Diao et al. J Nanobiotechnol (2021) 19:174 https://doi.org/10.1186/s12951-021-00916-2 Journal of Nanobiotechnology

RESEARCH Open Access Silicon dioxide nanoparticles induced neurobehavioral impairments by disrupting microbiota–gut–brain axis Jun Diao1†, Yinyin Xia1†, Xuejun Jiang2†, Jingfu Qiu3, Shuqun Cheng1, Junhao Su3, Xinhao Duan3, Min Gao3, Xia Qin4, Jun Zhang5, Jingchuan Fan5, Zhen Zou5,6* and Chengzhi Chen1,6*

Abstract

Background: Silicon dioxide nanoparticles ­(SiO2NPs) are widely used as additive in the food industry with contro- versial health risk. Gut microbiota is a new and hot topic in the feld of nanotoxicity. It also contributes a novel and insightful view to understand the potential health risk of food-grade SiO­ 2NPs in children, who are susceptible to the toxic efects of nanoparticles.

Methods: In current study, the young mice were orally administrated with vehicle or ­SiO2NPs solution for 28 days. The efects of ­SiO2NPs on the gut microbiota were detected by 16S ribosomal RNA (rRNA) gene sequencing, and the neurobehavioral functions were evaluated by open feld test and Morris water maze. The level of infammation, tissue integrity of gut and the classical indicators involved in gut–brain, gut–liver and gut–lung axis were all assessed.

Results: Our results demonstrated that ­SiO2NPs signifcantly caused the spatial learning and memory impair- ments and locomotor inhibition. Although ­SiO2NPs did not trigger evident intestinal or neuronal infammation, they remarkably damaged the tissue integrity. The microbial diversity within the gut was unexpectedly enhanced in ­SiO2NPs-treated mice, mainly manifested by the increased abundances of Firmicutes and Patescibacteria. Intriguingly, we demonstrated for the frst time that the neurobehavioral impairments and brain damages induced by SiO­ 2NPs might be distinctively associated with the disruption of gut–brain axis by specifc chemical substances originated from gut, such as Vipr1 and Sstr2. Unapparent changes in liver or lung tissues further suggested the absence of gut– liver axis or gut–lung axis regulation upon oral ­SiO2NPs exposure.

Conclusion: This study provides a novel idea that the ­SiO2NPs induced neurotoxic efects may occur through dis- tinctive gut–brain axis, showing no signifcant impact on either gut–lung axis or gut–liver axis. These fndings raise the exciting prospect that maintenance and coordination of gastrointestinal functions may be critical for protection against the neurotoxicity of infant foodborne ­SiO2NPs. Keywords: Silicon dioxide nanoparticles, Gut–brain axis, Gut microbiota, Neurobehavioral impairments

Introduction *Correspondence: [email protected]; [email protected] †Jun Diao, Yinyin Xia and Xuejun Jiang contributed equally to this work Silicon dioxide (SiO­ 2) is a natural chemical that most 1 Department of Occupational and Environmental Health, School widely used in structural materials, microelectronics, and of Public Health and Management, Chongqing Medical University, Chongqing 400016, People’s Republic of China as components in the food industry for various applica- 5 Molecular Biology Laboratory of Respiratory Disease, Institute of Life tions [1–4]. Te amorphous form of SiO­ 2, also known Sciences, Chongqing Medical University, Chongqing 400016, People’s as synthetic amorphous silica (SAS), is authorized as a Republic of China Full list of author information is available at the end of the article food additive coded E551 [3, 5]. It usually functions as

© The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​ iveco​ ​ mmons.org/​ licen​ ses/​ by/4.​ 0/​ . The Creative Commons Public Domain Dedication waiver (http://creat​ iveco​ mmons.​ org/​ publi​ cdoma​ in/​ ​ zero/1.0/​ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Diao et al. J Nanobiotechnol (2021) 19:174 Page 2 of 20

an anti-caking agent to maintain free-fowing properties [26]. In cultured neuroblastoma SH-SY5Y cells, treat- of powdery products, preserve food color during stor- ment of ­SiO2NPs causes deleterious efects on tau struc- age, and carry fragrances or favors. It is noteworthy that ture and cell integrity [27]. Tese results are verifed in ­SiO2 can exist in various particle sizes in the food addi- other type of neuronal cells, showing that exposure to tive depending on the manufacturing process [6]. For ­SiO2NPs induces pathological signs of Alzheimer’s dis- instance, the sizes of SiO­ 2 in the E551 contains primary ease, such as changed expression of amyloid precursor particles in the nano-size range, and most of the particles protein, increased phosphorylation of tau in neuro2a are normally greater than 100 nm. However, under the neuroblastoma cells [28]. Te evidence obtained from conditions present in the gastrointestinal tract, ­SiO2 par- cultured cells and animals together suggest that expo- ticles are capable of clumping together and subsequently sure to ­SiO2NPs can trigger neurotoxic efects and may degrading into small size particles, many of which are be considered as a risk factor facilitating the neurological nanoparticles with size less than 100 nm, approximately disorders onset and/or progression. Terefore, currently, 10–50 nm in size [3, 7, 8]. Te nano-size ­SiO2 in the E551 increasing concerns have been raised over the potential may have a completely diferent infuence on the uptake neurotoxicity of ­SiO2NPs on human health due to their and distribution of ­SiO2 within the body [4]. Although extensively use as food additive. However, whether oral ­SiO2 is used as a classical food additive for a long history exposure to SiO­ 2NPs may induce neurotoxic efects and without any detrimental health efects, the safety con- their underlying mechanisms remain largely unknown. cerns should be still paid on the potential health risks for Oral uptake is considered as the major route of expo- humans associated with nanoparticles in E551. sure to SiO­ 2NPs for general population [3, 6, 11]. Fol- Te potential toxic efects of nano-size ­SiO2 have been lowing ingestion, SiO­ 2NPs interact with the complex extensively studied for years [3, 4, 9, 10]. Despite these gastrointestinal microenvironment. Tey are able to nanoparticles are generally considered less harmful in accumulate in the gastrointestinal tract as a result of daily the past decade, excessive exposure to ­SiO2 nanoparticles consumption and afect the gut microbiota and mucus ­(SiO2NPs) has been recently reported to cause injuries layer directly [6]. A portion of ­SiO2NPs possibly show a to cells, tissues, and organs in vitro and in vivo [10–12]. signifcant impact on the enteric neurons when trans- Most of in vitro studies demonstrate that ­SiO2NPs are locating through the epithelial barrier, before reaching able to induce size-, shape- and dose-dependent cyto- systemic circulation [6]. It has been noted that the gut toxic efects in cultured human cell lines, such as glio- microbiota plays a critical role in the regulation of brain blastoma cells [13], A549 cells [14] and BEAS-2B cells functions as indispensable substrate for host health. Te [15] etc. Evidence from scarce in vivo investigations imbalance of intestinal microbial ecosystem may strongly illustrate that consumption of very large quantities of contribute to the neurobehavioral impairments via bidi- ­SiO2NPs can result in adverse efects in liver, kidney, rectional gut–brain communication [29]. Many neuro- and lung of animals [11, 16, 17]. Notably, if SiO­ 2NPs are logical diseases are closely related with changes along the injected directly into the bloodstream, even small quanti- microbiota–gut–brain axis, such as Alzheimer’s disease ties are harmful for the experimental animals [3, 11, 18, [30] and Parkinson’s disease [31]. However, whether oral 19]. Inhalation exposure to ­SiO2NPs is highly associated exposure to ­SiO2NPs facilitates the onset of neurologi- with human health, especially occurs in occupational cal disorders via microbiota–gut–brain axis has not been environment [20]. Moreover, either inhaled or ingested reported yet. ­SiO2NPs can penetrate cells and interact with cellular Terefore, in this study, we aimed to verify if oral expo- membrane or organelles to trigger mammalian cell death sure of ­SiO2NPs induced neurobehavioral impairments by induction of oxidative stress, endoplasmic reticulum through disruption of microbiota–gut–brain axis. Since stress and apoptosis [21–23]. children were usually susceptible to the neurotoxic efects Studies have demonstrated that exposure to SiO­ 2NPs of exogenous chemicals, young animals were subjected leads to the observable efects on the alterations of behav- to intragastric administration of SiO­ 2NPs for 28 days. ioral phenotypes in zebrafsh, such as disturbance on Herein, our data demonstrated for the frst time that, light/dark preference, abnormal exploratory behavior and the neurobehavioral impairments induced by ­SiO2NPs defcits in memory [24]. Even at low dose level, ­SiO2NPs treatment possibly occurred by distinctively damaged have also been shown to increase the apoptotic cells in the microbiota–gut–brain axis but showed no signifcant the central nervous system of zebrafsh embryos and dis- impact on either gut–liver or gut–lung axis. Tese fnd- rupt the axonal integrity [25]. Similarly, after treating of ings will provide us a novel insight that dietary exposure ­SiO2NPs by intranasal instillation for more than 30 days, to ­SiO2NPs may potentially disturb the intricate dialogue exposed mice exhibited obvious mood dysfunction, cog- between gut and brain functions, hence resulting in the nitive impairment and neurodegeneration-like pathology neurotoxicity. Te health risk of foodborne ­SiO2NPs Diao et al. J Nanobiotechnol (2021) 19:174 Page 3 of 20

