High maltose sensitivity of sweet taste receptors in the Japanese macaque (Macaca fuscata)

Taste sensitivity differs among animal species depending on feeding habitat. To humans, sucrose is one of the sweetest natural sugars, and this trait is expected to be similar in other primates. However, previous behavioral tests have shown that some primate species have equal preferences for maltose and sucrose. Because sweet tastes are recognized when compounds bind to the sweet taste receptor Tas1R2/Tas1R3, we evaluated the responses of human and Japanese macaque Tas1R2/Tas1R3 to various natural sugars using a heterologous expression system. Human Tas1R2/Tas1R3 showed high sensitivity to sucrose, as expected; however, Japanese macaque Tas1R2/Tas1R3 showed equally high sensitivity to maltose and sucrose. Furthermore, Japanese macaques showed equally high sensitivity to sucrose and maltose in a two-bottle behavioral experiment. These results indicate that Japanese macaques have high sensitivity to maltose, and this sensitivity is directly related to Tas1R2/Tas1R3 function. This is the first molecular biological evidence that for some primate species, sucrose is not the most preferable natural sugar, as it is for humans.

Scientific RepoRts | 6:39352 | DOI: 10.1038/srep39352 was higher than 10 mM, but not to maltose (Fig. 2a). Mf Tas1R2/Tas1R3 responded to sucrose and maltose, when the concentration was greater than 10 mM (Fig. 2b). This result suggests that the Japanese macaque is equally sensitive to maltose and sucrose, and can detect sweet tastes in > 10 mM solutions.
To confirm whether sweet taste sensitivity in monkeys is directly related to Tas1R2/Tas1R3 function, we conducted a two-bottle behavioral experiment using four Japanese macaques. Most monkeys chose sucrose or maltose significantly more often than water when the concentration was greater than 10 mM ( Figure S1). The average sweetener intakes of monkeys are shown in Fig. 2c. These thresholds were 10-30 mM, similar to the results of the Tas1R2/Tas1R3 functional assay, suggesting the contribution of Tas1R2/Tas1R3. Because the maltose sensitivities of Tas1R2/Tas1R3 in humans were very weak, these results suggest that monkey Tas1R2/Tas1R3 evolved sensitivity to maltose in addition to sucrose.
We then conducted a functional assay using chimeric Tas1R2/Tas1R3 between human and macaques to identify maltose binding site(s) (Fig. 3). We first measured the responses of Hs Tas1R2/Tas1R3 in which one side of  subunit was changed to that of Japanese macaques. Although the Hs Tas1R2/Mf Tas1r3 response was greater than that of Mf Tas1R2/Hs Tas1R3, both chimeras responded to maltose. Next, we measured the responses of the Mf T1R2 and Mf T1R3 chimeras to determine the precise part of Mf Tas1Rs that is critical. All cells with co-expression of the chimera responded to 30 mM maltose. These results might suggest that all parts of Mf Tas1R2 and Mf Tas1R3 are involved in the response to maltose. Even though it is difficult to generate site-directed mutants for each subunit, this finding suggests that Tas1R3 as well as Tas1R2 are involved in the evolution of sweet taste receptors.

Discussion
In the experiment, human and Japanese macaque Tas1R2/tas1R3 showed high responses to sucrose, as expected. Some amino acids critically related to sucrose binding are found in Hs Tas1R2. In Mf Tas1R2, amino acids at these positions correspond to those of Hs Tas1R2, except at position 40, which is serine in humans and threonine in Japanese macaques. Because both Mf Tas1R2/Tas1R3 and Hs Tas1R2/Tas1R3 responded to sucrose dose-dependently for concentrations of at least 10 mM, S40T seems to have no effect to sucrose binding to Tas1R2/Tas1R3. Maltose is a natural sugar constructed from two glucose molecules bound with α (1-4) glycosidic bonds. This sugar is produced from starch digested by α -amylase, an enzyme secreted from the salivary gland and pancreas. Pigtail macaques (Macaca nemestrina) and olive baboons (Papio anubis) also have higher sensitivity to maltose than humans 3 . Interestingly, only these Old World monkeys had similarly high sensitivity to maltose as sucrose 3 , while other primate species did not, including the spider monkey 6 and squirrel monkey 7 . Based on short-term behavioral experiments, mice 8 and rats 9 have maltose thresholds of 20 to 30 mM. Therefore, the Old World monkey lineage specifically gained high sensitivity to maltose. Old World monkeys are general herbivores with food repertoires that extend to leaves, fruits, and seeds. Japanese macaques spend half of their annual feeding time eating leaves and seeds 10 , which contain many starches 11 . Additionally, only Old World monkeys have a cheek pouch in which they can store food items; these food items are exposed to saliva containing α -amylase for extended durations. In macaques, the levels of saliva amylase are similar to those in humans 12 , though the mechanism explaining amylase levels is unknown. Therefore, sensitivity to starches is higher in macaques than other animals, suggesting an ecological advantage for starch consumption in macaques.
Primate species have various mutations in Tas1R2 and Tas1R3 13 . Generally, most sweet taste compounds bind to a huge external membrane domain called the venus flytrap domain (VFTD) of Tas1R2 14 . Some amino acid residues are critical for sucrose binding, but there is no information for maltose. The Mf Tas1R2 response to sucrose was similar to that of Hs Tas1R2, but differed from the response to maltose. In addition, these results suggested that the maltose response is related to both Tas1R2 and Tas1R3, indicating that the critical amino acids for binding are different from those for sucrose. In comparisons of Tas1R2 and Tas1R3 VFTD between Hominidae and Cercopithecinae, mutations at 45 and 43 amino acid residues have been observed, respectively 13 . We infer that these mutations affect sweet taste sensitivity to natural sugars by changing Tas1R2/Tas1R3 function. Characterizing Tas1R2/Tas1R3 function and sweet taste sensitivity may provide a basis for understanding Tas1R2/Tas1R3 evolution in primates and feeding habitat adaptations. It is possible that Tas1R2 and Tas1R3 of macaques evolved to bind to maltose, while the ability to bind to sucrose was maintained in Tas1R2. The binding of the Tas1R3 VFTD to sugars has been reported in the hummingbird 15 . In particular, hummingbird Tas1r2 was lost in the ancestor of the birds and Tas1R1/Tas1R3 became sweet receptors via a Tas1R3 binding subunit. Adding to the central region of the VFTD, amino acids at the edge of one subdomain were mutated in hummingbirds, causing Tas1R3 to function as a sweet taste receptor. In our case, chimeric Tas1R3 containing the VFTD and/or the cysteine-rich domain (CRD) plus the trans-membrane domain (TMD) of macaques exhibited a response to maltose. We speculate that not only the VFTD but also the CRD and/or TMD are effective for the maltose response, such as via conformational changes after maltose binding to Tas1R2/Tas1R3. Accordingly, Tas1R3 in animals may not only support the function of Tas1R1 or Tas1R2, but may also have the capacity to bind to sugars.

