Little is known about how musical stimuli impact on our gene expression. Navarro et al. (2021) have recently highlighted the importance of deeper exploration into this still poorly understood field of biological science [sensogenomics (Navarro et al., 2021; Gómez-Carballa et al., 2023, 2025; Navarro et al., 2023; Salas et al., In press; Cavenaghi et al., 2025; Castelo-Martínez et al., 2025)], taking advantage of new technologies emerging in the ‘-omic’ sciences, including genomics and transcriptomics. There have been only a few attempts to understand the gene expression mechanisms activated during music stimulation; the initial ones were carried out on healthy controls and professional musicians, with results indicating a few genes differentially expressed after exposure to classical music stimuli (Kanduri et al., 2015; Järvelä, 2018). Recently, the exploratory study by Navarro et al. (2023) provided suggestive evidence of the potential impact of music in the context of Alzheimer’s disease (AD) and reviewed the current evidence overall, concurring on the beneficial effect of music on neurodegenerative diseases. To the best of our knowledge, our most recent study, Gómez-Carballa et al. (2023), is the only attempt to date to examine the impact of music in a disease context, specifically in capillary blood samples collected from age-related cognitive disorder (ACD) patients. This study found an increased effect of music in ACD patients compared to healthy controls, but most interestingly, it revealed that brief musical stimuli can modify the way patients express genes typically altered in this condition, but in the opposite direction.

Building upon previous findings, the present study represents the first attempt to analyze the impact of music in ACD patients, but this time exploring salivary instead of blood transcriptomes. The interest in analyzing saliva stems from the growing importance of this non-invasive biological source in biomedical studies, with several salivary biomarkers being explored as proxies for diagnosing and monitoring brain health, stress, mental disorders, and neurological diseases (Farah et al., 2018; Bauduin et al., 2021; Ali and Nater, 2020; Engeland et al., 2019; Schepici et al., 2020). Saliva is primarily produced and secreted by the parotid, submandibular, and sublingual salivary glands and regulated by the autonomous nervous system (Chibly et al., 2022). These major salivary glands are innervated by both sympathetic and parasympathetic nerves, with compact fibers encircled by Schwann cells (Garrett and Kidd, 1993). The submandibular and sublingual glands are responsible for most unstimulated saliva production, whereas most of parotid gland saliva secretion occurs in response to stimuli. Previous studies have reported the presence of neurotransmitters in salivary glands extracted from mice and rats (Murai et al., 1998; Murai et al., 1995). Parasympathetic stimulation promotes the release of the neurotransmitter acetylcholine whereas sympathetic activation releases noradrenaline stimulating the secretion of proteins (Nakamura et al., 2004; Proctor and Carpenter, 2007). Given this connection between the salivary glands and the nervous system, the existence of a bidirectional oral-brain axis has been suggested (Sansores-Espana et al., 2021), through which an inflammatory response in the oral cavity may impact brain homeostasis and vice-versa. Since there is growing evidence indicating an impact of music on blood transcriptomes as well as its beneficial effects on many disease conditions, it seems imperative to explore if musical stimuli have also the potential to regulate salivary transcriptomes.

Here, we propose exploring gene expression patterns in saliva obtained from ACD patients and healthy controls before and after brief musical stimuli. The samples analyzed in this study partially overlap with those used in Gómez-Carballa et al. (2023); and all of them were collected during the same experimental concerts. Therefore, this scenario offers a unique opportunity to evaluate the potential of saliva analysis in capturing expression changes triggered by music, and to undertake a comparative analysis with the capillary blood signatures reported in our previous study (Gómez-Carballa et al., 2023).

Musical stimulation is a multifaceted cognitive phenomenon that intricately engages various brain regions, eliciting a spectrum of cognitive, emotional, and physiological responses. Elucidating the molecular changes triggered by music can help to understanding its effect on brain function and mental health in the general population, as well as its therapeutic potential in the context of various neurological and psychiatric conditions. Moreover, disentangling the complexity of gene networks (and molecular pathways involved in these networks) could reveal new targets for pharmacological interventions or provide guidance to develop new personalized music-based therapies.

Recently, we demonstrated that musical stimuli have an important impact on the capillary blood transcriptomes of ACD patients and healthy individuals, providing insights into the systemic gene expression response to music. We reported that music stimulation in ACD patients compensates for the expression of genes and pathways dysregulated due to cognitive impairment. Elaborating on this groundwork and given the known connection between the oral cavity and the brain, namely oral-brain axis, we aimed for the first time to investigate gene expression changes elicited by music in saliva samples in healthy donors and ACD patients. For this purpose, we have followed the same experimental design that has already been successfully used in our previous study on capillary blood samples, but this time employing a specific saliva collection device and a hybridization-based and PCR-free nCounter assay from NanoString. Currently, this is the most appropriate technology to deal with transcriptomes isolated from saliva samples, which are usually enriched with abundant genetic material from microbial species and poor in terms of quantity/quality.