should be reconsidered, especially for the infants and was 590–690 ­m2/g and the purity was 99.5% accord- children, who are more sensitive to the neurotoxic efects ing to the instruction from manufacture. Te morphol- of ­SiO2NPs than adults. ogy of SiO­ 2NPs was observed by a transmission electron microscope (TEM) (Hitachi-7500, Hitachi, Ltd, Tokyo, Materials and methods Japan). A total of 150 nanoparticles were counted and Chemical and reagents the average size of nanoparticle was calculated. Te aver- SiO2NPs nano-powder were obtained from Sigma age particle size of SiO­ 2NPs used in current study was Aldrich Chemical Co. (MO, USA, Cat Number: 637246). (27 ± 12.926) nm. Te feld emission scanning electron 4ʹ,6-Diamidino-2-phenylindole (DAPI) was from Bey- microscopy (Hitachi-SU8010, Hitachi, Ltd, Tokyo, Japan) otime Institute of Biotechnology (Shanghai, China). with energy-dispersive spectroscopy (Oxford X-MAN Hematoxylin–eosin, Toluidine blue O (TBO) and Alcian 50) (FE-SEM/EDS) was used to determine the chemical blue periodic acid schif (AB-PAS) staining kits were elemental composition of SiO­ 2NPs. Te characteristics all purchased from Solarbio Science & Technology Co., of ­SiO2NPs were shown in Fig. 1A–D. According to the Ltd. (Beijing, China). Antibodies against Hu protein C/D results of previous study, the hydrodynamic diameter of (HuC/D), neuron-specifc class III beta-tubulin (TuJ1) ­SiO2NPs used in this study was 471 ± 169 nm, the PDI and anti-lysozyme were purchased from Abcam Co., was 0.48 and the zeta-potential was 31.3 1.8 mV [32]. − ± 2 (Cambridge, UK). Mouse secretory immunoglobulin A Te surface area for ­SiO2NPs was 500–840 ­m /g detected (sIgA) and mouse diamine oxidase (DAO) enzyme-linked by Brunauer–Emmett–Teller nitrogen adsorption [32, immunosorbent assay (ELISA) kits were from Cusabio 33]. Biotech Co. Ltd. (Wuhan, China). Te commercial kits of determining sucrase, lactase, maltase, alkaline phos- Rationale for ­SiO2NPs’ dose selection and treatment phatase, γ-glutamyl transferase, superoxide dismutase Te dose of ­SiO2NPs used in this study were calcu- (SOD) activities and malondialdehyde (MDA) contents lated according to the Chinese Standard for Food Addi- were obtained from Nanjing Jiancheng Bioengineering tives (GB2760-2014). Te calculation methods were Institute Co., Ltd. (Nanjing, China). described in detail as follows. It declared that the upper limit of SiO­ 2 for using in food additives was 20 g/kg in Animals the national standard of GB2760-2014. Herein, we A total of 20 healthy male specifc pathogen-free assumed that the weight of each bag of infant food was C57BL/6J mice, aged 4 weeks and weighted 8–12 g, were 80 g and the weight of a 1-year-old infant was 10 kg. If provided by Experimental Animal Center of Chongqing each infant was fed with one bag of food each morning Medical University [Chongqing, China, License num- and evening twice a day. Te total daily intake of SiO­ 2 for bers: SCXK(Yu)2018-0003]. Animals were housed under each infant was 0.32 g/kg [(80 g × 2 ÷ 1000 g) × 20 g ÷ 1 constant conditions with room temperature at 23 ± 1 °C 0 kg = 0.32 g/kg]. Since the equivalent dose conversion and humidity at 55 ± 10%. Tey were all maintained in a from mouse to human was near 9.1-fold [34], the dos- standard 12 h:12 h light–dark cycle. Te mice were free age of SiO­ 2 was calculated from the following formula: access to laboratory mouse chow and tap water. Te 0.32 g/kg × 9.1 = 2.912 g/kg ≈ 3 g/kg. Tus, animals animals were randomly assigned into vehicle group and were treated with either vehicle solution or with 3 g/kg ­SiO2NPs-treated group according to the website https://​ ­SiO2NPs suspension solution once a day via intragastric www.​rando​mizer.​org. Each group had ten animals. Te administration between 9:30 A.M. and 10:30 A.M. and use of animals in experimental research were approved lasted for 28 days. Te number of cells in the bronchoal- by the Institutional Animal Care and Use Committee of veolar lavage fuid were counted using TC20­ TM Auto- Chongqing Medical University, and all eforts were made mated Cell Counter (Bio-Rad, Hercules, CA, USA). to minimize the pain or distress experienced by animals. Tis study received ethical approval from Te Ethical Hematoxylin–eosin staining Committee of Chongqing Medical University. Hematoxylin–eosin (H&E) staining was carried out according to the protocols described previously [35]. In Characterization and preparation of ­SiO2NPs brief, after designed treatment, the animals were sacri- Te powder of ­SiO2NPs were diluted in sterile physi- fced by cervical dislocation under anesthesia. Te intes- ological saline solution and sonicated with an ultrasonic tine and brain tissues were quickly dissected and fxed in cleaner (SB-5200DT, Ningbo Scientz Biotechnology Co., fresh prepared 4% paraformaldehyde. Te sections were Ltd, Ningbo, China) on ice at 20% of maximum ampli- subsequently dewaxed in xylene and dehydrated by etha- tude for 20 min. Te suspended solution of ­SiO2NPs nol, followed by staining with hematoxylin and eosin. was freshly prepared for each time use. Te surface area After mounting with neutral balsam, the sections were Diao et al. J Nanobiotechnol (2021) 19:174 Page 4 of 20

