**The Impact of Diet on Microbiota Evolution and Human Health. Is Diet an Adequate Tool for Microbiota Modulation?**

#### **Laura Moles \* and David Otaegui**

Multiple Sclerosis Group, Neurosciences Area, Biodonostia Health Research Institute,

C. P. 20014 Donostia-San Sebastián, Spain; david.otaegui@biodonostia.org

**\*** Correspondence: laura.moles@biodonostia.org; Tel.: +34-943-00-61-33

Received: 29 April 2020; Accepted: 31 May 2020; Published: 2 June 2020

**Abstract:** The human microbiome is emerging as an interesting field in research into the prevention of health problems and recovery from illness in humans. The complex ecosystem formed by the microbiota is continuously interacting with its host and the environment. Diet could be assumed to be one of the most prominent factors influencing the microbiota composition. Nevertheless, and in spite of numerous strategies proposed to modulate the human microbiota through dietary means, guidelines to achieve this goal have yet to be established. This review assesses the correlation between social and dietary changes over the course of human evolution and the adaptation of the human microbiota to those changes. In addition, it discusses the main dietary strategies for modulating the microbiota and the difficulties of putting them properly into practice.

**Keywords:** gut microbiota; Western diet; chronic disease; prebiotic; probiotic

#### **1. Introduction**

The human gut microbiota is a complex ecosystem formed by thousands of microorganisms that play an important role in human immune and metabolic functions, among others. It is estimated that more than 1000 species and 3 <sup>×</sup> 10<sup>13</sup> microbial cells live in or on us, being similar in number to human cells [1–3]. In terms of complexity and richness, the microbiota is even larger considering its genome (the microbiome). Specifically, the human microbiome has at least 100-fold more genes than the human genome; besides this, only 10% of the microbiome is shared between individuals. Therefore, the human microbiome is a unique fingerprint, and its richness and variability may explain its ability to adapt fast to environmental conditions [4–6].

The human microbiome is a relatively new field, but in recent years research into it has been increasing exponentially. The importance of the microbiota was underlined by it starting to be considered an organ in itself [4,5,7–9]. As the so-called "forgotten organ" [10] and considering its wide-ranging interaction with the host, research in this field could contribute to our understanding of many health problems that, so far, have proven difficult to tackle. In this context, the association of the human microbiota with health and disease is being intensively studied, and every day evidence emerges relating dysbiosis in the microbiota to more health problems, including diverse gastrointestinal and neurological disorders such as colitis, obesity, irritable bowel syndrome, Alzheimer's disease, autism, or multiple sclerosis; allergies; and some types of cancer (Figure 1) [11–17].

**Figure 1.** Host microbiota interactions and their relationship with disease.

The gut microbiota is especially moldable during infancy and notably stable in adulthood [18]. The limited microbiota present at birth undergoes dramatic changes before reaching the relative equilibrium that is characteristic of adulthood [7–9,19–21]. It is precisely in infancy when factors modulating the microbiota have the most marked influence [8,18,22]. Diet has been recognized as one of the strongest modulators of infant microbiota. In fact, numerous studies have described differences in the gut microbiota of breastfed and formula-fed infants [23,24]. It is believed that once the microbiota reaches an equilibrium (at 2–3 years of life), it is much more difficult to restore and modulate its composition. Once in adulthood, the gut microbiota remains relatively stable but diet continues to determine its composition. Studies carried out evaluating the microbiota associated with different diets in adulthood agree on the dominant presence of *Prevotella* in the gut of vegetarians, while levels of the genus *Bacteroides* and overall levels of the phylum Firmicutes are higher in people following diets high in protein and animal fats [25–27].

Consequently, it seems evident that our diet has the potential to modulate our microbiota. In this context, the aim of this study is to outline the challenges in dietary modulation of the microbiota, reviewing evolutionary changes in both diet and gut microbiota, their potential relationship, and consequences for human health and subsequently examining the strategies available for modulating the microbiota.

#### **2. Dietary Changes across Human Evolution**

Nutrition is one of the basic needs for a living being to survive and grow. Without a doubt, the human diet has dramatically changed from the time of first hominids to the present, and the changes have been especially fast and marked over the last 100 years [28]. Here, we describe the main features that have characterized these modifications in human diet.

#### *2.1. Ancient Diet*

The first hominids based their diet on plants gathered and animals hunted in the wild. Though relatively little is known about this time, it is believed that plants were the main foods eaten, while meat was limited to days on which hunting was successful. It is important to highlight that the ratios of plant to animal contributions of the hunter-gatherer diet are still controversial [29,30]; nevertheless, data from the current hunter-gatherer populations around the world suggest the predominance of

plant food [22,31–37]. The development of the ability to control fire had a great impact on many aspects of human life. It provided protection from predators and warmth and light and was a determinant factor in the development of cooking. Cooking contributes to food energy accessibility by the efficient denaturing of proteins and starch gelatinization; it also preserves foods for longer periods by substantially reducing foodborne pathogens [38–40].

Another feature that caused marked dietary changes was the domestication of plants and animals. Agriculture allowed the availability of food to, more or less, meet the demand, and is considered a key element in the emergence of community life and civilization [38]. The spread of agriculture had other consequences, however; in particular, it led to a reduction in nutritional intake diversity.

#### *2.2. First Civilization's Diet*

Populations from the first civilizations were able to produce their own food to meet their energy requirements. Dietary patterns were characterized by the development of the first fermentable foods, such as bread, beer, yoghurt, and wine. Furthermore, for centuries—though there were differences between civilizations and cultures—dietary habits were generally based on the consumption of carbohydrate-dominant foods (such as potato, rice, maize, wheat, and vegetables), probably because these were the most easily accessible [41,42].

Protein intake was primarily from legumes, as proteins of animal origin were only consumed occasionally [38,41–43]. Animal domestication facilitated access to meat and animal-derived products; nevertheless, cattle slaughter was commonly carried out only once or twice a year, and the meat obtained was used to supply whole families. On the other hand, the techniques used to preserve meat and fish were still quite limited; epidemics, famines, and wars marked civilizations for long periods, restricting access to valuable products, including animal-derived foods. Therefore, animal proteins remained a minor component of diets [38] (Figure 2).

**Figure 2.** Drivers of dietary trends and their relation to microbiota composition and changes in human health.

#### *2.3. Modern Diet*

Demographic and lifestyle changes, such as urbanization, the abandonment of rural areas, and the increase in women working outside the home, have marked current populations. Through the 1990s, the worldwide growth in the use of antibiotics and industrialization in ranching and agro industries led to a new revolution in food technology and production. Together, these changes had a huge impact on food production and dietary habits [38].

On the one hand, globalization and advances in agriculture have nearly eliminated the seasonality of foods in developed countries [44]. Indeed, food availability is such that individuals have a wide choice of what to eat. On the other hand, the adaptation of consumer behavior to modern lifestyles has led to the demand for safer and longer-lasting food; this, in turn, has driven industry to increase the use of additives and develop new preservation technologies. Cooking has become a secondary concern, as the consumption of and demand for pre-cooked and ready-to-eat products has exponentially increased. These products must be tasty as well as easy to prepare and store; in this context, the addition of fats, sugar, and salt is imperative in meeting these requirements [45]. In addition, competitive markets force producers to use cheap ingredients in processed foods, and these are hardly ever the healthiest ones [46]. All these changes underlie the food industry's transition from the natural products consumed by our ancestors to the processed products currently available, which tend to be high in artificial and added ingredients such as preservatives, colorants, fats, sugar, and salt (Figure 2) [44,47,48].

#### *2.4. What the Numbers Say*

Global industrialization has facilitated these changes in diet and, notably, similar changes are observed in many countries despite differences in culture, lifestyle, and culinary traditions. According to the annual food balance sheets published by the Food and Agriculture Organization (FAOSTAT database; http://www.fao.org/faostat/en/#home), calorie intake has been increasing over recent decades. From the 1960s to the present (last data from 2013), the world's average energy intake has increased by nearly 500 kcal per capita per day (Figure 3). The origin of this calorie increase is slightly different in developed and developing countries. While the consumption of meat, sugars, and vegetable oils has increased in developing countries, developed countries have seen rises in meat and fat intake [44]. The Food and Agriculture Organization data show that the increase in energy intake in European Union countries is approximately twice that observed in developing countries (Table 1).

**Figure 3.** Increase in calorie intake in different regions over the last 50 years.



The dietary pattern involving a high intake of saturated fats and sucrose and a low intake of fiber is commonly known as a "Western diet". Diet is one of the strongest modulators of chronic inflammation and Western diets represent a growing health risk, contributing to higher rates of metabolic diseases and inflammation [31,47,48].

#### **3. Human Gut Microbiota Evolution**

The human microbiota has been structured by its biological interaction with its host, and the resultant ecosystem is the consequence of thousands of years of evolution [49]. Furthermore, microbes have impressive abilities to spread, interact, and adapt to the environment; hence, microbial communities should not be considered in isolation, but rather as part of an interacting community [50]. The knowledge in this relatively new field is still quite limited. In fact, it is difficult to know the extent to which the human microbiome has been shaped by the selective pressure of modern diet, hygiene, antibiotic exposure, built environment, and lifestyle [51]. Despite these limitations, the following paragraphs attempt to outline the current knowledge on gut microbiota composition and its evolution.

#### *3.1. Defining "Healthy Microbiota"*

In general, it is accepted that only four bacterial phyla dominate the human microbiota (Firmicutes, Proteobacteria, Actinobacteria, and Bacteroidetes), while others (Chlamydiae, Cyanobacteria, Deferribacteres, Deinococcus–Thermus, Fusobacteria, Spirochaetes, or Verrucomicrobia) may be found at lower abundances [52–55]. Strict anaerobes, mainly represented by members of the phyla Bacteroidetes and Firmicutes, dominate the gut, outnumbering aerobe microorganisms by 100- to 1000-fold [56–60]. Facultative anaerobes account for less than 1% of the microbiota and are mainly represented by the family *Enterobacteriaceae* and the genera *Enterococcus* and *Lactobacillus* [19].

Despite research efforts, there is no consensus on a detailed description of a "normal" or "healthy" microbiota. The enormous complexity and inter-individual variability in the microbiota make this goal very hard to achieve with current tools. Microbial richness (number of species) and diversity (variety and relative abundance of the species in a niche) are global parameters associated with health. Stability has also been considered as a key feature of a healthy microbiota, and this is related to the concepts of resistance (ability of a community to resist change in the context of ecological stress) and resilience (its ability to return to an equilibrium state following a stress-related perturbation).

Nevertheless, the idea that there is an ideal composition of the microbiota seems too simplistic. In fact, it minimizes the importance of microbiota–host interactions, individual genomic differences, and variations in susceptibility to disease, all of which probably play a determining role in shaping the microbiota. An alternative concept consists of characterizing the collection of genes and metabolic pathways provided by the microbiome rather than just the microbiota composition. This approach is probably more appropriate, but also requires a greater in-depth knowledge of the human microbiome [59].

In any case, it should be a priority to reach a consensus on the definition of a healthy microbiota in order to clarify the goal of strategies for microbiota modulation.

#### *3.2. Clustering Individuals According to Their Microbiota Composition*

As a step towards defining the composition of a healthy gut microbiota, an interesting publication clustered the fecal microbiota of a healthy cohort into three so-called "enterotypes". Each cluster was characterized by the presence of some highly abundant genera that defined the group and many less abundant genera. Enterotype 1 was enriched in the genus *Bacteroides* and enterotype 2 in *Prevotella*, while enterotype 3 was dominated by *Ruminococcus*. The dominant genera tended to be observed together with other minority ones (*Parabacteroides*, *Desulfovibrio,* and *Akkermansia,* respectively) that, despite their low abundance, performed specialized functions beneficial to the host and are important for defining the enterotype. Though each enterotype preferred certain routes for generating energy, which suggests a specialization to their ecological niches [54], the data available support the idea

that the gut microbiota is characterized by a high functional redundancy. In fact, 25% to 43% of the enzymatic functions of the microbiota have been found to be shared, regardless of the enterotype to which the microbiota belonged [21,60].

Some years later, another publication associated these enterotypes with long-term diets. The *Bacteroides* enterotype was strongly associated with a variety of amino acids from animal proteins and saturated fats, and therefore with Western diets. In contrast, the *Prevotella* enterotype was closely associated with carbohydrates and simple sugars, indicating an association with typical diets of agrarian societies [25].

A recent publication uses a metagenomic approach to classify individuals according to the number of gut microbiota-encoding genes as "low gene count" (LGC) or "high gene count" (HGC), depending on whether their microbiota harbor fewer or more than 480,000 genes, respectively. This approach is based on the functionality of the microbiota and its relation with the microbiota composition. The difference in the mean number of encoding genes between groups is notably high, reaching some 40%, and is related to the microbial richness. Broadly, LGC individuals have a less rich microbiota, dominated by *Bacteroides*, *Parabacteroides*, *Ruminococcus*, *Campylobacter*, *Dialister*, *Porphyromonas*, *Staphylococcus, Anaerostipes* and most members of the phylum Bacteroidetes. In contrast, the phylum Firmicutes and the genera *Faecalibacterium*, *Bifidobacterium*, *Lactobacillus*, *Butyrivibrio*, *Alistipes*, *Akkermansia*, *Coprococcus,* and *Methanobrevibacter* are associated with HGC individuals [61,62].

#### *3.3. Ancient Microbiota*

Regarding the changes in our gut ecosystem over the course of human evolution, some studies suggest that these are both pronounced and worrying. The study of the ancient microbiota is not easy due to the low number of available ancestral biological samples. The gut microbiota of ancestral specimens was evaluated in mummies, revealing the predominance of species of the genera *Clostridium* and *Bacteroides* in the larger intestine [63–65]. These studies provided valuable information, however, in addition to the small sampling size, the storage conditions and the possible post-mortem alterations in the bacterial communities should be considered in the interpretation of the results. In this context, current hunter-gatherer populations are also being studied. Research with uncontacted Amerindians who continue to live a seminomadic hunter-gatherer lifestyle revealed that their fecal microbiota is the most diverse ever reported in humans, and the proportion of shared microbiota between them is also much higher than in other human populations [32]. This high microbial biodiversity was also observed in studies carried out in other hunter-gatherer populations, such as the Matses from the Peruvian Amazon [66], the Hadza from Tanzania [33], or indigenous ethnic groups from the Central African Republic [36]. The Ameridians' microbiota seems to be characterized by a high abundance of the phyla Verrucomicrobia and Mollicutes; the families *Aeromonadaceae*, *Oxalobacteraceae,* and *Methanomassiliicoccaceae;* and the genus *Prevotella*, while the abundance of the genus *Bacteroides* is lower [32]. The microbiota of the Matses is characterized by the abundance of the genera *Clostridium*, *Catenibacterium*, *Eubacterium*, *Lachnospira,* and *Treponema* [66]. The Hadza population presented a microbiota enriched in *Prevotella*, *Succinivibrio*, *Treponema,* and *Eubacterium* and impoverished in *Bacteroides*, *Blautia,* and *Dorea* genera [33]. The Central African Republic hunter-gatherer population's microbiota is characterized by the predominance of *Prevotella* and *Treponema* [36]. Despite the differences, it is worth noting that the abundance of *Prevotella*, *Treponema,* and *Eubacterium* and the scarcity of *Bacteroides* in the microbiota of these populations may be common characteristics of the ancestral microbiota.

