The ongoing COVID-19 pandemic has aroused some interest in taste and olfaction but even so, anosmia and dysgeusia are considered curious symptoms that at best may help in early differential diagnosis from benign respiratory tract infections (Menni et al., 2020) and, hopefully not, as a possible sign of subsequent central nervous system involvement, related to the so-called Long-COVID (Yong, 2021).

Chemical senses are commonly considered little more than evolutionary relics in humans. In everyday life, taste and olfaction are deemed relatively unimportant for many persons. As a matter of fact, often the impairment of olfaction was not even perceived by patients themselves (Croy et al., 2012), at least in pre-COVID-19 era. Olfaction or taste impairments, which may be quantitative or qualitative (Ercoli et al., 2021) are usually considered only with respect to the conscious awareness of these sensory inputs. Anyhow, these impairments may have important consequences on nutrition, social behavior, and mood (Croy et al., 2010, 2014a), as well as on work and quality of life (Brämerson et al., 2007). Actually, all the cells in our body may detect the presence of different molecules in the surrounding environment, yet only a few can use the detection of some chemicals to inform the central nervous system for organizing adaptive nervous or neuroendocrine responses that may impact the whole body. In addition to olfaction and taste, other cells monitor a wide range of chemicals inside our body: carotid bodies, solitary chemoreceptor cells, pulmonary neuroendocrine cells, and enterochromaffin cells. Furthermore, other ill-defined chemosensory cells are scattered throughout our body. Of all these inputs, we are rarely aware. In this review, we will briefly explore all these chemoreceptor systems and highlight their interconnections under physiological and pathological states. The following sections will introduce the chemical senses, their central connections and describe their involvement in human pathophysiology with the aim to highlight the intermingled contribution of different chemical receptors to health and homeostasis maintenance.

Afferent pathways from the peripheral chemosensory cells to the central nervous system can be structured in two sections, one entering from the brainstem and one from the olfactory bulb, but notably, inside the brain, they largely converge on the same areas (see Table 2 for a summary).

Chemosensory systems, including taste and olfaction, have many intricated subcortical connections also in the primate brain and, regardless of sensory information they convey, they are in a position to take part in shaping the integrated responses to external and internal stress conditions, under almost every instance. These responses involve cardiovascular, respiratory, endocrine, and immunologic functions that are essential for coping effectively with stress, and hence for survival. A brief outline of chemosensory central nervous pathways will highlight their complex connections.

