Professional phagocytes possess a powerful tool to aid in their innate immune activity: Nox2 (or gp91phox). Namely neutrophils, monocytes, macrophages, and their central nervous system (CNS) equivalent microglia are constitutive expressors of Nox2. While this enzyme is critically important in the normal defense against invading pathogens, it also has an apparent role in the development and progression of cardiovascular diseases, including endothelial dysfunction (Chan and Baumbach, 2013), hypertension (Murdoch et al., 2011), ischemic heart disease (Guzik et al., 2006), and atherosclerosis (Sorescu et al., 2002). Importantly, when reconstituting Nox2-containing wildtype monocytes back to LysM-positive cell ablated mice, the protection from angiotensin-II induced hypertension is absent—indicating that vascular impact of Nox2 expression is dominated by its abundance in phagocytic cells (Wenzel et al., 2011). Nox2 inhibition has also been shown to mitigate anxiety-like phenotypes and oxidative stress associated with chronic mild stress (Lv et al., 2019). Accordingly, our own studies demonstrate that upon noise exposure, Nox2 protein and mRNA is consistently upregulated in the murine aorta (Munzel et al., 2017) alongside activation mechanisms of Nox2, such as angiotensin-II dependent diacylglycerol-mediated protein kinase C activation with subsequent phosphorylation of the major cytosolic regulator of Nox2, p47phox, at serine 328 (Figure 2) (Kroller-Schon et al., 2018).

We also find that oxidative stress in the aorta, heart, and brains of noise-exposed mice is significantly increased over those of unexposed controls, which is entirely mitigated in mice with a genetic deletion of Nox2 (Kroller-Schon et al., 2018). These mice were similarly protected from increases in blood pressure, dysregulation of NO signaling, and endothelial dysfunction, which is in line with reports of NOX-derived superoxide being partially determinative in the endothelial dysfunction accompanying genetic, angiotensin, and deoxycorticosterone acetate (DOCA) salt hypertension (Laursen et al., 1997; Zalba et al., 2001; Harrison et al., 2003). Nox2 deletion also protected mice from microvascular dysfunction in the cerebral microvessels and proteomic analysis demonstrated that there was no noise-induced increase in inflammatory signaling in the plasma (Eckrich et al., 2021). In addition, Nox2 inhibition by GSK2795039 suppressed ROS signals in cerebral cryo-sections of noise-exposed mice (Kroller-Schon et al., 2018). It may be also speculated that noise-induced ROS formation promotes an inflammatory phenotype in the heart, vessels and the brain as central mediators of inflammatory reactions such as the NLR family pyrin domain containing 3 (NLRP3) inflammasome and high-mobility group box 1 protein (HMGB1) are activated under oxidative stress conditions via redox switches as well as redox-sensitive transcription factors such as nuclear factor kappa B (NFκB) (Wenzel et al., 2017; Steven et al., 2019). This is probably the reason, aside from stress hormone-dependent activation and infiltration of immune cells into the vasculature, for the observed noise-triggered inflammation in exposed mice (Munzel et al., 2017; Kroller-Schon et al., 2018; Steven et al., 2020; Eckrich et al., 2021; Frenis et al., 2021) but also the shift to a pro-atherothrombotic phenotype of the plasma proteome of train noise-exposed healthy human subjects (Herzog et al., 2019), epigenetic changes that promote immune cell activation and expression of CRP (Cai et al., 2017; Eze et al., 2020) and amygdala activation driven coronary atherosclerosis (Osborne et al., 2020; Hahad et al., 2021a; Osborne et al., 2021). Noise-mediated inflammation in mice was also prevented by genetic Nox2 deletion as shown by two independent preclinical studies (Kroller-Schon et al., 2018; Eckrich et al., 2021) and antioxidant pharmacological activation/induction of the Nrf2-HO1-axis (Bayo Jimenez et al., 2021).

