Regulation of Metabolic Health by an “Olfactory-Hypothalamic Axis” and Its Possible Implications for the Development of Therapeutic Approaches for Obesity and T2D

Abstract

The olfactory system is responsible for the reception, integration and interpretation of odors. However, in the last years, it has been discovered that the olfactory perception of food can rapidly modulate the activity of hypothalamic neurons involved in the regulation of energy balance. Conversely, the hormonal signals derived from changes in the metabolic status of the body can also change the sensitivity of the olfactory system, suggesting that the bidirectional relationship established between the olfactory and the hypothalamic systems is key for the maintenance of metabolic homeostasis. In the first part of this review, we describe the possible mechanisms and anatomical pathways involved in the modulation of energy balance regulated by the olfactory system. Hence, we propose a model to explain its implication in the maintenance of the metabolic homeostasis of the organism. In the second part, we discuss how the olfactory system could be involved in the development of metabolic diseases such as obesity and type two diabetes and, finally, we propose the use of intranasal therapies aimed to regulate and improve the activity of the olfactory system that in turn will be able to control the neuronal activity of hypothalamic centers to prevent or ameliorate metabolic diseases.

Introduction

Odor reception is an omnipresent primitive sensory mechanism in animals that allows the perception of volatile chemicals known as odorants. This system conveys information regarding the presence of familiar and unknown odors, conspecifics, potential mates, mother–offspring recognition, food sources, predators, and prey (Laska and Hudson 1993).

The central odor processing is achieved through a complex neuroanatomical network that includes the olfactory receptors in the nose, olfactory regions in the brain, higher integration cortexes and several hypothalamic nuclei (Nausbaum 1999; Klingler 2017; Meissner-Bernard et al. 2019; Nagayama et al. 2014).

In this review we describe how up-stream olfactory inputs affect hypothalamic activity and the regulatory mechanisms aimed to maintain energetic homeostasis. First, we discuss the neuroanatomy of the olfactory system to understand how odor information is processed and how it is able to reach the hypothalamus. Next, we analyze how the metabolic status of the body modulates olfaction, describing that the integration of humoral metabolic signals can also takes place in the olfactory system to adjust its olfactory capacity.

In the second part, we describe the influence of the olfactory system on hypothalamic activity, mentioning that the activation of the olfactory system promotes neuronal responses in several nuclei of the hypothalamus. Also, we describe the olfactory alterations found in metabolic diseases like obesity and type two diabetes highlighting the relevance of the olfactory system for the maintenance of metabolic balance. Finally, we discuss how the olfactory-hypothalamic relationship constitutes a potential target of study to develop noninvasive therapies obesity and diabetes, hence, enlisting the remaining questions that this research should answer to completely understand the exact mechanisms involved in the olfactory regulation of hypothalamic activity and to use this information for the development of therapies that target the interaction between both circuits.

Brief Neuroanatomy of the Olfactory System

There are important variations in the chemical structure, concentrations, and combinations of the odors able to activate specific mechanisms underlying odor detection and discrimination, which means that every element in nature has its own odorant stamp (Zou et al. 2009; Croy et al. 2015).

In mammals, the nose is the main olfactory organ, it consists of multiple olfactory subsystems including the main olfactory epithelium (OE) and the vomeronasal organ (VNO) (Trotier 2011). The OE has two types of cells: the microvillar cells and the olfactory sensory neurons (OSN), which express more than 1000 G-protein-coupled odor receptors arranged in a topological map (Nagayama et al. 2014). Each OSN responds to a single type of odorant because they only express one type of odorant receptor. Thus, complex odors, composed of more than two odorant molecules activate more than one olfactory neuron.

Olfactory neurons project to the olfactory bulb (OB), which integrates information from the OE and transmits it to other regions in the brain. The OB is organized into strict multiple layers of distinct cell types: juxtaglomerular (JG) cells that include the periglomerular (PG) cells, external tufted (ET) cells and superficial short-axon (sSA) cells, mitral cells, tufted cells, and granule cells (Tan et al. 2010; Kikuta et al. 2013). Both the OE and the OB are the primary olfactory regions.

In the OB, the first synapsis from the OE is established in the glomerular cell layer (GCL), which latter propagates to the cell bodies of tufted cells in the external plexiform layer (EPL) and mitral cells in the mitral cell layer (MCL), then horizontally back-propagates through the secondary dendrites in the EPL (Tan et al. 2010; Kikuta et al. 2013).

Finally, neurons in the EPL and the MCL send excitatory projections to the different areas of the olfactory cortex via the lateral olfactory tract where olfactory information is integrated (Nagayama et al. 2014; Zhou et al. 2019) (Fig. 1).

Fig. 1.