Humans have higher olfactory thresholds than macrosmatic mammals, although not so poor as commonly thought. Yet, when it comes to olfactory pathways, human subcortical connections are extensively developed, similar to other mammals. In other mammals, different sensory organs are hosted in the nasal cavity, including vomeronasal, septal organ, and Grueneberg ganglion, all projecting to the (main or accessory) olfactory bulbs (Weiler and Farbman, 2003; Matsuo et al., 2012; Mucignat-Caretta, 2021), while in humans, olfactory fibers originate almost exclusively from main olfactory mucosa and run through the first cranial nerve, the olfactory nerve, and contact olfactory bulb mitral cells, the main collector of afferents. From here, olfactory projections radiate mainly, yet not exclusively, to three amygdala subregions, possibly involved in the setup of motor response to olfactory stimuli (medial amygdala), reward and memory processes (cortical amygdala), and multisensory integration (periamygdaloid complex; Noto et al., 2021). Notably, amygdala expresses the acid-sensing ion channel ASIC1a, which centrally senses hypercapnia and acidosis and drives the subsequent fear responses in mice (Ziemann et al., 2009), while the human hortolog ACCN2 is associated with amygdala structure and function in panic disorders (Smoller et al., 2014). The human orbitofrontal cortex is the final recipient of sensory information and is involved in reward and emotional value of sensory stimuli, including olfaction and taste (Rolls, 2021), and decision-making in olfactory-driven food intake (Seabrook and Borgland, 2020). Other projections reach the olfactory nucleus, taenia tecta, entorhinal and piriform cortex, the so-called olfactory cortices, and olfactory tubercle, which appears to modulate odor-guided eating (Murata, 2020). Interestingly, also trigeminal stimuli can activate the piriform cortex, converging on the same chemosensory integrative center after entering the brain from the hindbrain (Hummel and Frasnelli, 2019). Moreover, all these areas receive modulatory dopaminergic inputs from the midbrain ventral tegmental area involved in reward (Oades and Halliday, 1987; Deutch et al., 1988). From the entorhinal cortex, olfactory information then reaches the hippocampus, while subsequent projections from the amygdala run to the hypothalamus (premammillary nuclei, medial preoptic area, and ventromedial nucleus) and bed nuclei of stria terminalis, thus involving, according to old terminology, the limbic system, well known for controlling emotional behavior. Interestingly, the bed nuclei of the stria terminalis are involved in anxiety behavior triggered by hypercapnia and acidosis (Taugher et al., 2014). The bed nuclei of stria terminalis and hypothalamus are in turn both connected to many other brain areas, including the brainstem nuclei (Mucignat-Caretta, 2021), and are involved in linking inflammation to complex homeostatic behaviors, like feeding and its dysregulation (Guillemot-Legris and Muccioli, 2017; Wang Y. et al., 2019). The hypothalamus appears thus as a chemosensory hub that regulates not only emotional and social, often competing, behavior but also visceral homeostatic functions like glycemia, appetite, energy expenditure, blood pressure, heart function, respiration, and immune system through neurohormonal and sympathetic activation to downstream hindbrain targets (Henderson and Macefield, 2021; Khodai and Luckman, 2021; Nakamura et al., 2022).

At variance to olfaction, most other chemosensory afferent fibers enter into the brainstem through the 5th (trigeminus), 7th (facial), 9th (glossopharyngeal), and 10th (vagus) cranial nerves. From the middle of the XIX century, brainstem is known to regulate mammalian metabolism, after Claude Bernard’s so-called piqûre diabétique, actually transient hyperglycaemia and glycosuria, triggered by the puncture of the floor of the fourth ventricle (Li, 1952; Feldberg et al., 1985); hence it is not surprising that the center that orchestrates visceral output according to sensory inputs resides in the brainstem. This center is the nucleus of the solitary tract.

Gustatory afferents converge in a rostro-caudal topographic way through facial, glossopharyngeal, and vagus nerves on the nucleus of the solitary tract and then relay (in rodents) to the pontine parabrachial nuclei, and the motor nuclei of cranial nerves to mediate taste-induced reflex responses (Rolls, 2019). Noteworthy, there is a large anatomical overlap between visceral chemosensory afferent fibers and taste afferents in peripheral nerves. Most afferent fibers from the carotid body and tongue posterior taste bud are already intermixed, both running through glossopharyngeal nerve and petrosal ganglion (Leonard et al., 2018). Afferent vagal fibers carry sensory information from pulmonary neuroendocrine cells (van Lommel et al., 1998) and express the purinergic receptor P2RY1, while other vagal sensory afferents from the lung express both NPY2 receptor and the capsaicin receptor TRPV1 (Chang R. B. et al., 2015). Afferent vagal fibers also carry information from some taste buds in the palate and epiglottis, solitary chemosensory cells in the lower airways (Krasteva et al., 2011), enterochromaffin cells (Hagbom et al., 2011), and glucose-sensing neurons (Browning, 2010; Grabauskas et al., 2010), and also from the aortic bodies (Piskuric and Nurse, 2013). Some afferent vagal fibers from the gut express olfactory receptors and may participate in the control of feeding with mechanosensitive vagal afferents (Bai et al., 2019). Notably, astrocytes within the nucleus of the solitary tract respond to glucoprivation and participate in increasing the gastrointestinal motility (McDougal et al., 2013). Afferent trigeminal fibers carry information from oral and nasal solitary chemosensory cells (Carey et al., 2016). Of note, the trigeminal spinal nucleus receives inputs from cranial nerves 5, 7, 9, and 10.

