The diet of hominoid primates is dominated by plant foods that are rich in protein and carbohydrates. In addition to the macronutrients and micronutrients, natural diets contain fibers and secondary compounds which dilute nutrient density, reduce digestibility, and can have toxic effects. A common challenge for consumers of plant foods is that macronutrients are usually associated with fibers, antifeedants, and other secondary compounds, and that the increase of the former enhances the intake of the latter (Gautier-Hion, 1990; Cipollini and Levey, 1997; Hohmann et al., 2006). Some secondary compounds such as alkaloids and cyanogenic glycosides have already toxic effects when consumed in small amounts (Glander, 1982), whereas others like tannins affect their consumers only when ingested in large amounts (Fowler, 1983). Tannins of various classes and structures are a common ingredient in the diets of wild herbivore animals (Provenza et al., 2003; Foley and Moore, 2005), and their functions in the context of herbivory are diverse and complex (Constabel et al., 2014). Although the ingestion of tannins can consolidate the health status of consumers (Mueller-Harvey, 2006; Patra and Saxena, 2011), it has been shown that tannins reduce palatability, digestibility, and the concomitant energy intake (Robbins et al., 1991; Simmen et al., 1999; Barbehenn and Constabel, 2011). One specific effect of tannins is its high binding affinity to proteins, which prevents conversion of dietary protein and enhances iron retention (Delimont et al., 2017). When tannins are ingested in excessive amounts, this interference can have toxic effects (Al-Shafi, 2002; Barbehenn and Constabel, 2011). In a review of 53 studies on 43 primate species, Windley et al. (2022) found that secondary plant compounds, especially condensed tannins, have a consistent deterrent effect on the food choice within Colobinae, whereas in Hominidae the response to tannins was highly variable and species specific. Table 1 summarizes information on tannin content in the diet of wild ape populations and the putative strategies of consumers to cope with tannin content. According to published information, gorillas (Gorilla spp.) consume consistently higher tannin quantities compared to the Pan species (Pan paniscus and Pan troglodytes). Orangutans (Pongo spp.) seem to avoid tannins but are exposed to seasonal fluctuation in tannin intake. Hominoid primates use various strategies to cope with the detrimental effects of tannins and other secondary compounds (Milton, 1981; Lambert, 1998; Hohmann, 2009). One is, to select food items with a high macronutrient/secondary compound ratio (Glander et al., 1989; Conklin-Brittain et al., 2002; Hohmann et al., 2010). Another is, to separate manually or orally the parts with high concentrations of macronutrients from those containing secondary compounds (Corlett and Lucas, 1990; Lambert, 1998). If the separation of different compounds is difficult (e.g., in leaves, herbs and pith), consumers might switch between different food species during feeding to avoid accumulation of toxic compounds (Bryant et al., 1991; McArthur et al., 2014).

In response to the omnipresence of tannins in plants, and the combination of macronutrients and secondary compounds, physiological adaptations have emerged which protect consumers from the harmful effects of this group of secondary compounds without compromising the intake of important macronutrients (Robbins et al., 1991; Makkar, 2003). Instead of reducing ingestion of tannins per se, consumers may enhance secretion of tannin-binding salivary proteins (TBSPs). Secretion of TBSPs serves as an immediate reaction to reduce the detrimental effects of dietary tannins (Mehansho et al., 1987; McArthur et al., 1995; Yan and Bennick, 1995; Bennick, 2002; Soares et al., 2012), and the capacity of this mechanisms as well as its plasticity varies with the phytochemical features of natural diets. Research on the effects of dietary tannin intake and related adaptations of the digestive system, in particular saliva composition, has been biased to herbivores in general and ruminants in particular (Kumar and Singh, 1984; Shimada, 2006; Lamy et al., 2011). Some studies found consistent links between general features of diet composition (e.g., grazers vs. browsers and the tannin-binding capacity of saliva [Clauss et al., 2005] but others studies did not come to the same conclusion [e.g., Ward et al., 2020]). Compared to the large body of research on herbivores, very little is known about the nutritional significance of tannins for species relying on a frugivore diet. Comparing data from humans with those from primates, Thamadilok et al. (2020) suggested that the natural diet is maybe a driving force for the differentiation of the salivary proteome. In mammals, including primates, saliva is produced by exocrine salivary glands located in the oral cavity. Across primate species, the morphology of salivary glands is comparable (Steiner, 1954; Gibbs et al., 2002; Ankel-Simons, 2007). The three paired major glands, parotid, submandibular, and sublingual glands, release 90–95% of saliva. Smaller amounts of saliva, 5–10%, are secreted from glands located in the mucous and submucous membranes (Shackleford and Klapper, 1962; Sreebny, 2000). The main constituents of saliva are approximately 99% water, and 1% are electrolytes, mucus, antimicrobial compounds, and various macromolecules which are almost exclusively proteins (Bennick, 1982; de Almeida et al., 2008). Salivary protein composition varies within and across species, and is closely correlated with age, morphological/anatomical structures, diet composition, and feeding behavior (Chauncey et al., 1963; Bennick, 2002). Analyzing the specific salivary compositions provides insights into the evolutionary adaptations of the species to their diet. Considering the amount of dietary tannin as a driving force for salivary composition, proline-rich proteins (PRPs; Bennick, 1982; Lu and Bennick, 1998; Soares et al., 2012) and histidine-rich proteins (HRPs or histatins; Yan and Bennick, 1995; Padovan et al., 2010) are of functional significance in binding tannins (Austin et al., 1989). In saliva of Homo sapiens, PRPs account for 70% of total protein content (Mehansho et al., 1987; Lu and Bennick, 1998). Proline constitutes 25–42% of the amino acids in PRPs, the other main components are glutamic acid and glycine (Bennick, 1982; McArthur et al., 1995). Taken together, these three amino acids account for 70–88% of all amino acids in PRPs (Bennick, 1982). Proline concentration contributes significantly to high tannin-binding affinity of proteins (Hagerman and Butler, 1981; Mehansho et al., 1987). The structural characteristics of proline, especially its strong conformational rigidity with fixed torsion angels caused by the cyclic structure of its side chain, as well as its functional role as structural disruptor or helix breaker, lead to open, loose protein structures that involve high tannin-binding capacities (Rinaldi and Moio, 2021). Along with serine and threonine, proline is a major component (on average 10%) of the central protein of salivary mucins, and responsible for protein-phenol complexation that leads to an astringent oral sensation (Gombau et al., 2019). Human nonglycosylated salivary α-amylase (SAA) contains 22 proline residues that form important polyphenol binding sites (Soares et al., 2007).

