Intra-amniotic Administration (Gallus gallus) of TiO₂, SiO₂, and ZnO Nanoparticles Affect Brush Border Membrane Functionality and Alters Gut Microflora Populations

Abstract

Metal oxide nanoparticles (NP) are increasingly used in the food and agriculture industries, making human consumption nearly unavoidable. The goal of this study was to use the Gallus gallus (broiler chicken) intra-amniotic administration of physiologically relevant concentrations of TiO₂, SiO₂, and ZnO to better understand the effects of NP exposure on gut health and function. Immediately after hatch, blood, cecum, and small intestine were collected for assessment of iron (Fe)-metabolism, zinc (Zn)-metabolism, brush border membrane (BBM) functional, and pro-inflammatory related proteins gene expression; blood Fe and Zn levels; cecum weight; and the relative abundance of intestinal microflora. NP type, dose, and the presence or absence of minerals was shown to result in altered mineral transporter, BBM functional, and pro-inflammatory gene expression. Metal oxide NP also altered the abundance of intestinal bacterial populations. Overall, the data suggest that the in vivo results align with in vitro studies, and that NP have the potential to negatively affect intestinal functionality and health.

Graphical Abstract

1. Introduction

Nanotechnology involves engineering, manufacturing, and using materials at a nanoscale to achieve specific properties, and nanomaterials are increasingly used in the food and agriculture industries. Nanoparticles (NP) are added to food to enhance texture, flavor, color, consistency, stability or nutrient bioavailability, and NP in food packaging can improve packaging flexibility, gas barrier properties, temperature and moisture stability, and act as antimicrobials or sensors for microorganism detection and identification (Chaudhry et al.,2008; Nanotechnology Applications in the Food Industry, 2018). In agriculture, nanomaterials are being developed for novel agrochemicals such as fertilizers and pesticides, and nanotechnology is being used to develop sensors for applications such as water or nutrient detection (Nanotechnologies in Food and Agriculture, 2015; Kah et al., 2018). Nanomaterial production and distribution for all applications results in emissions to the environment and landfills ( Keller et al., 2013, Keller et al., 2014). Human ingestion of nanomaterials though food, agricultural residue, or the environment is nearly unavoidable, which highlights the importance of understanding the effects of NP on gastrointestinal function.

Studies by Lomer et al. estimated that the average person in a developed country consumes 10¹²-10¹⁴ engineered fine (0.1–1 μm diameter) to ultrafine (<100 nm diameter) particles per day, primarily from the ingestion of nanosized titanium dioxide (TiO₂), silicates, and aluminosilicates (Lomer et al., 2002; Lomer et al., 2004). Titanium dioxide (TiO₂) nanoparticles, primarily the anatase form, are often added to processed food as whitening and brightening agents ( McClements et al., 2017). More recent studies on TiO₂ nanoparticle consumption from Weir at al. (Weir et al., 2012) and Rompelberg et al. (Rompelberg et al., 2016), have estimated TiO₂ intake to be 0.2 – 0.7 mg/kg body weight/day for US adults and nano-TiO₂ intake to be 0.55 μg/kg body weight/day for Dutch adults, respectively. Both studies estimated a higher intake in children when compared with adults, due to an increased intake of products with a high level of TiO₂ such as candy, gum, and toothpaste (Weir et al., 2012; Rompelberg et al., 2016). SiO₂ nanoparticles, generally amorphous solid spheres, are primarily used as a thickening agent, as an anticaking agent in powdered foods to maintain flow properties, and as a carrier for fragrances or flavors (Dekkers et al., 2011; Peters et al., 2012). Dekkers et al. have estimated nano-SiO₂ intake at 1.8 mg/kg body weight/day (Dekkers et al., 2011). Nano-sized zinc oxide (ZnO) is used as a source of zinc in supplements and functional foods, in food packaging as antimicrobial agents, in canned food lining to prevent food staining, and as an ultraviolet (UV) light absorber to protect foods sensitive to UV light (Wang et al., 2014; Sirelkhatim et al., 2015; Robertson et al., 2016; Espitia et al., 2016; The Wiley encyclopedia of packaging technology, 2009; Onwulata, 2014). Estimates for ZnO nanoparticle ingestion are sparse, but Moreno-Olivas et al. have estimated that approximately 2 mg of Zn is consumed from a canned food meal (Moreno-Olivas et al., 2019).