Overall, the results suggest that music significantly impacts on the salivary transcriptomes of patients and controls; with this impact being higher than in the capillary blood transcriptomes of donors (as evidenced by the number of DEGs captured, both non-adjusted and adjusted; Supplementary Table S2). This finding highlights, for the first time, the relevance of the host transcriptome response to music in saliva. Additionally, the results indicate that the relative impact of music on individual genes may vary considerably between tissues, leading to specific genes being upregulated in one tissue but downregulated in another; this differential behavior between tissues is consistent with responses to other significant external stimuli in the host, such as an infection (Gómez-Carballa et al., 2022).

Three significant observations from the present study on the impact of music on salivary transcriptomes have already been reported for capillary blood transcriptomes (Gómez-Carballa et al., 2023), namely, music elicits: (i) greater transcriptomic changes in ACD patients than in controls, (ii) a transcriptomic response toward upregulation in ACD patients compared to healthy donors, and (iii) music modifies the salivary expression of a few genes that are known to be altered in AD/MCI conditions. Despite the relatively low overlap existing between DEGs from saliva and blood capillary samples, KDM6B (Lysine Demethylase 6B) and TIMP2 (TIMP Metallopeptidase Inhibitor 2) genes emerged as the only common DEGs (adjusted p-value < 0.05) in both tissues from ACD patients. TIMP2 was downregulated in both tissues whereas KDM6B showed opposite regulation patterns between tissues after musical stimuli. Furthermore, these genes were also found significantly upregulated in MCI/AD patients compared to healthy controls (Supplementary Figure S1). As reported in the literature, both KDM6B and TIMP2 are required for a normal brain function, as alterations in these genes might lead to neurological conditions. TIMP2 protein regulates extracellular matrix (ECM) remodeling and is particularly enriched in the hippocampus in comparison to other TIMP proteins (Castellano et al., 2017). In the adult brain, TIMP2 participates in neurogenesis, neuronal differentiation and in hippocampus-dependent memory (Ferreira et al., 2023; Perez-Martinez and Jaworski, 2005). Moreover, TIMP2 level disturbances were found across several neurodegenerative disorders such as AD (Lorenzl et al., 2003; Wang et al., 2020; Aksnes et al., 2023). KDM6B cooperates with Tau in regulating synaptic plasticity and cognitive function (Wang et al., 2022) and is present in excitatory neurons. Deleting KDM6B in neurons led to impaired synaptic activity, resulting in learning and memory deficits in mice (Wang et al., 2022). Indeed, KDM6B has been recently reported as a risk gene for intellectual disability (Stolerman et al., 2019), highlighting its importance for an adequate brain activity.

In ACD patients, the CXCL8 and LGALS3 genes emerged as the most upregulated and downregulated genes in TP2, respectively. CXCL8 gene encodes the pro-inflammatory IL-8 cytokine and is primarily expressed in neurons, astrocytes, and microglia in the nervous system. Chronic inflammation has been reported to be a pivotal factor in AD development (Xia and Hyman, 1999), even from an early stage of the disease progression (MCI) and potentially preceding clinical symptoms. Significantly higher levels of CXCL8 have been reported in cerebrospinal fluid (CSF), brain and plasma from AD patients in comparison to levels from healthy controls (Alsadany et al., 2013; Ashutosh et al., 2011; Correa et al., 2011; Galimberti et al., 2006), suggesting a probable detrimental role of CXCL8 in AD. CXCL8 has been negatively correlated with cognitive scores from AD patients (Alsadany et al., 2013), and positively correlated with CSF amyloid beta (Aβ) levels (Correa et al., 2011). However, a neuro-protective role of CXCL8 in AD has also been suggested. While stimulation with Aβ triggers CXCL8 production, it exhibits neuroprotective effects against Aβ-induced toxicity, possibly through a CXCL8-induced intracellular signaling and the production of neurotrophic factors, such as brain-derived neurotrophic factor (Ashutosh et al., 2011). Thus, it is tempting to interpret that the upregulation of the CXCL8 gene observed in saliva might represent a neuroprotective response triggered by music in the brains of ACD patients. Nevertheless, the specific role of CXCL8 in AD pathogenesis is still unclear, and the disparity in results reporting both neuroprotective and detrimental effects may reflect issues related to, e.g., different experimental conditions and tissues.