Then, from the nucleus of the solitary tract, through the ventroposteromedial (VPM) nucleus of the thalamus, taste inputs reach the gustatory cortex, namely the anterior insula in the temporal lobe and the frontal operculum (Rolls, 2019), and terminate in the orbitofrontal cortex, which integrates multiple sensory inputs, including olfactory (see above, Rolls, 2021), and ultimately affects food intake and visceral modulation, besides memory (Vincis and Fontanini, 2019). However, a second bidirectional connection originates from the nucleus of the solitary tract to subcortical targets: the lateral hypothalamus, central nuclei of amygdala, and bed nuclei of the stria terminalis, and areas involved in feeding, autonomic modulation and emotional control, as mentioned above.

Also, the carotid bodies project to the nucleus of the solitary tract, which is bidirectionally connected with the retrotrapezoid nucleus in the rostroventral medulla oblongata: this nucleus also contains central chemoreceptors to detect local hypercapnia or acidification, including the proton-modulated TASK-2 potassium channels and the proton-activated GPCR GPR4 that drive the respiratory reflexes. Also, in these responses astrocytes are involved to modulate neuronal activation (Guyenet et al., 2016). Connexin 26 hemichannels are present in astrocytes within this nucleus: they are gated by CO₂ and participate in the centrally-driven ventilatory response to hypercapnia (van de Wiel et al., 2020). Electrical coupling driven by connexins was also found in the nucleus of the solitary tract under hypercapnic acidosis (Reyes et al., 2014).

The nucleus of the solitary tract, which is embedded in brainstem reticular formation, receives an astonishing number of different, not only chemosensory, afferents from basically the entire body through facial, glossopharyngeal, and vagus nerves. From a physiological point of view, it is a crucial hub, integrating incoming sensory signals with hypothalamic outflow, and being potentially connected to the whole brain and body, it is the ideal anatomical center to drive autonomic, neurohormonal, and behavioral responses to stress (Holt, 2021), both acute and chronic (Herman, 2018).

Dysfunction of carotid body/glossopharyngeal nerve has also been postulated in the potential pathogenesis of non-respiratory diseases like metabolic syndrome-type 2 diabetes and hypertension (Kim and Polotsky, 2020; Sacramento et al., 2020). In the past, vagal dysfunction has been hypothesized in luetic pain (Mylona et al., 2010; Klein and Ridley, 2014), gastroduodenal ulcer (Szabo et al., 2016; Zalecki et al., 2020), and more recently in obesity-metabolic syndrome (Berthoud and Neuhuber, 2019), but on the whole, in clinical practice, little attention is currently paid to the involvement of chemosensory systems in these diseases. A possible explanation is that chemosensory systems are not considered essential for human life, and hence have a minor position in medical training. Impairments in taste and olfaction in some cases are not even perceived by the patients themselves and do not affect body survival. With the exception of a few professionals, loss of olfaction or taste is rated very low by health insurance companies also.

Notwithstanding their importance, bilateral therapeutic ablation of carotid bodies in humans leads to apparently negligible side effects (Iturriaga et al., 2021). Therapeutic vagotomies have been performed in humans with only minor side effects (Szabo et al., 2016). The same is true for afferent trigeminal signals generated in the oral and nasal chemoreceptors (Barham et al., 2013), and a perturbation of this pathway seems to bear minor consequences. In mice, ablation of pulmonary neuroendocrine cells is apparently harmless (Song et al., 2012). These puzzling data suggest that the lack of a single chemosensory input is dispensable, or at least may be compensated by other inputs.