The increasing volume of nanomaterial production and the diversity of applications also result in environmental emissions during the nanomaterial life cycle. Keller et al. have estimated that TiO₂ NP emissions are 15,600 metric tons/year to water systems and 32,600 metric tons/year to landfills, SiO₂ NP emissions are 2,100 metric tons/year to water systems and 81,200 metric tons/year to landfills, and ZnO NP emissions are estimated to be 3,700 metric tons/year to water systems and 21,100 metric tons/year to landfills (Keller et al., 2013).

Our previous work with an in vitro model of the intestinal epithelium has shown that exposure to TiO₂, SiO₂, and ZnO NP changes the functionality of the intestinal cells (Moreno-Olivas et al., 2019; Guo et al., 2017; Gou et al., 2018; Pereira et al., 2018; Richter et al., 2018). Exposure to TiO₂ NP at physiologically realistic doses significantly decreased intestinal barrier function following chronic (5 day) exposure. Reactive oxygen species (ROS) generation, pro-inflammatory signaling, and intestinal alkaline phosphatase activity, which regulates the absorption of fatty acids across the apical intestinal epithelial membrane, all showed increases in response to nano-TiO₂. Iron, zinc, and fatty acid transport were significantly decreased following exposure to TiO₂ nanoparticles (Guo et al., 2017). SiO₂ NP exposure significantly affected iron, zinc, glucose, and lipid nutrient absorption (Guo et al., 2018). The brush border membrane intestinal alkaline phosphatase activity was increased in response to nano-SiO₂. The barrier function of the intestinal epithelium, as measured by transepithelial electrical resistance, was significantly decreased in response to chronic nano-SiO₂ exposure. Gene expression and oxidative stress formation analysis showed SiO₂ NP altered the expression levels of nutrient transport proteins, generated reactive oxygen species, and initiated pro-inflammatory signaling (Guo et al., 2018). Exposure of the intestinal in vitro model to ZnO NP can resulted in a significant decrease in glucose transport (Moreno-Olivas et al., 2019). TiO₂, SiO₂, and ZnO NP all damaged the physical structure of absorptive cells, which significantly decreased the surface of the absorptive cells covered with microvilli. The effects of TiO₂ NP on microvilli were ameliorated in the presence of beneficial bacteria (Richter et al., 2018), and the decreases in barrier function were eliminated when TiO₂ NP exposure was within a food matrix (Pereira et al., 2018).

In the current study, the intestinal effects of TiO₂, SiO₂, and ZnO ingestion were studied in vivo and by using the intra amniotic administration approach in Gallus gallus. The Gallus gallus is a fast-growing model, receptive to dietary manipulations, and sensitive to dietary deficiencies in trace minerals such as iron (Fe) and zinc (Zn) (Tako et al., 2010; Tako et al., 2014; Tako et al., 2014; Tako et al., 2015). There is >85% homology between human and Gallus gallus intestinal gene sequences for mineral transporters such as DMT1, DcytB, ZnT1, and Ferroportin (Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution, 2004). The broiler chicken gut microbiota is complex, dynamic, heavily influenced by the host’s genetics, environment, and diet, and there is considerable similarity at the phylum level between the gut microbiota of broilers chickens and humans (Yegani et al., 2008; Backhed et al., 2005). Further, in recent years, the the intra amniotic administration approach (in ovo feeding) was further developed and currently is widely applied in the evaluation process of the effects of functional foods and dietary factors on the functionality of the intestinal brush border membrane, as well as potential prebiotic properties and interactions with the intestinal microbial populations (Hou and Tako, 2018).

The goal of this work is to understand how exposure of metal oxide nanomaterials at physiologically relevant doses affects the expression of Fe and Zn metabolism-related genes, brush border membrane (BBM) development and functionality, and the relative abundance of probiotic populations of bacteria such as Bifidobacterium and Lactobacillus versus those of potentially pathogenic bacteria such as E. coli and Clostridium. Previous studies have demonstrated that in ovo feeding using intra-amniotic fluid administration is a fast and cost effective method to evaluate plant-origin prebiotic and bioactive compound effects on intestinal health and function, including mineral absorption, brush border membrane expression, and microflora populations (Hou and Tako, 2018). This will be the first study, however, to assess the effects of nanomaterials on gut health and function using in vivo, and by using the intra-amniotic administration approach.

2. Materials and Methods

2.1. Animals and and Study Design

Cornish cross-fertile broiler chicken eggs (n = 180) were obtained from a commercial hatchery (Moyer’s chicks, Quakertown, PA, USA). The eggs were incubated under optimal conditions at the Cornell University Animal Science poultry farm incubator. All animal protocols were approved by Cornell University Institutional Animal Care and Use committee.