We were able to further discern that Nox2-bearing cells were primarily responsible for noise-induced cardiovascular and cerebral damage through a selective ablation protocol targeting cells expressing lysozyme M. Monocytes and macrophages are generally LysM⁺, whereas microglia are only weakly LysM⁺ or even LysM⁻. As a result, we found that upon ablation, blood pressure, endothelial function, and oxidative stress parameters were largely protected in the periphery (Figure 2) (Frenis et al., 2021). However, an exaggerated stress response measurable through plasma corticosterone level was seen in mice whose monocytes/macrophages were ablated, accompanied by a neuroinflammatory phenotype. Markers of microglial activation, CD68, CD86, and MHC-II, were significantly elevated in flow cytometry analysis of noise-exposed murine brains and not normalized by genetic ablation LysM-positive cells (Figure 2) (Frenis et al., 2021). This apparent disparity somewhat implies that the blood-brain-barrier may be affected by noise, which has been reported in hypertension as well (Setiadi et al., 2018). Furthermore, the pro-oxidative and pro-inflammatory environment appears to have also affected the state of astrocytes in the brains of noise-exposed mice, as an increase in GFAP⁺ staining can be detected. These results are in line with reports of Nox2 activation in microglia in several pathologies affecting the cerebrovasculature (Simpson and Oliver, 2020) and may connect these studies in animals with data from the Gutenberg Health Study of 11,905 participants that demonstrates that annoyance to noise predicts depression and anxiety (Beutel et al., 2020).

The nitric oxide synthase (NOS) family is critically important for the normal functioning of vessels, due to their role in the production of bioavailable nitric oxide (^•NO) (Forstermann and Munzel, 2006; Daiber et al., 2019b). Because of the actions of ^•NO, the presence of normally functioning eNOS and nNOS is cardioprotective (Schulz et al., 2008). However, there is a substantial chink in NOS’s cardioprotective armor: eNOS requires a cofactor, tetrahydrobiopterin (BH4), to facilitate the transfer of electrons in order to produce ^•NO. The physiological consequence is that when BH4 levels are reduced, the rate at which this electron transfer occurs is slower than the rate of oxidative degradation, which effectively causes NOS to produce superoxide (Forstermann and Munzel, 2006). BH4 can be oxidized to an unusable form by ROS, which sets the stage for NOX-derived ROS to further “kindle” the production of other reactive intermediates by encouraging the uncoupling of NOS enzymes. In fact, superoxide is regarded as somewhat of a direct antagonist of nitric oxide (Gryglewski et al., 1986; Daiber et al., 2017b). Decreased BH4 levels in response to noise exposure were so far not reported.

In addition to cofactor BH4 availability, eNOS is tightly regulated through redox mechanisms (Schulz et al., 2014; Daiber et al., 2017a). The redox status of eNOS greatly impacts its synthase activity and can be modulated by the presence of oxidative stress. There are several sites for phosphorylation which can either enhance or decrease the synthase activity of eNOS, however, the most common readouts of eNOS activity are at Ser1177 (Akt-dependent positive effect Dimmeler et al., 1999) as well as Tyr657 and Thr495 (both negative effects) (Figure 3). Importantly, all phosphorylations are redox-sensitive and stimuli-dependent, which is well-established for the protein tyrosine kinase 2 (PYK-2)-dependent phosphorylation at Tyr657 (Fisslthaler et al., 2008; Loot et al., 2009) and the protein kinase C (PKC)-mediated phosphorylation at Thr495 (Fleming et al., 2001; Lin et al., 2003) as both kinases can be activated by hydrogen peroxide. In the presence of oxidative stress, eNOS can also undergo S-glutathionylation, leading to uncoupling (Chen et al., 2010; Karbach et al., 2014). Finally, peroxynitrite appears to have the ability to release zinc from the zinc-thiolate complex coordinating eNOS monomers in the active dimer, representing another mechanism for uncoupling via oxidative stress (Zou et al., 2002).

