olfactory nervous system, and the first time such rapid (72 h) deposition of Aβ in response to any bacterium in wild-type animals in vivo has been demonstrated.
The time-frame for infection of the CNS by C. pneumoniae was considerably faster than what has previously been shown (1 week–3 months16,17,18), which may be due to differences in the inoculation dose since we used a higher inoculation dose than two previous studies16,50 but lower than another 18.
Nevertheless, the time-frame is comparable to CNS invasion via cranial nerves by Burkholderia pseudomallei 32,43,44,52, Streptococcus pneumoniae 53, Neisseria meningitidis 54, Listeria monocytogenes 55 and now recently another Chlamydia species, C. muridarum24.
The amoeba Naegleria fowleri 56, as well as herpes simplex type virus type 1 (HSV-1)57, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)58,59,60 and other coronaviridae61 can also invade the CNS via these two paths (shown in humans and/or animals).
Within the olfactory bulb, C. pneumoniae inclusions were detected in OECs within the nerve fibre layer/glomerular layer. Another bacteria, Burkholderia pseudomallei also accumulated within the nerve fibre layer/glomerular layer after intranasal inoculation, suggesting that the glia limitans acts to restrict further progression of bacteria into the deeper regions.
However, with C. pneumoniae while we could easily detect the inclusion bodies, the much smaller infectious elementary bodies would likely be missed in our analyses of the tissue sections; thus it is possible that elementary bodies were present deeper in the olfactory bulb.
As inclusion bodies were detected in the olfactory piriform cortex, it suggests that C. pneumoniae did progress deeper into the olfactory bulb as previously reported 15,16,18,50.
Injury to the nasal epithelium has been shown to increase infection of the olfactory nerve and bulb by B. pseudomallei27 and to allow the entry of S. aureus, which does not normally invade cranial nerves, to enter the olfactory bulb28. We therefore hypothesized that epithelial injury may lead to increased C. pneumoniae invasion of the olfactory/trigeminal nerves, olfactory bulb and remaining parts of the brain.
We found that epithelial injury resulted in increased C. pneumoniae load in the olfactory mucosa (which contains the fascicles of the olfactory nerve), olfactory bulb and trigeminal nerve. In contrast, injury did not alter C. pneumoniae invasion of the brain after 7 days. We have previously observed a similar result for B. pseudomallei in some mice, in which the nasal infection in itself caused massive peripheral infection and destruction of the nasal epithelium (more pronounced than in our epithelial injury model used in the current study).
In these mice, B. pseudomallei invasion of the CNS was negligible20. We then hypothesized that this may be because glia in the olfactory nerve and outer layers of the bulb responded to both the injury and bacteria, secreting large amounts pro-inflammatory factors which limited CNS infection; this may also be the case for C. pneumoniae infection in the current study.
The ability to infect glia is considered key for CNS invasion via the cranial nerve paths20,27,28,34,35. We here showed that C. pneumoniae could infect, survive in and replicate (form inclusions) within glia from the PNS (OECs and TgSCs) and the CNS (astrocytes and microglia).
This is the first-time infection of OECs and TgSCs (or other Schwann cells) by C. pneumoniae has been reported, however, we have recently shown that C. muridarum can infect OECs and TgSCs 24. Whilst C. pneumoniae infection of cultured primary astrocytes and microglia has not been described, infection of astrocyte and microglial cell lines has been demonstrated 62,63,64,65,66. Most relevantly, however, C. pneumoniae antigens have been detected inside both astrocytes and microglia in post-mortem human brains9,11,67,68.
OECs, Schwann cells and astrocytes are all innate immune cells which can respond to and phagocytose bacteria, and microglia (the macrophages of the CNS) are well characterized professional phagocytes31,69,70. The fact that C. pneumoniae can form inclusions in these cells suggest that the bacteria, at least to some extent, can overcome phagocytic destruction; this may be one important mechanism by which this bacterium can invade and establish long-term infection of the CNS.
We also detected localized deposition of Aβ adjacent to C. pneumoniae IBs and in the olfactory bulb after 7 days and 28 days post inoculation. Diffuse/scattered Aβ immunoreactivity was also present in these tissues of control mice, however, the co-localisation of Aβ deposits and C. pneumoniae inclusions in inoculated mice was clear and distinct.
One study reported that whilst there were not necessarily more Aβ deposits in the cortex of C. pneumoniae-infected animals, Aβ deposits in infected animals were morphologically different from those in control animals18. A previous long-term study showed that C. pneumoniae infection of the cerebral cortex preceded the peak of Aβ deposition17. In combination with the findings of the current study, it appears that Aβ secretion occurs in response to the infection.
One reason may be that Aβ is secreted as an antimicrobial agent12 but alternatively it may be secreted in response to infection because of pathway activation for the processing of the APP protein into Aβ which is then secreted; future work can clarify the secretion and role of Aβ in this context.
The secretion of Aβ may thus be a normal immune response to any microbe that may invade the nervous system, and if infection clears, the deposited Aβ can be cleared by phagocytic glia71. It is, however, possible that if bacteria are not cleared and instead become persistent or latent in neural cells, continued Aβ deposition may occur, contributing to late-onset dementia and/or accelerating Aβ deposition in familial Alzheimer’s disease 7.
In the case of C. pneumoniae, one study in wild-type mice demonstrated that Aβ deposits resulting from infection were subsequently cleared17, whilst another study showed that the deposits did not disappear over several months16.
It is interesting that we observed Aβ deposits in the olfactory nerve earlier than in the bulb, as one study in an Alzheimer’s disease mouse model (APP/PS1 mice) showed that the terminal end of the olfactory nerve within the nasal olfactory epithelium is the first nervous system area to exhibit Aβ deposition, which then progresses to the olfactory bulb and other CNS areas 72.