A B

Mean=27.00 nm Std.Dev=12.926 25 N=150

e 20

15

Percentag 10

5

0 10 20 30 40 50 60 nm Size

CD

Fig. 1 Characteristics of ­SiO2NPs used in this study. A Transmission electron microscope (TEM) image of SiO­ 2NPs was shown. Scale bar, 200 nm. B The average size of ­SiO NPs was around 27 12.926 nm calculated from 150 nanoparticles in TEM images. C Scanning electron microscope image 2 ± of ­SiO2NPs. Scale bar, 500 nm. D Chemical composition of ­SiO2 was detected by FE-SEM/EDS. Scale bar, 5 μm

observed under an Olympus light microscope (IX53, mounted with neutral balsam. Finally, the sections were Tokyo, Japan). observed under a light microscope (Olympus, IX53, Tokyo, Japan). Alcian blue periodic acid schif (AB‑PAS) staining Goblet cells were stained by AB-PAS according to Toluidine blue staining the procedures described previously [36]. Briefy, Mast cells were stained by Toludine blue according to after treatment, the intestine tissue was collected and the procedures reported previously [37]. In brief, the immersed into fresh prepared 4% paraformaldehyde. intestine tissue was dissected and fxed in fresh pre- Te sections were then dehydrated in the ethanol fol- pared 4% paraformaldehyde. Te sections were then lowed by staining with alcian blue solution, Schif Rea- deparafnized by xylene, dehydrated in the ethanol fol- gent and hematoxylin, respectively. Te sections were lowed by staining with toluidine blue solution. After then subjected to the ethanol dehydration process and mounting with the neutral balsam, the sections were Diao et al. J Nanobiotechnol (2021) 19:174 Page 5 of 20

observed under an Olympus light microscope (IX53, shown as follows, 95 °C for 2 min, followed by amplifca- Tokyo, Japan). tion in 40 cycles of 95 °C for 5 s, 15 s at 60 °C and 20 s at 72 °C, then 65 °C and 95 °C for 5 s. Te relative mRNA Immunofuorescence assay expressions of target genes were normalized by the rela- Immunofuorescence assay was performed according to tive amount of β-actin mRNA. the protocols reported previously [38]. Briefy, the intes- tine tissue was collected immediately and fxed in fresh Measurement of sucrase, lactase, maltase, alkaline prepared 4% paraformaldehyde. Hereafter, the sections phosphatase, γ‑glutamyl transferase, SOD activities were washed with phosphate bufer solution and blocked and MDA contents in the normal serum for 30 min. After incubation with Te tissues of intestine, lung and liver were quickly col- the primary antibodies against TuJ1 (1:250) and HuC/D lected from each mouse at the end of treatment and (1:250) overnight at 4 °C, the sections were then incu- stored at − 80 °C before determination. Te tissues bated with fuorescent dye-conjugated secondary anti- were then homogenized in 0.9% sodium chloride solu- bodies for 1 h. Finally, the sections were stained with tion using an electric glass homogenizer. Te activi- DAPI, and observed under a fuorescence microscope ties of sucrase, lactase, maltase, alkaline phosphatase, (Olympus, IX53, Tokyo, Japan). Te fuorescence inten- γ-glutamyl transferase, SOD and MDA contents were all sity was measured by using Image-Pro Plus 6.0 software determined using commercial kits. (Bethesda, MD, USA). Enzyme linked immunosorbent assay (ELISA) Immunohistochemistry assay Te ELISA assay was conducted according to the proto- Immunohistochemistry assay was carried out accord- cols described previously [35]. Briefy, the intestine and ing to the procedures reported previously [39]. In brief, lung tissues were obtained and stored at − 80 °C until the intestine tissue was collected and immersed into the assays were performed. Mouse sIgA and DAO ELISA fresh prepared 4% paraformaldehyde. Te sections were kits were placed in room temperature for at least 20 min deparafnized in xylene and subjected to conventional before use. Te samples were added into each well fol- gradient ethanol dehydration. Subsequently, the fresh lowed by incubation for 2 h at 37 °C. Biotin-antibody was 3% hydrogen peroxide was used to inhibit the activity of hereafter added after removing the liquid of each well, endogenous tissue peroxidase. After washing thoroughly and incubated at 37 °C for 1 h. After washing with bufer, with phosphate bufer solution, the sections were incu- the horseradish peroxidase-conjugate was added into bated in the block solution containing serum for 30 min each well and incubated for additional 1 h at 37 °C. Sub- followed by incubation of anti-lysozyme primary anti- sequently, the chromogenic substrate was incubated for body (1:1000) at 4 °C overnight. Next day, the sections 15 min at 37 °C in the dark for color development. Te were incubated with biotinylated secondary antibody absorbance was measured at the wavelength of 450 nm and horseradish enzyme-labeled streptavidin for 10 min, by a microplate reader (Termo Fisher Scientifc Inc., respectively. Te positive reactions in the tissues were Waltham, MA, USA). visualized by a freshly prepared diaminobenzidine. Te sections were then observed under a light microscope 16S ribosomal RNA gene sequencing (Olympus, IX53, Tokyo, Japan). Te 16S ribosomal RNA (rRNA) gene sequencing was performed according to the protocols described previ- Quantitative PCR assay ously [41]. Briefy, at the end of 28 days’ treatment, the Quantitative PCR assay was conducted according to the fecal samples of animals were collected under sterile con- protocols reported previously [40]. Briefy, the tissues of ditions and stored at − 80 °C before use. Total genomic cortex, lung, intestine, liver were quickly collected and DNA was extracted from samples and verifed by 1% aga- stored at − 80 °C. Total RNA was isolated by using TRizol rose gel electrophoresis. Te bacterial 16S rRNA were method. Complementary DNA (cDNA) was synthesized amplifed by the forward and reverse primers designed using Perfect Real Time PrimeScript™ RT Master Mix. by adding a barcode to primer. Te polymerase chain Quantitative RT-PCR was performed with TBGREEN reaction (PCR) amplifcation reaction was performed on Premix Ex Taq™ II (TliRNaseH Plus) on the CFX Con- the ABI GeneAmp 9700 (Termal cyclers from Applied nect™ Real-Time System (Bio-Rad, Hercules, CA, USA). Biosystems, CA, USA) with TransStart Fastpfu DNA Te specifc primers were synthesized by Sangon Bio- Polymerase. Te PCR products were excised from aga- tech, Co., Ltd. (Shanghai, China), and the sequences of rose and purifed by AxyPrep DNA Gel Extraction Kit target genes were listed in Additional fle 2: Table S1. Te (Axygen Biosciences, CA, USA). Subsequently, the PCR amplifcations were carried out according the conditions products were quantifed by QuantiFluor™-ST Blue Diao et al. J Nanobiotechnol (2021) 19:174 Page 6 of 20

Fluorescence Quantifcation System (Promega Co., WI, activity of each mouse in the apparatus was observed USA). Te MiSeq library was constructed for preparation for 5 min. Te apparatus was carefully cleaned with 75% of the fragment DNA by TruSeq™ DNA Sample Prep Kit. ethanol between each trial. During the test, the total dis- At last, the raw sequence reads were obtained by Illumina tance, distance moved in center and central square dura- MiSeq platform at Majorbio Bio Tech Co. Ltd (Shanghai, tion were all obtained from the tracking system with China). video camera mounted above the apparatus.

