*3.4. Results*

#### 3.4.1. TiO2 and gu<sup>t</sup> Microbial Diversity

Alpha-diversity variations were measured in five studies [40,43–45,50]. Chao1—an estimate of species richness based on a vector or matrix of abundance data—did not significantly vary between exposed groups and controls groups in mice exposed to 100 mg/kg bw/day of TiO2 NPs for eight weeks [40], in pregnan<sup>t</sup> rats exposed to 5 mg/kg bw/day of TiO2 NPs for 12 weeks [43], but decreased in mice exposed to 150 mg/kg bw/day of TiO2 NPs for 30 days (*p* = 0.0052) [50]. In regards to Shannon's diversity—another index accounting for both abundance and evenness of the species with equal weighting given to abundant and rare species—no significant variations were observed between groups in mice exposed to 100 mg/kg bw/day of TiO2 NPs for eight weeks [40], in mice exposed to a diet containing 0.1% TiO2 NPs for three months [44], in mice exposed to 2, 10, 50 mg/kg bw/day of TiO2 NPs for three weeks [45], and in pregnan<sup>t</sup> rats exposed to 5 mg/kg bw/day

of TiO2 NPs for 13 days [43]. However, in the study of Zhang et al. [50], Shannon's diversity decreased in mice exposed to 150 mg/kg bw/day of TiO2 NPs for 30 days (*p* = 0.0036) [50]. Finally, applying Simpson's diversity index—another diversity index measuring richness and evenness in which more weighting is given to abundant species—in four out of the same studies [43–45,50], no significant variations were found except for the study of Zhang et al. [50] showing a significant increase after TiO2 NPs exposure (*p* = 0.0180).

#### 3.4.2. TiO2 and Abundance of Individual Microbial Species

In rodents, four studies showed an increase in Firmicutes abundance after TiO2 exposure compared with controls [35,39,49,51]. *Lactobacillus* was the most studied genus and significantly decreased in four studies [35,36,39,44] but increased in one study [45] after TiO2 NPs exposure compared with controls. Moreover, an increase in *Allobaculum* abundance was reported in one study [45] while a decrease was observed in another mice model [35]. Other variations in genera and family abundance after TiO2 exposure compared with controls are observed such as an increase in *Oscillospira* [35,51], *Turicibacter* [36], and Clostridiales [43], and a decrease in *Veillonella* [36], *Prevotella* [40,51], and *Dehalobacteriaceae* [43].

Bacteroidetes abundance could also be influenced by TiO2 exposure in rodent models. Three studies showed a decrease in Bacteroidetes levels [35,49,51] while one study reported an increase in Bacteroidetes levels [44]. Especially, TiO2 exposure could lead to an increase in *Bacteroides* [40], *Parabacteroides* [45], and a decrease in *Barnesiella* [49].

Actinobacteria phylum could decrease in abundance [35] after TiO2 exposure with a decrease in *Bifidobacterium spp* reported in two rodent studies [35,44]. Moreover, an increase in *Rhodococcus* abundance [40] and a decrease in *Adlercreutzia* levels [45] were observed.

In regards to other phyla, Proteobacteria could increase after TiO2 exposure, as reported in three studies [40,50,51], and Desulfovibrionaceae [51] and Verrucomicrobia could decrease, in particular in the *Akkermansia* genus [51].

All these findings observed in rodent models showed that TiO2 exposure could impact gu<sup>t</sup> microbiota composition, although the variations in specific phyla and genera abundances remain to be elucidated with large sample size animal studies using the same dose and duration of TiO2 exposure.

In regards to animal models other than rodents, a model organism *Drosophila melanogaster* [42] showed that the exposure of 1, 2, and 200 mg/mL dietary of three different sizes of TiO2 NPs for five days did not inhibit the growth of gu<sup>t</sup> bacteria in *Drosophila* larva or adults. On the other hand, a silkworm model [41] showed different gu<sup>t</sup> microbiota compositional variations after intake of mulberry leaves soaked in 5 mg/L TiO2 NPs and naturally dried from the third day of fifth instar larvae until morning.

#### 3.4.3. TiO2 and SCFAs Levels

A total of six rodent studies reported between-group differences in fecal SCFA concentrations after different TiO2 NPs dose exposure and length of exposure. Three studies showed no significant variations in SCFAs levels [36,37,48] while two studies observed a decrease in SCFAs levels in mice treated with 0.1 weight percent of TiO2 NPs for eight weeks [35] and in mice treated with 50 mg TiO2/kg bw/day for three weeks [45]. Interestingly, one study [39] reported an increase of SCFAs in stools in mice exposed to 1 g/kg bw TiO2 for 14 days. This can be explained by an increase in SCFAs production or a decrease in absorption.