TiO₂ (30 nm anatase, US Research Nanomaterials, Inc., Houston, TX), SiO₂ (20–30 nm amorphous, US Research Nanomaterials, Inc., Houston, TX), and ZnO (10 nm, Meliorum Technologies, Inc., Rochester, NY) NP were used in this study. All particles have previously been characterized with electron microscopy and dynamic light scattering (DLS) in 18 MΩ resistivity deionized (DI) water (Moreno-olivas et al., 2019; Guo et al., 2017; Guo et al., 2018). Physiologically relevant NP concentrations were chosen for this study based on previous results in vitro and to mimic potential human NP ingestion exposures. Nanoparticle formulations were prepared in sterile 18 MΩ DI water. The osmolality of all solutions was measured using a VAPRO Vapor Pressure Osmometer (Wescor, Logan, UT, USA) to ensure the osmolality was less than 320 OSM, which is necessary to prevent the chicken embryos dehydration upon injection of the solution. Doses were the following: 1.4×10⁻⁶ mg TiO₂ NP/mL (medium) and 1.4×10–4 mg TiO2 NP/mL (high), 2.0×10⁻⁵ mg SiO₂ NP/ml (medium) and 2.0×10⁻³ mg SiO₂ NP/ml (high), and 9.7×10⁻⁴ mg ZnO NP/mL (medium) and 9.7×10⁻² mg ZnO NP/mL (high). NP injection solutions with ⁵⁸Fe or ⁶⁷Zn were prepared identically to previously publish in vitro studies (Moreno-olivas et al., 2019; Guo et al., 2017; Guo et al., 2018).

The intra-amniotic administration procedure was previously described by Tako et al. (Tako et al., 2014). On Day 17 of embryonic incubation, eggs with viable embryos were weighed and divided into 18 groups (n = 10) with an approximately equal weight distribution. The intra-amniotic injection solution (1 mL per egg) was injected with a 21-guage needle into the amniotic fluid, which was identified by candling. Following the injection, the injection sites were sealed with cellophane tape. Eggs were then placed in hatching baskets, with each treatment equally represented at each incubator location.

2.2. Blood and Tissue Collection

Immediately following hatching (Day 21, hatchability rate was 92%), birds were weighed and euthanized with CO₂ exposure. Blood was collected using micro-hematocrit heparinized capillary tubes (Thermo Fisher Scientific, Waltham, MA, USA). The small intestines, ceca, and livers were then quickly removed from the carcasses and placed in separate sterile cryovials (Simport, Beloeil, QC, Canada) for storage. Ceca were weighed before storage. The samples were immediately frozen in liquid nitrogen and stored at −80 °C until analysis. Blood sample ⁵⁸Fe and ⁶⁷Zn concentrations were determined with Inductively Coupled Plasma Mass Spectrometry (ICP-MS, ICAP Model 61E Trace Analyser; Thermo Jarrell Ash Corporation, Franklin, MA, USA).

2.3. Isolation of Total RNA From Chicken Duodenum

Total RNA was extracted from 30 mg proximal duodenal tissue using a Qiagen RNeasy Mini Kit (Qiagen Inc., Germantown, MD) according to the manufacturer’s protocol. Total RNA was eluted in 50 μL of RNase-free water. All steps were carried out under RNase free conditions. RNA was quantified with a NanoDrop 2000 (ThermoFisher Scientific, Waltham, MA) at A _260/280. RNA was stored at −80°C until use.

2.4. Real-Time Polymerase Chain Reaction

The primers used in the real-time polymerase chain reactions (RT-PCR) were designed using Real-Time Primer Design Tool software (IDT DNA, Coralvilla, IA) based on 11 gene sequences from GenBank database. The sequences are shown in Table 1. The amplicon length was limited to 90 to 150 bp, the length of the primers was from 17- to 25- mer, and the GC content was between 41 and 55%. The specificity of the primers was tested by performing a BLAST search against the genomic NCBI database.

Table 1:

Real-time polymerase chain reaction (RT-PCR) primer sequences.