Neuroanatomy of the olfactory system. Volatile chemicals (odorants) enter the nasal cavity and are detected by olfactory sensory neurons (microvillar cells, blue, and olfactory sensory neurons, purple) in the main olfactory epithelium. Olfactory sensory axons enter the cribriform plate into the glomerular layer of the olfactory bulb where they synapse with periglomerular cells. The signal is propagated to the cell bodies of tufted cells in the external plexiform layer and mitral cells in the mitral cell layer. Finally, the neurons in the external plexiform and mitral layer send excitatory projections to the different areas of the primary olfactory cortex (Amygdala, Lateral entorhinal cortex, Olfactory tubercle, Piriform cortex, Anterior olfactory nucleus), via the lateral olfactory tract where the integration of olfactory information takes place. The system is mainly connected with the hypothalamus via poly-synaptic pathways from the olfactory bulb to the preoptic area of the hypothalamus. Other poly-synaptic pathways connecting the olfactory system with the paraventricular nucleus, dorsomedial nucleus, ventromedial nucleus and lateral hypothalamic area have also been described (see the text). Created with BioRender.com

The integration and organization of the olfactory information is quite complex and it implies different integration levels that have already been described in detail (Nausbaum 1999; Klingler 2017; Meissner-Bernard et al. 2019; Nagayama et al. 2014). In the next sections we will discuss the link between the olfactory and the hypothalamic system and its implication for the regulation of energy balance.

Metabolic Status Directly Modulates Olfactory Sensitivity

There is a fair amount of literature stating that the olfactory system is controlled by the metabolic status of the body (Price et al. 1991). The most clear example is that food deprivation can increase the firing rate of the neurons in the OB, increasing its response to food-related odors; this increase in activity is immediately damped after food intake (Apelbaum and Chaput 2003; Badonnel et al. 2012; Palouzier-Paulignan et al. 2012).

Many neuropeptides and metabolic hormones such as Gonadotropin-Releasing Hormone (GnRH), Neuronal Peptide Y (NPY), insulin, leptin, adiponectin and orexins are known to modulate the sensitivity of olfactory sensory neurons in different species (Martin et al. 2009).

Both the OE and OB express leptin (Ob-R), insulin (IR), ghrelin and adiponectin receptors (Baskin et al. 1983; Elmquist et al. 1998; Hass et al. 2008; Miranda-Martínez et al. 2017), suggesting that hormonal signals not only regulate energy balance by signaling into hypothalamic regions but may also control metabolism through direct modulation of olfactory perception (Riera and Dillin 2016).

On the other hand, insulin also modulates most steps of odor detection; at the level of the olfactory mucosa, and the increase of insulin levels in the OB reduces food odor-induced sniffing behavior in rats and olfactory sensitivity in humans (Aimé et al. 2012; Brünner et al. 2013).

The OB has the highest concentrations of insulin and insulin receptors (IR) in the brain (Edwin Thanarajah et al. 2019; Kleinridders et al. 2014). The loss of one allele of IR was reported to modify the electrical phenotype of mitral cells, without changing olfactory ability (Das et al. 2005). Ablation of the insulin like growth factor receptor (IGF1R) in OSNs enhances olfactory performance and increases adiposity (Riera et al. 2017). Demonstrating that insulin reduces olfactory function.

Another peptide crucial for energetic homeostasis is leptin. This hormone is produced by adipocytes in proportion to the fat content (Houseknecht et al. 1998; Mantzoros 1999; Friedman 2002; Pinto et al. 2004), and has a role in various physiological functions, including food intake, body weight regulation, reproduction, bone formation, and angiogenesis; these actions appear to be mediated through signaling into hypothalamic regions (Coppari et al. 2005; Pandit et al. 2017; Kwon et al. 2016).

Leptin exerts its metabolic actions through the leptin Ob-R receptor (Meister 2000; Meister and Håkansson 2001; Gorska et al. 2010; Liu et al. 2007). As previously mentioned the olfactory system expression of these receptors have been reported in the olfactory mucosa, OB, piriform cortex and entorhinal cortex, and there is also local synthesis of leptin in the olfactory mucosa (Caillol et al. 2003; Baly et al. 2007; Shioda et al. 1998). Similar to insulin, leptin has an inhibitory role in olfactory perception. Leptin administration in the CNS reduces food odor exploration (Prud’homme et al. 2009) and decreased the performance in odor discrimination tasks associated with decreased mitral/tufted cells firing (Sun et al. 2019). Savigner et al., demonstrated that leptin induces changes in olfactory sensitivity (Savigner et al. 2009), since ob/ob leptin deficient mice are hyper-osmic and obese, and leptin replacement in these animals decreases olfactory perception and food intake (Getchell et al. 2006).

Adiponectin is an adipokine almost exclusively secreted by the white adipose tissue. This hormone has several pleiotropic effects, it regulates lipid and glucose metabolism, increases insulin sensitivity, modulates immune activation in different immune cells and it has antioxidant properties. Adiponectin secretion is inversely correlated with BMI meaning that obese and diabetic patients with increased adiposity present hypoadiponectinemia (Mojiminiyi et al. 2005). The effects of adiponectin are mediated by the adiponectin receptors 1 and 2 (AdipoR1 and AipoR2) and are generally expressed simultaneously, these receptors are highly expressed in the brain (Yamauchi et al. 2007). There is a third cell surface molecule that binds to adiponectin, T-cadherin that is mainly expressed in endothelial cells (Akingbemi 2013).