Interest in taste and olfaction as diagnostic and/or prognostic tools has been fuelled by the recent COVID-19 pandemic, during which many persons developed chemosensory symptoms. The loss of sense of smell apparently had a prevalence of 47%, ranging from 11 to 84%, due to differences in symptom detection (for a discussion see Karamali et al., 2022), with psychophysical testing being more reliable than self-report (Bordin et al., 2021). The actual prevalence of these disorders is hardly estimated because patients may not report paucisymptomatic disease or may not be aware of their disturbances. Actually, psychophysical testing detects more subjects with impaired sensitivity than subjectively reported (Hummel and Podlesek, 2021). In addition, with the spreading of vaccination, a difference in prevalence between pandemic waves could not be attributed to differences in virus variants since the appearance of symptoms in vaccinated people may depend on their immunological state. Moreover, virus sequencing requires expansion in cell cultures to obtain enough viruses for sequencing, which may introduce variations in genomic sequence, an issue rarely taken into account in the literature (World Health Organization [WHO], 2021, p. 18). Another critical point is the process of recovery, which may benefit from olfactory training (Addison et al., 2021; Denis et al., 2021), yet may be not smooth, since some olfactory symptoms like parosmias and phantosmias may emerge well after the end of the infection, and in some cases, olfactory recovery does not take place (Cecchetto et al., 2021; Ohla et al., 2022). The different chemosensory systems (taste, olfaction, carotid body, solitary chemosensory cells, pulmonary neuroendocrine cells) share common molecular characteristics, including some receptors and integrative centers (see above). Hence, it is conceivable that, in addition to taste, olfaction, and trigeminal chemesthesis (Parma et al., 2020), other chemosensory systems may also be affected by SARS-CoV2 virus. As a matter of fact, the virus has been detected in autoptical carotid bodies of patients with COVID-19 (Lambermont et al., 2021). Notably, ACE2, TMPRSS2, and Neuropilin-1, the three entry sites for SARS-CoV2, have been recently found in the glossopharyngeal and vagus nerves near the medulla (Vitale-Cross et al., 2022). Apparently, SARS-CoV2 enters the sustentacular cells in olfactory mucosa and destroys it (Bryche et al., 2020), opening its way to the endothelial and nervous tissues up to the central nervous system (Meinhardt et al., 2021), even if the direct involvement of the olfactory nerve seems implausible (Butowt et al., 2021).

It can be suggested that some peculiar symptoms of COVID-19 patients may be explained by the impairment of chemosensory systems. The “silent hypoxia,” that is the absence of dyspnea in the presence of life-threatening hypoxemia, may depend on the damage of carotid bodies; moreover, the impairment of other oxygen sensor systems like the pulmonary neuroendocrine cells may be involved, causing a lack of hypoxic pulmonary vasoconstriction (Caretta and Mucignat-Caretta, 2021). The gastrointestinal symptoms of COVID-19 patients are related to an increase in serotonin levels (Ha et al., 2021; Jin et al., 2021), presumably dependent on the dysregulation of enterochromaffin cell function, which usually produces the vast majority, roughly 90%, of gut serotonin (Koopman et al., 2021). An increase in blood serotonin concentration, again linked to impairment of enterochromaffin cell function, is reported in inflammatory bowel disease (El-Salhy et al., 1997) and in rotavirus diarrhea (Hagbom et al., 2011).

Also, the so-called cytokine storm, that is the dysregulation of innate immune response leading to COVID-19 ARDS catastrophic outcome (Hue et al., 2020), may be related to the impairment of vagal immune reflexes (De Virgiliis and Di Giovanni, 2020). The brain fine-tunes the anti-inflammatory pathway via cholinergic vagal efferents, whose stimulation may prevent fatal endotoxemia (Tracey et al., 2001; Andersson, 2005; Chavan et al., 2017). This is in addition to the extensive regulation of respiration and lung defensive reflexes constantly operated via mechano- and chemosensory vagal inputs (Mazzone and Undem, 2016).