Gene	Forward Primer (5′-3′)	Reverse Primer (5′-3′)	Base Pairs	GI Identifier
DcytB	CATGTGCATTCTCTTCCAAAGTC	CTCCTTGGTGACCGCATTAT	103	20380692
Ferroportin	CTCAGCAATCACTGGCATCA	ACTGGGCAACTCCAGAAATAAG	98	61098365
DMT-1	TTGATTCAGAGCCTCCCATTAG	GCGAGGAGTAGGCTTGTATTT	10l	2.07E+08
Zip1	TGCCTCAGTTTCCCTCAC	GGCTCTTAAGGGCACTTCT	144	XM_015298606.1
ZnT1	GGTAACAGAGCTGCCTTAACT	GGTAACAGAGCTGCCTTAACT	105	54109718
SGLT-1	GCATCCTTACTCTGTGGTACTG	TATCCGCACATCACACATCC	106	8346783
IL-8	TCATCCATCCCAAGTTCATTCA	GACACACTTCTCTGCCATCTT	105	Y14971.1
TNF-α	GACAGCCTATGCCAACAAGTA	TTACAGGAAGGGCAACTCATC	109	53854909
SI	CCAGCAATGCCAGCATATTG	CGGTTTCTCCTTACCACTTCTT	95	2246388
NFκBI	CACAGCTGGAGGGAAGTAAAT	TTGAGTAAGGAAGTGAGGTTGAG	100	2130627
18S	GCAAGACGAACTAAAGCGAAAG	TCGGAACTACGACGGTATCT	100	7262899

DcytB, Duodenal cytochrome B; DMT1, Divalent metal transporter 1; ZIP1, Zinc Transporter 1; ZnT1, Zinc Transporter 1; SGLT-1, Sodium-glucose transporter 1; IL-8, Interleukin 8; TNF-α, Tumor necrosis factor-alpha; SI, Sucrose isomaltase; NF_ΚB1, Nuclear factor Kappa B Subunit 1; 18S rRNA, 18S Ribosomal subunit.Isolation of DNA from Chicken Cecum.

cDNA was generated using a C1000 Touch thermocycler (Biorad, Hercules, CA) and a Promega-Improm-II Reverse Transcriptase Kit (Catalog #A1250) 20 μL reverse transcriptase reaction. The reverse transcriptase reaction consisted of 1 μg total RNA template, 10 μM random hexamer primers, and 2 mM of oligo-dT primers. All reactions were performed under the following conditions: 94°C for 5 minutes, 60 minutes at 42°C, 70°C for 15 minutes, and then hold at 4°C. The concentration of cDNA obtained was determined with a NanoDrop 2000 at A _260/280 with an extinction coefficient of 33 for single stranded DNA.

RT-PCR was performed with a Bio-RadCFX96 Touch (Hercules, CA, USA). The 10 μL RT-PCR mixtures consisted of cDNA (2 μg), 2X BioRad SSO Advanced Universal SYBR Green Supermix (Cat #1725274, Hercules, CA, USA), forward and reverse primers, and nuclease-free water (for the no template control). The no template control of nuclease-free water was included to exclude DNA contamination in the PCR mix. All reactions were performed in duplicates and under the following conditions: initial denaturing at 95°C for 30 seconds, 40 cycles of denaturing at 95°C for 15 seconds, various annealing temperatures according to IDT for 30 seconds and elongating at 60°C for 30 seconds. After the cycling process was completed, melting curves was determined from 65.0°C to 95.0°C with increments of 0.5°C for 5 seconds to ensure amplification of a single product. RT-PCR efficiency values for the eleven genes were as follows: DcytB, 1.046; Ferroportin, 1.109; DMT1, 0.998; Zip1, 0.921; ZnT1, 1.09; SGLT1, 0.994; IL-8, 0.998; TNF-α, 1.046; SI, 1.032; NK-κβ1, 1.113; 18s rRNA, 0.934. Gene expression levels were obtained from Ct values based on the ‘second derivative maximum’ as computed by the Biorad CFX Maestro Software (BioRad, Hercules, CA, USA). Gene expression was normalized to the expression of 18S.

The contents of the ceca were placed into a sterile 50 mL tube (Corning, Corning, NY, USA) containing 9 mL of sterile 1X phosphate buffered saline (PBS) and homogenized by vortexing with glass beads (3mm diameter) for 3 minutes. Debris was removed by centrifugation at 700 x g for 1 minute, and the supernatant was collected and centrifuged at 12,000 x g for 5 minutes. The pellet was washed twice with 1X PBS and stored at −20°C until DNA extraction.

To extract DNA, the pellet was re-suspended in 50 mM EDTA and treated with 10 mg/mL lysozyme (Sigma Aldrich CO., St. Louis, MO) for 45 min at 37°C. The bacterial genomic DNA was then isolated using a Wizard Genomic DNA purification kit (Promega Corp., Madison, WI) following the manufacturer’s instructions.