AdipoR1 and AdipoR2 are expressed in the sensory neurons of the OE and in the OB (Hass et al. 2008; Miranda-Martínez et al. 2017). Furthermore, AdipoR1 is expressed in the OE (Prud’homme et al. 2009) and the septal organ, but not in the vomeronasal organ, which specializes the perception of social odors, suggesting that adiponectin signaling in the olfactory system is mainly involved in the perception of food odors (Hass et al. 2008). In mice, adiponectin increases the activation of juxtaglomerular interneurons and enhances the electrical response in the olfactory epithelium to odor stimulation (Loch et al. 2013); although the exact effects in olfactory perception of this adipokine are unknown.

On the other hand, ghrelin receptors (GHSR-1a) are also expressed in the glomerular, mitral and granular cell layers in the OB; intraperitoneal (i.p.) administration of ghrelin increases the olfactory detection of food and exploratory sniffing in both rodents and humans (Loch et al. 2015; Tong et al. 2011).

Interestingly, there is evidence that in the postprandial phase an important elimination of new cells in the granular cell layer of the OB takes place (Yamaguchi et al. 2013; Yokoyama et al. 2011; Komano-Inoue et al. 2014), suggesting that the metabolic status might not only be involved in the modulation of neuronal excitability but also has a direct influence in the organization of the olfactory neuronal circuits and in olfactory sensitivity.

Glucose and lipids are also sensed in the olfactory circuits, changes in hepatoportal glucose levels result in the neuronal activation of several areas in the brain including the OB (Delaere et al. 2013). Similar to other tissues, the OE and the OB express glucose transporters GLUT-1 and GLUT-2 (Hichami et al. 2007; Leloup et al. 1994) and GLUT-1 expression levels in the OB are determined by both metabolic status and the performance of olfactory tasks (Hichami et al. 2007; Soria-Gomez et al. 2014). Furthermore, sodium-coupled glucose transporters 1 (SGLT1) and insulin dependent glucose transporters 4 (GLUT4) are also expressed in the OB (Aimé et al. 2014; El Messari et al. 1998).

Furthermore, mitral cells in the OB increase their firing rate in response to changes in glucose concentrations (Tucker et al. 2013), suggesting that this region might have glucose sensor properties. In contrast, the lack of glucose and lipid sensing mice null for the voltage-gated potassium channel Kv1.3 leads to hyper-osmia, suggesting that sensing low reservoirs levels enhance the olfactory perception, and that this is interpreted as a signal to seek for food sources. Although this data should be taken with great reserve since Fadool et al., also reports structural changes in the glomerular portion of the OB in the KO-Kv1.3 mice (Fadool et al. 2011).

Besides from the evidence demonstrating that the metabolic status of the body modulates olfactory responses to odors, this is not a one-way street, and olfactory activation can also promote neuronal responses in the hypothalamus.

The Olfactory System Modulates Neuronal Hypothalamic Activity

One of the under-studied aspects of olfactory systems is the control of hypothalamic outputs. The most studied neuroanatomical connection between the olfactory system and the hypothalamus is the connection of the olfactory cortex with the hypothalamic preoptic area. Via this pathway, olfactory cues like pheromones can control sexual behavior, female ovulation inter-male aggression, among other behaviors (Yoon et al. 2005; Gascuel et al. 2012; Edwards et al. 1993; Guillot and Chapouthier 1996).

In addition, the autonomic functions that regulate energy expenditure are also modulated by olfactory signals. For example, a series of experiments using grapefruit oil, which induces olfactory stimulation, significantly increases the sympathetic output to the brown interscapular and white epididymal adipose tissue, elevates renal sympathetic nerve activity and blood pressure (Niijima and Nagai 2003; Shen et al. 2005; Tanida et al. 2005), suggesting that the olfactory areas of the brain modulate the activity of hypothalamic nuclei involved in energy balance.

Different poly-synaptic pathways connect the olfactory system with the hypothalamus. Tracing studies performed by injecting the multi-synaptic neuronal tracer wheat germ agglutinin-horseradish peroxidase (WGA-HRP) in the lateral hypothalamus (LH), showed labeled cells in the anterior olfactory nucleus, the piriform cortex, the olfactory tubercle, and the anterior cortical nucleus of the amygdala. Furthermore electrical stimulation of the OB or the olfactory cortex induces neuronal responses of the LH (Gascuel et al. 2012; Velozo and Almli 1992; Anand and Brobeck 1951; Price et al. 1991).

A study by Murata et al., demonstrated that GABAergic neurons in the olfactory peduncle project to the LH, suggesting that this pathway could be inhibiting processes like food intake and alertness (Murata et al. 2019). In addition, Kondon et al., observed that CRH neurons in the paraventricular nucleus (PVN) receive innervation from multiple olfactory cortical areas, and that this neuroanatomical connection regulates CRH production in response to predator odors (Kondoh et al. 2016).