The main pathological characteristics of AD include the formation of Aβ plaques and neurofibrillary tangles, neuronal loss, inflammation, oxidative stress, and microglial activation. Galectin-3, encoded by LGALS3 gene, has been extensively associated with the activation of microglial cells around Aβ plaques in AD, indicating a major involvement in disease pathogenesis (Boza-Serrano et al., 2019; Holtman et al., 2015; Krasemann et al., 2017). Molecular signatures of microglial activation in AD and aging have been described LGALS3 as one of the most upregulated genes (Krasemann et al., 2017; Keren-Shaul et al., 2017). Although there are no data on the Galectin-3 measurements in microglia from MCI patients, elevated galectin levels have also been reported in the serum of both AD and MCI patients (Ijsselstijn et al., 2013; Ma et al., 2020; Yazar et al., 2021; Wang et al., 2015), suggesting a role of Galectin-3 in the disease and/or a risk factor for disease development. These pieces of evidence support the usefulness of Galectin-3 as a biomarker for AD, and its inhibition could have significant therapeutic benefits (Boza-Serrano et al., 2019). Therefore, the under-expression of LGALS3 might suggest a beneficial effect of music in ACD by compensating for the pathological effect of LGALS3 over-expression due to the disease condition.

The top DEG in healthy controls, the Thimet Oligopeptidase (THOP1) gene, showed a significantly lower expression after the musical stimuli. THOP1 is responsible for encoding a metallopeptidase, which participates in the metabolism of different neuropeptides expressed in neurons and glial cells, and plays a role in the brain neuropeptide degradation (Kim et al., 2003). Dysregulation of THOP1 has been associated with an unbalance in dopamine and serotonin turnover (Ferro et al., 2020) and it is widely understood that listening to music can influence both dopaminergic and serotoninergic pathways. Thus, under-expression observed in the THOP1 gene may be indicative of a music-mediated regulation of these neurotransmission systems. In addition, some studies have reported an over-regulation of THOP1 in brain and CSF of AD patients (Del Campo et al., 2023; Shi et al., 2020), indicating a potential association with Aβ-mediated toxicity. However, this over-expression of THOP1 in AD patients has been attributed to a protective response against Aβ toxicity (Pollio et al., 2008).

Pathways analysis of DEGs suggests an influence of musical stimuli on unsaturated fatty acid metabolism in ACD patients. The brain, predominantly composed of lipids, necessitates proper lipid homeostasis for a normal brain function and development. Within the brain, unsaturated fatty acids, particularly polyunsaturated fatty acids (PUFAs) govern critical processes such as cell survival, neurogenesis, brain inflammation and synaptic function (Bazinet and Laye, 2014). With normal aging, there is a decline in cholesterol and PUFAs levels in lipid rafts, affecting cell–cell communication, signal transduction, and synaptic plasticity. However, this reduction is significantly more prominent in AD and other neurodegenerative diseases, leading to dysregulation in unsaturated fatty acid metabolism, increased Amyloid Precursor Protein (APP) processing, and rapid formation of Aβ aggregates (Grassi et al., 2020; Snowden et al., 2017). Reduced levels of unsaturated fatty acids have been detected in the brain and plasma of AD patients (Snowden et al., 2017; Cunnane et al., 2012). Interestingly, various therapeutic approaches targeting lipid metabolism are being considered in the context of AD (Tong et al., 2024); our finding indicating a role of music in lipid homeostasis deserves further exploration.

Co-expression modules analysis pointed out to an impact of music on the salivary transcriptomes, higher in ACD than in healthy controls (both in correlation values and lower p-values). Overall, the molecular response to music was characterized by a stronger involvement of both adaptative and innate immune systems in healthy controls than in ACD patients. Specifically, the over-regulated GZMA module was found to be engaged in several T-cell adaptative related responses whereas the under-regulated SERPINA1 module participates in different innate processes, such as IL-1 and TLR signaling. Interactions between the immune and nervous systems are bidirectional, with each directly influencing the behavior of the other. Furthermore, both the adaptive and innate immune systems play complex and dynamic roles in learning and memory, brain function, and neurostimulation and collaborate closely to preserve immune homeostasis (Filiano et al., 2015). Dysregulation of these immune components can have significant repercussions on brain function and development, underscoring the importance of proper immune regulation in maintaining neurological health. For instance, microglia are macrophage-like innate resident immune cells of the central nervous system (CNS) with essential functions in the brain, ranging from immune surveillance and response to synaptic pruning and neuroprotection (Paolicelli et al., 2011; Chen et al., 2014; Ransohoff and Cardona, 2010). Microglia interact with neurons modulating synaptic transmissions and, therefore, directly influencing synaptic plasticity and neuronal excitability (Pascual et al., 2012; Zhan et al., 2014). Dysregulation of microglial activity has been implicated in neurological and neurodegenerative disorders, such as AD (Scheltens et al., 2021). Another innate immune component, the TLR family, is expressed in microglia, astrocytes and oligodendrocytes, but neurons also express TLRs, regulating proliferation, differentiation, outgrowth and neuron survival (Okun et al., 2011). TLRs cascades are involved in different brain-related functions contributing to neurogenesis, modulation of CNS plasticity and learning (Rolls et al., 2007; Okun et al., 2010). Disturbances in TLR signaling might have either a negative or positive impact on nervous system homeostasis. Studies also provide evidence of the important role of cytokines considered pro-inflammatory, such as the innate immunity mediators IL-1 and TNF, for normal synaptic function (Ben Menachem-Zidon et al., 2011; Yirmiya and Goshen, 2011). Similarly, adaptative immunity T-cells have showed to be important for normal brain functioning with a beneficial role in cognition and behavior (Kipnis et al., 2012; Brynskikh et al., 2008; Radjavi et al., 2014). A decreased number or dysfunction of T-cells may contribute to the etiology of different neurological disorders like autism or AD.