16S ribosomal RNA gene sequencing analysis Statistical analysis Te raw sequence reads were clustered into opera- All the experimental data were reported as mean ± stand- tional taxonomic units (OTUs) with 97% similarity on ard error of the mean (S.E.M). Independent student-t test Majorbio’s cloud website at https://​cloud.​major​bio.​com or non-parametric Mann–Whitney U test were applied (Majorbio, Shanghai, China) based on the Usearch soft- to detect the statistically signifcant diferences between ware programs (version 7.1). Te alpha diversity indi- two groups. Te repeated-measure analysis of variance ces, Shannon, Simpson, Chao, Ace and the observed (ANOVA) was use to assess the statistical signifcance on species were calculated by Mothur software programs the escape latency, swim distance and swim speed in the (version v.1.30.1). β-diversity was obtained by principal Morris water maze. All the statistical analysis was carried component analysis (PCA) and partial least squares dis- out using GraphPad Prism 8.0 (GraphPad Software, La criminant analysis (PLS-DA) in the R software. Te dom- Jolla, CA), and the statistical signifcance level was set at inant phylotypes responsible for signifcant diferences p value less than 0.05. between two groups were assessed by linear discriminant analysis efect size (LEfSe). Linear discriminant analysis Results (LDA) score was set as 2.0. Te enterotypes of microbiota Oral exposure to ­SiO2NPs led to spatial learning were analyzed in the ade4, cluster and clustersim pack- and memory impairment and locomotor inhibition ages of R software. Te relationship between sample and Morris water maze is a widely used and well-validated microbial community was visualized by the Circos-0.67-7 neurobehavioral test for the assessment of spatial learn- software. Phylogenetic tree on the Genus level was boot- ing and memory ability [43]. In this study, Morris water strapped using Mega software (version 10.0). Te net- maze was carried out after treated with ­SiO2NPs for work analysis on the OTU level and network correlation 28 days. As shown in Fig. 2A, the results demonstrated analysis were performed on the software of NetworkX. that the escape latencies of mice exposed to ­SiO2NPs Te PICRUSt was used to predict the functional compo- were signifcantly higher than those of vehicle controls, sition of microbial community. indicating that the mice took more time to make frst contact with the hidden platform. Similarly, the swim Morris water maze distances of ­SiO2NPs-treated animals were also much Morris water maze was used for the assessment of spa- longer than those of vehicle controls (Fig. 2B). Tese data tial learning and memory function [42]. Briefy, Morris suggest that the spatial learning function of animal is water maze was performed after indicated treatment. Te remarkably impaired by ­SiO2NPs. In the probe test, the circular pool was divided into four equal quadrants and results revealed that the number of platform crossings flled with water. Te mice were put into the water maze was obviously reduced in ­SiO2NPs group as compared for adaptation with a 60 s free swim without the platform with vehicle group (Fig. 2C). Te time spent in target before trials. Te hidden platform test was conducted for quadrant was also slightly decreased in SiO­ 2NPs-treated 4 consecutive days. Te swim path of each mouse was mice, but it did not reach the statistical signifcance recorded by the video camera mounted above the pool. (Fig. 2D). No signifcant changes were observed on the Te swim distance, escapes latency and the swim speed swim speed between two groups (Fig. 2E). Representa- were all obtained from the tracking system. Finally, the tive track maps of each group in Morris water maze were hidden platform was removed for post-training probe depicted in Fig. 2F. Tese obtained results together imply tests at the last day, the time spent in the target quadrant that oral exposure to ­SiO2NPs partially disrupt the spatial and the number of platform crossings were recorded dur- learning and memory function of mice. ing test. Open feld test is a classical test used for evaluation of locomotor activity in animal models [44]. To further Open‑feld test evaluate the efect of ­SiO2NPs on the locomotor activity, Open-feld test was applied to assess the locomotor activ- the open feld test was conducted. Te results illustrated ity of animals [42]. In brief, the mouse was placed in that the total distance in the open-feld apparatus was the center of apparatus facing the same direction. Te sharply declined in ­SiO2NPs-treated mice as compared Diao et al. J Nanobiotechnol (2021) 19:174 Page 7 of 20

A BC Vehicle Vehicle SiO NPs 150000 60 2 SiO NPs 2 5 * * 100000 * 4 40 3

2 20 50000 1 Escape latency (sec ) Swim distance (mm) 0 0 0 Number of platform crossings (n ) 1 234 12 3 4 Vehicle SiO2NPs Days Days

DE F Vehicle SiO2NPs N.S. Vehicle 25 0.4 SiO2NPs 20 0.3 15 0.2 10 0.1 Place navigation trial

5 Swim speed (m/s)

0 0.0 me spent in target quadrant (sec) Vehicle SiO NPs 1234

Ti 2 Days

Spatial probe trial GH I 1500 * 25 * 200 *

20 150 1000 15 100 10 500 50 otal distance (cm) 5 T

0 Central square duration (sec ) 0 0 Distance moved in center (cm )

Vehicle SiO2NPs Vehicle SiO2NPs Vehicle SiO2NPs J

Vehicle SiO2NPs

Fig. 2 Oral exposure to ­SiO2NPs led to spatial learning and memory impairment and locomotor inhibition. After indicated treatment, Morris water maze and open fled test were used to evaluate the spatial learning and memory function and locomotor activity of mice, respectively. Efects

of ­SiO2NPs on the escape latency were shown in (A). B Efects of ­SiO2NPs on the swim distance. C Efects of ­SiO2NPs on the number of platform crossings. D Efects of ­SiO2NPs on the time spent in target quadrant. E Efects of ­SiO2NPs on the swimming speed. F Representative track maps of vehicle mice and ­SiO2NPs-treated mice in the place navigation trial and spatial probe trial. G Efects of SiO­ 2NPs on the total distance in the open feld test. H Efects of ­SiO2NPs on the central square duration. I Efects of ­SiO2NPs on the distance moved in the center. J Representative track maps of two groups in the open feld test. Data were shown as mean S.E.M. Statistical analysis was conducted by using repeated-measure ANOVA or ± independent student-t test or Mann–Whitney test. Asterisk * indicated P < 0.05 Diao et al. J Nanobiotechnol (2021) 19:174 Page 8 of 20

with vehicle controls (Fig. 2G). Meanwhile, the central contributed to 95.16% (531/558) and 93.81% (531/566) of square duration and distance moved in center were both vehicle group and SiO­ 2NPs-treated group, respectively. notably declined in ­SiO2NPs-treated group in compari- Principal component analysis (PCA) was employed to son to control group (Fig. 2H, I). Representative track detect the dissimilarities in microbial composition on maps of each group in open feld test were depicted in OTU level. As shown in Fig. 3E, ­SiO2NPs-treated samples Fig. 2J. Tese results together specify that oral exposure were primarily concentrated on the left side, whereas the to ­SiO2NPs signifcantly inhibits the locomotor activity of samples in the vehicle control group presented mainly on animals. the right side. Partial least squares discrimination analysis (PLS-DA) further clearly distinguished ­SiO2NPs-treated Oral exposure to ­SiO2NPs caused the disturbance of gut samples from vehicle samples, indicating that there were microbiota and their associated biological functions signifcant diferences on the gut microbial community After exposure to ­SiO2NPs, the bacterial DNA in feces compositions between two groups (Fig. 3F). Two ente- were extracted and analyzed by 16S rRNA gene sequenc- rotypes were observed in this study on the family level, ing. Te sequences were clustered into OTU based on one was f_Muribaculaceae, the other one was f_Prevotel- a 97% similarity threshold. Obtained total number of laceae. No signifcant diference was shown on the enter- OTU was 593. Te coverage indices of vehicle group otype between two groups (Fig. 3G). Linear discriminant and ­SiO2NP-treated group were 0.9982 ± 0.0003 and analysis (LDA) coupled with efect size (LEfSe) estab- 0.9979 ± 0.0002. Te rarefaction curves of two groups lished 41 bacterial clades showing statistically signifcant both showed clear asymptotes (Additional fle 1: Figure and biologically consistent diferences (LDA score > 2.0) S1A). Tese results together suggest a near-complete from phylum to genus level (Fig. 3H, I). sampling of the community. α-diversity is an intuitive Circos analysis was carried out to visualize the correla- index for evaluation of microbial diversity. In this study, tions between microbiota and samples in vehicle group the results showed that the Sobs, Ace, and Chao were all and ­SiO2NPs group (Fig. 4A). Te evolutionary relation- signifcantly enhanced in SiO­ 2NPs-treated group, but no ships among bacteria on the genus level were displayed in signifcant changes were observed on two other indices, the phylogenetic tree (Fig. 4B). Network analysis was per- Simpson, and Shannon (Fig. 3A). Tese data together formed at the OTU level and shown in Fig. 4C. Te rela- imply that exposure to SiO­ 2NPs can partially afect tionships among microbial communities were displayed α-diversity of gut microbiota. in Fig. 4D. Te red line indicated the negative correlation Te percent of community abundance on phylum and the green line presented positive correlation between level in two groups was presented in Fig. 3B. In the vehi- each species at genus level. Te size of dot indicated the cle group, the dominant bacteria were Bacteroidetes degree of relevance with other species. Te PICRUSt was (62.68%), Firmicutes (28.58%), Verrucomicrobia (3.11%), further used to predict the functional composition of a Epsilonbacteraeota (2.09%), Proteobacteria (1.79%), microbial community’s metagenome. In the analysis on whereas the ­SiO2NPs group displayed a high relative the clusters of orthologous groups (COGs), the results abundance of Bacteroidetes (54.30%), followed by Firmi- revealed that the top three increased functional abun- cutes (39.17%), Epsilonbacteraeota (1.82%), Proteobacte- dances of microbial community were cytoskeleton, cell ria (1.80%) and Verrucomicrobia (1.03%). Te abundances motility and signal transduction mechanisms, and the of Firmicutes and Patescibacteria in the ­SiO2NPs-treated top three decreased functional abundances of micro- group were signifcantly higher than those in the vehicle bial community were extracellular structures, RNA pro- control group (Fig. 3C). As shown by the Venn’s diagram cessing and modifcation and chromatin structure and in Fig. 3D, two groups shared the compositional overlap dynamics (Fig. 4E). In the Kyoto Encyclopedia of Genes of 531 core microbiota. Tese overlapping phylotypes and Genomes (KEGG) function analysis, the top three