One of the most convincing evidence regarding the influence of the olfactory system in hypothalamic neuronal activity was performed by Chen et al., who demonstrated that the olfactory detection of food rapidly modulates the neuronal activity of the hypothalamic arcuate nucleus (ARC) (Chen et al. 2015), a nucleus that senses blood borne metabolic cues (Cone et al. 2001). In this study the sole olfactory perception of food significantly increased the firing rate in the proopiomelanocortin (POMC) neurons, a neuronal population known to inhibit food intake and increase energy expenditure thought the mobilization of metabolic reservoirs and activation of brown adipose tissue thermogenesis (Coll et al. 2004). In contrast, decreased the neuronal firing rate of the orexigenic neurons expressing agouti related peptide (AgRP) in the ARC, which are antagonists to POMC (Chen et al. 2015). These data suggest that the communication between the olfactory system can modulate the neuronal activity in the hypothalamus involved in food intake and energy expenditure.

Olfactory perception can also modulate metabolic health, since anosmic mice without either OE receptor cells or OB are resistant to diet-induced obesity and exhibit lower body weight gain, decreased blood glucose levels, increased locomotor activity, heart rate and body temperature (Getchell et al. 2006; Chen et al. 2015; Riera et al. 2017). These data suggest that the olfactory system can regulate the hypothalamic neurons controlling endocrine and autonomic functions not only involved in food intake but also in energy expenditure.

In the next section we discuss how olfactory function becomes altered in metabolic diseases, to further suggest the possibility of intervening this system to ameliorate the metabolic impairments observed in diseases like obesity and diabetes.

Olfactory Sensitivity in Obesity and Diabetes

Altered odor sensitivity has been reported in disorders such as anorexia, bulimia, obesity and diabetes (Enck et al. 2014). As previously mentioned, there is compelling evidence demonstrating a bidirectional relationship between the olfactory and the metabolic system and that this interaction regulates the metabolic balance of the body (Julliard et al. 2017; Soria-Gomez et al. 2014).

The relationship between olfactory sensitivity and body mass index (BMI) is controversial, as both negative and positive associations have been reported (Peng et al. 2019). For example, a study comparing the BMI scores in patients with olfactory dysfunction shows a significant decrease in the olfactory function of subjects with high BMI (Patel et al. 2015). Furthermore, a gene methylation analysis of 474 individuals showed that BMI and waist circumference were associated with 13 CpG sites at olfactory genes including the olfactory receptors OR4D2, OR51A7, OR2T34 and ORDY1, and downstream signaling molecules that regulate odor detection such as SLC8A1, ANO2 and CAMK2D (Ramos-Lopez et al. 2019), demonstrating that obesity induces epigenetic changes in olfactory related genes, that could impair olfactory reception in the OE. Other studies have also determined that obese individuals present poor olfactory identification and discrimination (Fernandez-Garcia et al. 2017; Skrandies and Zschieschang 2015; Pastor et al. 2016).

In contrast, Jacobson et al., showed a positive correlation between greater BMIs and the activation of the primary olfactory regions (Jacobson et al. 2019). Another study demonstrated that obese patients presented increased levels of sensitivity and preference for odors associated with palatable and highly caloric food (Stafford and Whittle 2015). In addition, the hippocampus of obese individuals is significantly more activated after the exposure to food-related odors, suggesting that processing of feeding related odors differs in these subjects and may be part of what causes overeating (Stafford and Whittle 2015). Interestingly, Fardone et al., demonstrated a differential neuronal activation in the juxtaglomerular cells in mice fed a HFD (Fardone et al. 2019), suggesting that only a particular set of neurons are activated by food-related odors, if these glomerular cells are those responding to odorants from energy dense foods remains unknown.

In animal models, obesity-prone rats present decrease odor thresholds, meaning higher olfactory sensitivity, but poor olfactory memory and learning (Lacroix et al. 2015). Furthermore, ob/ob and db/db mice, as well as Zucker rats, present a better performance in olfactory tests and increased food‐seeking behaviors in response to food-related cues (Thanos et al. 2013; Badonnel et al. 2014; Getchell et al. 2006); this is related to hyperactivation of the OB, since its gamma range waves present higher frequencies and the beta activity was longer and stronger in ob/ob mice (Chelminski et al. 2017).

The hypothesis that metabolic health is determined by the OB is supported by bulbectomy studies in rodents showing an increase in the total amount of food intake and meal frequency (Meguid et al. 1993; Miro et al. 1980, 1982). Furthermore, mice with selective ablation of mature sensory neurons (OSN) are resistant to DIO (Fadool et al. 2011; Chelminski et al. 2017; Riera et al. 2017).

These studies suggest that initially a HFD might increase olfactory perception, without being able to inhibit food intake (Fig. 2a). The main complication with human studies and its interpretation, is that many obese patients also present other comorbidities such as vascular diseases and type 2 diabetes (T2D).

Fig. 2.