In our own noise studies, we consistently reported an overexpression and overactivation of NADPH oxidase (NOX) enzymes (Munzel et al., 2017; Kroller-Schon et al., 2018; Steven et al., 2020). It is most likely that due to this overexpression of superoxide-producing enzymes, eNOS in the aorta (and nNOS in the brain) uncouples following noise exposure, as shown by dihydrethidium staining with eNOS inhibitor N^G-nitro-L-arginine methyl ester (L-NAME) (Kroller-Schon et al., 2018; Steven et al., 2020; Frenis et al., 2021). The paradoxical increase in eNOS protein expression and activating Ser1177 phosphorylation in mice exposed to 4 days of aircraft noise can be best explained by the presence of a largely uncoupled eNOS enzyme. Upregulation of an uncoupled eNOS and increased Ser1177 phosphorylation of an uncoupled eNOS would be detrimental through enhanced superoxide, largely compatible with the observed diminished NO bioavailability (Munzel et al., 2017; Kroller-Schon et al., 2018). However, we also found a reduction in activating phosphorylation at Ser1177 in hypertensive mice who were also exposed to 7 days of noise [mean 72 dB(A)] (Steven et al., 2020). Since oxidative stress was normalized in noise-exposed mice with Nox2 deletion (Kroller-Schon et al., 2018), it is presumptive that extinguishing the initiating spark of superoxide production from Nox2 was sufficient to prevent the uncoupling of eNOS in these mice. In the brains of noise-exposed mice, however, nNOS appeared to be downregulated and uncoupled, which was also preventable through the deletion of Nox2 (Kroller-Schon et al., 2018). In addition, we found eNOS S-glutathionylation in aorta and heart of noise exposed mice (Munzel et al., 2017) that was normalized in Nox2 knockout mice (Kroller-Schon et al., 2018) and was aggravated in an additive manner in noise-exposed hypertensive mice (Steven et al., 2020). Increased eNOS phosphorylation at Thr495 or Tyr657 in response to noise exposure was so far not reported but could be expected due to the redox-sensitivity of the kinases PKC and PYK-2 that confer these phosphorylations. Monomerization of eNOS due to zinc-sulfur complex oxidation in noise-exposed animals was so far also not observed.

In our own studies, noise exposure of mice resulted in decreased nNOS protein expression and triggered uncoupling of nNOS in cerebral tissue. Noise caused phosphorylation of nNOS at serine 847 (Kroller-Schon et al., 2018), which was previously reported to be associated with inhibited (Komeima et al., 2000) or even uncoupled nNOS enzyme (Kasamatsu et al., 2014). Of note, phosphorylation at serine 847 of nNOS is mediated by the redox sensitive calcium/calmodulin-dependent protein kinase (Kasamatsu et al., 2014). The oxidative stress signal in brains of noise-exposed mice could be also partially blocked by specific inhibition of nNOS by ARL-17477, which was in support of nNOS-derived ROS generation and compatible with uncoupling of nNOS enzyme (Kroller-Schon et al., 2018). Oxidative depletion of the protective neurotransmitter ^•NO also provides an explanation for the observed noise-induced neuroinflammatory phenotype, loss of the protective antioxidant transcription factor Foxo3, all of which contributes to the noise-induced cerebral oxidative stress (Kroller-Schon et al., 2018; Frenis et al., 2021). In addition, suppression of nNOS signaling and shift to a pro-oxidative/inflammatory phenotype of noise-exposed brains provides a feasible explanation for impairment of cognitive development (memory/learning) of school children exposed to high noise levels (Stansfeld et al., 2005). In line with this, impaired learning and memory in adult rats was also found to be associated with Nox2 activity (Kan et al., 2015).