(See fgure on next page.)

Fig. 3 Oral exposure to ­SiO2NPs caused the disturbance of gut microbiota and their associated biological functions. The fecal samples of two groups were collected and subjected to 16S ribosomal RNA (rRNA) gene sequencing. A Sobs, Ace, Chao, Simpson, and Shannon were determined to assess the α-diversity of gut microbiota. Data were reported as mean S.E.M. Statistical analysis was calculated by using independent student-t ± test. Asterisk * indicated P < 0.05. B The percent of community abundance on phylum level was detected in two groups. The diferences on the community abundance on phylum level were analyzed and shown in (C). D Venn diagram showed the overlap of core microbiota between vehicle

group and ­SiO2NPs-treated group. E The dissimilarities in microbial composition on OTU level was determined by principal component analysis (PCA). F Partial least squares discrimination analysis (PLS-DA) was used to detect the gut microbial community compositions between two groups. Bacterial clades and biologically consistent diference (LDA score > 2.0) were assessed by Linear discriminant analysis (LDA) coupled with efect size (LEfSe) (H, I). Typing analysis on family level was shown in (G) Diao et al. J Nanobiotechnol (2021) 19:174 Page 9 of 20

A 3 N.S. B Vehicle

SiO2NPs

2 Community barplot analysis

Bacteroidetes Firmicutes Verrucomicrobia N.S. Epsilonbacteraeota * * * Proteobacteria Actinobacteria

ehicle Deferribacteres

V Patescibacteria 1 others Fold change to control NPs 2 SiO 0 00.2 0.40.6 0.81 Sob Ace Chao Simpson Shannon Percent of community abundance on Phylum level

C D E Vehicle Wilcoxon rank-sum test bar plot on Phylum level PCA on OTU level SiO2NPs 95% confidence intervals 25 Vehicle Bacteroidetes 0.1212 20 SiO2NPs Firmicutes * 0.03121 Vehicle 15 Verrucomicrobia 0.7333 SiO2NPs Epsilonbacteraeota 0.1304 10 Proteobacteria 0.65

Actinobacteria 0.7913 Pvalue 5

Deferribacteres 0.5705 27 531 35 0 Patescibacteria * 0.02109

Tenericutes 0.3254 -5 PC2(9.88%) Cyanobacteria 0.241 -10 unclassified_k__norank_d__Bacteria 0.1974

Chloroflexi 0.5428 -15 Spirochaetes 1 -25 -20 -15 -10 -5 0 5 10 15 0 5 10 15 20 25 30 35 40 45 50 55 60 -20 -15 -10 -5 0 5 10 15 20 PC1(13.01%) Proportions(%) Difference between proportions(%)

F G Typing analysis on Family level

0.25

SiO2NPs PLS-DA on OTU level 0.2 Vehicle

14 0.15 Vehicle 12 0.1 SiO2NPs 10 0.05 8

6 0

4 -0.05 2

0 -0.1

-2 -0.15 -4 -0.2 -6 COMP2(6.1%) -8 -0.25

-10 -0.3 Type -12 f__Muribaculaceae -0.35 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 f__Prevotellaceae COMP1(12.4%) -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 H I

LEfSe Bar SiO2NPs Vehicle p__Firmicutes c__Clostridia Vehicle o__Clostridiales f__Lachnospiraceae SiO2NPs f__Ruminococcaceae g__Peptococcus e g__Rikenella f g__Papillibacter d v : g__Anaerofustis g__Anaerovorax o a : p__Firmicutes 1 h w : f__unclassified_o__Clostridiales g___Eubacterium__nodatum_group i b : c__Clostridia n j c : o__Clostridiales x : g__unclassified_o__Clostridiales g__Candidatus_Saccharimonas 1 k l d : f__Lachnospiraceae y : g__Peptococcus f__Saccharimonadaceae m z : g__Erysipelatoclostridium 1 m e : g__Anaerovorax p__Patescibacteria g n o f : g___Eubacterium__nodatum_group a1 : g__unclassified_f__Erysipelotrichaceae o__Saccharimonadales l1 c p g : f__Ruminococcaceae b1 : g__unclassified_f__Prevotellaceae c__Saccharimonadia b q h : g__norank_f__Ruminococcaceae c1 : g__Rikenella k1 r g__UBA1819 s i : g__Caproiciproducens d1 : f__unclassified_o__Bacteroidales g__norank_f__Ruminococcaceae t j : g__UBA1819 e1 : g__unclassified_o__Bacteroidales f__Christensenellaceae u v k : g__Papillibacter f1 : g__unclassified_f__Eggerthellaceae a l : g__Ruminococcaceae_NK4A214_group g1 : o__Bifidobacteriales g__Desulfovibrio w g__Ruminiclostridium_9 x m : g__Ruminococcaceae_UCG_014 h1 : f__Bifidobacteriaceae n : g__Oscillibacter i1 : g__Bifidobacterium g__Ruminococcaceae_UCG_014 y o : g__Ruminiclostridium j1 : g__Desulfovibrio g__unclassified_f__Eggerthellaceae j1 p : g__Ruminiclostridium_9 k1 : p__Patescibacteria g__unclassified_o__Bacteroidales q : g__Ruminiclostridium_5 l1 : c__Saccharimonadia f__unclassified_o__Bacteroidales r : g__Ruminococcaceae_UCG_005 m1 : o__Saccharimonadales 1 g__unclassified_o__Clostridiales g s : g__Ruminiclostridium_6 n1 : f__Saccharimonadaceae f__unclassified_o__Clostridiales t : f__Christensenellaceae o1 : g__Candidatus_Saccharimonas 1 g__Bifidobacterium h u : f__Eubacteriaceae

o__Bifidobacteriales z f__Bifidobacteriaceae i1 a

d1 1 g__Caproiciproducens b c 1 1 g__Ruminiclostridium_6 1 f

g__Ruminococcaceae_NK4A214_group e1 g__Ruminiclostridium g__Oscillibacter g__Ruminiclostridium_5 g__Ruminococcaceae_UCG_005 g__Anaerofustis f__Eubacteriaceae g__unclassified_f__Prevotellaceae3.30000937841 g__unclassified_f__Erysipelotrichaceae3.2705946243 g__Erysipelatoclostridium 2.9941504682