2.5. PCR Amplification of Bacterial 16s rDNA

Primers for Lactobacillus, Bifidobacterium, Clostridium, and E. coli were designed according to previously published data by Zhu et al. (Zhu et al., 2002). The universal primers, which identify all known strains of bacteria in the intestine, were prepared with the invariant region in the 16S rRNA of bacteria and used as internal standard to normalize the results. PCR products were separated by electrophoresis on 2% agarose gel, stained with ethidium bromide, and quantified using the Quantity One 1-D analysis software (BioRad, Hercules, Ca, USA).

2.6. Statistical Analysis

Results are expressed as mean ± standard error, n ≥ 8. Results were analyzed by one-way multiple analysis of variance (ANOVA) using JMP software (SAS Institute Inc., Cary, NC).Differences between treatments were compared with Tukey’s post hoc test, and results were considered statistically different at p < 0.05.

3. Results

3.1. Body Weight, Cecum Weight, and Cecum –to-Body Weight Ratio

There were no significant differences in body weight across all treatment groups (Figure 1A). For groups with NP and ⁵⁸Fe, the high TiO2 and mid SiO2 groups had a cecum weight and cecum weight/body weight ratio that was significantly higher than the non-injected controls (Figure 1B, D). For groups injected with NP and ⁶⁷Zn, the mid TiO₂ and high TiO₂ groups had a cecum weight and cecum weight/body weight ratio that was significantly higher than the non-injected controls (Figure 1C, E). Based on these results, only medium and high TiO₂ and medium SiO₂ treatments were further analyzed.

Figure 1:

Effects of day 17 intra-amniotic in ovo injections on day 0 chick body weight, cecum weight, or cecum weight/body weight ratio. Day 17 embryos were exposed to water, stable isotope (⁵⁸Fe or ⁶⁷Zn), or medium or high doses of TiO₂, SiO₂, or ZnO nanoparticles (NP) with ⁵⁸Fe or ⁶⁷Zn via in ovo, intra-amniotic injections and day 0 chicks were assessed for body weight (A), cecum weight (B, C), and cecum weight to body weight ratio (D, E). Values are mean ± SEM. Bars that do not share any letters are significantly different according to a one-away ANOVA with Tukey’s post hoc test (p ≤ 0.05).

3.2. ⁵⁸Fe and ⁶⁷Zn Absorption

Figure 2A shows ⁵⁸Fe detected via ICP-MS in the chick bloodstream, and Figure 2B shows ⁶⁷Zn detected in the chick blood. There are no statistically significant differences in ⁵⁸Fe or ⁶⁷Zn absorption from the amniotic fluid into the bloodstream for all conditions tested.

Figure 2:

Effects of day 17 intra-amniotic in ovo injections on day 0 chick ⁵⁸Fe or ⁶⁷Zn absorption. 58Fe (A) or 67Zn (B) in the day 0 chick bloodstream following day 17 in ovo injections with water, stable isotope (⁵⁸Fe or ⁶⁷Zn), or medium or high doses of TiO₂, SiO₂, or ZnO nanoparticles (NP) with ⁵⁸Fe or ⁶⁷Zn. Values are mean ± SEM. There are no statistically significant differences between means according to a one-way ANOVA with Tukey’s post hoc test (p ≤ 0.05).

3.3. Nutrient Transport, Enzyme Function, and Pro-Inflammatory Gene Expression

Figure 3 shows intestinal gene expression results. The relative expression of Zn metabolism genes (ZnT1, ZIP1) were assessed in response to the intra-amniotic injections. ZnT1 was significantly upregulated (p < 0.05) in response to medium SiO₂ + Zn when compared with water injected controls. ZIP1 was significantly upregulated (p < 0.05) in response to high TiO₂ + Fe, high TiO₂ + Zn, and a medium dose of TiO₂ NP when compared 255 with water injected controls.

Figure 3.

Effects day 17 intra-amniotic in ovo injections on day 0 chick intestinal gene expression. Values are means ± SEM. Per gene, treatments groups that do not share any letters are significantly different according to a one-way ANOVA with Tukey’s post hoc test (p ≤ 0.05). SI = Sucrose isomaltase.

The relative expressions of Fe metabolism genes (DcytB, DMT1, and ferroportin) were assessed in response to injection with water, minerals, and/or NP. DMTI and ferroportin were significantly upregulated (p < 0.05) in the presence of high TiO₂ + Fe, high TiO₂ + Zn, medium SiO₂ + Fe, and medium SiO₂ + Zn when compared with water injected controls. DcytB was significantly upregulated (p < 0.05) following exposure to high TiO₂ + Zn, 261 medium SiO₂ + Fe, and medium SiO₂ + Zn when compared with water injected controls.