Notably, CD59 gene emerged as both one of the top downregulated DEGs and the hub gene of the under-regulated module in ACD patients. CD59 is a glycoprotein that plays a crucial role in regulating the complement system by preventing the formation of the Membrane Attack Complex (MAC). In the brain, complement proteins regulate neurodevelopment, neural migration, proliferation and synaptic pruning. Downregulation of CD59 (Yang et al., 2000) and upregulation of the complement system have been reported in several studies involving AD mouse models and brain tissue from AD patients (Hammond et al., 2019; Shi et al., 2017; Reichwald et al., 2009). However, it is still unclear if the changes in complement activity observed in AD are harmful or beneficial, as studies report both neuroprotective and neurodegenerative roles (Shi et al., 2017; Maier et al., 2008; Benoit et al., 2013; Toledo et al., 2014; Brucato and Benjamin, 2020). Therefore, the biological interpretation of the expression changes induced by music is complex. The two downregulated modules in ACD patients and controls (CD59 and SERPINA1) showed some functional commonalities, with intracellular transport and autophagic pathways emerging as common processes associated with the musical stimuli in both modules. Autophagy promotes the clearance of misfolded proteins and pathological aggregates, and in the brain, helps to maintain neuronal cellular morphology and physiological activities for a proper CNS function, prevents cellular toxicity and plays a crucial role in synaptic plasticity (Haynes et al., 2015). In fact, deficient autophagic machinery is one of the most relevant hallmarks found in neurodegenerative diseases like AD (Filippone et al., 2022; Zhang et al., 2021). The effect of music on the downregulation of modules involved in autophagy could be related to a local response, as these pathways are crucial for the homeostasis of most, if not all, tissues. However, this issue deserves further investigation in future studies due to the significant role of autophagy in neurodegenerative processes.

After our initial and recent attempt to investigate the impact of music on neurodegenerative diseases, the present follow-up study is pioneering in revealing several aspects: (i) It demonstrates that music has the ability to influence the expression patterns captured from saliva donors, (ii) It establishes parallels between gene expression observed in saliva and blood; (iii) It highlights that music has a stronger impact on the transcriptome of ACD patients compared to healthy individuals, as, e.g., measured by the number of DEGs altered by music, and (iv) It reveals that music overall triggers the upregulation of gene expression in patients compared to healthy controls.

Among the limitations of the present study, we echo those already discussed in our previous study on sensogenomics22 experimental concerts carried out on capillary blood samples (Gómez-Carballa et al., 2023). Additionally, we acknowledge the challenge posed by the analysis of saliva in RNAseq analysis. Fortunately, the methodology employed in the present study, although it evaluates a lower number of genes for expression, offers the advantage of being a gold standard for gene expression studies. Therefore, the high quality of the gene expression results provided by NanoString compensates somewhat for the limitation of analyzing fewer genes.

The present study has demonstrated the power of short-duration musical stimuli in modifying the salivary transcriptome of ACD patients and healthy donors. The impact of music on saliva tissue is comparable to, or even greater than, that observed in blood with a more pronounced effect seen in patients than in healthy controls. Of note is the discovery that music influences the expression of genes and modules commonly altered in neurodegenerative diseases, a finding that may help to elucidate the known beneficial effects of music as reported by specialists in neuroscience and cognitive sciences. Further efforts to validate these findings in larger cohorts and other disease scenarios and to explore the impact of music not only on the gene expression level but also on other ‘-omic’ layers are warranted.