Olfactory-hypothalamic axis role in metabolic balance. a Food odors are perceived through the sensory neurons in the olfactory epithelium (OE) which project to the olfactory bulb (OB). Here the mitral (black) and tufted (pink) cells send their projections to the anterior olfactory nucleus (AON), the olfactory tubercle (OT) and piriform cortex (PC), and to other parts of the olfactory cortex, in this region the olfactory information is processed for odor detection, discrimination, learning and recalling, these processes also include other cortexes and the hippocampus (not shown here). In a separated pathway, many regions in the olfactory cortex (see text) send axonal projections to the preoptic (POA) and the lateral (LH) areas of the hypothalamus, that in turn establish connections with nuclei involved in the regulation of food intake and energy balance, like the dorsomedial hypothalamus, the paraventricular nucleus, the ventromedial hypothalamus and the arcuate nucleus (ARC). Hypothetically, olfactory information could reach the ARC, where food perception activates the POMC anorexigenic neurons to reduce food intake and increase energy expenditure, and at the same time the olfactory pathway would also inhibit the orexigenic AgRP/NPY neurons. Peripheral hormones like insulin and leptin, are known signals that activate POMC neurons in hypothalamus thus inhibiting food intake and increasing energy expenditure, and in the olfactory system it is known that these hormones are able to suppress olfactory sensitivity. The pathway activated when the individuals have a normal weight is represented in blue and its alteration in obesity is represented in faded red, in which hormones like insulin and leptin do not exert its actions nor in the olfactory system neither in the ARC thus the olfactory perception remains hyperactivated and the ARC circuits are unable to suppress food nor to elicit energy expenditure. b Hypothetical model of how the “olfactory-hypothalamic axis” could be intervened intranasally through the direct activation or desensitization of the olfactory system to induce the activation of neurons in the ARC that inhibit food intake and promote the activation of the autonomic mechanisms involved in the increase of energy expenditure. Created with BioRender.com

The relationship between T2D and olfactory dysfunction has also been described. Diabetic patients show low scores in odor identification and discrimination related with both macro and microvascular impairments in the olfactory system (Weinstock et al. 1993; Zaghloul et al. 2018; Hassing et al. 2004; Watson and Craft 2004). In a cross-sectional study, insulin dependent diabetic subjects presented higher prevalence of perceiving phantom odors, in addition the patients with more aggressive treatments were also hyposmic or anosmic (Chan et al. 2017).

Impaired olfactory ability in diabetic subjects has been associated with neuropathic pain and retinopathy, suggesting that the olfactory screening can be an early predictor of microvascular complications of diabetes (Brady et al. 2013; Gouveri et al. 2014). However, other studies did not find olfactory impairment in diabetic individuals with or without micro and macroangiopathy (Naka et al. 2010).

In T2D rodent models, olfactory performance is impaired and is associated with decreased IRS phosphorylation in the main olfactory bulb and piriform cortex (Rivière et al. 2016; Lietzau et al. 2018).

The proposed mechanisms involved in the olfactory dysfunction in T2D include macro and microvascular causes, olfactory nerve damage, central insulin resistance and low-grade inflammation may play a key role in olfactory modulation (Zaghloul et al. 2018). Recently, our group demonstrated that T2D rats present olfactory dysfunction associated with IL-1β and miR-146a overexpression in the OB, suggesting the upregulation of inflammatory mechanisms (Jiménez et al. 2020).

IR is also expressed in the OE (Lacroix et al. 2008; Marks et al. 2009), and its expression is decreased in the olfactory mucosa and OB of obese rats (Lacroix et al. 2015). Furthermore, olfactory dysfunction was associated with insulin resistance in humans (Palouzier-Paulignan et al. 2012; Min and Min 2018).

The previous information highlights the relevance of the olfactory system in the metabolic balance and metabolic diseases, therefore, the development of strategies aimed to modulate the olfactory function should be considered to improve the treatment of diseases such as obesity and diabetes.

Central Insulin and Leptin Resistance

We have summarized how the olfactory-hypothalamic bidirectional interactions may contribute to the regulation of energy intake and expenditure, and that the olfactory system is a key player for the development of metabolic diseases. This suggests that the modulation of the olfactory-hypothalamic axis might be an important new target area to develop noninvasive therapies aimed to prevent/revert the metabolic impairments resulting from obesity or diabetes.

One of the main characteristics of obesity and T2D are insulin and leptin resistance, which involve an impairment in the capacity of these two hormones to exert their biological effects.

Insulin resistance is a state in which insulin dependent tissues increase their response threshold, thus they require increased concentrations of this hormone to achieve the biological effects normally elicited by lower concentrations.

Impairments in the IR intracellular cascade are believed to be the origin of peripheral insulin resistance. The IR is part of the tyrosine kinase superfamily, following activation its kinase phosphorylates tyrosine residues creating binding sites for signaling protein partners. For more detailed reviews see refs (Taniguchi et al. 2006; White 2003). Most of the physiological effects of the IR are mediated through the activation of IRS-1 and IRS-2 dependent pathways, linking the signaling cascades of the InsR to other pathways such as the cytokine receptors and TNF-α receptors, whose intracellular pathways are well known negative regulators of IRS (Greene et al. 2010).