The relative expressions of BBM functional genes (SI and SGLT1) were significantly (p < 0.05) upregulated in the presence of high TiO₂ + Fe, high TiO₂ + Zn, medium SiO₂ + Fe, and medium SiO₂ + Zn when compared with water injected controls. Finally, the expression of gut inflammatory genes (TNF-α, IL-8, NF-κβ) were assessed. Exposure to high TiO₂ + Fe and medium SiO₂ + Zn significantly upregulated (p < 0.05) TNF-α expression when compared with water injected controls. Exposure to high TiO₂ + Fe, high TiO₂ + Zn, medium SiO₂ + Fe, and medium SiO₂ + Zn significantly upregulated (p < 0.05) IL-8 expression when compared with water injected controls. Exposure to medium SiO₂ + Fe or medium SiO₂ + Zn significantly upregulated (p < 0.05) NF-κβ expression when compared with water injected controls.

3.4. Intestinal Content Bacterial Genera- and Species-Level Analysis

Figure 4 shows the chick cecal genera and species-level bacterial populations. The relative abundance of Bifidobacterium or Lactobacillus, which are considered to be probiotics, significantly increased (p < 0.05) in the presence of medium TiO₂ + Zn, high TiO₂ + Zn, and medium SiO₂ + Zn when compared with non-injected or water injected controls. Bifidobacterium or Lactobacillus abundance was significantly decreased (p < 0.05) by medium TiO₂ + Fe, high TiO₂ + Fe, and medium doses of TiO₂ NP or SiO₂ NP only when compared with non-injected or water injected controls. Bifidobacterium abundance was significantly decreased (p < 0.05) following exposure to water + Zn and medium SiO₂ + Fe when compared with non-injected controls.

Figure 4.

Effects day 17 intra-amniotic in ovo injections on cecal genera and species-level bacterial populations (day of hatch). Values are means ± SEM. Per bacterial category, treatments groups that do not share any letters are significantly different according to a one-way ANOVA with Tukey’s post hoc test (p ≤ 0.05).

Clostridium, which is an opportunistic or potentially pathogenic bacteria, was significantly increased (p < 0.05) in the presence of medium doses of SiO₂ NP when compared with non-injected or water injected controls. E. coli, which is also considered opportunistic and potentially pathogenic, was significantly increased (p < 0.05) following exposure to medium SiO₂ + Zn when compared with water injected controls. Clostridium and E. coli were not significantly decreased by any of the water, mineral, or NP treatments.

4. Discussion

The intra amniotic administration procedure (“In ovo feeding”) of fertilized Gallus gallus eggs has recently become an accepted method for evaluating how ingestion of functional, plant-origin compounds affects the functionality of the small intestine and resident microbial populations (Hou and Tako, 2018), as previous work has demonstrated that broiler chickens are an appropriate in vivo model for human iron and zinc dietary bioavailability (25, 34). The current study used the in ovo Gallus gallus model to assess the effect of metal oxide nanoparticles, a common food additive, on the animal model body weight, ceca weigh, blood iron and zinc content, BBM gene expression (used as a biomarker of tissue functionality), and the relative abundance of representative microbial species immediately after hatch. The results show that intra-amniotic-delivered metal oxide NP affected cecum size, BBM functional gene expression, and cecum bacterial populations.

The gastro intestinal (GI) tract of poultry is composed of the esophagus, crop, proventriculus, gizzard, small intestine (duodenum, jejunum, and ileum), cecum, colon, and cloaca (Pan and Yu, 2014). The ratio of GI length: body length in poultry is shorter than in mammals, and this results in a faster passage of digesta in poultry and selection for fast growing or mucosal-layer adherent bacteria (Pan and Yu, 2014)). The exception is the ceca, which are two blind pouches with a much slower passage rate (Pan and Yu, 2014)). There is a diverse, well-studied microbiome within the ceca that has been shown to effect host nutrition and health (Pan and Yu, 2014). As shown in Figure 1, the ceca of animals that received ⁵⁸Fe and high TiO₂ or mid SiO₂, or ⁶⁷Zn and mid TiO₂ or high TiO₂ groups had a significant increase in cecum weight and cecum weight/body weight ratio. As was previously demonstrated, an increase in cecal weight indicates an overall increase in the bacterial content (Wang et al., 2019). Unlike germfree rodents, which display an enlarged cecum (Gustafsson et al., 1971), germfree chickens have lower cecum weights with a cecal thinner wall than conventional birds (Rolls et al., 1977). Gut microbiota can affect the intestinal morphology of poultry, germ free broilers have shorter intestinal villi and shallower crypts when compared with conventionally-raised birds (Forder et al., 2007; Gabriel et al., 2006), but an overabundance of bacteria could also compete for nutrients in the developing animals (Wang et al., 2019). In the current study the larger cecum size, and the higher bacterial loads, did not affect Fe or Zn absorption (Figure 2).