We also found induction of endothelin-1 expression in the aorta of noise-exposed mice and also exacerbation of endothelin-receptor signaling as envisaged by more pronounced endothelin-1 dependent vasoconstriction (Munzel et al., 2017; Kroller-Schon et al., 2018). Importantly, endothelin-1 is not only one of the most potent endogenous vasoconstrictors but also a potent activator of Nox2 activity, by induction of gene expression (Duerrschmidt et al., 2000; Chen et al., 2012) and direct endothelin-receptor-dependent NADPH oxidase derived ROS formation—demonstrated by ex vivo stimulation with endothelin-1 or ROS suppression by ET_A-receptor blockade of vascular cells (Cerrato et al., 2012; Chen et al., 2012; Steven et al., 2018) or white blood cells (Steven et al., 2017). Endothelin-1 triggered NADPH oxidase-dependent ROS formation was also observed in different models of hypertension (Li et al., 2003a; Li et al., 2003b; Li et al., 2003c). Vice versa, it is also well established that oxidative stress conditions in general and Nox2-derived ROS formation in particular may increase the activity of the endothelin-1 promoter and thereby increase endothelin-1 expression (Kahler et al., 2000; Kahler et al., 2001). Given the cross-activation of Nox2 and endothelin-1, the stimulation of either pathway may lead to a vicious circle that contributes significantly to the cardiovascular oxidative stress and damage (Daiber et al., 2017a). Mitochondrial ROS can also stimulate the release of endothelin-1 as shown in pulmonary artery cells (Ouyang et al., 2012). Endothelin-1 shares also several cross-activation mechanisms with the renin-angiotensin-aldosterone system as evident from higher endothelin-1 expression levels in angiotensin-II treated hypertensive rats (Rajagopalan et al., 1997) and by decreased blood pressure as well as lower plasma angiotensin-II levels in hypertensive animals with bosentan (ET_A/B receptor blocker) therapy (Tran et al., 2009). Noise-driven renin-angiotensin-aldosterone system activation by stress hormones can lead to endothelin-1 upregulation or vice versa noise-triggered oxidative stress can stimulate endothelin-1 release and subsequently higher renin-angiotensin-aldosterone system activity. By these mechanisms, endothelin-1 may also contribute to the pronounced toxic effects of noise on Alport (Col4a3^-/-) mice who display glomerular dysfunction and hearing loss (Meehan et al., 2016).

The circadian clock regulates a number of essential biological functions such as sleep, body temperature, appetite, cognitive functions via time-dependent hormone release such as cortisol or melatonin (Van Laake et al., 2018). Circadian disruption has been identified as a risk factor for cardiovascular disease independently of noise (Crnko et al., 2019), but has also been associated with high (night-time) noise exposure burden (Eze et al., 2017; Munzel et al., 2020; Munzel et al., 2021a) or disrupted sleep pattern such as in shift workers (Furlan et al., 2000; Morris et al., 2016; Thosar et al., 2018). Importantly, given the context of the previous sections detailing the importance of oxidative stress in the adverse effects of noise exposure, redox mechanisms have also been implicated as important in the “redox control of cellular timekeeping” (Putker and O'Neill, 2016). Direct redox modifications of circadian components cryptochrome (CRY), period (PER), and F-box/leucine rich-repeat protein 3 (FBXL3) arise as thiol oxidation/reduction and the formation or disruption of zinc-sulfur complexes, which then control the binding of these components to the regulators circadian locomotor output cycles protein kaput (CLOCK) and brain and muscle Arnt-like protein 1 (BMAL1) complex, an essential part of the feedback mechanism inherent to circadian control (Figure 4) (Schmalen et al., 2014). While the direct redox modifications of clock components in the context of noise exposure have yet to be realized, other regulatory redox mechanisms also exist. Redox-sensitive kinases, histone deacteylases, stress-response proteins, and transcription factors can be modulated by the presence of ROS with further impact on the clock system (Figure 4) (Li et al., 2019a). The impact of various environmental stressors, including mental/social isolation stress, air pollution, heavy metals and pesticides on the circadian clock, especially its adverse redox regulation, was reviewed in (Li et al., 2020).

Bridging the concepts of circadian disruption by noise and redox control of the clock system, there is evidence of an important role in environmental cues and stressors in the regulation of the circadian rhythm (Li et al., 2019a). In mice exposed to continuous aircraft noise for 4 days [mean sound pressure level (SPL) of 72 dB(A)], expression patterns of key components of the circadian pathway in aorta and kidney were altered in comparison to unexposed controls (Kroller-Schon et al., 2018), including downregulation of Per1 and REV-ERB-α/β (Nr1d1/2) or RORα and upregulation of Bmal1, Cry1, Cul1, Prkag1/2, Parp1. In total, more than 30 circadian genes were altered in their expression levels. Downregulation of forkhead-box-protein O3 (FoxO3), a transcription factor that seemed to function as a central signalling hub regulating the circadian genes in the vascular tissue, was also reported. Pharmacological activation of FoxO3 by bepridil successfully prevented noise-induced oxidative stress in the aorta and the endothelial dysfunction that arises from it (Kroller-Schon et al., 2018). In a study of the transcriptomics of neurons within the inferior colliculus, a brain structure that has an important role in sound processing, distinct profiles between day and night-time exposure appeared in clock genes (Park et al., 2016).