The metabolic impairments involved in insulin resistance are not exclusive to peripheral organs, this is also observed in the brain. There are two hypotheses regarding the mechanisms of insulin resistance in the brain. The first one being due to an impaired transport into the brain (Banks et al. 2012; Chen et al. 2017; Urayama and Banks 2008) and the other includes decreased basal activation of IR due to reduced insulin binding (Rivera et al. 2005), increased serine phosphorylation of IRS-1 and reduced levels of PI3K (Talbot et al. 2012; Moloney et al. 2010; Liu et al. 2011) both mechanisms imply that the effects of insulin in the CNS including the hypothalamus are damped.

Obese individuals present hyperleptinaemia and a diminished response to leptin, condition known as leptin resistance (Mazor et al. 2018). The Ob-R is part of the cytokine receptors associated with the Janus kinase 2 (JAK2) (Ihle and Kerr 1995). Upon binding, the Ob-R undergoes a conformational change that activates JAK2, thus phosphorylating tyrosine residues that in turn phosphorylates the signal transducer and activator of transcription 3 (STAT3) promoting its translocation into the nucleus and the transcription of several target genes, including the suppressor of cytokine signaling 3 (SOCS3), that inhibits the Ob-R–JAK2 signaling (Couturier and Jockers 2003).

Importantly, in most cases the impairment in leptin signaling is not related to genetic modifications of this hormone or its receptors, instead, similar to insulin there is an impaired transport into the brain, thus preventing the activation of its corresponding signaling pathways (Banks 2001). Restoring leptin signaling in the brain via intracerebroventricular administration (i.c.v.) reduces food intake, enhances energy expenditure and improves glycemic indexes (Fliedner et al. 2006; Van Heek et al. 1997; Halaas et al. 1997).

In addition, central leptin resistance has been related to the overactivation of Ob-R through the increased activation of SOCS-3 resulting from hyperleptinemia and the mild inflammation caused by obesity (Engin 2017).

On the other hand, there is a direct relationship between hypoadiponectinemia and the development of insulin resistance and diabetes (Mojiminiyi et al. 2005).The intracellular domain of AdipoR1 and AdipoR2 binds an adaptor protein containing pleckstrin homology domain, phosphotyrosine binding domain and leucine zipper motif (APPL1). This adaptor protein mediates the downstream effects of adiponectin via AMPK activation (Zhou et al. 2009; Wen et al. 2010).

There is an important crosstalk between insulin and adiponectin pathways, APPL1 forms a complex with IRS1/2 under basal conditions, thus facilitating its union with the insulin receptor (Ryu et al. 2014). Upon stimulation with either insulin or adiponectin the complex APPL1/IRS1/2 is recruited to the insulin receptor therefore increasing insulin sensitivity (Yamauchi et al. 2007) (Combs et al. 2001; Berg et al. 2001).

I.c.v. adiponectin administration enhances both insulin and leptin intracellular signaling pathways, reduces food intake, and increases the expression of IRS1/2, JAK2 and STAT, furthermore adiponectin depolarizes Ob-R expressing POMC neurons and even potentiates its response to leptin (Sun et al. 2016). Furthermore, in genetically and diet-induced obese mice i.c.v adiponectin enhances glucose tolerance (Koch et al. 2014). It has also been suggested that in obesity adiponectin transport into the brain is also impaired (Kos et al. 2007).

One common trait of the hormonal resistances in the brain is that if administered into the CNS, their biological actions are restored (El-Haschimi et al. 2000; Woods et al. 1979; Brown et al. 2006). This might be because the access of these hormones into the brain is a saturable process, therefore, conditions like hyperinsulinemia and hyperleptinemia impair this mechanism reducing the access into the entire brain (Rhea et al. 2018; Banks 2004; Kos et al. 2007; Banks et al. 1996). In agreement, it has been demonstrated that dextrans of various molecular weights administered intravenously do not access the hypothalamus in mice fed a high fat diet (Dodd et al. 2019).

The fact that i.c.v. administration of insulin or leptin in obesity animal models is able to reduce food intake, and improve blood glycaemia suggests that therapies should aim to restore central levels of these hormones or to mimic their effects (Arase et al. 1988; Rahmouni et al. 2002; Heni et al. 2014).

The Intranasal Pathway and Its Therapeutic Potential

The intranasal pathway (i.n.) is a noninvasive method of drug delivery into the CNS (Hanson et al. 2013). This administration route offers the bypassing of the blood–brain barrier (BBB) and extends drug bioavailability, thus offering a reliable and promising pathway to deliver a wide range of therapeutic agents including small and large molecules, including peptides (Scheibe et al. 2008).

The i.n. administration of drugs allows for a rapidly access into brain tissue via two routes: 1) through direct delivery to regions of the peripheral olfactory system that connect the nasal conduits with the OB and 2) associated with the peripheral trigeminal system which connect to the brainstem and spinal cord (Thorne et al. 2004).

The OB glomerular layer is a highly irrigated network (Chaigneau et al. 2007; Yang et al. 1998) with the presence of fenestrated blood vessels (Ueno et al. 1996). Thus, i.n. administration should represent a good conduit for increasing OB activity and modulating hypothalamic functions (Fig. 2b).