Metagenomic approaches allow us to analyze the genetic composition and function of complex communities. The application of these tools to the ancient microbiota provide further evidence to support the view that it has a higher functional diversity, characterized by increased metabolic pathways involving amino acid metabolism; glycosyltransferases; and the biosynthesis of lipopolysaccharides, terpenoid-quinones, and vitamins [32]. These findings suggest that not only is microbial diversity

being lost, but also some of the functionality of gut microbials. As a consequence, it is not surprising that there is growing interest in ancient microbiome research and recovery [22,51,67].

#### *3.4. Western Diet Microbiota and Its Consequences*

Evidence suggests that lifestyle changes, including poor diet, urbanization, scarce physical activity, built environment, wide-spread antibiotic exposure, and better hygiene, have impacted the composition of our microbiota and also the emergence of the so-called diseases of modern civilization. These changes are included in the concept of "Westernalization" and contribute to microbiota alteration and disease [68]. Even if all aspects of Western lifestyle should be considered in this process, diet is accepted as one of the most potent ones shaping microbial communities [68–70].

There is a tendency to lose the overall diversity of the gut microbiota in people following Western diets. The gut microbiota composition has also undergone specific changes, characterized by an increase in the abundance of the phylum Firmicutes and the family *Enterobacteriaceae* and a decrease in the phylum Actinobacteria and the genus *Prevotella*. The presence of some bacterial species associated with anti-inflammatory conditions and the capacity to produce beneficial metabolites is diminishing in our guts. These species include *Akkermansia muciniphila*, *Faecalibacterium prausnitzii, Roseburia spp*., *Eubacterium hallii, Clostridium clusters XIVa* and *IV,* and *Ruminococcus*, among others. Indeed, some research has revealed the extinction of several bacterial groups from the guts of people following Western diets. It is difficult to assess the significance of that loss, but we are probably witnessing just the beginning of its consequences [12,31,47,71].

Furthermore, the gut microbiome's circadian rhythm is influenced by factors such as light–dark cycles, sunlight exposure, sleep, and dietary patterns; some of them are common stressors of the modern lifestyle [72]. The consumption of food in an undisturbed daily rhythm coincides with the light phase of the light–dark cycle and the activity phase of the day, which has consequences on the regulation of the hosts' intestinal cell transcription, the rhythms of the circulating metabolites, and the gut microbiota composition and function [72,73]. Some studies evidenced exacerbated effects on the gut microbiota of people with the circadian disruption of high sugar and fat diets; these effects were characterized by a drastic reduction in bacterial diversity and the Firmicutes/Bacteroidetes ratio [74,75].

The aforementioned changes in microbiota composition have been associated with a greater tendency to develop inflammation and, in turn, with a higher incidence of obesity; diabetes; allergies; cardiovascular disease; and metabolic, gut, and neurological disorders. Microbiota dysbiosis in these diseases may be involved in the alteration of certain specific microbial groups; nevertheless, in most cases the overall loss of microbial biodiversity is an important factor defining the dysbiosis [76]. The misalignment of the rhythms that control our energy metabolism also increases the risks of suffering diseases such as metabolic syndrome, including type 2 diabetes mellitus and obesity [77,78]. The growing incidence of these diseases in contemporary, industrialized populations over recent decades is believed to be associated, among other factors, with a lack of adaptation of our metabolism to the rapid dietary and lifestyle changes that have occurred over the course of human evolution [22,32,51,79–81].

It is likely that several different factors are contributing to the changes in microbiota composition and the increased prevalence of associated diseases. In any case, the impact of these changes on human health underlines the urgent need to find effective tools to halt this trend.

#### **4. Modulation of Human Gut Microbiota with Diet**

While it has already been described that geographical localization, culture, and genetic background all affect the microbiota composition, some authors consider that diet is responsible for more than 50% of the variability in the microbiota [82,83]. Even though it is difficult to determine this value accurately, there is evidence that dietary interventions, with significant changes in content, are able to exert modulatory effects on microbiota composition that may be seen within 1–4 days and are strong enough to shift the enterotype [84,85]. Nevertheless, dietary modulatory effects are diluted over time when the diet is discontinued, and there is a tendency to return to the original state [82].

The capacity of microbiota to recover its original status has also been observed after a course of antibiotics. Several studies have evidenced that some weeks after the use of antibiotics (one of the treatments that most dramatically alter the microbiota), the microbiota has nearly completely returned to its original composition, though this recovery is treatment- and age-dependent [86–88]. The frequent use of antibiotics or the requirement for prolonged treatments has a more marked effect on the microbiota composition [89,90]. Similarly, the modulatory effect of probiotics (live microorganisms which, when administered in adequate amounts, confer a health benefit to the host [70]; dead microbes, microbial products, or microbial components do not come under the probiotic classification [91]) is believed to disappear progressively together with the loss of the beneficial strains.

All this evidence supports the idea that treatments to modulate gut microbiota must be maintained over time. In line with this, the use of diet as a modulatory tool could be ideal whenever diet is considered as a long-lasting change in everyday habits.

#### *4.1. Strategies for Modulating the Microbiota: Prebiotics*

Possibly the most widely explored strategy for modulating the microbiota is the use of prebiotics. Prebiotics are defined as "substrate that is selectively utilized by host microorganisms conferring a health benefit" [92]—that is, nutrients resistant to gastric acid secretion and digestive enzymes that once in the gut stimulate the growth of beneficial microbes or their activity. Certain dietary components such as inulin, fructooligosaccharides (FOS), galactooligosaccharides (GOS), and resistant starch (RS) have been studied as prebiotics, and their efficacy is commonly indirectly measured by the production of short chain fatty acids (SCFAs) or the decrease in intestinal pH [76,93]. Inulin is a fructan carbohydrate that may vary its polymerization degree and whose fructose chains ranges from 2 to 60 monomers [94]. Inulin stimulates the growth of lactobacilli and bifidobacteria; besides this, an increase in *F. prausnitzii* and *A. muciniphila* populations in the gut has been described and seems to produce early satiety by modulating the gut endocrine function. Nevertheless, it is still difficult to determine the mechanisms underlying these effects [93].

The FOS are oligosaccharides of glucose and fructose that differ from inulin in their polymerization degree that is under 10; whereas, GOS are oligosaccharides of glucose and galactose with a polymerization degree of 2 to 8. As typical prebiotics, FOS and GOS have been used to stimulate the growth of the beneficial bacteria, bifidobacteria and lactobacilli. The administration of FOS in a culture-dependent study resulted in an increase in *Bifidobacterium* and *F. prausnitzii*, while culture-independent studies based on high-throughput sequencing have revealed changes in more than 100 bacterial taxa. The most marked changes in abundance were an increase in *Bifidobacterium;* reductions in the genera *Phascolarctobacterium*, *Enterobacter*, *Turicibacter*, *Coprococcus,* and *Salmonella*; an overall increase in Bacteroidetes; and a decrease in the phylum Firmicutes [95]. Other genera that could be increased by FOS administration are *Lactobacillus* and the butyrate producers *Faecalibacterium*, *Ruminococcus,* and *Oscillospira* [96]. On the other hand, GOS administration resulted in increases in *Bifidobacterium* levels and decreases in the levels of *Ruminococcus*, *Dehalobacterium*, *Synergistes,* and *Holdemania* [95]. The effect of GOS on the gut microbiota could also improve the butyrate production and the presence of butyrate producers, such as *Eubacterium rectale* [97].

RS is defined as the total amount of starch and the products of starch degradation that resists digestion and has been shown to be composed of a linear molecule of α-1, 4-D-glucan, derived from the retrograded amylose fraction. Various classifications have been proposed for RS based on four or five types, and the content of RS in foods is influenced by the physical form of the food, the size and composition of the starch granules (amylose–amylopectin ratio), and the food processing methods and conditions [98,99]. Notably, it has been found that an increment in RS in the diet is associated with colonization by higher levels of the phylum Bacteroidetes and the genera *Bifidobacterium*, *Akkermansia,* and *Allobactum* [98].

#### *4.2. Strategies for Modulating the Microbiota: Probiotics*

The field of probiotics—which, as stated above, are live microorganisms which when administered in adequate amounts confer a health benefit to the host [70,91]—has notably grown in recent years. The microorganisms commonly used as probiotics are the yeast *Saccharomyces cerevisiae* and members of the bacterial genera *Lactobacillus* and *Bifidobacterium*, though some formulations may also include some *Streptococcus*, *Enterococcus*, *Pediococcus*, *Propionibacterium*, *Bacillus,* or *Escherichia* strains. Most *Lactobacillus* and *Bifidobacterium* species have been assigned "Generally Recognized As Safe" status by the US Food and Drug Administration and "Qualified presumption of safety" status by the European Food Safety Authority, facilitating their preferential use as probiotics. On the other hand, their long history of use as probiotics means that there is a substantial body of evidence for a wide range of beneficial properties [100], though we must recall that probiotic properties are strain-specific—that is, they are not a characteristic of a species [76].

Nevertheless, it can be expected that, in the near future, other species will be used as probiotics—ones that are more commonly found in the human gut and play important functions in mitigating intestinal inflammation, inducing immune regulation, or enhancing the intestinal barrier function. These are likely to include species with anti-inflammatory properties (*A. muciniphila*, *F. prausnitzii*) and butyrate-producing bacteria [91,101].

Currently, probiotics are used in a wide range of contexts, normally being indicated for healthy people in a special situation (e.g., during infancy, pregnancy, breastfeeding, and old age) and for preventing or treating several specific health problems. As a consequence, the assessment of the safety of probiotics must pay attention not only to the selected strain, manner and frequency of administration, and dose and treatment duration but also to the potential vulnerability of the consumer and the physiological function that the strain may play in the host [102].

Despite the heterogeneity of clinical studies making it difficult to determine the most suitable strains and therapeutic guidelines, there is clear evidence of probiotic effectiveness in the prevention or treatment of diseases such as necrotizing enterocolitis, antibiotic-associated diarrhea, colitis, or acute gastroenteritis [103,104]. Nonetheless, to date there is a lack of evidence of the effectiveness of prebiotics or probiotics in achieving long-term changes in the microbiota.

#### *4.3. Strategies for Modulating the Microbiota: Controlling the Gut Environment*

The concept of gut environment modulation considers not only the composition of the microbiota but also its function and interaction with the host. Nowadays, it is known that fat-rich diets enhance bile secretion to facilitate lipid digestion. Bile acids have antimicrobial properties; their detergent effect produces damage to bacterial membranes and exerts a strong selective pressure on microbiota composition. Studies carried out in rats reveal the strong resistance of the phylum Firmicutes to bile acids, in particular, the classes *Clostridia* and *Erysipelotrichia* and the family *Enterobacteriaceae* [105]. These findings are in agreement with the predominant presence of Firmicutes in people following Western diets [79].

On the other hand, intestinal pH may oscillate between 5 and 7 under normal physiological conditions, depending on the fermentation products present (such as SCFAs) and the metabolite absorption by host epithelial cells and their level of bicarbonate secretion. Intestinal pH affects not only the microbiota composition but also its metabolism. Some Firmicutes species, especially those belonging to *Clostridium cluster XIVa,* are tolerant of a low pH; however, many Bacteroidetes and Actinobacteria members are more sensitive to pH changes [12,106].

The Mediterranean diet is associated with a higher production of SCFAs in the gut [107,108]. These substances play important roles maintaining the integrity of the large bowel and small intestinal barrier, providing energy to epithelial cells and reducing inflammation [108] and support higher microbial richness in the gut [107,109].

#### *Nutrients* **2020**, *12*, 1654

The modulation of the gut microbiota through the induction of environmental changes has been less explored; but the gut environment remains closely related to dietary pattern and therefore should not be disregarded in future dietary interventions.

#### *4.4. Challenges in Microbiota Modulation: Interindividual Variability*

The complexity of gut microbiota modulation lies in interpersonal variability not only in the microbiota composition and functioning but also in differences in lifestyle and genetic predisposition to disease. It has been described that changes in fiber consumption lead to various changes in the gut microbiota. In general, the abundance of the family *Lachnospiraceae* increases with the incorporation of insoluble fiber or wheat bran to the diet, while enrichment in RS causes an increase in *Ruminococcaceae*, but changes are individual specific [62]. Along the same lines, a recent study revealed that changes in oral glucose tolerance and the microbiota induced by a given prebiotic intervention vary between individuals. Notably, a close positive correlation was found between glucose tolerance and the presence of some butyrate-producing bacteria [95].

These data indicate the enormous importance of knowledge of the microbiota composition to predict the effects of modulatory treatments. The available data suggest that parameters such as age, diet, and lifestyle, body mass composition, and the presence of specific diseases could help to define some key features of the microbiota composition, but treatments to modulate the microbiota are still based on estimates and it is likely that some effects are being masked.

#### *4.5. Challenges in Microbiota Modulation: Microbial Metabolic Redundancy*

To estimate the effects of a modulating treatment, in addition to microbiota composition there is a need to consider metabolic redundancy. The use of next-generation sequencing and metagenomics is greatly helping to advance our understanding of this factor. In this context, recent publications describe the metabolic capacity of some bacterial groups. It seems that members of the genus *Bacteroides* have a wide metabolic arsenal allowing them to utilize different polysaccharides, whereas Firmicutes members have less metabolic diversity for polysaccharide degradation and greater nutritional specialization [62].

Metabolic analyses of LGC and HGC individuals suggest a great capacity to handle oxidative stress but also a higher production of detrimental metabolites and a predisposition to inflammation in LGC individuals. On the other hand, HGC individuals' microbiota has a greater capacity to produce organic acids, including SCFAs. Further, it seems that there is a higher incidence of obesity and metabolic syndrome in LGC individuals and dietary interventions for weight loss increase microbiota diversity and the overall gene counts [61,62,110].