Zn is a charged, hydrophobic ion, and cannot cross biological membranes by simple diffusion. Specialized protein transporters exist to move Zn across the plasma membrane for cellular uptake and release. ZIP1 mediates Zn uptake from the gut lumen, and overexpression of ZIP1 results in upregulated enterocyte uptake of Zn (Michalczyk and Ackland, 2013). ZnT1 codes for a basolateral intestinal epithelial Zn transporter that shows differential mRNA expression in response to the amount of zinc in the diet (McMahon et al., 1998a,b). The intra-amniotic administration of TiO₂, SiO₂ and/or Fe or Zn led to the significant upregulation of duodenal Zn transport genes when compared with water-injected controls (Figure 3). ZIP1 was significantly upregulated in response to TiO₂ + Fe, TiO₂ + Zn, or TiO₂ NP alone when compared with water injected controls. Similar to our previous results in vitro (Guo et al., 2018), ZnT1 in was upregulated in vivo, and in response to a medium dose of SiO₂ + Zn exposure. Exposure to TiO₂, SiO₂, and/or Zn or Fe generally upregulated the pro-inflammatory genes TNF-α, IL-8, and NF-κβ (Figure 3). The changes in Zn metabolism gene expression are likely due to the pro-inflammatory conditions, which have previously been shown to increase Zn absorption ( Pekarek et al., 1975; Sas et al., 1979).

Similar to Zn, Fe requires specialized protein transporters to pass biological membranes. DMT1, the primary apical intestinal enterocyte Fe transporter, DcytB, Fe reductase, and Ferroportin, the primary basolateral intestinal enterocyte Fe transporter were generally upregulated by TiO₂, SiO₂, and/or Zn or Fe exposure. In previous intra amniotic administration studies, treatment groups with improved Fe status showed a downregulation of Fe metabolism genes (Tako et al., 2011). These results suggested that increased Fe bioavailability led to Fe sufficient conditions, resulting in an environment where additional Fe transporters were not required as a compensatory mechanism. Although ⁵⁸Fe levels were unchanged in the bloodstream of NP-exposed chicks (Figure 2), the upregulation of Fe metabolism genes suggests that NP exposure could lead to Fe insufficient luminal conditions. The decrease in Fe metabolism genes with in ovo administration of NP mirrors previous results with TiO₂ and SiO₂ in vitro (Guo et al., 2017; Guo et al., 2018).

The intestinal functional proteins that were analyzed in the current study, were sucrase-isomaltase (SI) which is an intestinal BBM glucosidase (Gericke et al., 2016), and sodium glucose transporter 1 (SGLT1) which is an intestinal enterocyte apical glucose transporter (Gorboulev et al., 2012). The gene expression of these BBM functional genes was used as a biomarker of tissue functionality, including BBM digestive and absorptive capabilities (Reed et al., 2015; Uni et al., 2003). SI and SGLT1 were significantly upregulated in the presence of TiO₂, SiO₂ with Fe or Zn (Figure 3). Prior to hatching, Gallus gallus embryos have low intestinal expression of SI and SGLT1 and a limited ability to digest and absorb nutrients (Uni and Ferket, 2003). Feeding immediately post hatch is critical for broiler chicken intestinal development (Noy and Sklan, 1998), and nutrients that are supplied in ovo can support intestinal development during embryonic development (Kadam et al., 2013). An increase in BBM functional gene expression suggests an improvement in Gallus gallus hatchling intestinal development, digestive capabilities, and the potential for increased absorption of nutrients as Fe or Zn. The upregulation of BBM functional genes with exposure to NP could support intestinal development, or could be a compensatory mechanism to improve absorption due to intestinal damage.