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 LDA SCORE(log10) Diao et al. J Nanobiotechnol (2021) 19:174 Page 10 of 20

elevated functional abundances were KO03406 (methyl- mRNA expressions of Tjp1, Ocln, Cldn 7 were meas- accepting chemotaxis protein), KO02529 (LacI family ured. As shown in Fig. 5F, the results demonstrated transcriptional regulator), KO02003 (putative ABC trans- that the mRNA expressions of Tjp1 and Ocln were both port system ATP-binding protein), whereas the top three sharply reduced in SiO­ 2NPs-treated group as compared reduced functional abundances were KO6142 (outer with vehicle group. But no signifcant alteration was membrane protein), KO2014 (iron complex outer mem- observed on the mRNA expression of Cldn 7. Tese brane receptor protein), KO3773 (FKBP-type peptidyl- results together indicate that oral exposure to ­SiO2NPs prolyl cis–trans isomerase FklB) in the top 50 highest for 28 days can damage the integrity of tissue and cause abundances of KEGG orthology (KO) (Additional fle 1: inconspicuous infammation response in the intestine. Fig. S1B). On the KEGG pathway level 2, carbohydrate metabolism, global and overview maps and amino acid Oral exposure to ­SiO2NPs partially afected the activities metabolism were the top increased functional abun- of intestinal digestive enzymes and immune functions dances, while circulatory system and substance depend- Digestive enzymatic activity plays a critical role in main- ence were the most decreased functional abundances taining the normal microbial ecology of the gastroin- (Additional fle 1: Fig. S1C). Taken together, these results testinal system [47]. Terefore, the activities of typical suggest oral exposure to SiO­ 2NPs may lead to the distur- digestive enzymes, including sucrase, lactase, maltase, bance of gut microbiota and their associated biological alkaline phosphatase and γ-glutamyl transferase, were functions. evaluated after SiO­ 2NPs administration. Te results revealed that the activity of sucrase was dramatically Oral exposure to ­SiO NPs did not result in infammation 2 elevated in ­SiO2NPs-treated group when compared in the intestine but remarkably damage the tissue integrity with vehicle control group (Fig. 6A). However, no sig- Recently, studies have suggested a new role for gut nifcant changes were observed in other kind of diges- microbiota in the regulation of intestinal infammation tive enzymes’ activities (Fig. 6B–E). Te obtained results [45, 46]. To investigate if ­SiO2NPs-induced dysbiosis of imply that oral exposure to SiO­ 2NPs may partially afect gut microbiota causes to the infammation in the intes- the activity of intestinal digestive enzyme, but this infu- tinal tract, the mRNA expressions of Il-6 and Tnf were ence is very limited. determined. As shown in Fig. 5A, unapparent alterations Te gastrointestinal tract is a highly complex system on these two genes after treating of mice with ­SiO2NPs that has distinct functions not only in digestion, but also were spotted. Next, the tissue integrity of intestine was in immune homeostasis [10, 48]. sIgA is an essential part assessed by histopathological examination. Te obtained of intestinal barrier. It can bind to antigens and increase results in H&E staining assay illustrated that, exposure the capture of antigens, thereby strengthening the to ­SiO2NPs caused the abnormal villous shortening, immune function of intestinal barrier [49]. In response to and the tips of villi appeared ragged, irregularly shaped, exogenous damage, the gastrointestinal tract commonly or completely destroyed (Fig. 5B). In the AB-PAS stain- generates sIgA immune response [10]. In this study, the ing assay, we found that the number and size of goblet results showed that treatment of ­SiO2NPs, the levels of cells were both reduced signifcantly in ­SiO2NPs-treated sIgA were sharply declined as compared with vehicle animals (Fig. 5C). Paneth cells were further identifed controls (Fig. 6F). Since DAO is closely associated with using a specifc anti-lysozyme antibody. Similar trend microbial induction of intestinal sIgA [49, 50], the con- was observed in Paneth cells after exposure to SiO­ 2NPs. tent of DAO was further detected using ELISA. Never- Our results demonstrated that the size of the Paneth cell theless, we did not fnd any signifcant alteration in the compartment and the overall granule content were obvi- DAO content between two groups (Fig. 6G). Collectively, ously reduced in the mice treated with ­SiO2NPs (Fig. 5D). these results identify sIgA as previously unrecognized On the contrary, the number of toluidine blue-positive immune mediators of microbe–host interplay in the mast cells were increased signifcantly in response to intestine after ­SiO2NPs administration. ­SiO2NPs exposure (Fig. 5E). To further investigate the efects of SiO­ 2NPs on the intestinal tight junction, the

(See fgure on next page.)

Fig. 4 Oral exposure to ­SiO2NPs caused the disturbance of gut microbiota and their associated biological functions. A The correlations between microbiota and samples in the vehicle group and ­SiO2NPs group was detected using Circos analysis. B The phylogenetic tree was used to show the evolutionary relationships among bacteria on the genus level. Network analysis at the OTU level was shown in (C). The relationship of microbial communities each other was displayed in (D). E The clusters of orthologous groups (COGs) analysis displayed the top increased and the top decreased functional abundances of microbial community Diao et al. J Nanobiotechnol (2021) 19:174 Page 11 of 20

e A B Phylogenetic tree on Genus level

aculacea

ib 01

_Mur Bar Groups _ 0.97 norank_Muribaculaceae Muribaculum Vehicle 1 Prevotellaceae_UCG-001 0.15 SiO2NPs 0.4 unclassified_Bacteroidales norank_f 0.37 Alloprevotella otellaceae_UCG-0 Taxa v 0.3 0.07 0.15 Bacteroides unclassified_Prevotellaceae Bacteria Pre s NPs unclassified_Muribaculaceae % % % 0% 0 % 2 0% 10% 20 bacillu 1 3 Parabacteroides 4 50 60 70% 80% SiO 90% Alistipes Lacto 0.96 0k 0.91 0.22 Rikenella 0k 0.51 Rikenellaceae_RC9_gut_group Lachnospiraceae_NK4A136_group Odoribacter 0.19 0k Parasutterella Alloprevotella Candidatus_Saccharimonas s 0.12 0.9 Enterorhabdus Bacteroide 0.2 0.59 unclassified_Atopobiaceae Bifidobacterium others 0.2 Helicobacter Dubosiella 0.12 Mucispirillum 0.84 norank_Desulfovibrionaceae 0.15 unclassified_f__Ruminococcaceae Desulfovibrio Lactobacillus Muribaculum 0.78 0.64 Dubosiella Akkerm 0.54 1 Allobaculum ansia Faecalibaculum unclassi 0.27 fied_f__L Quinella He Akkermansia licob achnospirac norank_f__Lachnospiraceaeacte 0.05 Lachnospiraceae_NK4A136_group r 0.06

Al eae 0.45 Blautia Vehicle isti Ty pes 0.06 Lachnoclostridium zzerel Rikenell 0.030.24 unclassified_Lachnospiraceae P norank_Lachnospiraceae a la_ 0.11 Parasutterra ella 3 0.05 Roseburia Odoribacterbacte a 0.3 Roseburia Lachnospiraceae_UCG-001