In accordance with this, the analysis of the metabolic specialization of the enterotypes indicated that individuals with the *Bacteroides* enterotype are better able to digest lipids, proteins and carbohydrates of animal origin, while the *Prevotella* enterotype showed a plant fiber hydrolysis specialization. In contrast, the *Ruminococcus* enterotype has no such marked specialization, but has been linked to a higher microbiota diversity and a lower inflammatory status in the host [84].

The aforementioned results suggest a healthy microbiota should be closer to that of HGC individuals and the *Ruminococcus* enterotype. Nonetheless, no consensus has yet been reached on the definition of a healthy microbiota. In fact, all enterotypes have been associated with some types of disease and it is believed that each enterotype has a different susceptibility to given illnesses. In any case, a greater knowledge of individuals' microbiota could be a good tool to guide treatment decisions and estimate patient response.



An alternative more affordable approach to exploring the microbiome function without resorting to metagenomics is the measurement of SCFA production. These metabolites are considered important for their influence on intestinal homeostasis, pH, and preventing the growth of potentially pathogenic bacteria. Moreover, SCFAs have anti-inflammatory and anti-apoptotic properties, contribute to the gut barrier integrity, and are key elements of the gut–brain axis. These fatty acids seem to be necessary for the proper maturation and functioning of microglia, also known as brain macrophages, that participate in brain physiology and homeostasis have phagocytic capacity and also participate in the integrity of the blood–brain barrier [53,111–113].

Microbiota-derived metabolites such as SCFAs depend directly on diet; mainly on its content of non-digestible carbohydrates, lipids, and proteins and the metabolic pathways available in the microbiome [12,62]. Some common microorganisms of the gut microbiota and the SCFAs that they produce are listed in Table 2. Individual dietary choices influence not only gut microbiota composition but its function, but we still lack an understanding of how the microbiome interacts with nutritional and host-genomic axes to confer predisposition to disease.

#### **5. Conclusions**

Diet and lifestyle habits have undergone dramatic changes since the origin of humanity. The first hominid's diet was based on the consumption of raw vegetables collected from the wild and a low intake of protein of animal origin. After thousands of years of evolution, the modern diet, influenced by globalization and consumerism, is characterized by an excessive lipid and energy intake and the introduction of processed and refined foods that are rich in lipids, sugars, salt, and preservatives. These foods have facilitated the accelerated pace of today's life, but it is likely that their negative consequences for human health are just beginning to be noticed.

Despite the relative youth of microbiome research, microbiota dysbiosis has already been associated with diverse health problems. Numerous international projects are focusing their research efforts on the study of human microbiome in different contexts, including the Human Microbiome Project (HMP: https://www.hmpdacc.org/hmp/), the International Cancer Microbiome Consortium (ICMC: https://www.icmconsortium.org/), the International Multiple Sclerosis Microbiome Study (iMSMS: http://imsms.org/home/), and the Inflammatory Arthritis Microbiome Consortium (IAMC); more are likely to be launched in the near future. Diet has been identified as one of the factors with the greatest influence on microbiota acquisition (in newborns) and modulation. Therefore, detailed studies of the complex interactions that occur between diet and microbiota are necessary to effectively direct desirable changes in human microbiota or alter its abnormal composition in disease. So far, making simple predictions of dietary effects remains extremely difficult.

Microorganisms, especially prokaryotes, are characterized by their amazing capacity to adapt to the environment, as evidenced by their ability to remain stable in an "altered microbiota". Human beings seem to be much less efficient at environmental adaptation. Certainly, the growing rate of diseases related to compromised immune or nervous system function and alterations in metabolism may be a response to a lack of adaptation to microbiota dysbiosis.

Most of the strategies designed to modulate the microbiota are based on one-off treatments that could be effective during their administration, but do not seem to produce stable changes in the microbiota, maybe with the exception of a fecal transplant (although there is still challenges with this practice [114]). In this context, diet will probably be the most powerful tool for microbiota modulation, but to achieve that a better understanding of diet–host–microbiota interactions is necessary. Learning how to use diet to generate a healthy microbiota must be a priority for society, and arguably represents one of the key steps to achieving real preventive and personalized medicine.

**Author Contributions:** Both authors contributed equally to this manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding

**Acknowledgments:** We would like to acknowledge the BIOEF (Basque Foundation for Health Innovation and Research) language service for the thorough revision of the text.

**Conflicts of Interest:** 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.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Review* **A Revolutionizing Approach to Autism Spectrum Disorder Using the Microbiome**

**Dinyadarshini Johnson 1, Vengadesh Letchumanan 1, Sivakumar Thurairajasingam <sup>2</sup> and Learn-Han Lee 1,\***


Received: 29 May 2020; Accepted: 30 June 2020; Published: 3 July 2020

**Abstract:** The study of human microbiota and health has emerged as one of the ubiquitous research pursuits in recent decades which certainly warrants the attention of both researchers and clinicians. Many health conditions have been linked to the gut microbiota which is the largest reservoir of microbes in the human body. Autism spectrum disorder (ASD) is one of the neurodevelopmental disorders which has been extensively explored in relation to gut microbiome. The utilization of microbial knowledge promises a more integrative perspective in understanding this disorder, albeit being an emerging field in research. More interestingly, oral and vaginal microbiomes, indicating possible maternal influence, have equally drawn the attention of researchers to study their potential roles in the etiopathology of ASD. Therefore, this review attempts to integrate the knowledge of microbiome and its significance in relation to ASD including the hypothetical aetiology of ASD and its commonly associated comorbidities. The microbiota-based interventions including diet, prebiotics, probiotics, antibiotics, and faecal microbial transplant (FMT) have also been explored in relation to ASD. Of these, diet and probiotics are seemingly promising breakthrough interventions in the context of ASD for lesser known side effects, feasibility and easier administration, although more studies are needed to ascertain the actual clinical efficacy of these interventions. The existing knowledge and research gaps call for a more expanded and resolute research efforts in establishing the relationship between autism and microbiomes.

**Keywords:** clinician; autism spectrum disorder; microbiome; aetiology; comorbidities; diet; prebiotics; probiotics; faecal microbial transplant

#### **1. Introduction**

Autism spectrum disorder (ASD) is identified with persistent deficit in social communication and phenotypic behaviours which are typically repetitive and restrictive [1]. This disorder affects more boys than girls in a ratio close to 3:1 [2]. There has been a significant increase in the prevalence of ASD over the decades, it is currently estimated to affect 1% of the general population [3]. The most recent Global Burden of Diseases, Injuries, and Risk Factors Study in 2016 estimates 62.2 million individuals live with ASD globally [4]. It was also demonstrated that the prevalence of ASD based on special education enrolment data within the United States (US) over 11 years, from 2000 (1.2 per 1000) to 2010 (5.2 per 1000) alone shows an increase of 331%. It was inferred that the diagnostic recategorization may be the possible explanation for this significant rise [5]. However, it will be an understatement if it is solely attributable to the changes introduced in the diagnostic criteria of autism and increased awareness

which will be explored in this review. It has been speculated that the evolving environmental influence which contributes significantly to the occurrence of ASD could be partly responsible for the rise in the ASD prevalence globally [6]. A rising prevalence in ASD is a matter of concern and calls for a more effective management of this disorder.

The study of human microbiomes has become a prime hope in understanding this disorder and catering to the growth of a more clinical-based intervention more than to merely augment the widely advocated behavioural therapies. The understanding of microbial–human host relationship which was once thought to be purely commensal in nature, if not pathogenic, skewing to a one-sided relationship has now evolved into a complex interaction holding imperative roles in key physiological processes in the human body [7]. Determining the role of the microbiome in human health has become an intriguing quest in recent decades. A plethora of health conditions have been associated with microbiome across a wide range of populations identifying a distinctive spread of microbiome in a selected patient cohort compared to a healthy cohort pointing to its potential etiological role in the occurrence of a commonly identified disorder. It is not an understatement if this could potentially invent a breakthrough management approach in curbing a wide variety of health disorders and pandemics which include obesity and cardiometabolic diseases, infections, respiratory, allergic, gastrointestinal, neuropsychiatric disorders and even cancers [7–14]. The National Institutes of Health (NIH) Human Microbiome Project, launched in 2007, was one of the large-scale projects initiated to support and catalyst the growth of this emerging field of research. The microbiome in five major habitats in the human body, which include the gastrointestinal tract, airway, skin, oral cavity and vagina, were explored in this pursuit [15]. The gut microbiome has been the most widely studied area, accounting for four-fifths of total publications in microbiome over the last four decades [9]. Germ-free rodents and controls—conventionally colonized rodents with specific pathogen free (SPF) rodent models—have been utilized to ascertain the microbial influence in behavioural outcomes which have been grouped into four domains including, ASD-mimicking behaviours, stress and anxiety-related behaviours, learning and memory and motor controls. The germ-free rodents exhibited significant deficits in social interactions, cognitive and motor functions compared to SPF rodents suggesting the imperative role of microbiome in neurobehavioral outcome [16]. The gut microbial composition, if disturbed, has an impact on the various physiological activities regulated by these microbes principally through its metabolites and has a bidirectional communication with the brain involving autonomic nervous system. Neuronal, neuroendocrine and immunologic pathways have been described through which the microbes contribute to the bidirectional signalling between the gut and the brain [17,18]. The bidirectional transfer of information between the gut and the brain is principally controlled by the vagus nerve. The gut microbiota communicate to the brain via endocrine and neurocrine pathways while the brain impacts the microbial composition via immune and humoral systems mediated by autonomic nervous system, thus establishing the gut–brain–microbiota axis [18,19]. In the context of ASD, the exploration of other habitats of the microbiome, in the vagina and oral region, highlights the influence of maternal factors in the development of ASD. The vertical transmission of disrupted maternal vaginal microbes to the offspring at birth predisposes the offspring to the prenatal risk of developing ASD [20]. The direct relation of the oral microbiome and ASD has yet to be established; however, the resemblance of infants' oral microbiome to the maternal microbiome during the first six months of birth ascertains the significant contribution of maternal microbiome in early stage of oral microbial colonization [21].

Although the aetiology of ASD largely remains unanswered, the emerging microbial knowledge may be a key finding in explaining the etiopathogenesis of ASD; however, with more extensive work needed to understand its involvement at the molecular level. In an attempt to understand the molecular involvement in ASD, a study on post-mortem brain tissue and small intestines of ASD subjects revealed that blood–brain barrier (BBB) and gut barrier were disrupted with significant neuroinflammation evident by increased expression of genes and markers associated with brain inflammation. It was further inferred that the gut–brain axis disruption may be associated with non-self-antigens which triggers neuroinflammatory reaction by crossing the damaged gut barriers, thus leading to ASD in genetically susceptible subjects [22]. The BBB has a pivotal role in early phase of brain development and neuronal functions [23–25]. It was found that adult germ-free mice models and the foetuses of germ-free mice's mothers had more permeable BBB compared to mice with pathogen-free gut microbiota. The faecal transfer from a pathogen- to germ-free mice models and administration of short-chain fatty acid (SCFA) producing bacteria were able to restore the permeability of BBB, emphasizing the roles of intestinal microbes and SCFA in guarding the integrity of BBB [26]. The genetic component has primarily conditioned ASD as a highly heritable disorder through twin studies and large population-based studies which assess familial risk [27]. Nevertheless, the influential role of environment may potentially supersede the sole genetic involvement in the development of ASD, especially with the recent data pointing to the integrative approach of epigenetics in the study of pathogenesis of ASD [28]. In a study involving 192 twins, genetic factors were made accountable for only 38% of ASD risk, whereas the remaining 68% was attributed to environmental factors [29]. The microbial composition is regulated and influenced by many factors which could be broadly classified into extrinsic and intrinsic factors. The extrinsic factors consist of mainly environmental elements which include diet, lifestyle, microbial exposure in early developmental phase and infection, whereas intrinsic factors are naturally occurring elements within an individual which include genetic make-up, metabolites, immunologic and hormonal aspects [30]. The understanding of microbial involvement as part of an interplay between the intrinsic and extrinsic factors gives rise to an immense possibility of associating various factors in explaining its relation to ASD; however, the research progress in understanding the actual etiological mechanism involved in ASD development is still at an infant stage with a lack of defined explanations [18,31,32]. The co-existing health conditions in ASD is another major challenge in both research and clinical fronts. At least one comorbidity exists in about 70% of children with ASD, while 41% have two or more, reflecting the onerous disease burden in this population [33]. There are limited numbers of published studies on comorbidities alone in ASD due the complex and large spectrum of heterogeneity associated with this disorder. The widely reported and studied medical and psychiatric comorbidities in ASD include gastrointestinal (GI) disorders, epilepsy, depression and anxiety disorder [33–35]. These comorbidities can be individually linked to the microbiome which either overlaps with a similar ASD pattern of microbial dysbiosis, or that the related use of medication induces microbial dysbiosis which could contribute to ASD occurrence. This relation points towards the need to include the assessment of associated comorbidities to further understand the possible shared aetiology in ASD development, as many microbial studies using ASD subjects often do not take into account their existing comorbidities.

This review therefore focuses on autism spectrum disorder and microbiome in general while examining its relation to the hypothetical aetiology of ASD and its commonly associated comorbidities.

#### **2. Evolving Conceptualization of Autism Spectrum Disorder**

ASD is a behaviourally defined, neurodevelopmental disorder which lacks specific clinical biomarkers and has seen an evolving conceptualization over the last seven decades since it was first described [36,37]. The Centers for Disease Control and Prevention (CDC) reports that the earliest known median age of ASD diagnosis is at 52 months in the United States (US), whereas in the United Kingdom (UK), the median age of ASD diagnosis is reported at 55 months [38,39]. In general, ASD is diagnosed by three years of age in most children, although roughly 39% of children are not first evaluated until after four years of age [38]. A psychiatric diagnosis which has behaviour as its basis of definition heavily relies upon a meticulous observation and clinical expertise as it lacks standardised biomarkers [40]. Therefore, it is crucial to understand the core concept of a disorder to facilitate a more effective diagnostic process. The idea of autism emerged as early as in the 19th century when a Swiss psychiatrist, Eugen Blueler described the aloofness he observed in individuals with schizophrenia as a form of schizophrenic trait itself [41,42]. It was in the 1943, when two child psychiatrists, Leo Kanner and Hans Asperger, garnered the attention of scientific community through their publications, which

eventually led to recognising autism as a distinct category of diagnosis in children years later [42]. They individually published clinical cases of children with distinctive behaviours, primarily reflecting social deficits and echoing the descriptive term, autism coined by Blueler [43,44]. However, autism was often identified as a form of schizophrenia until a more refined and stand-alone diagnostic classification was first introduced in 1980 by American Psychiatric Association in Statistical Manual of Mental Disorders (DSM-III) and subsequently by the World Health Organization in International Statistical Classification of Diseases and Related Health Problems (ICD-10) in 1990 [36]. These are the two broadly adopted sources of reference for both research and clinical purposes with revised versions to date. The most recent definition, based on the DSM-5 published in 2013, includes subtypes of autistic disorders, Asperger's syndrome and Pervasive developmental disorder not otherwise specified (PDD-NOS) under one umbrella term known as ASD. DSM-5 allows classification of ASD based on its severity and takes into account intellectual ability and other comorbidities [1]. Despite an evolving conceptualization from a disorder to a spectrum highly denoting its clinical diversity, the core definition of autism has remained central to deficits in social interaction and stereotypic behaviours. It took several decades to define ASD within a smaller framework of classification to facilitate diagnostic process. However, it still remains a debate to perfectly fit ASD in a confined framework of definition due to the heterogenous expression of the core symptoms which are further influenced by age and development factors. The progressive work to define ASD within a narrower framework of theoretical concept inadvertently reflects the complexity that revolves around this neurodevelopmental disorder, which makes it even more challenging to conclusively define other aspects pertaining to ASD.