The investigation of gut microbiota in the pathomechanisms of disease has experienced an explosion in recent years. The microbiome affects fundamental processes such as inflammation and redox signalling in the gastro-intestinal tract (Figure 5) (Campbell and Colgan, 2019). As a result, significant associations between the state of gut microbiota and cardiometabolic diseases have been made (Jones and Neish, 2017; Campbell and Colgan, 2019). Additionally, the existence of a gut-brain axis appears to be a central player for mood and behavior regulation as well as for the development of neuropsychiatric disorders and intestinal inflammatory disease (Collins et al., 2012; Cryan and Dinan, 2012). This may also be of particular interest for the present review as transportation noise is obviously also associated with a higher incidence of all-cause dementia, namely Alzheimer’s disease (Cantuaria et al., 2021). These states are noteworthy in the current context due to their known cardiovascular and mental risk (Hahad et al., 2019). Relatively few studies explicitly probe the relationship between noise exposure and alterations of the gut microbiome, but those that have been conducted show notable effects of noise. In one study of chronic exposure for 4 h/d during the sleeping phase over the course of 30 days [88–98 dB(A)], alterations of the microbiome-gut-brain axis were reported (Cui et al., 2018). The mice of the study were a model for Alzheimer’s disease, and chronic noise exposure was associated with cognitive impairment and amyloid beta peptide (Aβ) accumulation. The mice correspondingly had decreased neurotransmitter levels (5-HT and GABA), increased markers of neuroinflammation, and impaired intestinal and brain endothelial tight junction protein expression (e.g., claudins and occludin). Underlying these changes, alterations of the intestinal flora were revealed by 16S ribosomal RNA sequencing, which was supported by additional experiments utilizing fecal transplantation (Figure 5). Feces from mice exposed to 98 dB(A) noise were transplanted into unexposed mice, who subsequently developed an Alzheimer-like phenotype (Cui et al., 2018). Changes in the gut microbiome in a mouse model for Alzheimer’s disease were associated with an imbalance between intestinal pro-oxidative and antioxidant pathways as well as low-grade systemic inflammation in response to noise exposure (Chi et al., 2021).

Results of altered gut composition were also reported for noise-exposed rats in a similar experimental exposure (Cui et al., 2016), as well as an alteration in the balance of health-compromising proteobacteria and health-promoting actinobacteria as measured by 16S rRNAseq (Zymantiene et al., 2017). These noise-induced changes in microbial balance were accompanied by increased TNF-α and IL-1β as well as alterations of body weight, and haematological parameters as well as histopathological changes in the organs (Zymantiene et al., 2017). Anxiety-like behavior arose following noise exposure in rats, where higher serum corticosterone levels reflecting the increased stress response. Probiotic treatment alleviated these symptoms by apparently restoring the gut-brain axis (Figure 5) (Hadizadeh et al., 2019). Though the data up until now is rather sparse, these early findings indicate that noise disruption of the gut-brain axis through disturbance of the gut microbiota could be exacerbating the inflammatory phenotype that arises following noise exposure, which could potentially lead to cardiometabolic disease development (Karl et al., 2018).

Metabolic syndrome comprises a cluster of co-occurring conditions which lead to or complicate cardiometabolic disease. These conditions include hypertension, hyperglycemia, insulin resistance, dyslipidemia, type 2 diabetes, nonalcoholic fatty liver disease, and dementia. These conditions are all associated with oxidative stress via the increased production of ROS (Spahis et al., 2017a; Spahis et al., 2017b; Carrier, 2017). Notably, metabolic syndrome has known associations with the two modes of disturbance outlined in the noise reaction scheme: sleep and stress. Metabolic syndrome has a positive correlation with both abnormal sleep patterns, such as overly long or short duration (Smiley et al., 2019), sleep apnea (Borel, 2019), as well as circadian disruptions (Depner et al., 2014).