R 0.83

cterium]_copros

[Euba Riken umi r [Eubacterium]_xylanophilum_group Rhodococcus oid

Bifidobacteriu Mu

GCA-900066575 norank_f__Ruminococcaceae GCA-900066575

A Quinella e

F norank n s

6

ASF35 unclas Enterorhabdu c

l Ruminococcaceae_UCG-014

[Eub Blautia D Candidatus i lobaculum

a is e c 0.95 Tyzzerella_3 ec l e pirillumll o

a 0.22 ASF356

s s a ceae t

u ri

acterium]_xylanophilum_group

m lu libacu Candidatus_Arthromitus l _ d

f s

o ium_9 ified_o__Bacteroidalef__Desulfovis unclassified_Ruminococcaceae

v _ 0.55

i

b RC9_g

m 0.82 Caproiciproducens

r i _ 0.87

o Ruminiclostridium_5

S s

accharim

t Anaerotruncus

p nes_grou anolige u t_group 0.75 Ruminiclostridium_9

b 0.98 norank_Ruminococcaceae rionacea 0.81 Oscillibacter 0.15 onas Ruminiclostridium 0.64 Ruminococcaceae_UCG-014

e norank_Clostridiales_vadinBB60_group

0 20 40 60 80 100 120 140 160 180 200 0k 20k 40k 60k 80k100kk120kk140k C D

SiO NPs Network analysis on OTU level 2 Vehicle p__Firmicutes p__Bacteroidetes p__Proteobacteria species(OTU) p__Patescibacteria g__Mucispirillum p__Actinobacteria p__Deferribacteres p__Verrucomicrobia OTU325 g__unclassified_f__Ruminococcaceae OTU493 p__Epsilonbacteraeota OTU83 g__Caproiciproducens g__Rikenella OTU245 OTU480 g__unclassified_f__Atopobiaceae OTU19 g__Desulfovibrio OTU487 OTU551 OTU492 g__unclassified_o__Bacteroidales OTU9 OTU348 g__Tyzzerella_3 OTU443 OTU281 g__norank_f__Ruminococcaceae g__Rikenellaceae_RC9_gut_group OTU204 OTU457 OTU360 OTU561 g__unclassified_f__Prevotellaceae OTU504 g__Parasutterella g__Ruminiclostridium OTU26 OTU428 g__Ruminiclostridium_9 OTU299 OTU384 OTU422 OTU402 g__Ruminiclostridium_5 g__norank_f__Muribaculaceae OTU313 OTU588 OTU488 g__Candidatus_Saccharimonas OTU14 OTU146 g__Candidatus_Arthromitus OTU439 g__[Eubacterium]_xylanophilum_group OTU169 OTU451 OTU304 OTU552 g__Dubosiella OTU347 OTU429 g__Ruminococcaceae_UCG-014 OTU104 OTU414 g__GCA-900066575 OTU484 OTU497 OTU509 OTU177 g__Oscillibacter OTU62 OTU283 OTU534 g__Prevotellaceae_UCG-001 g__unclassified_f__Muribaculaceae OTU273 g__norank_f__Clostridiales_vadinBB60_group g__Lachnospiraceae_UCG-001 OTU366 OTU306 g__Enterorhabdus OTU361 g__Roseburia OTU499 OTU545 OTU52 g__Lachnospiraceae_NK4A136_group OTU441 OTU75 g__Lachnoclostridium OTU5 g__norank_f__Desulfovibrionaceae OTU421 OTU471 OTU506 g__Bacteroides OTU495 OTU292 g__Allobaculum OTU15 OTU543 OTU486 OTU184OTU167 g__unclassified_f__Lachnospiraceae g__Alistipes OTU160OTU558 g__Quinella OTU496 OTU463 OTU587 OTU100 OTU31 OTU219 OTU490 OTU82 OTU397 OTU483 g__Helicobacter g__ASF356 g__Parabacteroides OTU4 OTU113 OTU510 OTU394 OTU498 OTU138 OTU355 g__norank_f__Lachnospiraceae OTU262 g__Faecalibaculum OTU30 OTU522 g__Alloprevotella OTU501 OTU556 OTU271 OTU161 OTU110 OTU560 OTU401 g__Odoribacter OTU578 OTU445 g__Anaerotruncus OTU10 OTU35 OTU553 OTU87 OTU338 g__Akkermansia OTU500 OTU39 OTU90 OTU58 OTU548 OTU155 OTU101 OTU532 OTU317 OTU3 OTU209 OTU291 OTU502

E

COG function classification

A : RNA processing and modification B : Chromatin structure and dynamics C : Energy production and conversion D : Cell cycle control, cell division, chromosome partitioning E : Amino acid transport and metabolism F : Nucleotide transport and metabolism G : Carbohydrate transport and metabolism H : Coenzyme transport and metabolism I : Lipid transport and metabolism J : Translation, ribosomal structure and biogenesis K : Transcription L : Replication, recombination and biogenesis ehicl e M : Cell wall/membrane/envelope biogenesis V N : Cell motility O : Posttranslational modification, protein turnover, chaperones P : Inorganic ion transport and metabolism Q : Secondary metabolites biosynthesis, transport and catabolism S : Function unknown T : Signal transduction mechanisms U : Intracellular trafficking, secretion, and vesicular transport V : Defense mechanisms W : Extracellula structures Z : Cytoskeleton NPs 2 O Si

00.2 0.40.6 0.81 Relative Abundance Diao et al. J Nanobiotechnol (2021) 19:174 Page 12 of 20

A Intestine B

Vehicle 3 N.S. SiO2NPs H&E (Intestine)

2 N.S. Vehicle SiO2NPs

1 Relative mRNA expression 0 Il-6 Tnf

CD

AB-PAS staining (Intestine) IHC-lysozyme (Intestine)

SiO NPs Vehicle SiO2NPs Vehicle 2

EF Toluidine blue staining (Intestine) Intestine Vehicle 3 N.S. SiO NPs Vehicle SiO2NPs 2

2 * *

Zoom-in 1 Relative mRNA expression

0 Tjp1 Ocln Cldn7

Fig. 5 Oral exposure to ­SiO2NPs did not result in infammation in the intestine but remarkably damage the tissue integrity. A Efects of SiO­ 2NPs on the Il-6 and Tnf mRNA expressions in gut were shown. B The morphological alterations of gut tissues were determined by H&E staining

after oral exposure to ­SiO2NPs. C AB-PAS staining assay was used to evaluate the number and size of goblet cells between vehicle control and ­SiO2NPs-treated animals. D Efects of ­SiO2NPs on the Paneth cell compartment and the overall granule content. E The number of toluidine blue-positive mast cells were evaluated by Toluidine blue staining. F Efects of SiO­ 2NPs on the intestinal tight junction was measured by the mRNA expressions of Tjp1, Ocln and Cldn 7. Scale bar, 100 μm or 50 μm. Data were reported as mean S.E.M. Statistical analysis was performed by using ± independent student-t test. Asterisk * indicated P < 0.05 Diao et al. J Nanobiotechnol (2021) 19:174 Page 13 of 20

A BC

Sucrase (Intestine) Lactase (Intestine) Maltase (Intestine) N.S. N.S. 300 * 15 600 500 200 10 400

100 5 300 Maltase activity (U/mg prot) Sucrase activity (U/mg prot) 0 Lactase activity (U/mg prot) 0 200 Vehicle SiO2NPs Vehicle SiO2NPsSVehicle iO2NPs

D EF

Alkaline phosphatase (Intestine) γ-glutamyl transferase (Intestine) sIgA (Intestine)

250 15 400 * N.S. N.S. 200 10 300

200 150 5 100 concentration (ng/mL)

100 0 sIgA Alkaline phosphatase activity (King unit/g prot) 0 γ-glutamyl transferase activity (U/g prot) Vehicle SiO2NPs Vehicle SiO NPs 2 Vehicle SiO2NPs

G DAO (Intestine)

35 N.S.

30

25

DAO concentration (mU/mL) 20

Vehicle SiO2NPs

Fig. 6 Oral exposure to ­SiO2NPs partially afected the activities of intestinal digestive enzymes and immune functions. After indicate treatment, the efects of ­SiO2NPs on the activities of typical digestive enzymes, including sucrase, lactase, maltase, alkaline phosphatase and γ-glutamyl transferase were shown in (A–E). F Efects of ­SiO2NPs on the levels of sIgA in the gut tissues were detected using ELISA assay. G Efects of SiO­ 2NPs on the DAO contents were determined by ELISA. Data were reported as mean S.E.M. Statistical analysis was performed by using independent student-t test. ± Asterisk * indicated P < 0.05

