Why low concentrations of salt enhance sweet taste

A long-standing culinary practice is to sprinkle a little salt on fruits such as watermelon, tomatoes or cantaloupe to make them taste sweeter. In the current issue of Acta Physiologica, Yasumatsu et al¹ describe a likely mechanism for this phenomenon. They report for the first time that the sodium-glucose cotransporter 1 (SGLT1) is expressed in the apical membrane of sweet-sensitive taste cells and is an important contributor to the taste of glucose-containing sugars.

Among the five taste qualities, sweet, a strongly appetitive taste quality that drives ingestion, indicates the presence of carbohydrates in food and a source of energy. The detection of sweet compounds takes place in the mouth at the level of the taste bud. Taste buds contain between 50 and 100 taste receptor cells that are commonly divided into three types based on morphological, molecular and physiological properties. Type I cells represent about half of the total number of cells in a bud and are thought to have a glial-like function. They express enzymes and transporters for the degradation and reuptake of various neurotransmitters. Type III cells represent only 5%–10% of the taste bud cells and respond to sour and some salty stimuli. They form conventional synapses with afferent gustatory nerve fibres and release serotonin as a neurotransmitter. Finally, Type II cells possess the G protein coupled receptors and transduction machinery for the detection of sweet, bitter and umami stimuli. They do not form conventional synapses with nerve fibres but release ATP via a non-vesicular synapse to communicate with the nervous system. Most Type II cells respond to only one taste quality since they express only one type of taste receptor. These receptors include the T2R family, which serve as receptors for bitter stimuli and the T1R family, which detect sweet and umami stimuli. The T1R2/T1R3 heterodimer functions as a receptor for all sweet compounds, including natural sugars as well as artificial sweeteners (for recent review on taste cell types and transduction, see²).

An initial study reported that ablation of either T1R2 or T1R3 completely abolished behavioural and neural responses to all sweet stimuli,³ while subsequent studies have shown significant responses remain to glucose⁴ as well as sucrose, which is broken down to glucose and fructose by salivary enzymes.⁵ These studies suggest that other transduction pathways may be involved in the detection of sweet stimuli, particularly sugars. Consistent with these later studies, the rapid cephalic-phase insulin release (CPIR) which occurs when glucose is applied on the tongue is not abolished in T1R3 knockout mice and requires KATP channels,⁶ similar to glucose sensing in the pancreas.

A candidate receptor for T1R-independent responses to glucose is the sodium-glucose cotransporter 1 (SGLT1). SGLT1 plays an important role in intestinal L and K cells where it transports glucose into the cells by coupling with sodium. The transport of glucose into the cell produces ATP, which subsequently blocks the ATP-inhibited K⁺ channels (KATP channels) causing membrane depolarization and release of GLP-1 or GIP depending on the cell type. SGLT1 has been detected in sweet-responsive taste cells using RT-PCR and immunohistochemistry,⁷ however, its functional role in sweet taste transduction has remained unclear.

In the present study, Yasumatsu et al¹ provide strong evidence that SGLT1 is the basis of the T1R-independent nerve response to glucose-containing sugars. The authors used gustatory nerve recordings in wild-type and T1R3 knockout mice to show that low concentrations of NaCl potentiate the response to glucose and glucose analogues, reflecting the required sodium to provide the driving force for SGLT1 function. The Na⁺-dependent potentiation is blocked by the specific SGLT1 antagonist, phlorizin. Interestingly, responses to artificial sweeteners (saccharin and SC45647) were unaffected by NaCl or phlorizin, suggesting a different mechanism for these sweeteners. To test whether functional glucose transporters were found on the apical surface of the tongue, a fluorescent glucose analogue, 2-NBDG, was applied to the apical surface of the tongue and sections of taste buds were observed for fluorescence. Several taste cells showed fluorescence, which was enhanced by the presence of NaCl and blocked by phlorizin, confirming that the transporter is located on the apical surface of the tongue, as SGLT1 is in other systems. Additional behavioural experiments confirmed that NaCl potentiates preference for glucose in T1R3 knockout mice and that the potentiation is blocked by phlorizin.

Single fibre recordings from sweet-best taste fibres of the chorda tympani nerve show three different types of sweet responsive fibres, presumably reflecting the properties of the taste cells they innervate. One type responds only to sucrose and synthetic sweeteners such as SC45647 but responses are completely unaffected by NaCl or phlorizin, reflecting the properties of the T1R receptors (T1R-Type) in the taste cells that connect to these fibres. A second type responds only to glucose and glucose analogues. These responses are potentiated by NaCl and the potentiation is inhibited by phlorizin (SGLT1-Type), suggesting that the taste cells that drive these fibres contain only SGLT1 as a sweet taste receptor. The third type of response shows features of both other fibre types as they respond to both sugars and artificial sweeteners (mixed type fibres) and responses to glucose are potentiated by NaCl, similar to the SGLT1 only fibres. The taste cells that drive these fibres apparently contain both T1Rs and SGLT1, as illustrated in Figure 1. The presence of fibres that respond selectively to glucose and synthetic sweeteners may represent a mechanism for discriminating between natural sugars and artificial sweeteners, and the authors suggest that mixed type fibres, which also respond to fatty acids and umami stimuli, may be a mechanism for detection of calorie-rich foods.

FIGURE 1.

Three types of sweet-sensitive taste cells, based on recordings from sweet-best single chorda tympani nerve fibres. T1R-Type (also called Suc SC-Type) have only the T1R2/T1R3 heterodimer and bind all sweet stimuli, including sucrose (Suc) and SC45647 (SC), the stimuli that were used in this study. Responses are not potentiated by sodium and are unaffected by phlorizin. SGLT1-Type (also called Glc-Type) have only the SGLT1 transporter and respond only to glucose (Glc) and Na⁺, which is needed for glucose transport into the cell. The glucose chorda tympani responses are potentiated by low concentrations of NaCl and the potentiation is blocked by phlorizin. The third sweet cell type is the Mixed Type, which have both T1R2/T1R3 and SGLT1 as sweet taste receptors. These cells respond to all sweet stimuli, including sugars and sweeteners, and responses to glucose are potentiated by NaCl and sensitive to phlorizin, as with the SGLT1-Type cells. The Mixed-Type cells also have receptors for umami stimuli and fatty acids (not shown)

Taken together, this new study provides a mechanistic explanation for the remaining responses to glucose in the T1R3 knockout mice. Evidence for this type of pathway was actually suggested well before the cloning of sweet receptors by studies in dogs, which have a strong appetitive taste for sugars, similar to humans. Isolated canine lingual epithelia exhibit a large sugar-activated short-circuit current that is potentiated by NaCl,⁸ and their chorda tympani responses to sugars are similarly potentiated by low concentrations of NaCl.⁹ This raises the intriguing possibility that this mechanism will also underlie the transduction of sugars in humans, but clearly further studies are needed.

In conclusion, sweet taste detection involves more components than the originally proposed T1R2/T1R3 heterodimer. While SGLT1 is an important component of the T1R-independent nerve response to glucose in mice, important questions remain. For instance, are the KATP channels required to mediate depolarization and transmitter release as in the small intestine? Do additional glucose transporters play a role in the taste bud? The glucose transporter type 4 (GLUT4), which transports glucose by diffusion has also been detected in sweet-responsive taste cells.⁷ Does it play a role in either sweet taste or CPIR? And finally, does this mechanism explain why, in humans, that adding a pinch of salt makes most fruits taste sweeter?