Lastly, solitary chemosensory cells (Saunders et al., 2014) and enterochromaffin-pulmonary neuroendocrine cells may be involved in the inflammatory pathway or the immune modulation (Koopman et al., 2021; Pavón-Romero et al., 2021).

Loss of taste and olfaction are easily and non-invasively tested in vivo. It is not specific of COVID-19 disease, since it may be caused by many different chemicals, infections, traumatic, or neurodegenerative processes. However, also during physiological aging, a decrease in chemosensory performance is well-documented (Landis et al., 2004; Seubert et al., 2017), and increasing age is associated with the rising prevalence of smell disturbances (Bhattacharyya and Kepnes, 2015). Notably, dysfunction of olfaction correlates with increased mortality and with patient frailty in the general population, possibly mediated by inflammation (Liu et al., 2019; Laudisio et al., 2019). Conversely, dysfunction of olfaction and/or taste has been reported in many diseases encompassing dysregulation of multiple visceral functions. In addition to head trauma (Howell et al., 2018), tumors affecting the olfactory areas and epilepsy of the temporal lobe affecting olfactory areas, for which olfactory involvement is apparent, given below is a non-exhaustive list of diseases and conditions in which an impairment in olfaction (most often), taste, or both, has been demonstrated in human patients (see Table 3):

Of note, some of the above-mentioned diseases that may correlate with taste/olfaction impairment are risk factors for a fatal outcome of the SARS-CoV2 virus infection, for example, obesity (Khan et al., 2020), older age, heart disease, dementia, diabetes, and chronic liver or kidney disease (Singh et al., 2020; Kim et al., 2021; CDC, 2022, and references therein).¹

With few exceptions, only a minor diagnostic value has been attributed to taste/olfaction alterations. At best, a role has been recognized in modifications of food intake because of changes in food hedonic value. As a matter of fact, this variegated list of supposedly unrelated diseases might suggest that chemosensory impairment is an uninteresting side effect potentially caused by many vaguely defined agents.

Conceivably, it makes sense to reverse this interpretation: alterations of taste and olfaction may signal impairment also of other chemosensory systems, and hence have a causal part in worsening the clinical course of diseases because of (up or down) dysregulation of homeostatic mechanisms, which commonly rely on chemosensory inputs. As an example, it may be worth considering two strikingly opposite diseases, cachexia (which includes increased resting energy expenditure and anorexia) and obesity.

These multifactorial syndromes share common biochemical characteristics. In addition to alteration of food intake, similar metabolic changes are present as insulin resistance, wasting of muscle tissue, altered energy expenditure (in both diseases, elevated or reduced depending on the patient), increased lipolysis, unregulated excess protein catabolism, chronic inflammation (Mantovani et al., 1997; Zwickl et al., 2019), deregulated activation of the innate immune system (Tisdale, 2001; Laviano et al., 2017; Antuna-Puente et al., 2021); in most patients similar multiple endocrine dysfunctions are present and, notably, the elevation of peripheral serotonin levels (Crane et al., 2015; Minami et al., 2003). Both central and peripheral serotonin, from the enterochromaffin cells, appear as a major player in the regulation of metabolism (Yabut et al., 2019). All these parameters are, to a large extent, under control of chemosensory systems. A causal relationship is not yet established, but is worth considering.

Under ordinary life conditions, chemosensory impairment of a single system does not seem to be detrimental for survival. Its malfunctioning presumably is to a large extent compensated by the other chemosensory mechanisms. However, under stress conditions, COVID-19 pneumonia for example, and also altered glucose homeostasis (Bentsen et al., 2019), every chemosensory system may become vital for enabling the brain to organize an efficient functional homeostatic response that could effectively increase the survival rate. As an instructive example, the lateral parabrachial nuclei respond to hypoglycemia and are connected to the hypothalamic ventromedial nucleus and subsequently to the bed nucleus of the stria terminalis (Meek et al., 2016), all areas involved in chemosensation, and part of the central autonomic network that orchestrate the body integrated responses (Benarroch, 1993).