Materials and Methods
Sweet taste compounds. The natural sugars maltose (Wako, Osaka, Japan) and sucrose (Wako) were used for sweet taste solutions. Each compound was dissolved in deionized water to make 1, 3, 10, 30, and 100 mM solutions for the behavioral tests. To prevent potential osmotic effects, we used a maximum concentration of 100 mM. For the in vitro functional analysis, the compounds were dissolved in assay buffer (10 mM HEPES, 130 mM NaCl, 10 mM glucose, 5 mM KCl, 2 mM CaCl 2 , 1.2 mM MgCl 2 , pH 7.4).  Table S1) higher than those to 0 mM maltose.

DNA.
To obtain Japanese macaque (Mf) Tas1R2/Tas1R3, RNA was extracted from Japanese macaque tongues stored at the institute. RT-PCR was conducted to obtain Tas1R2 and Tas1R3 cDNA. Amplified Tas1R2 and Tas1R3 sequences were checked by sequencing using the ABI 3130xl. Human (Hs) Tas1R2 and Tas1R3 were purchased from Kazusa DNA Res. Inst. (FHC07060 and FHC10750). Tas1R2 and Tas1R3 were inserted into the mammalian expression vector pEAK10 (Edge BioSystems, Inc., Gaithersburg, MD, USA) and transfected to HEK293T cells with Gα 16-gust44 16 using Lipofectamine 3000 (Life Technologies, Inc., Carlsbad, CA, USA). Chimeric receptors were made using In-fusion kits (Takara Bio Inc., Shiga, Japan). For the functional analysis, Calcium 4 (Molecular Devices, Inc., Eugene, OR, USA) was used as an intracellular Ca 2+ indicator. Fluorescence was measured at 525 nm following excitation at 485 nm using the FlexStation 3 Microplate Reader (Molecular Devices Japan, Inc., Tokyo, Japan). The calcium response amplitudes were expressed as Δ F/F, which is the ratio of the ligand-dependent increase in fluorescence to the fluorescence before ligand addition. The response of cells that were transfected with the empty pEAK10 vector and Gα 16gust44 was defined as the mock response (TAS2R-independent response) and subtracted from all responses. Δ F/F values were fitted to the Hill equation (y = (max-min)/(1 + (x/EC50)rate)). Comparisons across concentrations were made by one-way ANOVA followed by Welch's tests (p < 0.05). Dunnett's tests were performed to determine the concentration at which significantly higher responses were observed compared to that at 0 mM (p < 0.05). Monkeys were kept in individual cages in the same room. Access to food and water was not restricted. To measure the sweet taste threshold of monkeys, the two-bottle test 3 was conducted. In this test, one of the bottles contained water and another contained the sweet taste solution (at various concentrations). In a trial, monkeys were exposed to two bottles at the same time for 1 min and solution intake was measured. This experiment was conducted twice a day, 1 h after feeding. The position of bottles was changed in each experiment. Four trials were performed for each concentration. The intake rates of sweet taste solutions were calculated as follows: Intake rate of sweet taste solution = Intake of sweet taste solution/Total intake. The order of exposure to solutions was randomized, except in the first trial, in which 100 mM solutions were used to motivate monkeys. A threshold concentration was calculated for each monkey according to previous methods 3 as the lowest concentration for which a significant increase in sugar water drinking compared with a sugar concentration of 0 was detected (p < 0.05).