#### 3.4.4. TiO2 and Metabolism

A total of seven studies [36,37,39,42,43,49,50] showed significant metabolic variations in TiO2 exposed animals compared with controls. Lipopolysaccharides (LPS) proportionally increased in mice exposed to 2, 10, and 50 mg/kg bw/day of TiO2 for 30 days [36], in mice exposed to 10, 40, and 160 mg/kg bw/day of TiO2 for 28 days [49], and in mice exposed to 10 μL/g bw/day for eight weeks [51]. Interestingly, in TiO2-treated mice fed

with a high-fat diet (HFD), LPS significantly increased compared with TiO2-treated mice fed with a high fiber diet (CHOW diet) [51]. Triglycerides levels (TG) levels increased in mice after exposure to 160 mg/kg bw/day of TiO2 for 28 days, while TG levels reduced in rats exposed to 10 and 50 mg/kg bw/day of TiO2 for 90 days. Moreover, glucose levels could increase after TiO2 exposure, as reported in two rodent model studies [43,49]. Interestingly, in Sprague-Dawley pregnan<sup>t</sup> rats, exposure of 5 mg/kg/day of TiO2 NPs for 12 weeks could strengthen genes about type 2 diabetes mellitus related to function and lipid biosynthesis, compared with controls [43].

The two *Drosophila* model studies [42,46] reported contradictory results. One study showed no alterations of pupation cycle, weight, and lipid levels after 1, 2, and 200 mg/mL dietary TiO2 NPs of different sizes for five days while Richter and colleagues [46] demonstrated alterations of metabolic gu<sup>t</sup> homeostasis with significant changes in pupation, time to pupation, reduction of body size, and glucose levels.

#### 3.4.5. TiO2 and Gut Barrier Permeability

Bettini et al. observed no significant changes in epithelial paracellular permeability in the E171 group in comparison to the controls [8]. Additionally, a previous study [48] found no effect compared with controls on mucin O-glycosylation in the small intestine of the rats following 7- or 60-day TiO2 exposure, regardless of TiO2 type (E171 and NM-105) or E171 dose tested (0.1 mg/kg bw/day and 10 mg/kg bw/day). Another study [39] showed that at 24 h post-gavage, MUC2 gene expression was lower in TiO2 NP-treated mice (1 g/kg bw/day) compared with controls but this trend was reversed from 48 h post-gavage to seven days with an elevated expression of mucin-2 for the rest of the study.

On the other hand, in mice exposed to 0.1 weight percent of TiO2, goblet cells and crypts significantly decreased compared to controls. Furthermore, three studies [45,49,51] reported a decrease in MUC2 gene expression in mice treated with TiO2 NPs. Yan et al. [49] also reported a reduction of mucus thickness in all exposed mice compared with controls. Interestingly, MUC2 gene expression and cryp<sup>t</sup> length significantly decreased in TiO2- treated mice fed with HFD compared with TiO2-treated mice fed with CHOW diet [51].

#### 3.4.6. TiO2 and Inflammatory Responses

A total of ten studies have assessed levels of different gu<sup>t</sup> microbiota associated biomarkers of mucosal immunity and intestinal inflammation such as interleukins (IL) levels, number of T reg cells, macrophages, and T helper cells. A reduction of T reg cells numbers was found in food-grade E171 treated mice after 100 days [8] and in mice exposed to a diet containing 0.1% TiO2 NPs for three months [44]. Inflammatory cytokines levels increased in exposed rodents compared with controls in the majority of studies: IL1 [49,51], IL2 [38], IL6 [8,36,45,49,51], IL10 [45], IL12 [35], IL17 [8,35], IL18 [8], as well as TNF α levels [45,49,51]. The production of macrophages and the expression of β defensin gene are also stimulated [45]. Interestingly, TiO2 NPs decreased the CD4+ T cells, T regs, and macrophages in the mesenteric lymph nodes and increased neutrophil gelatinase-associated lipocalin (LCN2) levels in mice aggravating the DSS-induced chronic colitis [44]. Moreover, IL-1 levels, IL-6 levels, TNF α levels increased in TiO2-treated mice fed with HFD compared with TiO2-treated mice fed CHOW diet [51]. All these results showed the potential involvement of TiO2 in the imbalances in intestinal and systemic immune responses.