The intestinal microbiota are becoming recognized as a key aspect in human and animal nutrition. A number of recent studies have also shown that adult and embryonic Gallus gallus cecal microbial populations are a useful indicator of gut and host overall health (Hou and Tako, 2018). Frequent exposure to metal oxide NP that are incorporated into consumer products or released into the environment has led to a number of studies focused on understanding how NP can change the composition and function of intestinal microbial communities. Many of these studies investigating the effects of metal oxide NP on individual species or communities of bacteria have reported conflicting results, most likely due to complex experimental factors such as such as the model microbiome community or animal model used, the experimental microenvironment, or the differing physicochemical properties of the NP (54).

Previous in vitro studies have shown that TiO₂ NP display photocatalytic-mediated toxicity in both Gram positive and Gram negative bacterial species (Qui et al., 2018), and can inhibit the growth of commensal intestinal bacteria such as lactobacillus in the absence of light (Liu et al., 2016). In vitro microbiome models or in vivo studies with exposure to TiO₂ NP have shown that food grade TiO₂ NP can significantly shift the abundance of bacterial species (specifically Bacteroides and Clostridium) (Dudefoi et al, 2017). Although these compositional changes did not result in measurable physiological changes to the GI tract in one study, a separate study showed that TiO₂ NP did alter the release of bacterial metabolites in vivo, promote biofilm formation of commensal bacterial in vitro, alter intestinal epithelial mucin gene expression, and promote inflammatory colon conditions (Dudefoi et al., 2017; Pinget et al., 2019). A third study found no obvious disturbance of gut microbiota in mice that ingested TiO₂ NP for 7 days (Chen et al., 2017). SiO₂ NP have been shown to induce reactive oxygen species (ROS) and to adsorbed to bacterial membranes, both of which caused bacterial toxicity in vitro (Zhang et al., 2013; Jiang et al., 2009). SiO₂ NP fed to mice for 7 days increased colon pro-inflammatory cytokine expression, and increased the abundance of Lactobacillus (Chen et al., 2017).

As shown in Figure 4, doses of medium TiO₂ + Zn, high TiO₂ + Zn, and medium SiO₂ + Zn significantly increased the relative abundance of the probiotics Bifidobacterium or Lactobacillus. Water + Zn, medium TiO₂ + Fe, high TiO₂ + Fe, medium SiO₂ + Fe, and medium doses of TiO₂ NP or SiO₂ NP significantly decreased the relative abundances of Bifidobacterium or Lactobacillus, however. The potentially opportunistic Clostridium or E. coli were significantly increased following exposure to medium SiO₂ + Zn or SiO₂ NP only. Previous work with TiO₂ and SiO₂ NP have shown that these particles possess antibacterial properties (Qui et al., 2018), and Zn is also a known, concentration-dependent antimicrobial (Seiler et al., 2012). Gram positive bacteria, such as Bifidobacterium, Lactobacillus, and Clostridium have been shown to be more susceptible to metal oxide NP toxicity; bacterial NP toxicity is thought to result from oxidative stress via the generation of ROS on surfaces of the nanoparticles (Azam et al., 2012; Dizaj et al., 2014). In the current study, gene expression results (Figure 3) showed an increase in pro-inflammatory gene expression, which may also be due to ROS generation. The data additionally show that combinations of TiO₂, SiO₂, and Fe or Zn resulted in a significant increase in cecal bacterial populations. The combined antibacterial and inflammatory effects of the NP and heavy metals may foster “blooms” of otherwise low-abundance, opportunistic bacteria (Zeng et al., 2016). Enterobacterial blooms, for example, are commonly observed in gut dysbiosis related to inflammation in the gut (Zeng et al., 2016).

5. Conclusions

The results of this study show that NP type, dose, and the presence or absence of minerals can affect broiler chicken intestinal functional gene expression, inflammatory gene expression, cecum bacterial content and abundance. Overall, the data suggest that TiO₂ and SiO₂ NP have the potential to negatively affect intestinal function and health. The results obtained with intra-amniotic administration of NP and minerals also overlaps well with previous in vitro studies with the same types and concentrations of NP, which validates the use of these methods as a relatively fast and low-cost screening method for NP-biological interactions in the gut following ingestion.

• NP type, dose, and the presence or absence of minerals altered mineral transporter, BBM functional, and pro-inflammatory gene expression.

• Metal oxide NP altered the abundance of intestinal bacterial populations.

• TiO2 and SiO2 NP have the potential to negatively affect intestinal function and health.

Abbreviations:

BBM

Brush Border Membrane

DMT1

Divalent Metal Transporter 1

Fe

Iron

SGLT1

Sodium Glucose Transporter 1

SI

Sucrose Isomaltase

SiO₂

Silicon Dioxide

TiO₂

Titanium Oxide

Zn

Zinc

ZnO

Zinc Oxide