#### **3. Microbiome and Autism Spectrum Disorder**

The terms microbiota and microbiome have a slight semantic difference, however, collectively refer to the entire microbial community residing on and in human body encompassing bacteria, eukaryotic viruses, fungi, protozoa, archaea and bacteriophage. The number of bacteria is overwhelmingly larger than the other taxa to an extent where microbes interchangeably could simply refer to bacteria alone. Our human body is home to trillions of microbial cells which encode 100-fold more genes than human genome, with the latest revision estimating a ratio of 1.3 bacterial cells for every human cell, showing a reduction from widely quoted 10:1 and 100:1 ratios, respectively [15,45,46]. The enormous spread of the microbiome in the human body has both therapeutic and pathogenic roles in health depending on the microbial composition [47]. In the context of ASD (Figure 1), the link between gut, vaginal and oral microbiomes and ASD have been studied thus far using animal models and human subjects.

Figure 1 illustrates the possible mechanism involved in the microbiome–brain interaction in the context of autism spectrum disorder (ASD). The neural, neuroendocrine, and immunologic and humoral pathways are the potential mediators in the bidirectional communication between the gut microbiome and the brain. The maternal contribution is significant in determining the early intestinal colonization in the offspring while in general, the environmental factors that significantly alter maternal microbiome during the prenatal and perinatal periods influence the microbial composition in the offspring. Ectopic transfer and dissemination of pathogenic oral bacteria mediated by the olfactory nerve via the blood, disrupted blood–brain barrier (BBB), perivascular space and circumventricular organs to the gut and brain, respectively, are plausible mechanisms resulting in neuroinflammation and metabolic disruption in the brain, thus indicating the influence of oral microbiota and dysbiosis in ASD occurrence. Another plausible exchange pathway in the gut–brain axis is thought to be mediated by the oropharynx which has a significant role in the pathology of ASD. In general, the interaction between microbiome and the brain in the context of ASD involves a complex mechanism and interplay between the genetic and various environmental factors, in which, some could be explained through epigenetic mechanisms involving short-chain fatty acids (SCFAs) and brain-derived neurotrophic factors (BDNFs).

**Figure 1.** Illustration of autism spectrum disorder (ASD) and its association with microbiome.

#### *3.1. Gut Microbiome*

Gut microbiome has been the most extensively explored area in relation to ASD compared to microbiome in other habitats of the body. The years between 2013 and 2017 saw the largest number of publications focusing on gut microbiota alone accounting for more than 80% of total publications on microbiota in the last four decades, implying the immense possibility of gut microbe in relation to human health [9]. The human gut is the largest reservoir of bacteria in human body which has imperative physiological roles in metabolism, digestion, immunity, endocrine and neurological activities [48]. The gut microbes interact closely with multiple human cells and any imbalance in the microbes, otherwise known as dysbiosis, will have an impact on host key physiological processes and has a potential aetiological role in many health conditions [47]. The immense possibility of gut microbes in human health has prompted researches to highlight the importance of recognizing it as an individual organ system in the human body [49].

In ASD, the relation between gut microbes and central nervous system resulting in manifestation of ASD behaviour has been established using rodent models and clinical subjects, although the latter is still largely inadequate to strongly affirm the causative relation between gut microbes and ASD symptomatology. *Bacteroidetes* (e.g., *Bacteriodes* and *Prevotella*), *Firmicutes* (e.g., *Lactobacillus*, *Clostridium*, *Ruminococcus*), *Proteobacteria* (e.g., *Enterobacter*) and *Actinobacteria* (e.g., *Bifidobacterium*) are four major phyla that are constituent of a healthy adult gut with *Bacteriodes* and *Firmicutes,* representing more than 90% of gut microbes [45,50]. It was found that, in the ASD population, the ratio of *Bacteroidetes* to *Firmicutes* phyla were increased compared to neurotypical cohort [51]. This pattern is also observed prominently in Western ASD children and has been inferred to be influenced by environment and dietary habits [52]. There is also another study which found increased *Firmicutes* compared to *Bacteroidetes*in the ASD group compared to neurotypical group where both groups have GI symptoms [53]. A large number of gram-negative bacteria (e.g., *Bacteroidetes*), exhibit pathogenic nature because of their cell wall which contains lipopolysaccharide (LPS) that has malefic effect on the immune system of the host. LPS confers the ability to breach the blood–brain barrier, increasing the mercury level in the cerebrum and decrease glutathione level which is a key antioxidant in the detoxification of heavy metals. Two species of bacteria with LPS in their cell walls, namely *Bacteriodes vulgatus* and *Desulfovibrio*, have been significantly raised in ASD children compared to neurotypical children. The pathogenic feature of these bacteria may

possibly contribute to ASD development. Another noteworthy genus of gram-negative bacteria is *Prevotella*, a healthy-gut biomarker which is at a greater abundance in neurotypical group while almost absent in the autistic group. *Prevotella* is found in abundance in individuals whose diet is rich in plant-based carbohydrates and includes fish oil which is vital to normal brain development. It has a metabolic role in vitamin B1 production which is known to alleviate ASD symptoms. The lack of *Prevotella* suggests a distinctive dietary habit in autistic children which significantly alters the gut microbial composition and has influence on neurodevelopment which points to its therapeutic potential if restored [48,54]. Another concerning microbe is *Clostridium*, a gram-positive genus in the *Firmicutes* phyla, which has been found higher in ASD group. *Clostridum boltae, C. histolyticum*, and subgroups of *I* and *XI* are the associated species. These gram-positive microbes release enterotoxins which damage the intestinal tissue resulting in diarrhoea and may cater to the increased absorption of large molecules like casein and gluten. On the other hand, beneficial bacteria *Bifidobacterium* were found less abundant in the ASD group [48]. The increased abundance of potentially pathogenic bacteria and decreased beneficial bacteria affirms the existence of gut dysbiosis in the ASD population. In another perspective, the general lack of gut microbial richness and diversity in ASD group, predisposes them to a vulnerable gut environment which could lead to GI disturbances, infections and autistic behaviours [48]. In general, a significant alteration in gut microbial composition interferes with key physiological processes which have an influence on the neurobehavioral manifestation of ASD symptoms either through the absence of beneficial microbial metabolites, release of harmful microbial endotoxins, pathogenic invasion of the intestinal wall and/or through immune mediators catering to neuroimmune inflammation [6,47,48,50,55].

Many factors have been associated with the dysbiosis of gut microbes which include diet, medication and hygiene as well as numerous maternal factors which include maternal stress, infection and a high-fat diet during pregnancy [6,56–59]. The frequent use of oral antibiotics in ASD children during the first three years of life is another factor that has also been hypothesized to disturb the natural balance of the gut microbes while some antibiotics confer benefits [48,55,60]. For instance, the use of macrolides except penicillin within the first six months has been associated with decreased *Actinobacteria* and increased *Bacteroidetes* and *Proteobacteria* and this microbial alteration persisted for a year [61]. On the other hand, the administration of oral vancomycin in a small eight-week clinical trial, which targets gram-positive bacteria including *Clostridium*, temporarily but significantly relieves gastrointestinal and ASD symptoms in 8 out of 11 children with regressive onset autism [62].

#### *3.2. Vaginal Microbiome*

The vaginal microbiome denotes the influence of maternal factors in the occurrence of ASD. As it has been understood that we are born germ-free, the first colonization of microbes in human gut is thought to begin at birth while passing through the vagina, although there are emerging data suggesting it occurs even earlier in utero via placental colonization [63,64]. In cases of caesarean delivery, the colonization occurs after in contact with the maternal skin and environmental microbes [65]. The passing of microbial community through vertical transmission determines the child's gut microbial composition, thus any significant disturbance to maternal vaginal microbe (e.g., reduced *Lactobacillus*) inadvertently interferes with the normal neurodevelopment in offspring due to high metabolic demand during the critical time of early brain development [20,66].

The vaginal tract houses more than fifty microbial species, dominantly populated by *Lactobacillus* in a healthy woman [67]. Maternal stress during early pregnancy has been found to exert a suppressive effect on the vaginal immune response and the abundance of *Lactobacillus*, thus resulting in gut microbial dysbiosis in offspring via vertical transmission [66]. Bacterial vaginosis, a common infection among women characterized by a marked reduction in *Lactobacillus*, predisposes a pregnant woman to the risk of preterm delivery [68]. It has been reported that infants born extremely preterm have 10-fold higher risk of developing ASD compared to infants born at term [69,70]. Further, numerous clinical studies and large epidemiological studies have ascertained that prenatal maternal infection and elevated level of pro-inflammatory cytokines increases the risk of ASD in offspring. The injection of antigens activating such maternal immune response in pregnant mice, rats or monkeys have resulted in neurobehavioral deficits mimicking ASD in their offspring [71,72].

#### *3.3. Oral Microbiome*

Oral cavity is home to more than 700 bacterial species or phyla of which more than 50% are yet to be cultivated and these microbes have an influence on individual oral health. Poor oral health is one of the concerning issues amongst ASD children and it is more prevalent in this group compared to neurotypical group [73]. In the context of ASD, only a handful of studies have explored the differences between oral microbiota in autism and neurotypical children. Notable differences in the distribution of oral microbes were detected in ASD children compared to neurotypical children [54,74,75]. In one of the largest cross-sectional studies which studied the oral microbiome profiles in ASD and typically developed children, eight oral taxa that could distinguish children with ASD from typically developed children were identified. Further, 28 taxa that distinguish ASD children with and without GI disturbances were also identified. It was inferred that the gut microbial disruption could potentially extend to the oropharynx. The analysis of oral microbiome to aid the clinical diagnosis of ASD was also suggested [74].

ASD children often have speech-related difficulties and are very selective with food choices where each of these have motor and sensory involvement, respectively. Oropharynx serves as an important bridge to the GI tract and is innervated by five cranial nerves of both motor and sensory origins. It is thought to play a significant role in the pathology of ASD and has a plausible exchange pathway in the gut–brain axis [74,76–78]. Other plausible pathways in which the oral bacteria could reach the brain have also been hypothesized resulting neuroinflammation and metabolic disruption in the central nervous system. It has been hypothesized that the olfactory nerve in the olfactory tract may act as a potential mediator in bacterial dissemination to the brain through blood, disrupted blood–brain barrier, perivascular spaces or circumventricular organs [79–81]. Further, similar distribution pattern of gut and oral microbiota were demonstrated in a study of oral microbiota in ASD children with significantly higher amount of *Proteobacteria* and *Firmicutes* and lower amount of *Bacteroidetes* and *Actinobacteria* suggesting a potential interaction between gut and oral microbiota leading to a shared pathway in the etiopathology of ASD [54,75,81]. It was also found that oral bacteria resemble 45% of the stool bacteria in the Human Microbiome Project, suggesting a possible interaction between the gut and the oral microbiome [82]. Another example is the increased oral *Bacilli* genus (*Firmicutes* phylum) in ASD children which is also found in abundance in the gut of ASD children and those with inflammatory bowel disease [75,83]. This intestinal colonization by oral bacteria has been hypothesized to occur through ectopic transfer of pathogenic oral bacteria (e.g., *Porphyromonas gingivalis* in chronic periodontitis) to the gut which could induce gut microbial dysbiosis and trigger systemic inflammation [81,84,85]. Poor oral health and hygiene, dental caries and lack of dental care have been found more prevalent in ASD children, implying that ectopic transfer of pathogenic bacteria to the gut associated with oral dysbiosis may possibly explain its relation in ASD occurrence [86]. A marked increase in potential pathogens in analysis of oral samples of ASD children which include *Streptococcus* and *Haemophilus* and reduced abundance of beneficial bacteria like *Prevotella* also point to a significant oral microbial dysbiosis in the ASD group. It is hypothesized that *Haemophilus parainfluenza*, a gram-negative bacteria associated with oral diseases, and its metabolites could potentially cross the blood–brain barrier and impose a detrimental effect on brain development which results in ASD [54,87].

There are specific bacterial species found in subjects with periodontal disease which were otherwise not detected in individuals with a healthy oral cavity [88]. Periodontal disease is associated with an increased risk of preterm birth by 2 to 7 times [89]. Microbial species which were detected in the oral cavity and not in the urogenital tract have been found as causative organisms in intrauterine infection which confers a high risk for preterm birth [90]. More interestingly, the microbiome of 48 term placentae were found to resemble oral microbiome more compared to vaginal microbiome with theory suggesting hematogenous dissemination, especially with underlying periodontal disease, and oral sex may be possible route for such colonization resulting in intrauterine infection [91–93]. The prenatal maternal infection which increases the risk of preterm birth inevitably pave the way for ASD occurrence risk in infants born to mothers with the associated conditions which implies a possible but indirect association with the maternal oral microbiome. The maternal influence was more significant during the infant stage and early childhood. During the first six months after birth, 85% of the oral microbiota of infants resembled their mother's [21]. The pregnancy term, mode of delivery and feeding method were other identified factors that influence the development of oral microbiota in early childhood [94].

#### **4. Aetiology of Autism Spectrum Disorder**

The aetiology of ASD has always remained a puzzle that has yet to be fully understood. The diverse expression of ASD symptomatology, its associated risk factors and comorbidities make it difficult to pinpoint the exact mechanism involved in the etiopathology of ASD. Nevertheless, the aetiology of ASD can be broadly classified into genetic and environmental causes. Although these are not direct causative factors, strong associations have been established and linked to the development of ASD. Individually, genetic and a myriad of environmental factors have been identified to contribute to the occurrence of ASD through various mechanisms. The microbes and microbiome have both environmental and genetical origins in the manifestation of ASD. More interestingly, a complex interaction between gene and environment through an epigenetic mechanism has also been implicated in the pathogenesis of ASD [95–97].