While we are unaware of any translational animal studies explicitly investigating metabolic syndrome as a cluster, there are several studies in mice and rats that focus on type 2 diabetes and insulin resistance. One such study found that diabetes induced by high fat diet was worsened by merely 4 h/d of 85 dB SPL noise exposure, as measured through glucose intolerance, insulin resistance, fasting hyperglycemia, and apparent dyslipidemia (Liu et al., 2018b). Another study found that in male mice, 4 h/d of 95 dB SPL noise exposure caused insulin resistance accompanied by phosphorylation of Akt, IRS1, and JNK, increased levels of circulating inflammatory cytokines TNF-α and IL6, and increased SOD and catalase activity, indicative of oxidative stress (Liu et al., 2018a). Insulin resistance was also documented in noise exposure of 1, 10, and 20 days (Liu et al., 2016). Rats exposed to 28 days of 95 dB noise were also found to have increased corticosterone, triglycerides, total cholesterol, and altered the balance of lipoproteins (Morakinyo et al., 2019). These findings are also in agreement with observational studies on the prevalence and incidence of metabolic syndrome conducted in humans, both during occupational and other noise exposures (Huang et al., 2020; Khosravipour et al., 2020; Yu et al., 2020). Road traffic noise was also associated with incident diabetes in the population-based Danish Diet, Cancer and Health cohort comprising 57,053 participants (Sorensen et al., 2013). Overall, there is evidence of a possible link between high noise exposure and several components of metabolic syndrome, with a possible mechanistic link through stress and sleep disruption prompting the production of ROS, though more investigation is certainly required.

Epigenetic changes can modulate the development and progression as well as the severity of cardiovascular diseases by control of atherosclerotic processes (Ordovas and Smith, 2010; Kuznetsova et al., 2020). Epigenetic processes are largely redox-regulated (Mikhed et al., 2015; Kietzmann et al., 2017; Leisegang et al., 2017) and thereby noise-induced oxidative stress will most likely change the epigenetic landscape at multiple layers. We and others reported noise-induced changes of coding RNA by next-generation sequencing in models of non-auditory noise effects (Munzel et al., 2017; Kroller-Schon et al., 2018) and studies on hearing loss (Wei et al., 2020; Lavinsky et al., 2021). However, noise exposure, sleep deprivation and mental stress can also lead to altered expression patterns of non-coding RNA, e.g., in microRNAs that have significant health impact (Miguel et al., 2018; Miguel et al., 2020). The dysregulation of microRNAs can be mediated by the indirect pathway, e.g., the known stress response, but also via direct mechanical damage of the inner ear during hearing loss (Miguel et al., 2018). Higher expression levels of miR-134/183 in the central amygdala were observed after acute stress exposure (Meerson et al., 2010). Both microRNAs seem to have significant health impact as they were found at higher concentrations in patients with coronary artery disease and depression. Numerous of these microRNAs that are associated with environmental risk factors such as noise exposure or mental stress are either regulated by oxidative stress or themselves influence gene transcription encoding for antioxidant defense or pro-oxidative proteins (Miguel et al., 2018; Miguel et al., 2020). Methylation of DNA bases is another epigenetic regulatory process with large impact on cardiovascular risk (Greco and Condorelli, 2015). Alterations of the DNA methylome, the sum of all methylated DNA bases with significant effect on transcriptional activity of DNA, were demonstrated in the brain of rats after chronic noise exposure, pointing towards epigenetic regulation of metabolic pathways by stress signalling in the form of noise exposure (Guo et al., 2017). Of note, a cohort study (SAPALDIA) conducted in Switzerland found an association between long-term exposure to transportation noise and DNA methylation patterns indicating activation of inflammatory pathways, alterations of cellular development and changes of immune responses (Eze et al., 2020). Epigenetic effects observed by human and animal studies on hearing loss but also epigenetic changes in non-auditory models were reviewed in (Leso et al., 2020).