As the mice in that study were kept in a standard animal holding facility (not specific pathogen free), perhaps exposure to infectious agents may have contributed to this early, peripheral deposition of Aβ (which likely would be much more pronounced in an Alzheimer’s disease model than in wild-type mice).
Chlamydia pneumoniae infection also resulted in up-regulation of key pathways involved in Alzheimer’s disease pathogenesis. The pathologic features of Alzheimer’s disease like activated microglia, production of inflammatory mediators and reactive oxygen species (ROS) were highly regulated in infected brain tissue at 28 days post inoculation as compared to 7 days post inoculation.
Theses neuroinflammatory responses are considered a major driving factor in patients with neurodegeneration and Alzheimer’s disease pathology, which starts early in the course of the disease, prior to the formation of Aβ plaques in the brain 73.
Previous studies have shown that microglia and astrocytes act as host cells of C. pneumoniae in Alzheimer’s disease brain9. It has been shown that following infection, activated microglia and astrocytes secrete pro-inflammatory cytokines, including IL-1β, TNFα and IL-6 which are neurotoxic and may directly increase Aβ production via activation of β-secretase (BACE)66,74. BACE cleaves amyloid precursor protein and initiates the amyloid cascade.
Microglia activation reduces the accumulation of Aβ in the brain by increasing its phagocytosis, clearance, and degradation75. However, the neuroinflammation associated with Alzheimer’s disease could be a double-edged sword because persistent microglia activation stimulated by the binding of microglia to Aβ can increase the production of inflammatory mediators and reactive oxygen species (ROS), which further amplifies the neuroinflammatory response causing chronic inflammation and neurodegeneration76.
Disturbance of endoplasmic reticulum (ER) function is emerging as a relevant factor driving neurodegeneration in Alzheimer’s disease77. Several reports have described manifestations of ER stress in post-mortem brain samples from Alzheimer’s disease patients78. Protein folding in the endoplasmic reticulum (ER) is an essential cell function and to safeguard protein production and ensure quality control, ER-stress triggers the activation of several biochemical pathways collectively referred to as the unfolded protein response (UPR).
Chlamydia infection can induce cellular stress that impacts protein folding, thus inducing UPR activation however it is also proposed to modulate the UPR to promote their survival and replication 79. Interestingly, we found UPR pathway being up-regulated in infected cortical tissues at 28 days post inoculation as compared to 7 days post inoculation. Intracellular pathogens like Chlamydia would benefit from UPR since increase in folding capacity and activation of lipid biosynthesis can sustain bacterial replication.
However, if the ER stress due to infection is sustained and misfolded protein cannot be refolded or degraded, the cells can also directly increase Aβ production and associated neuroinflammation80. Conversely, Aβ oligomers have also been proposed to cause ER dysfunction leading to UPR mediated neurotoxicity and neuronal cell death77.
We have also observed similar trends in our study where molecular pathways related to cell death like autophagy and apoptosis were up-regulated in cortical tissues at 28 days post inoculation.
In addition to considering key pathways, it is also useful to consider changes in individual gene expression. Long term C. pneumoniae infection (day 28) triggered down-regulation of most other key genes involved in AD pathogenesis. Most importantly there was downregulation of key protective heat shock protein (Hspa1b or Hsp70-2), associated with increased oxidative stress and initiation of AD pathology81.
In addition, Bag2, a Bcl-2 associated co-chaperone gene which controls Hsp70 functionality was also downregulated leading to further failure of the system to protect cells from oxidative damage 82. The long term infection also depressed the 26S proteasome ubiquitination system by downregulation of Psmd8 83 and Psmc6 84 leading to persistence of stress-induced protein aggregates. At a sub-cellular level, infection led to mitochondrial dysfunction evident by downregulation of Ndufa5 (a structural subunit of complex I)85 and Atp5j2 86.
Nevertheless, all these gene modulations led to increased unfolded protein response, oxidative stress, and had higher disease association as evident by the biological processes heat map. In fact, the long-term infection was also associated with low expression of Cd2ap which has been previously associated with AD pathology aggravated by increased deposition of Aβ and Tau-induced neurotoxicity 87.
In contrast to the downregulated genes, long-term infection was also associated with some repairing mechanisms, limiting the spread of further neuroinflammation. It led to higher expression of Cdk18 which is a cyclin-dependent kinase and usually functions to clear DNA damages 88.
Thus, mechanisms inducing chromatin modification, transcription and splicing were highly reduced in them. Amongst the heat shock proteins, we found Hspa4 to be upregulated which ensured that it maintained the disaggregating property of any misfolded proteins, thereby, preventing further damage and inducing tissue integrity 89. Additionally, it also helped in sulfatase gene (Sumf2) maintenance which is associated with modulation of tissue homeostasis90.
Interestingly, long term infection also induced expressions of Ezr91 and Cldn592 which are associated with maintenance of actin cytoskeletal structure, synapse and tight junctions, respectively. This ensured that long-term infection induced greater organisation of neuronal cytoskeleton and/or dendritic structure while also maintaining synaptic transmission and reuptake.
Additionally, further protection was provided by increased Axl activity, a phagocytotic gene (autophagy) important in maintaining homeostatic levels of major AD related lipoprotein, ApoE93. Long term infection also induced Mbp gene, which maintained myelination and prevented further neuroinflammation 94.
In contrast, upregulation of Flt1, a key gene in vascular endothelial growth factor (VEGF) regulation was a key feature in long term infection which could reflect cognition impairment due to higher Aβ and Tau deposition in AD etiopathogenesis, as observed in a previous study 95. As a result of increased growth factor signalling, angiogenesis was induced with increased trophic activity demonstrating ongoing inflammatory activity.