#### *4.1. Genetic Factors*

Heritability confers a large accountability in cases of ASD, with an estimate of 83% of familial risk in a meta-analysis of twin and family studies. This reanalysis sees a drop in the percentage risk compared to previously published meta-analysis of twin studies which estimates heritability ranging between 64% and 90% with minimal influence of shared environmental factors between the twins [27,98]. The diverse clinical phenotype in ASD is due to the genetic heterogeneity in the ASD population, particularly when comorbid conditions exist. Both common and rare genetic variations which are either heritable or occurs newly as de novo mutation (DNM) contribute to the occurrence of ASD [99,100]. Rare genetic variants have a larger effect compared to the smaller effect of common variants in ASD phenotypes and can combine to create an ASD risk [101]. DNM is a rare variant and newly occurs during gamete formation or at the early phase of embryonic development which are not inherited from either of the parents and unique to the child, resulting in the sporadic occurrence of ASD. DNM is more frequently associated with ASD subjects with co-occurring intellectual disability or developmental delay [102,103]. DNM and common genetic variants provoke the idea of environmental influences which could potentially surpass the etiological weightage of genetic factors and heritability in ASD.

Despite the high accountability to genetic origins, there is no individual genetic marker that has been identified to date. However, continuous attempts are being made to identify novel genes which are significantly attributable to ASD. In a meta-analysis which was attempted to identify the genetic risk of ASD, two novel ASD risk genes, namely, YBX3 and HSPA1A, were found to be associated with the pathogenesis of ASD through an indirect regulation of neuronal pathways involved in behavioural manifestation of ASD [104]. These genes confer protective mechanism against the development of ASD, but it was later found to be of a weak association [105]. This further ascertains the lack of homogeneity inevitably challenges the genetic evaluation in ASD populations.

#### *4.2. Environmental Factors*

A vast number of environmental factors have been linked to the development of ASD. The role of environment has been speculated to begin since pre-conception and extends to post-natal period in a child. Pre-conceptionally, advanced paternal age and maternal age at more than 50 and 40 years old, respectively, confers a significant risk for ASD development, apart from low level of education status in parents [106,107]. The greatest number of environmental factors have been linked during the prenatal period, particularly during the first and second trimester of pregnancy which include maternal infections, comorbid cardiometabolic conditions, certain antidepressant and antiepileptic medications, toxic exposures, diet and lifestyle. Perinatal factors which include mode of delivery, obstetrics complications, prematurity, hypoxia, low birth weight and low Apgar score have been associated with ASD risk [108,109]. Postnatally, congenital infection, the use of steroid therapy in very low birth weight infants, birth asphyxia and neonatal jaundice have all been found more prevalent in children with ASD [110,111]. The environmental factors may be explained by direct biological impact on neuronal activities of developing brain of the foetus or foetal activation of neuroinflammatory responses and gene dysregulation which are hypothesized to result from maternal immune activation which crosstalk through the placenta [109,112]. Largely, these environmental factors have been studied retrospectively in cohorts of ASD children and mothers of ASD children, while only limited number of studies done using rodent models. There are studies which deem environmental factors to play a more prominent role in the liability and genetical variance of ASD, thus paving way for the eminent role of epigenetics in ASD [96,113,114].

#### *4.3. Epigenetic Factors*

While traditional methods of identifying genes and environmental factors independent of each other has seen a lack of integrative approach in studying the aetiology of ASD, epigenetics has become the prime pursuits in recent decades incorporating the role of environmental influence on genes. Epigenetic refers to non-genetical influences which changes the phenotypic expression of a gene without actual mutation or alteration occurring in the original DNA sequence and these changes are heritable [97]. It is an epitomic description of a gene–environment interaction in ASD which can be explained by integrating various environmental factors at a cellular level in the occurrence of ASD, particularly through gut microbiota-derived metabolites. At the molecular level, the epigenetic modification occurs via two broadly studied mechanisms, which include DNA-methylation and histone modification besides Ribonucleic acid (RNA) interference [28,115–117]. Epigenetic programming of the brain explained through stress-regulating pathways and reduced expression of brain-derived neurotrophic factor (BDNF) by environmental factors during pre-natal and postnatal periods have been thought to exert a long-lasting effect on the neural functioning and behavioural outcome [118,119]. BDNF has a crucial role in the regulation of neurodevelopment, neuronal functions and neuroplasticity and it is frequently associated with depressive disorder and neuroinflammation [120,121]. However, recent findings on new-borns later diagnosed with ASD have demonstrated a significantly reduced level of BDNF in the blood samples [122]. In mice models, decreased levels of BDNF transcript variants were observed in germ-free mice models and antibiotic-treated SPF mice models in the amygdala regions [123–125]. These associations of BDNF with the epigenetic mechanism, ASD subjects as well as microbiome, warrant the need to further study and explore the role of BDNF in relation to ASD as currently there are no available studies to affirm this relationship.

The gut microbiome is thought to modulate gene expression through an epigenetic mechanism which may either institute the etiopathogenesis of ASD and/or aggravate ASD symptomatology primarily [126]. It has also been hypothesized that the epigenetic regulation is diet-dependent where the role of microbial metabolite, SCFAs come into play. SCFAs are the product of the ASD-associated bacterial (such as *Clostridia, Bacteriodetes, Desulfovibrio*) fermentation of dietary carbohydrates and are regulated by the gut microbial composition [115,127,128]. However, an overproduction of SCFAs (e.g., propionic acid, butyrate, acetate) have been indicated in ASD based on analysis of faecal samples of ASD children and it was inferred that this possibility could be due impaired fermentation process and utilization of its by-products without eliminating the possibility of increased faecal SCFA could also be due to an overall increased SCFA production in ASD children, which is indeed regarded beneficial [129,130]. Therefore, the microbial dysbiosis (reduced SCFA-producing, ASD-associated

bacteria) manipulates the level of nutrients and metabolites like SCFA which regulate the DNA methylation and histone modification resulting in immune activation that have a potential role in ASD development. These metabolites either directly inhibit the enzymes which catalyse these epigenetic processes or alter the substrate availability required for the enzymatic process, thus highlighting the possible epigenetic mechanism in the occurrence of ASD [127].

#### **5. Comorbidities in Autism Spectrum Disorder**

Gastrointestinal comorbidities have been found to be significantly higher and more common in the ASD group compared to neurotypical children, with 46.8% of ASD children exhibited at least one GI symptom [131–133]. GI disorders have been associated with gut dysbiosis, although whether underlying GI disturbances result in dysbiosis or the dysbiosis results in GI disturbance largely remains a chicken-or-the-egg conundrum. Gastrointestinal symptoms (e.g., constipation, diarrhoea, bloating) have been correlated to gut dysbiosis and severity of autistic symptoms [6,65,134,135]. It was hypothesized that the gut dysbiosis gives rise to the pathogenic invasion of intestinal wall, thus resulting in the breach of the gut barrier which ultimately results in neuroinflammation responsible for ASD's behavioural manifestation [9,48]. However, some of the recent findings demonstrated reduced diversity and altered microbial pattern in autistic group compared to a neurotypical group; however, no significant correlation was found between GI symptoms and severity of ASD symptoms [31,48,136,137]. It was further inferred that GI disturbances alone could possibly trigger ASD symptoms possibly due to heightened sensory perception and experience in ASD children [134]. An interesting correlation between GI symptoms, intestinal mucosal dysbiosis, gene expression and ASD was highlighted in a study which looked at the molecular mechanism involved in explaining the GI disturbances in ASD children by analysing the intestinal biopsies obtained from 15 children with ASD and GI disease, and seven children with GI disease alone. The messenger ribonucleic acid (mRNA) deficiencies in genes encoding disaccharidases and hexose transporters responsible for carbohydrate digestion and transport across the intestinal epithelium was linked to the intestinal dysbiosis (decreased *Bacteroidetes*, increased *Firmicutes* to *Bacteroidetes* ratio, and increased *Betaproteobacteria*) in the ASD group. The pre-existing impaired carbohydrate digestion and absorption manifests as GI disturbances with dietary intake of carbohydrates which leads to fermentation, increased gas production and osmotic load in the gut of ASD children. Although no dietary evaluation was done in the children involved in this study, its finding provides a significant insight regarding the GI disturbances in ASD children, especially on how gluten-free/low-carbohydrate diets may benefit some of the ASD children [53]. In another study, the restoration of microbial imbalance with significant and lasting improvement in terms of GI symptoms and autistic behaviours in children diagnosed with ASD was made possible through microbiota transfer therapy (MTT) [138]. This highlights the influence of microbiome in both GI symptoms and autistic behaviours; however, no distinctive microbial patterns specific to ASD subjects, with or without the presence of GI disturbances, have been elicited so far.

Epilepsy and ASD have often been associated with one another; however, the possible pathophysiological link has yet to be fully understood. It has been reported that ASD subjects possess a seven-fold increased risk of developing epilepsy compared to general population, with an estimated prevalence of epilepsy in ASD up to 50% [139,140]. The prenatal use of anti-epileptics, mainly valproate, has a teratogenic effect on the neurodevelopment of the offspring and predisposes them with a risk of developing ASD [141,142]. The valproate has been demonstrated to cause microbial disruption in mouse models of ASD in the context of in utero usage [143]. Interestingly, ketogenic diet in ASD cases with refractory epilepsy has shown significant improvement in both ASD symptoms and seizure episodes through gut microbial modulation [144]. *Akkermansia* and *Parabacteroides* are two ketogenic diet associated microbes which confer the anti-seizure effect [145].

Depression and anxiety are part of neuropsychiatric disorders which have been elucidated through the gut–brain–microbiota axis [146]. These disorders are associated with the alteration of gut microbial composition, thus impacting the neurobehavioral outcome as narrated in the context of ASD [147,148]. Similar to gut microbial observation in ASD subjects, the ratio of Bacteroidetes to Firmicutes phyla were markedly increased in both animal models and human subjects with depression [149,150]. Further, a recent review on psychotropics and the microbiome has highlighted gut dysbiosis and anti-microbial exertions of antipsychotic, antidepressant and antianxiety drugs which are commonly prescribed in ASD subjects with associated psychiatric comorbidities [151,152].

#### **6. Microbiota-Based Interventions and ASD**

#### *6.1. Dietary and Supplementary Interventions*

Diet has an influential role in determining the intestinal microbial composition and in rodent models; it has been demonstrated that significant dietary change has the potential to alter the microbiome as rapidly as over one day [153]. The microbiota evaluation of ASD children differed across countries, suggesting the influence of geographical location in microbiota profile, and diet has been thought to be an important factor to explain such differences [32,154]. SCFA, the colonic by-product of bacterial fermentation of dietary fibres and resistant starch, is gaining attention as one of the important mediators of the gut microbiota–brain communication, although the exact mechanism on how this metabolite influences the brain physiology and neurobehavioral outcome is yet to be fully understood [155–159]. An intervention using a diet that promotes increased SCFA levels (Mediterranean diet, plant-based proteins, dietary fibre) may potentially exert a positive impact on ASD behavioural outcomes, although no studies have been done so far to support this relation in human subjects [158]. Dietary intervention which has been commonly advocated amongst ASD children include gluten- and casein-free diets which have been reported to confer a beneficial outcome in ASD symptoms, while more recent study shows no significant changes in ASD symptoms after a six-month trial [160–162]. This intervention is recommended only when there is a clinically diagnosed intolerance to gluten and casein [162]. In terms of maternal aspect, rodent models have significantly demonstrated that offspring born to mothers fed with high-fat diet eight weeks before mating had impaired social interaction and repetitive behaviours mimicking ASD, with altered gut microbial composition and notable nine-fold decrease in *Lactobacillus reuteri* [163]. The Western diet, also referred to as the high-fat diet, has been associated with reduced microbial diversity and richness [164]. Dietary intervention should be exercised with caution to avoid malnourishment and nutritional deficiency. Vitamin D deficiency has been commonly reported in ASD children and low prenatal vitamin D has been associated with an increased risk of ASD occurrence [165–167]. It was also found that GI problems are more evident in ASD subjects with vitamin D deficiency than those without this deficiency [166]. The supplementation of vitamin D3 shows a remarkable improvement in ASD core symptoms [168].

#### *6.2. Prebiotics, Probiotics, Synbiotics and Antibiotics*

Prebiotics are mainly fibres consisting of non-digestible diet components which benefit the host's health by selectively promoting microbial proliferation in the colon, generally *Lactobacilli* and *Bifidobacteria*. The probiotics are live, non-pathogenic microorganisms that confer health benefits to the host when adequately administered, whereas synbiotics are formulations consisting of both prebiotic and probiotic which were meant to improve the efficiency of probiotics [169]. These interventions have been recommended as adjuvant therapies given their promising benefits in terms of improving ASD behavioural symptoms and ameliorating GI disturbances in ASD children [170,171]. The administration of probiotics consisting of three strains (*Lactobacillus acidophilus*, *Lactobacillus rhamnosus* and *Bifidobacteria longum*) for a duration of three months significantly improved autistic behaviours and GI symptoms in ASD children with increased count of the beneficial bacteria *Bifidobacteria* and *Lactobacilli* [170]. In a pre-clinical study, *Bacteroides fragilis* has also been demonstrated to restore the microbial dysbiosis, improve gut permeability and autism behaviours in the offspring of maternal immune activation (MIA) mouse models which displayed ASD behavioural features [172]. In ASD children, probiotics—namely, *Lactobacillus reuteri, Lactobacillus plantarum, Lactobacillus acidophilus, Lactobacillus rhamnosus* and *Bifidobacterium longum*—have

been utilized in clinical trials and have significantly improved autistic and GI outcomes. These periods of trial ranged from three weeks to six months [170,173–175]. In a randomized control trial, which was carried out on 75 infants, where *Lactobacillus rhamnosus* was administered in 40 infants and a placebo in the remaining infants in control group for the first six months, the outcome revealed that the early administration of probiotic may potentially reduce the risk of developing ASD. These infants were followed over 13 years and 6 out 35 infants in the control group were diagnosed with either Asperger's syndrome (now part of ASD) or attention deficit hyperactivity disorder (ADHD), while none were diagnosed in the probiotic group [175]. It is also worth mentioning that the a causal link between maternal diet, altered gut microbiome and social behaviour was demonstrated in a breakthrough study where the administration of *Lactobacillus reuteri* over four weeks in rodent offspring born to mothers fed with a high-fat diet with reduced gut microbial diversity, particularly *Lactobacillus reuteri*, was able to significantly improve social behaviour [163]. However, one of the recent systematic reviews has pinpointed the lack of evidence in supporting the beneficial roles of probiotics and prebiotics in ASD and ascertained that more studies are needed to validate the benefits of these interventions in ASD subjects [176].