Even though PRPs and other TBSPs are of particular interest in great apes as fruit and leaf eaters, their salivary protein and amino acid composition is not well explored. In our study, we investigated the salivary amino acid profiles of zoo-housed Gorilla gorilla, P. paniscus, P. troglodytes, and Pongo abelii to explore possible relationships between diet and salivary composition. Although we did not measure tannin content of food and/or tannin-binding capacity of saliva, we interpreted the data in the context of the feeding strategies of the four ape species, assuming that high levels of proline in combination with glutamic acid and glycine correlate with high levels of PRPs. We also presumed a similar correlation for levels of histidine and HRPs. Furthermore, proline could originate from other salivary proteins like mucins or SAA that also exhibit high tannin-binding capacities. The proline and histidine concentrations together with the levels of total salivary protein of zoo-housed apes were used as an indicator to assess whether the hominid species have adapted to the common low-tannin zoo diet or if they maintain adaptations to their natural diet. In addition, we compared the measures of salivary profiles obtained from hominid primates with corresponding values from H. sapiens. Measures from H. sapiens samples are used as a benchmark against which we compared measures in samples from ape species. Although the diet of H. sapiens contains meat and other foods of animal origin that are rarely consumed by zoo-housed apes, evidence suggests that human populations maintain adaptations to the intake of secondary compounds (Lu and Bennick, 1998). We compared this preservation of adaptations to those of apes feeding on low-tannin diets.

It has been shown that diet composition affects patterns of feeding behavior, food processing, and digestive physiology (Dew, 2005; German et al., 2010; Clauss et al., 2013), and it is therefore reasonable to hypothesize (H1) that zoo-housed apes adapted to a low-tannin diet. In this case, we predict that (1) salivary amino acid profiles are similar across species, and (2) similar to the profiles of H. sapiens living on a similar diet. An alternative hypothesis (H2) is that physiological adaptations to natural diets persist in spite of life-long exposure to artificial diets. Comparing passage rate across species, Clauss et al. (2008) found that interspecific variation corresponds with dietary patterns of natural foods. Another study found that variation in SAA content of hominoid primates matches with phytochemical properties of natural diets (Behringer et al., 2013). The key prediction of H2 is that salivary amino acid profiles reflect signatures that correspond to patterns of natural diets (see Table 1). In this case, we assumed that proline concentrations are highest in samples of G. gorilla and lowest in samples of P. troglodytes. All types of PRPs are also rich in glycine and glutamine (McArthur et al., 1995), and therefore, high proline concentrations are likely to be associated with higher concentrations of glycine and glutamine. We also predict higher concentrations of histidine in tannin-consuming species as HRPs are known to have high tannin-binding capacities (Yan and Bennick, 1995; Padovan et al., 2010). We did not expect sex and age differences in individuals if diet composition drives amino acid composition.