Under this perspective, loss of taste and olfaction may be considered not only as a diagnostic tool but also as a warning signal of the impairment of visceral chemosensory systems. Defective chemosensory information and the subsequent dysregulation of CNS homeostatic responses should be considered an important contributor to the higher death rate of COVID-19 in comparison to other ARDS (Hue et al., 2020). Similarly, impairment of chemical senses may have a crucial role in the above-mentioned non-infectious chronic diseases.

Smell (both orthonasal and retronasal), taste, and also chemesthesis can be easily tested in patients with different techniques, ranging from psychophysical to electrophysiological (Hummel and Podlesek, 2021), validated in humans (Doty, 2018; Whitcroft and Hummel, 2019). The most used psychophysical tests for olfactory function include the Sniffin’ Sticks (Hummel et al., 1997, 2007) and the University of Pennsylvania Smell Inventory Test – UPSIT (Doty et al., 1984) and its cross-cultural version (Doty et al., 1996). More recent tests include the food-associated olfactory test (Denzer-Lippmann et al., 2017) and the culturally independent retronasal test of olfactory function (Croy et al., 2014b). Taste is usually tested with taste strips (Landis et al., 2009; Doty et al., 2021). Psychophysical testing is implemented for trigeminal functions, also in an automated way (Hummel et al., 2016; Huart et al., 2019). The central processing of chemosensory stimuli has been investigated with chemosensory event-related potentials for olfactory and trigeminal stimuli (Kobal and Hummel, 1988) and with the more recent neuroimaging techniques, including PET and fMRI (Small et al., 2004; Gottfried and Zald, 2005; Barresi et al., 2012; Mantel et al., 2019), also using computerized pulse-ejection systems (Kumazaki et al., 2016; Mazzatenta et al., 2016).

When it comes to visceral chemosensory afferents, non-invasive validated procedures are largely lacking, perhaps because their dysfunctions are generally rated low on the list of clinical symptoms and they are not usually examined in ordinary clinical practice.

Procedures that may evaluate (not exclusively) carotid body function are “6 mins walking test” and hypoxic ventilatory response (Grove et al., 1995; Niewinski et al., 2021), which are well-validated, and newer tests are being proposed (Bhattacharyya et al., 2020). In addition, hypoxic pulmonary vasoconstriction and hypoxic or high-altitude bronchoconstriction depend (also) on serotonin release and the activation of pulmonary neuroendocrine cells, for which a specific test is lacking. For enterochromaffin cells and other visceral chemosensory systems, no functional test is available for humans. At present, for visceral chemosensory systems, examination of bioptical or autoptical material is the only available option. Nevertheless, putting chemosensory dysfunction on the clinical spot may help in understanding the patient’s symptoms and the underlying pathophysiology of complex diseases, involving multiple systems’ integrated response.

The different chemosensory systems should be considered as sharing common characteristics and being involved in the regulation of vital functions, also in humans, under physiological and pathological conditions. Many vital visceral functions are under the control of multiple, apparently overlapping mechanisms, as an example, oxygen partial pressure is measured in the plasma by carotid body cells, in bronchial air by pulmonary neuroendocrine cells, in extracellular fluid by yet unknown tissue receptors, while oxygen binding to hemoglobin is sensed by the aortic bodies. Conversely, none of these sensory systems are actually redundant, each measuring the same parameter under different conditions, fostering the proper functional adaptive responses. Furthermore, these chemosensory systems are not specific for respiratory gases, they measure also different chemical parameters acting as polymodal sensors. Further examples may be quoted from blood pressure control, glucose control, and food intake. Apparently, all these functions rely on a set of chemosensory inputs that are necessary for the brain to properly implement a coordinated homeostatic response.

AC and CM-C conceived and wrote the manuscript and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.