In terms of antibiotics, vancomycin and metronidazole have been used in treating ASD symptoms. However, metronidazole is not a preferred choice due to its possible risk of causing systemic adverse effects [177]. There is one case report on the administration of amoxicillin over a 10-day course which deemed to improve autism symptoms in a child as reported by his parents [178]. However, in rodent models, the maternal use of oral antibiotics (non-absorbable sulfonamide, neomycin, bacitracin, pimaricin) either during preconception or early gestation period yielded offspring with anxiety-like behaviours and impaired social interactions [179,180]. Further analysis of the faecal samples of the offspring exposed to antibiotics also showed a 50% reduction in the abundance of the *Lactobacillales* and increased *Clostridium*, thus implying that early exposure of antibiotics have a negative impact on the behavioural outcome in offspring [179].

#### *6.3. Fecal Microbial Transplant (FMT) and Microbiota Transfer Therapy (MTT)*

In a study using mice models, it was demonstrated that the colonization of germ-free mothers with microbiota obtained from SPF mice 30 days prior to mating was able to "normalize" the behavioural outcome in the germ-free mice while the same produced no changes in adult germ-free mice. The elevated stress hormones level was also reversed through colonization of the germ-free mice with the microbiota of SPF mice before 6 weeks of age [124]. This does not only indicate the importance of maternal microbiome at the time of conception but also the need for an earlier intervention for an improved behavioural outcome, which may be challenging in human subjects as the ASD-like behaviour may not be apparent during the perinatal period. In an open-label study involving 18 children diagnosed with ASD, the MTT involving antibiotic treatment for an initial two weeks, followed by extended faecal microbiota transplant (FMT) over a period of 7–8 weeks, significantly improved GI disturbances (constipation, diarrhoea, abdominal pain, indigestion) as well as ASD's behavioural symptoms. Further, improved bacterial diversity with significant increase in *Bifidobacterium*, *Desulfovibrio* and *Prevotella* were observed and all these changes persisted for eight weeks after the cessation of treatment [138]. Despite its promising benefits, the faecal transplantation may possess the risk of transmitting norovirus and some autoimmune conditions (rheumatoid arthritis, Sjogren's syndrome), aspiration and even inducing obesity in recipients [181,182]. However, these are theoretical speculations which should not be simply disregarded.

#### **7. Discussion**

While the relation between microbiome and ASD is extensive, the immense possibility of the microbial role in ASD is indisputable, inclusive of whether viewed from an aetiological point of view or its association with various comorbidities affecting the ASD population. However, research efforts in understanding the exact mechanism involved in the context of ASD are still largely inadequate; relationships are yet to be established as causal and are rather associative at this stage, mainly due to the

heterogeneous nature of this disorder. Although germ-free and animal models have been utilized to a great extent to demonstrate the clear associations of microbes in host physiology, coupled with the fact that human genome resembles 85% of mouse genome, it is near impossible to create germ-free human samples and eliminate all the confounding factors in establishing the same outcome [183]. The genetic makeup, age, sex, life exposure, environmental influences, medication and comorbidities are all possible confounding factors which need to be carefully defined in human samples [16]. Further, a larger sample size is required to affirm the clinical outcomes especially those involving microbiota-based interventions such as dietary, probiotics and faecal experiments. The influence of geographical locations and diet in ASD and its associated microbiota profile are becoming more relevant; therefore, studies fostering larger cohorts from diverse geographical locations with standardized specifications (e.g., age, gender, comorbidities) are needed to understand if the findings are consistent across the different locations and dietary habits.

Clinically, ASD patients are managed by a multi-disciplinary healthcare team primarily by a child psychiatrist, paediatrician, clinical psychologists and other physicians as per the associated comorbidities, although a regular follow-up and monitoring remains largely questionable. The role of parents and caregivers are of utmost importance in ASD. Although many maternal factors and few paternal factors have been associated with the risk of ASD, a proper parental counselling should be provided to ensure a rational understanding and awareness about ASD development. A prenatal counselling to avoid potentially teratogenic factors especially in mothers with underlying medical or psychiatric comorbidities and those with existing ASD children should not be neglected. However, it has be acknowledged that the maternal influence has been mainly described using animal models and no defined relationship between maternal factors and ASD occurrence have been established. This gap in the existing studies need to be understood to ensure the proper channelling of information to the parents, for this may possibly trigger a detrimental emotional response (anxiety, paranoia), especially in mothers who are either planning to conceive or have given birth to children with ASD. The knowledge about the availability of genetic testing should also be given to parents who wish to conceive or those with a known familial risk of neurodevelopmental disorders.

The incorporation of microbial knowledge in clinical context of ASD management provides a promising insight into defining the neurodevelopmental disorder using potential clinical biomarkers, although no definite classification of biomarkers have been established and widely adapted to date [40,75]. Clinical trials revolving around the microbial field are inadequate with a smaller sample size, lack of large population-based studies, and this is certainly attributable to the highly heterogenous and complex nature of ASD itself. The potential clinical intervention targeting gut microbiome which include prebiotics, probiotics, antibiotics, dietary and supplementary interventions, faecal microbial transplant (FMT) and a modified protocol of FMT, namely microbiota transfer therapy (MTT), are important breakthrough interventions which need to be carefully evaluated and validated using double-blinded, randomized, controlled trials involving larger sample size and standardized treatment regimens and durations. FMT and MTT seem the most promising treatments in restoring gut dysbiosis in ASD, but the safety and tolerability in a long run is still questionable [138]. The various clinical interventions have to be carefully evaluated and catered based on individual suitability and requirements. In the management of highly heterogenous disorder like ASD, personalized treatment based on appropriate clinical judgement will be more beneficial.

Although ASD has been largely identified as a childhood disorder, it should not be forgotten that the adult population is not spared from the gigantic wave of ASD. Epidemiological studies on adult autism is very limited and still remains a poorly explored topic. A survey on psychiatric morbidity in adults carried out in England revealed the prevalence rate of autism is about 1% in adults, affecting more males than females, consistent with the findings involving the cohort of children [184]. Adults with autism has also been reported to be at greater risk of socially deprived and isolated, lack of financial security and poor access to specialist care other than having poor physical and mental health

outcomes [185]. The children with ASD who will eventually grow into adults should be examined and followed-up closely and consistently.

#### **8. Conclusions**

ASD has a major health and economic burden to the affected population and care providers. It is becoming an alarming global epidemic which calls for a greater attention and effort. The wide range of serious health comorbidities associated with ASD is often a huge concern when it comes to clinical interventions. The possible explanation and relation of microbes in terms of aetiology and ASD comorbidities point to the need for an integrative approach in treating ASD subjects. Newly emerging fields of research, such as epigenetics, provide more integrative insights and approaches in directing future research, incorporating knowledge of the microbiome. The study of epigenetics incorporating the microbiome reflects the highly dynamic process which is involved in shaping and reprogramming the growth and development of an individual. The systematic identification of environmental factors which interfere with gene regulation could possibly create new avenues in the clinical management of ASD. The treatment approach in ASD should be more individualized to ensure the best clinical outcomes. Diet and probiotics are important and promising breakthrough microbiota-based interventions in the context of ASD. These interventions have better feasibility and an easier method of administration with less known side effects compared to other interventions. Research efforts should be expanded and focused in this direction to ascertain the clinical efficacy of these interventions.

**Author Contributions:** D.J. performed the literature search, critical data analysis as well as the manuscript writing. Technical supports and proofreading were contributed by V.L., S.T., and L.-H.L. L.-H.L. set up the research project. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by PVC Award Grant (Project No. PVC-ECR-2016), External Industry Grant (Biotek Abadi—Vote No. GBA-808138 and GBA-808813) awarded to L.-H.L.

**Acknowledgments:** Mohamed Shajahan Yasin, Professor and Head of School, Jeffrey Cheah School of Medicine and Health Sciences, Monash University Malaysia.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Glossary**


#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Gut Health-Promoting Benefits of a Dietary Supplement of Vitamins with Inulin and Acacia Fibers in Rats**

**Malén Massot-Cladera 1,2, Ignasi Azagra-Boronat 1,2, Àngels Franch 1,2, Margarida Castell 1,2, Maria J. Rodríguez-Lagunas 1,2 and Francisco J. Pérez-Cano 1,2,\***


Received: 7 June 2020; Accepted: 20 July 2020; Published: 23 July 2020

**Abstract:** The study's objective was to ascertain whether a nutritional multivitamin and mineral supplement enriched with two different dietary fibers influences microbiota composition, mineral absorption, and some immune and metabolic biomarkers in adult rats. Nine-week-old Wistar rats were randomly assigned into four groups: the reference group; the group receiving a daily supplement based on a food matrix with proteins, vitamins, and minerals; and two other groups receiving this supplement enriched with inulin (V + I) or acacia (V + A) fiber for four weeks. Microbiota composition was determined in cecal content and mineral content in fecal, blood, and femur samples. Intestinal IgA concentration, hematological, and biochemical variables were evaluated. Both V + I and V + A supplementations increased *Firmicutes* and *Actinobacteria* phyla, which were associated with a higher presence of *Lactobacillus* and *Bifidobacterium* spp. V + A supplementation increased calcium, magnesium, phosphorus, and zinc concentrations in femur. V + I supplementation increased the fecal IgA content and reduced plasma total cholesterol and uric acid concentration. Both fiber-enriched supplements tested herein seem to be beneficial to gut-health, although differently.

**Keywords:** inulin fiber; acacia fiber; immune system; microbiota; mineral absorption; IgA

#### **1. Introduction**

The intake in an appropriate dose (20–35 g/day for healthy adults) of dietary fiber (DF) has long been linked to reduction of metabolic diseases incidence, including diabetes, cardiovascular disease and obesity, among others, due to its capacity to lower blood cholesterol and C Reactive Protein (CRP), to attenuate glucose absorption and to improve insulin response [1–3]. Moreover, when non-digestible fiber reaches the colon unaltered and is selectively metabolized by microbiota, it induces specific changes, both in the composition and/or functionality of one or a limited number of bacteria potentially associated with health and well-being [4–6]. Meeting all these criteria, non-digestible fiber is considered a prebiotic as defined by Gibson and Roberfroid [1,4–6].

Prebiotic consumption is also believed to improve the immune system in both humans and animals [7]. The most examined mechanism involved in this effect is the indirect modulation of the immune response by changing the microbiota composition [8,9] and, therefore, its crosstalk with the immune system. Moreover, the enrichment of beneficial bacteria induced by prebiotic intake can result in the modulated release of pro- and anti-inflammatory cytokines [7] as well as in increasing the intestinal and fecal immunoglobulin (Ig) A content [10]. In addition, although little information is available, direct effects of prebiotics on the immune system such as changes of the intestinal gene expression, such as toll-like receptors (TLRs), have also been reported [11,12].

The favorable shift in the gut microbiota composition after prebiotic fiber consumption [1] is proposed as a potential mechanism by which prebiotics improve mineral absorption [13,14]. In this regard, the most widely accepted theory supporting this effect is associated with the microbial fermentation of the prebiotic fiber into short-chain fatty acids (SCFAs) in the colon. This metabolic process acidifies the intestinal compartment; thereby, preventing the formation of complexes between minerals and negatively charged metabolites. Consequently, it increases the extent of mineral absorption [15]. Alternatively, prebiotic consumption may also influence tissue morphology by increasing cell density, intestinal crypt depth, and blood flow in the large intestine, a mechanism that is believed to increase the intestinal surface area and lead to a higher mineral absorption [16,17].

There are several well-documented prebiotic fibers, such as the inulin-type fructans [18,19]. Inulin is part of everyday human diet. It can be found naturally among others in a range of plants such as chicory, garlic, tomato, and banana [20]. Its bifidogenic effects have been widely described in vitro, in vivo, and in clinical studies [19,21]. In the last few years, a new prebiotic fiber has emerged: acacia gum. It is a soluble DF obtained from the stems and branches of *Acacia senegal* and *Acacia seyal* and it is mainly composed of complex polysaccharides [22]. It resists digestion in the upper gastrointestinal tract; thus, reaching the large intestine and it can induce an increase in *Bifidobacterium* spp. in vitro [23,24] and in human [25] studies. However, unlike the inulin, little is known about the impact of acacia gum on health benefits.

On the other hand, the maintaining of a healthy diet, defined as an appropriate balance of energy, macro- and micronutrients and water, is important for adults but it is particularly relevant for members of the elderly population who are more vulnerable to malnutrition. Moreover, the efficiency of nutrient absorption may be impaired in this population; thus, involving different nutritional requirements. Moreover, the presence of oral problems, for example with dentition, together with a decrease in smell and taste perception induce a change in dietary patterns. Additionally, there is a concomitant decline in the normal function of the immune system (immunosenescence) that may contribute to an increase in the risk of infection and frailty [26]. On this basis, the hypothesis of the present study is that the intake of a nutritional supplement containing proteins, vitamins, minerals, and fiber is beneficial for the adult population and if so, its impact should be important to take into account for the elderly. Therefore, the aim of the current study was to ascertain whether a nutritional multivitamin and mineral supplement enriched with two different DFs influences microbiota composition, mineral absorption, and some immune and metabolic biomarkers in adult rats.

#### **2. Materials and Methods**

#### *2.1. Animals and Supplements*

Nine-week-old female and male Wistar rats, purchased from Janvier Labs (Saint Berthevin Cedex, France), were housed individually in polycarbonate cages with large fibrous-particle bedding and tissue papers as enrichment, in a controlled environment of temperature and humidity and in a 12/12 h light/dark cycle at the Faculty of Pharmacy and Food Science animal facility. All rats were fed a commercial diet corresponding to the American Institute of Nutrition 93 M formulation [27] (Teklad Global 14% Protein Rodent Maintenance Diet, Envigo, Indianapolis, IN, USA), which contains 5% of cellulose, and water ad libitum throughout the study.

After the acclimation period, animals were randomly assigned into four experimental groups (*n* = 10/each, 5 females and 5 males). One constituted the reference (REF) group which did not receive any supplement; another group received a daily supplement based on a food matrix with proteins, vitamins and minerals (V) (Table S1); and two other groups received this supplement enriched with inulin (V + I) or acacia fiber (V + A), containing 4.5 g of fiber/100 g of product each (La Piara S.A, Manlleu, Barcelona). The chow consumption was measured every other day to adjust the dose of the

supplements that were administered during 4 weeks in daily small portions. With this aim, the weights of each serving in V + I and V + A were readjusted periodically for each animal in order to receive an extra 20% of fiber daily. Accordingly, the same amount of supplement was used for the V group.