The use of intranasal administrations of hormones has shown metabolic benefits in obese and diabetic patients, and animal models. Obese rats administered intranasally with leptin show decreased food intake and body weight loss (Schulz et al. 2012). These effects are similar to those observed in leptin i.c.v. administrations (El-Haschimi et al. 2000). Importantly, Fliedner et al. showed that i.n. leptin enters several areas in the brain even in hyperleptinemic rats (Fliedner et al. 2006). In addition it significantly attenuates sleep-disordered breathing, improves glucose impairments and decreases bodyweight gain in obese rats (Khafagy et al. 2020; Yuan et al. 2017; Santiago and Hallschmid 2019; Reger and Craft 2006; Schulz et al. 2012; Berger et al. 2019), suggesting that i.n. leptin is metabolically effective in obese animals, This suggests that using the nose-to-brain route bypasses defective leptin transport through the BBB that has been observed in metabolic diseases, thus restoring the central effects of leptin without the need of invasive intracerebral procedures.

Similarly, i.n. insulin administrations have also demonstrated promising therapeutic effects in obese and diabetic patients, decreasing food intake and body weight (Hallschmid et al. 2004; Jauch-Chara et al. 2012; Khafagy et al. 2020; Yuan et al. 2017; Santiago and Hallschmid 2019; Reger and Craft 2006), increasing postprandial energy expenditure and decreasing insulin secretion (Benedict et al. 2011). These effects are similar to those observed with i.c.v. administrations in obese rodents and primates (Woods et al. 1979; Brown et al. 2006).

The improvements in metabolism observed after i.n. administration of hormones could have two possible mechanistic explanations: (1) The direct increase of these hormones in the brain parenchyma including the hypothalamus allows them to access their target areas. (2) Because the OB contains many of the hormone receptors and has the highest transport rate across the blood–brain barrier (BBB) (Banks et al. 1999), signaling to this area activates neurons connecting with the hypothalamus. Even if the hypothalamus presents a certain degree of hormonal resistance, this pathway would be able to activate it because it constitutes a neuronal and not a humoral connection.

As mentioned before there is an important lack of information regarding receptor expression in the olfactory system in obese and diabetic patients. In Table 1, we summarize some of the studies with data regarding the expression of IR and Ob-R, most of which are unchanged, except for the olfactory epithelium that presents a downregulation of the insulin receptor. Interestingly, most of the reports in the hypothalamus show a marked down regulation of the receptors, suggesting that there is a selective resistance to insulin and leptin in this area and that the olfactory areas are not as affected.

Table 1.

Leptin and insulin receptor expression in the brain in obesity models

Receptor	Species	Tissue	Metabolic condition	Expression	References
IR	Sprague–Dawley rats	Olfactory Bulb	Diet-induced obesity	Unchanged	Lacroix et al. (2015)
	Zucker rat	Hippocampus	Obesity genetic model	Unchanged	Livingston et al. (1993)
	Zucker rat	Piriform cortex	Obesity genetic model	Unchanged	Gisslinger et al. (1993)
	Zucker rat	Olfactory bulb	Obesity genetic model	Unchanged	McFarlane et al. (1993)
	Zucker rat	Brain	Obesity genetic model	Unchanged	Livingston et al. (1993)
	Sprague–Dawley rats	Olfactory mucosa and Olfactory bulb	Diet-induced obesity	Down regulated	Lacroix et al. (2015)
	Neuroblastoma cell line	Differentiated human neuroblastoma	Leptin pre-treatment	Down regulated	Benomar et al. (2005)
Ob-R	Sprague–Dawley rats	Olfactory bulb Olfactory mucosa	Obesity genetic model	Unchanged	Lacroix et al. (2015)
	New Zealand Obese mice	Isolated cerebral microvessels	Obesity genetic model	Unchanged	Hileman et al. (2002)
	Diet-induced obese mice	Isolated cerebral microvessels	Diet-induced obesity	Unchanged	Hileman et al. (2002)
	Long-Evans rats	Hypothalamus	Chronic leptin administration	Down regulated	Martin et al. (2000)
	Brown Norway rats	Hypothalamus	Diet-induced Obesity	Down regulated	Wilsey and Scarpace (2004)
	Mice	Hypothalamus	Diet-induced Obesity	Down regulated	Zhai et al. (2018)
	ob/ob mice	Hypothalamus	Obesity genetic model	Up regulated	Huang et al. (1997)
	ob/ob mice	Piriform cortex	Obesity genetic model	Up regulated	Hassall and Hoyle (1997)
	ob/ob mice	Olfactory cortex	Obesity genetic model	Up regulated	Huang et al. (1997)
	Agouti viable yellow (Avy) mice	Hypothalamic microvessels and astrocytes	Obesity genetic model	Up regulated	Pan et al. (2008)
	B6 mice	Hypothalamic microvessels	Diet-induced Obesity	Up regulated	Pan et al. (2008)

In this sense the effects of intranasal therapies observed on metabolism may imply the stimulation of the olfactory-hypothalamic axis. Therefore, since insulin and leptin decrease olfactory sensitivity this stimulation will ultimately activate an until now unknown pathway that will increase POMC and decrease AgRP neuronal activity. This hypothesis is based on the fact that many of the i.n. administrations of insulin and leptin show decreased food intake, enhancements in postprandial thermogenesis, reduced glucose production and improved insulin sensitivity, all of these functions are modulated by several neuronal types in the ARC especially the POMC and AgRP/NPY, that in obesity and diabetes present a certain degree of insulin and leptin insensitivity, thus these autonomic effects might be promoted by the olfactory system.

As previously mentioned the other possible mechanism through which hormones like insulin and leptin do not exert a biological effect is the downregulation or desensitization of their receptors in different areas of the brain (Chen et al. 2017; Yarchoan and Arnold 2014; De Felice and Ferreira 2014; Zemva and Schubert 2014; Bomfim et al. 2012) suggesting that if the underlying cause of the resistance is due to impairments in the receptors (Table 1), then the use of intranasal hormonal therapies might not show as much benefit as expected, therefore the employment of other peptides or drugs capable to mimic or improve receptor sensitivity could be an alternative option.

A study conducted by Seelke et al., showed that chronic intranasal oxytocin administration reduce fat mass in voles without altering lean mass, (Seelke et al. 2018), this hormone has demonstrated a beneficial effect in olfactory memory and sensitivity (Oettl et al. 2016; Oettl and Kelsch 2018).

On the other hand, the human fibroblast growth factor 1 (FGF1) is a peptide secreted by adipose-derived microvascular endothelial cells (MVECs) with insulin sensitization properties, this peptide has been used to prevent the progression of neurodegenerative diseases in rodents (Lou et al. 2012) the brain damage in stroke models (Cheng et al. 2011; Fan et al. 2019) and its central and peripheral administration confers notable improvements in diabetic mice (Gasser et al. 2017).

The thiazolidinedione (TZD), pioglitazone, is an insulin sensitizing molecule for the management of insulin resistance. Its enhances peripheral insulin sensitivity, and increases plasma adiponectin concentrations (Miyazaki et al. 2004; Miyazaki et al. 2001; Kemnitz et al. 1994; Tozzo et al. 2015) this molecule has also been used intranasally in Alzheimer’s disease (AD) patients, a neurodegenerative disorder associated with impaired brain insulin signaling proving a notable improvement in cognitive function (Jojo et al. 2019; Wong et al. 2020).

These studies suggest that the intranasal use of other molecules in which their receptors do not undergo insensitivity or downregulation in metabolic diseases might be a possible option therapies to enhance hypothalamic activity through modulation of the olfactory-hypothalamic pathway.

Future Directions

Definitely more studies are needed to determine if there is a mechanism that could allow the use of intranasal therapies to ameliorate the metabolic impairments caused by metabolic diseases like obesity and diabetes that nowadays are the most common diseases in industrialized countries.

Metabolic diseases like obesity and T2D are known to impair both the olfactory and hypothalamic systems. There is no doubt of the existence of the “olfactory-hypothalamic axis”, moreover it is clear that the bidirectional relationship is able to control metabolic processes like food intake, autonomic activity of tissues and more, thus modulating metabolic balance.

In addition, there is also compelling evidence that the olfactory system is a major player for the development of metabolic diseases. The current review proposes that intervening the olfactory system could enhance the activity of the hypothalamus through its multi-synaptic connections. We propose a noninvasive approach as the i.n. administration of drugs to improve the metabolic impairments in obese and diabetic patients.

Although there are some important questions to be answered before betting for the olfactory system as a conduit to improve hypothalamic activity: (1) there are no reports describing the neuronal and genetic profiles of the olfactory regions in postmortem brains, therefore, the mechanisms underlying the olfactory impairments observed in these diseases remain unknown (2) Although there is evidence indicating that the activity of the olfactory system can change blood glucose levels, enhance lipid oxidation in the adipose tissue and other metabolic functions, there is no evidence that this effects are carried out through a direct input to hypothalamic neurons. The only evidence of this was provided by Chen et al., but this study did not demonstrate that the sensory activation of the ARC has an implication in metabolism. Therefore, future studies should demonstrate that the stimulation of the primary olfactory regions can in fact modulate insulin sensitivity, glucose metabolism, lipid oxidation, through a hypothalamic pathway.

Conclusion

The possible anatomic pathways involved in the modulation of energy balance regulated by the activity of the olfactory system have been reported. Evidence also suggests that the olfactory system could be involved in the development of metabolic diseases such as obesity and type two diabetes. Here we propose a model to explain its implication for the maintenance of the metabolic homeostasis of the organism. The use of alternative therapies such as the intranasal administration of drugs, aimed to regulate and improve the activity of the olfactory system that in turn will be able to control the neuronal activity of hypothalamic centers, could represent a suitable approach to deliver drugs and prevent/ameliorate the metabolic impairments observed in obese and diabetic patients.

Funding

This study was supported by grants DGAPA-PAPIIT IN215716 and IN221819 to RGG and IA204121 to MAGR.

Declarations

Conflict of interest

The authors declare that they have no competing interests.