All experimental procedures were approved by the Ethical Committee for Animal Experimentation of the University of Barcelona and the Government of Catalonia (CEEA UB ref. 351/17 and CG 9735, respectively), in accordance with the EU Directive 2010/63/EU.

With regard to sample size estimation (*n* = 10/group), the Appraising Project Office's program from the Universidad Miguel Hernández de Elche (Alicante, Spain) was used to calculate the minimum number of animals providing statistically significant differences among groups, assuming that there is no dropout rate, a beta risk of 0.2 (80% power) and a type I error of 0.05 (two-sided). We used the IgA-coating bacteria percentage data from a previous study [11] with similar design: mean values in the REF group were 25.3%, the estimated common standard deviation was 13 and the minimum expected difference was 12. In addition, the sample size was adjusted to the minimum needed to follow the University Ethical Committee guidelines.

#### *2.2. Monitoring, Sample Collection and Processing*

Body weight and food and water intake were monitored three times per week throughout the study. Fecal samples were collected weekly in order to determine changes in the fecal wet weight, humidity, and pH. Additionally, fecal samples collected at the end of the study allowed the concentration of IgA and the proportion of IgA-coated bacteria to be quantified, as previously described [28], as well as the mineral elimination.

At the end of the nutritional intervention, animals were anesthetized intramuscularly with ketamine (90 mg/kg) (Merial Laboratorios, S.A. Barcelona, Spain) and xylazine (10 mg/kg) (Bayer A. G., Leverkusen, Germany) in order to obtain tissue samples. The body weight and naso–anal length were measured to calculate the body mass index (BMI) as body weight/length<sup>2</sup> (g/cm2). Urine samples for mineral quantification were obtained by direct puncture of the bladder. The weight of stomach, duodenum, jejunum, ileum, cecum, colon and rectum, spleen, liver, thymus, kidneys, heart, submandibular gland, and the length of the small and large intestines were recorded. Blood samples were collected in heparin-treated tubes to determine the hematological and biochemical variables, and plasma mineral content. For IgA quantification, the gut wash (GW) from the distal part of the small intestine was obtained as previously described [11]. Finally, the central part of the left femur was excised for mineral quantification.

#### *2.3. Fecal Variables*

Fresh feces collected weekly were used to determine fecal pH using a surface electrode (Crison Instruments, S.A., Barcelona, Spain). Afterwards, fecal samples were dried for 24 h at 60 ◦C. Fecal humidity of each sample was calculated considering the weight difference between before and after the drying process.

#### *2.4. Mineral Analysis in Biological Samples*

Firstly, blood, feces, femur, and urine samples underwent a chemical cleavage process to obtain aqueous solutions without precipitation or colloids. For that, the same amount of each type of sample was introduced into a tared high-pressure vessel made of polytetrafluoroethylene. Then, 2 mL of concentrated nitric acid (HNO3) and 2 mL of concentrated hydrogen peroxide (H2O2) were also added in order to ensure a better oxidation of the organic matrix. All the high-pressure vessels were incubated overnight at 90 ◦C. After digestion, the samples were diluted with 16 mL of ultra-pure water, and then the vessels were weighed again in order to determine the weight of the aqueous sample solutions. Additionally, for each digestion cycle, triplicates of digestion blanks containing only HNO3, H2O2 and ultra-pure water were also prepared. Finally, the digested solutions were transferred into the test tubes. The concentrations of calcium (Ca), magnesium (Mg), phosphorous (P), and zinc (Zn) were

determined using an inductively coupled plasma-optical emission spectrometer (ICP-OES, Optima 3200 RL, Perkin-Elmer, Massachusetts, USA), whereas iron (Fe) and zinc (Zn) concentrations were measured by an inductively coupled plasma-mass spectrometer (ICP-MS, Nexlon 350 D, Perkin-Elmer, Massachusetts, MA, USA) using standard conditions. The analysis was carried out at the Unit of Metal Analysis of the Scientific and Technological Centers of the University of Barcelona (CCiT-UB). Results are expressed as mg/g of sample.

#### *2.5. Hematological and Biochemical Analysis*

Heparin-treated blood was immediately used to count platelets and white and red blood cells and related variables using an automated hematology analyzer (Spincell3, MonLab, Barcelona), following the manufacturer's instructions.

Plasma samples were used to quantify total cholesterol (TC) by cholesterol oxidase (CHOD)-peroxidase (POD) method; high-density lipoprotein cholesterol (HDL-C) by colorimetric method; triglycerides (TG) by glycerol phosphate oxidase (GPO) method; glucose by glucose oxidase (GOD)-POD method and uric acid by Uricasa-POD method using kits provided by Química Clínica Aplicada, S.A. (Química Clínica Aplicada, S.A., Tarragona, Spain) and following the manufacturer's instructions. Low-density lipoprotein cholesterol (LDL-C) was assessed according to the formula by Friedewald et al. in 1972, in which cLDL = TC – (cHDL + TG/5).

#### *2.6. Immunoglobulin A and IgA-Coated Bacteria Quantification*

The concentration of IgA in GW, fecal homogenates, and plasma was quantified at the end of the nutritional intervention by ELISA as previously described [11,29]. Moreover, the proportion of bacteria coated with IgA in feces was determined and analyzed by flow cytometry, as previously established [28]. The results concerning IgA are expressed as ng/g of tissue, ng/mg of fecal sample, and ng/mL of plasma, whereas those related to the IgA-coated bacteria are expressed as percentage.

#### *2.7. Analysis of Cecal Microbiota Composition by 16S rRNA Sequencing*

Cecal samples from all the animals at the end of the study (*n* = 10 animals/group) ranging from 500–1000 mg collected were used to extract genomic DNA using QIAmp DNA Stool Mini Kit (Qiagen) with some previous modification as has been previously reported [30]. Briefly, extra purification and concentration were performed following the cleaning protocol from QIAmp Micro Kit (Qiagen, Madrid, Spain). Then, for massive sequencing, the hypervariable region V3-V4 of the bacterial 16s rRNA gene was amplified using key-tagged eubacterial primers (forward: S-D-Bact-0341-b-S-17, 5 -CCTACGGGNGGCWGCAG-3 and reverse S-D-Bact-0785-a-A-21, 5 -GACTACHVGGGTATCTAATCC-3 ) [31] and sequenced with a MiSeq Illumina Platform (Illumina Inc., San Diego, CA, USA) by the Life Sequencing facilities (ADM Life Sequencing, Valencia, Spain), following the Illumina recommendation for Library preparation and sequencing for metagenomics studies.

The software Paired-End read merger (PEAR v 0.9.6, Exelixis Lab, Heidelberg, Germany) was used to merge raw sequences forward and reverse. Using this approach, the ends of the obtained sequences were overlapped in order to get complete sequences. The amplification primers from the sequences obtained in the sequencing step were trimmed with Cutadapt v1.8.1 [32], using parameters by default, in order to reduce the bias in the annotation step. Once the primers were removed, sequences lower than 200 nucleotides were excluded from the analysis because short sequences have a higher chance of generating wrong taxonomical group associations. After obtaining the clean complete sequences, a quality filter was applied in order to delete sequences of poor quality. The resulting sequences were inspected for PCR chimera constructs (Uchime, USEARCH) [33], that may occur during the different experimental process, which were removed from further analysis. Later, each group of sequences was compared to a database of rRNA using an alignment BLAST strategy to associate taxonomic groups. The relative proportions of phyla, families, and genera were calculated. Moreover, to estimate the specific genus biodiversity, the Shannon–Wiener and CHAO1 indexes were calculated.

Results of the qualitative analyses relative to the most abundant phyla, families, and genera are represented with stacked bars separated by gender. The category "others" represented in each graph includes those phyla whose presence was lower than 0.05% in the REF group; and those families and genera whose presence was lower than 0.8% in the same group.

To estimate the presence or absence of certain bacterial genera in the experimental groups, it was agreed that all bacterial genera present in all animals belonging to the same group with a proportion higher than 0.01% were computed as "present". Otherwise, they were computed as "absent" in such groups. Based on that, the Venn diagrams were created for all groups together allowing the way the genera were distributed among the groups to be seen numerically, in order to compare their coincidences and differences.

#### *2.8. Principal Components Analysis*

The principal components analysis (PCA) was performed to evaluate the dimensionality of microbiota with regard to the supplements. The model was done using Simca v14.1 (Umetrics, Umeå, Sweden) as previously reported [30].

#### *2.9. Statistical Analysis*

The Statistical Package for the Social Sciences (SPSS v22.0) (IBM, Chicago, IL, USA) and the R software (R-3.6.3) were used for statistical analysis. Data were tested for homogeneity of variance and normality distribution by the Levene's and Shapiro–Wilk tests, respectively. When data were homogeneous and normally distributed, a two-way ANOVA test was applied. When no differences between genders were observed, the data were analyzed together using a conventional-one-way ANOVA. Otherwise, they were analyzed separately. When significant differences among groups were detected, Bonferroni's post hoc test was performed. Kruskal−Wallis test was used when results were neither equally nor normally distributed, followed by Nemenyi post hoc test in the case of significant difference among groups. To compare variables along the study, a repeated-measures ANOVA or Friedman test were applied followed by Student's t-test or Nemenyi post hoc test, respectively. Significant differences were considered when *p* < 0.05, except regarding repeated comparisons, when *p* value was corrected, dividing it by the number of applied tests.

#### **3. Results**

#### *3.1. E*ff*ects of Supplements on Morphometry and Food and Water Intake*

Throughout the study period, male rats from all groups had a higher body weight (447.51 ± 5.80 g), chow intake (28.09 ± 0.68 g/day/rat), and water consumption (29.23 ± 0.56 mL/day/rat) than female rats (255.25 ± 2.5 g/day/rat, 17.73 ± 0.33 g/day/rat, and 21.66 ± 0.61 mL/day/rat, respectively) (*p* < 0.05), and none of the dietary supplementations modified these variables (Figure S1a,b).

Regarding the body mass index (BMI) at the end of the nutritional intervention, only sex-associated significant differences were observed within all experimental groups. In particular, BMI in female rats was 0.64 ± 0.01 whereas it was 0.85 ± 2.5 in male rats, considering all animals independently of the experimental group, and no effects due to supplementation were detected (Figure S1c).

Moreover, whereas male rats showed significantly lower relative weight in most of the organs analyzed than female rats in the same group (*p* < 0.05) (Table S2), supplementation did not result in changes. No differences between groups were found when the small intestine and large intestine were measured, their mean length being 80.69 ± 1.43 cm and 16.76 ± 0.34 cm, respectively, for all experimental groups considered together at the end of the study.

#### *3.2. E*ff*ects of Supplements on Fecal Variables*

No sex-associated differences were detected in all the fecal variables studied; thus, these results were analyzed considering both female and male rats together (Figure S2).

Fecal weight registered was similar throughout the study and this was around 0.25 ± 0.01 g/day for all experimental groups, without being influenced by the dietary supplementations (Figure S2a**)**. Similar results were observed when humidity of the feces was measured. All samples had around 53.24 ± 0.48% of water independently of the experimental group (Figure S2b).

When pH was measured in fecal samples, it was similar during the first two weeks of supplementation, but higher pH was observed at the end of the study. This was not associated with any of the nutritional interventions given that the pH was similar (5.94 ± 0.04) in all experimental groups (Figure S2c).

#### *3.3. E*ff*ects of Supplements on Mineral Concentration*

Mineral content measured in blood, feces, femur, and urine samples at the end of the nutritional intervention for all experimental groups is summarized in Table 1.

**Table 1.** Mineral concentration (mg/g) in blood, feces, femur, and urine samples at the end of the study for all experimental groups considering both sexes together.


Results are expressed as mean ± SEM (*n* = 10/group). Calcium and magnesium blood concentrations are expressed as the mean ± SEM of mg of each mineral × 10−3/g of sample. REEF: animals no receiving supplement; V: animals receiving a daily supplement based on a food matrix with proteins, vitamins and minerals; V + I: inulin-enriched supplement-fed animals, and V + A: acacia-enriched supplement-fed animals.\* *p* < 0.05 vs. REF group; <sup>β</sup> *p* < 0.05 vs. V group; <sup>ε</sup> *p* < 0.05 vs. V + I group. Ca: calcium; Fe: iron; Mg: magnesium; P: phosphorus; Zn: zinc.

In blood samples the most abundant mineral studied was iron, followed by phosphorus, calcium, and magnesium, whereas zinc was the one detected in the lowest concentration. Regarding the nutritional intervention, V + I and V + A-supplemented animals had lower calcium concentration compared to the group receiving non-fiber-enriched supplement (V group) (*p* < 0.05). Moreover, the supplement containing acacia fiber resulted in a lower magnesium blood concentration compared to the REF and V groups (*p* < 0.05).

The most abundant mineral detected in feces was calcium, followed by potassium, magnesium, and iron. No effects due to dietary intervention were identified in this compartment.

When mineral content in femur was analyzed, the most abundant was calcium, followed by phosphorus, magnesium, and zinc. Iron was detected in a very low proportion. Interestingly, only acacia-enriched supplement significantly increased the concentration of calcium, magnesium, phosphorous, and zinc in comparison with both the V and V + I groups (*p* < 0.05).

In urine samples the mineral found in most abundant concentration was magnesium, followed by calcium and potassium. Both zinc and iron were detected in very low concentrations. No effects due to dietary intervention were observed in this compartment.

#### *3.4. E*ff*ects of Supplements on Hematological and Biochemical Variables*

From all the parameters related to the leucocytes, erythrocytes, and platelets, only punctual differences in the erythrocyte parameters were seen (Table S3). The inulin-enriched supplement-fed animals (V + I) showed a slightly lower mean corpuscular hemoglobin (MCH) and a reduction in its concentration (MCHC) when compared to that of the REF and V groups (*p* < 0.05).

The inulin-enriched supplement (V +I) intake resulted in a significantly lower plasma concentration of total cholesterol and uric acid in comparison to that of the REF group and to the acacia-supplemented animals (*p* < 0.05) (Figure 1a,f). Moreover, the acacia-enriched supplement intake reduced the glucose concentration compared to the supplement without fiber (*p* < 0.05) (Figure 1e).

**Figure 1.** (**a**) Total cholesterol; (**b**) high-density lipoprotein cholesterol (HDL-C); (**c**) low-density lipoprotein cholesterol (LDL-C); (**d**) triglycerides (TG); (**e**) glucose; and (**f**) uric acid concentration in blood samples at the end of the study for all experimental groups considering both sexes together. Results are expressed as mean ± SEM (*n* = 10/group). Statistical significance: \* *p* < 0.05 vs. REF group; <sup>β</sup> *p* < 0.05 vs. V group; <sup>ε</sup> *p* < 0.05 vs. V + I group.

No effects either on HDL-C, LDL-C, or triglycerides were found due to supplementation with the DF.

#### *3.5. E*ff*ects of Supplements on IgA Concentration*

The mean fecal IgA content was 1.5–2-fold higher in animals receiving both the inulin- (V + I) and acacia-enriched (V + A) supplements, compared to the REF and V groups (Figure 2). However, only the increase caused by inulin was statistically significant compared to both the REF and V groups (*p* < 0.05).

**Figure 2.** (**a**) IgA concentration in gut wash (GW) and (**b**) fecal samples, (**c**) proportion of fecal IgA-coating bacteria and (**d**) plasma IgA concentration quantified at the end of the study for all experimental groups considering both sexes together. Results are expressed as mean ± SEM (*n* = 10/group). Statistical significance: \* *p* < 0.05 vs. REF group; <sup>β</sup> *p* < 0.05 vs. V group.

No changes due to dietary intervention were detected either in the gut wash and plasma IgA concentrations or in the proportion of fecal IgA-coated bacteria.

#### *3.6. E*ff*ects of Supplements on Cecal Microbiota Composition*

#### 3.6.1. Diversity and Taxonomic Analysis

No changes on the Shannon–Wiener Index (3.52 ± 0.06) and CHAO1 (448.13 ± 10.69), as indicators of the diversity and richness of the microbial community, respectively, were observed after any dietary supplementation for all experimental groups at the end of the study.

Although it is well established that *Firmicutes* and *Bacteroidetes* are the most abundant phyla in the cecal microbiota, male rats had a lower proportion of *Bacteroidetes* in favor of that of Firmicutes. This increase in *Firmicutes* in males was also associated with a higher proportion of the family *Lactobacillaceae* spp. and in particular, the genus *Lactobacillus* (Figure 3).

**Figure 3.** Main taxonomic ranks showing the proportion of bacterial populations in the cecal content at the end of the study in males and females. The relative proportion of the bacteria was calculated in each taxonomic rank: (**a**) phylum, (**b**) family, and (**c**) genus. Results are expressed as mean (*n* = 5 female or male/group). Significant differences not shown.

Regarding the nutritional intervention, the acacia-enriched supplement (V + A) increased the proportion of the *Firmicutes* (up to 81.76%) and decreased *Bacteroidetes* (up to 12.92%), compared to the REF group whose proportions were 63.28% and 33.29%, respectively (*p* < 0.05) (Figure 3a). These changes were more evident in female than in male rats. These changes after acacia fiber-enriched supplement (V + A) intake were associated with an increase of *Lactobacillaceae* family, this effect being stronger in female than in male rats (Figure 3b), whose *Lactobacillaceae* proportion was already increased at the baseline. Moreover, acacia supplementation significantly increased genera belonging to the *Firmicutes* and *Actinobacteria* phyla (Figure 3c). In particular, a significant increase in *Bifidobacterium* spp. (up to 0.07% and 0.06% in females and males, respectively) and *Lactobacillus* spp. (up to 28% and 33% in females and males, respectively) was observed in V + A-fed animals in comparison with those in the REF (< 0.015% and < 15% for *Bifidobacterium* spp. and *Lactobacillus* spp., respectively, in both females and males) and V (< 0.03% and < 30% for *Bifidobacterium* spp. and *Lactobacillus* spp., respectively, in both females and males) groups (*p* < 0.05) (Figure 4a,b).

**Figure 4.** The relative proportion of the (**a**) *Bifidobacterium* and (**b**) *Lactobacillus* genera in cecal content at the end of the study differentiating between sexes. Results are expressed as mean ± SEM (*n* = 10/group). Statistical significance: \* *p* < 0.05 vs. REF group; <sup>β</sup> *p* < 0.05 vs. V group. (**c**) A representation of Venn diagrams showing the diversity in all genera differentiating between sexes. Results derived from *n* = 10/group.

#### 3.6.2. Venn Diagrams and Principal Components Analysis: Genera

The analysis of the genera distribution in Venn diagrams revealed that there was a core of 24 and 22 genera, in female and male rats, respectively, that persisted in all four experimental groups when considering both sexes separately (Figure 4c). Moreover, when comparing the microbiota depending on the supplementation, it could be observed that both the inulin (V + I) and acacia-enriched supplements (V + A) were able to exclusively promote the colonization of new genera in both female and male rats. On the one hand, inulin was able to promote the colonization of one new genus (*Longibaculum*) in female rats and five new genera in male rats (*Frisingicoccus, Erysipelatoclostridium, Gordonibacter, Parvibacter,* and *Enterorhabdus*), two of which also appeared with acacia supplementation. On the other hand, acacia supplementation resulted in the colonization of five new genera (*Bifidobacterium, Asaccharobacter, Extibacter, Enterorhabdus,* and *Enterococcus)* in female rats and three new genera in male rats (*Streptococcus*, *Parvibacter*, and *Enterorhabdus*), two of which also appeared in the inulin. In addition, some particular genera present in both the REF and V groups were absent in the V + I and V + A groups, this being the case in eight for females and eleven for males.

The PCA score plot revealed that the microbiota of the acacia-fiber supplemented animals (V + A) clustered differentially compared to those of both the REF and V groups in the genera analysis (Figure 5a). Moreover, the loading plot revealed that the *Bifidobacterium* spp. and *Lactobacillus* spp. were variables involved in the clustering of the V + A group (Figure 5b).

**Figure 5.** (**a**) Representation of Principal Components Analysis (PCA) for all experimental groups in a score plot and (**b**) a loading plot. Results derived from *n* = 10/group.

#### **4. Discussion**

Changes in dietary pattern, implying a lower intake of vitamins and minerals, may lead also to changes in microbiota composition. This situation, besides having a poor nutrient absorption and reduced bacterial diversity, is even more evident in the elderly [34]. In the current study, adult rats have been used to mimic the feasible impact of the fiber-enriched nutritional supplements tested herein on the immunological, hematological, and biochemical variables. Moreover, due to the disparity existing in the immune response, microbiota composition and the susceptibility to disease between gender [35–37], both female and male rats have been included in the present interventional study.

Herein, we demonstrate that the supplementations with inulin, a well-known prebiotic, and acacia gum fibers modify the adult rat microbiota composition with different intensity. Rats aged nine weeks that received acacia gum supplementation daily for four weeks showed an increased proportion in *Firmicutes* and *Actinobacteria.* In this regard, the acacia supplementation resulted in a significantly higher presence of *Lactobacillus* and the appearance of *Bifidobacterium* in both genders. In fact, this is not the first time that acacia gum's potential as a prebiotic agent has been described in both in vitro [38] and clinical studies [22,25]. Indeed, an interventional study carried out in human volunteers demonstrated that consumption of 10 and 15 g/day of acacia gum for 10 days increased the counts of both lactic acid-producing bacteria and *Bifidobacterium* in feces [22]. Moreover, its ability to selectively prevent the overgrowth of unwanted bacteria such as *Clostridium di*ffi*cile* or *Clostridium histolyticum,* has also been studied in vitro [38,39], although some controversy exists.

Moreover, acacia fiber supplementation produced the appearance of the genus *Asaccharobacter* (which belongs to the *Actinobacteria* phylum) in female rats, a single species of which has been reported to be a powerful equol producer [40]. Therefore, older people receiving the acacia fiber-enriched supplement may benefit from equol's health-promoting benefits, as has been reported, for example on osteoporosis, prostate cancer, and cardiovascular diseases [41,42].

In this study, the appearance of the genus *Enterorhabdus*(*Actinobacteria* phylum) has been associated with the consumption of acacia supplementation in both female and male rats and also with the inulin supplement in male rats. Although little is known about the possible role of this *Actinobacteria* genus, its higher relative abundance has been negatively correlated to serum TC, TG and LDL-C and hepatic TC, TG, bile acids, and non-esterified fatty acids in *Grifola frondosa* polysaccharide-chromium III-treated type 2 diabetes mellitus (T2DM) mice [43]. Although further, more in-depth studies should be carried out into this association, it seems that *Enterorhabdus* genus could exert some kind of hypoglycemic and hypolipidemic activities in T2DM, one of the multifactorial chronic metabolic disorders affecting mainly adults worldwide. Therefore, this result suggests that the inclusion of acacia or inulin fiber in the diet of adult people would be beneficial for them.

One of the objectives of the present study was to compare the prebiotic activity of both inulin and acacia fibers in adult rats. Surprisingly, the effects observed on microbiota composition after acacia fiber intake in female rats were significantly stronger than those exerted by inulin in the same gender. These differential results agree with those reported by Calame et al. who evidenced that acacia gum was able to produce a higher increase in both bifidobacteria and lactobacilli than an equal dose of inulin in healthy men [25].

The shift in microbiota composition has been proposed as a potential mechanism by which prebiotics improve mineral absorption [13,14]. In the current study, increased calcium, magnesium, phosphorus, and zinc concentrations were observed in femur from both female and male acacia-supplemented animals; thus, suggesting that acacia-enriched supplement could be beneficial for bone mineralization. Similar findings to those described herein have been reported after galactooligosaccharides (GOS), fructooligosaccharides (FOS), a mixture of GOS/FOS, and inulin supplementation in vitro, in vivo, and in human studies [15,44]. However, the fact that in our study no effects were observed in the inulin-supplemented group could be due to an insufficient dose, the type of inulin used, or the duration of the treatment. On the other hand, either the promotion of the lactic-acid bacteria or the production of SCFA may result in an acidification of the colon compartment; thus, preventing the formation of complexes between mineral and negatively charged metabolites, and therefore improving the bioavailability of minerals [15]. In the present study we found no changes in either fecal or cecal pH after the dietary supplementations that would explain this mechanism. The lack of intestinal acidification after inulin intake has already been reported in younger rats fed a diet with inulin for three weeks [28]. Further studies should be carried out in order to elucidate the impact of both fiber supplementations on SCFA production and their relationship with the mineral absorption in rats.

Most research has only been focused on the indirect effects of prebiotics over a considerable period of time. However, it has recently been evidenced that prebiotics may also cause direct effects, such as immunomodulation in the gastrointestinal tract [45]. In this regard, it is well known that prebiotic administration, such as inulin, generally results in increased fecal IgA concentration [18]. This fact is in line with those results obtained here because the intake of inulin-enriched supplement for four weeks increased fecal IgA content; thus, enhancing the intestinal immune system, with no difference between genders. With regard to the acacia-enriched supplement-fed animals, although a relevant tendency to increase fecal IgA levels was observed, it did not reach statistical significance. None of the dietary supplementations tested herein modified the proportion of IgA-coated bacteria, contrary to what was observed in the youngest rats whose proportion increased after three weeks of inulin intake [28]. Further studies should confirm these results and should be aimed at understanding them.

On the other hand, fiber and prebiotic intake in appropriate doses is associated with less incidence of metabolic diseases due to its indirect capacity to modulate the blood lipid profile and other metabolic variables [46]. In the current study, the intake of the inulin-enriched supplement for four weeks exerted lipid-lowering effects by significantly reducing the cholesterol and the uric acid in plasma. These results are partially in line with those reported in animal models [47] and in hypercholesterolemic [48] and healthy [49,50] subjects receiving an inulin supplementation for 3–16 weeks. Nevertheless, the modulation of biochemical variables after acacia fiber intake is quite controversial. Whereas some authors did not observe significant effects either in hypercholesterolemic [51] or in healthy [52] subjects, others have attributed to acacia fiber significant benefits for metabolic disorders [53,54]. In particular, the intake of 30 g of acacia for three months resulted in a significant reduction of blood triglyceride and fasting plasma glucose concentrations in type 2 diabetic patients [53]. However, the conditions (dose and length) tested within the study evidenced a tendency to reduce the glucose and uric acid concentrations. This lack of effect after the acacia-enriched supplement on biochemical variables agrees with that reported in hypercholesterolemic [51] or in healthy [52] subjects. Further studies are required to clarify the protective effects of acacia gum on cardiometabolic diseases.

#### **5. Conclusions**

Overall, both fiber-enriched supplements tested in the present study show the potential to be beneficial to gut-health, although differently. Whereas inulin-enriched supplement shows intestinal immune enhancement, acacia fiber supplement has stronger prebiotic activity, which may lead to increasing mineral absorption.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2072-6643/12/8/2196/s1, Figure S1: (**a**) Body weight and (**b**) food intake registered throughout the study and (**c**) body mass index (BMI) measured at the end of the study differentiating between sexes. Results are expressed as mean ± SEM (*n* = 5/group). No significant differences were observed. Figure S2: (**a**) Fecal wet weight, (**b**) humidity, and (**c**) pH registered throughout the study considering both sexes together. Results are expressed as mean ± SEM (*n* = 10/group). No significant differences were observed. Table S1: Composition and content of macronutrients and micronutrients (fiber, vitamins, and minerals) of the experimental supplements. Table S2: Relative weight of organs expressed as percentage (%) with respect to the body weight at the end of the study for all experimental groups differentiating between sexes. Results are expressed as mean ± SEM (*n* = 5 female and male/group). Statistical significance: <sup>α</sup> *p* < 0.05 vs. REF. Table S3: Hematological parameters in blood samples at the end of the study for all experimental groups considering both sexes together. Results are expressed as mean ± SEM (*n* = 10/group). Statistical significance: \* *p* < 0.05 vs. REF group; <sup>β</sup> *p* < 0.05 vs. V group; <sup>ε</sup> *p* < 0.05 vs. V + I group.

**Author Contributions:** Conceptualization, M.M.-C., À.F., M.C., and F.J.P.-C.; methodology, M.M.-C., I.A.-B., and M.J.R.-L.; formal analysis, M.M.-C. and F.J.P.-C.; writing—original draft preparation, M.M.-C. and F.J.P.-C.; writing—review and editing, I.A.-B., À.F., M.C., and M.J.R.-L.; supervision, F.J.P.-C.; and funding acquisition, F.J.P.-C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded through "Identificació de Nous Ingredients Moduladors de la Microbiota Humana i Alimentació Fent Ús de la Biotecnologia Industrial, les Tecnologies Òmiques i les Tecnologies de Big Data–MICROBIOTA (COMRDI-15-1-0029)" project included in the FEDER funding (Fondo Europeo de Desarrollo Regional 2014-2020).

**Acknowledgments:** The authors gratefully acknowledge the Units of Metal Analysis and Cytometry of the "Centres Científics i Tecnològics" of the university of Barcelona (CCiT-UB) for their help and advice with the ICP-MS and ICP-OES and flow cytometry analysis, respectively. In addition, we also would like to thank the Faculty of Pharmacy and Food Science animal facility's workers for their technical assistance and the students involved in the laboratory work (Alex Llorca, Mara Carmona, and Sergi Casanova).

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