Oral exposure to ­SiO2NPs did not result in brain neurobehavioral impairments in animals via gut–brain infammation but might cause the brain damage via gut– axis, the levels of its related indicators were assessed. At brain axis frst, the mRNA expressions of infammation indicators, To investigate whether exposure of SiO­ 2NPs triggers Il-6 and Tnf, were both detected in the cerebral cortex, Diao et al. J Nanobiotechnol (2021) 19:174 Page 14 of 20

where motor and memory functions were initially pro- were the indicators involved in the regulation of gut–lung cessed. As shown in Fig. 7A, the results elaborated that axis or gut–liver axis (Fig. 8E, F). Collectively, these fnd- either Il-6 or Tnf mRNA expression did not change sig- ings indicate that exposure of ­SiO2NPs does not obvi- nifcantly after exposure of SiO­ 2NPs, indicating that oral ously afect the gut–lung axis and gut–liver axis. administration of ­SiO2NPs might not result in the neu- roinfammation in the cerebral cortex. Intriguingly, our Discussion results revealed that oral exposure to ­SiO2NPs remark- Oral route of exposure to SiO­ 2NPs is a vital considera- ably increased the expressions of HuC/D and TuJ1 in the tion due to their deliberate addition to food and unin- gut, indicating the excitement of enteric neurons induced tentional ingestion from contaminated environment [3, by ­SiO2NPs (Fig. 7B, C). Gut–brain peptides are the key 11]. In the past decade, awareness has been raised on the signaling molecules that are involved in the regulation behavior and interaction of ­SiO2NPs in the gastrointesti- of gut–brain axis. Alterations of gut-derived peptides nal tract [51]. However, in absence of robust evidence on are highly related with the neurobehavioral dysfunction. the gastrointestinal uptake and distribution of ­SiO2NPs, Te results demonstrated that the mRNA expression of it is now challenging to evaluate the potential gut health Vipr1 was obviously reduced in both gut and cortex tis- risk of foodborne SiO­ 2NPs accurately. Terefore, in the sues of ­SiO2NPs-treated group as compared with control present study, the young animals were orally treated with group (Fig. 7D). Similar trends were observed on the Sst2 ­SiO2NPs, and the potential efects of ­SiO2NPs on the gas- expressions in the gut and cortex tissues, showing that trointestinal environment and its associated functions the mRNA expressions of Sstr2 was remarkably reduced were assessed. No signifcant changes were observed on in mice exposed to SiO­ 2NPs (Fig. 7D). However, the levels the body weight between two groups (data not shown). of Bdnf, Sst and Ghsr expressions did not change signif- Intriguingly, our results illustrated that, in addition to cantly (Fig. 7D). Te pathological changes were observed evoking dysfunctions and injuries in the gut, ingested under light microscope in SiO­ 2NPs-treated animals, ­SiO2NPs exhibited a moderate to severe impact on the manifested by reduced neuronal cells, cell shrinkage, gut microbial communities, and thereby resulting in rupture and deformation, chromatin condensation and spatial learning and memory impairments and locomo- nuclear fragmentation (Fig. 7E). Tese results together tor inhibition via distinctive microbiota–gut–brain axis. suggest that oral administration of SiO­ 2NPs is capable Considering the long-term exposure of ­SiO2NPs via of inducing brain damage via gut–brain axis by specifc food, the potential efects of ­SiO2NPs on the gut micro- chemical substances originated from gut, but does not biota and its related brain functions should be seriously trigger neuroinfammation in the cerebral cortex of mice. concerned in health risk assessment, especially for the infants or children who are susceptible to the neurotoxic- Oral exposure to ­SiO2NPs did not afect the gut–lung axis ity of nano-sized particles. and gut‑liver axis Gut microbiota plays important roles in regulation of To determine whether oral exposure to SiO­ 2NPs distinc- gastrointestinal functions [29, 45, 46, 48]. For instance, tively disrupt the gut–brain axis and therefore resulting the microbes within the gut are essential to facilitate in the neurobehavioral impairments such as spatial learn- the digestion and produce vitamins, which provide a ing and memory and locomotor inhibition, the gut–lung good foundation for human health. Te gut microbiome axis and gut–liver axis related indicators were all evalu- is also indispensable for the development of intestinal ated. As shown in Fig. 8A, the cell count number in epithelium as well as the host immune defense against bronchoalveolar lavage fuid did not alter after ­SiO2NPs pathogens. All of these biological functions of gut micro- administration. Meanwhile, we did not fnd notably biota contribute greatly to lifelong maintenance of gas- alteration on the level of sIgA in the lung tissue collected trointestinal microenvironment homeostasis [29, 48]. from ­SiO2NPs-treated mice (Fig. 8B). Both the activities Conversely, a shift in the gut microbiota is related to of SOD and contents of MDA in the lung and liver tis- the onset and/or progression of various diseases [30, 31, sues did not change signifcantly in SiO­ 2NPs-exposed 45, 46]. Herein, our results demonstrated that exposure animals in comparison to vehicle controls (Fig. 8C, D). to ­SiO2NPs not only resulted in a signifcant shift in the Te mRNA expressions of Il-6 and Tnf in the two groups composition of microbial communities, but also drasti- did not show any signifcant diference. Tere was no sig- cally enhanced the microbial species and diversity within nifcant alteration on the mRNA expression of Ccl2 in the the gut. Similar phenomenon was observed in the CD-1 liver between two groups. No signifcant changes were mice orally administrated with SiO­ 2NPs for 7 days [52]. observed on the mRNA expressions of Col1a2, Tgfbr2 Such an unexpected efect may due to the low absorption and Serpine1in the lung and liver tissues, all of which rate of gastrointestinal tract for precipitated or fumed Diao et al. J Nanobiotechnol (2021) 19:174 Page 15 of 20

A B Cortex 2.5 Vehicle N.S. SiO NPs DAPI TuJ1 Merged N.S. 2 2.0

1.5

1.0

0.5 Relative mRNA expression 0.0 Il-6 Tnf

DAPIHuC/D Merged C Intestine Vehicle 2.0 SiO2NPs l 1.5 * *

1.0

0.5 Fold change to contro

0.0 Intestine TuJ1 HuC/D

D E N.S. Vehicle 3.0 SiO2NPs

2.5

H&E (Cortex) 2.0 N.S. Vehicle SiO2NPs N.S. * N.S. N.S. 1.5 * * N.S. *

1.0 Relative mRNA expression Zoom-i n 0.5

0.0 Bdnf Ghsr Vipr1 SstSstr2 Bdnf Ghsr Vipr1 Sst Sstr2 Intestine Cortex

Fig. 7 Oral exposure to ­SiO2NPs did not result in brain infammation but might cause the brain damage via gut–brain axis. A Efects of SiO­ 2NPs on the mRNA expressions of infammation indicators, Il-6 and Tnf, in the cerebral cortex tissues. The expressions of HuC/D and TuJ1, both of which were specifc enteric neuron markers, were detected by immunofuorescence assay. The representative images and fuorescence intensity of HuC/D and

TuJ1 expression were displayed in (B and C), respectively. Scale bar, 100 μm. D Efects of SiO­ 2NPs on the mRNA expressions of Vipr1, Sstr2, Bdnf, Sst and Ghsr in the cortex and gut. E The pathological changes of ­SiO2NPs-treated animals were measured using H&E staining. Scale bar, 100 μm and 50 μm. Data were reported as mean S.E.M. Statistical analysis was performed using independent student-t test. Asterisk * indicated P < 0.05 ± Diao et al. J Nanobiotechnol (2021) 19:174 Page 16 of 20

silicate [10]. Another explanation is that the interactions ultimately causing the neurodegeneration-like changes in between nanomaterials and microbiome may depend behaviors, such as spatial learning and memory impair- on the sampling area. Te microbial composition can be ment [19, 26]. Other previous studies have demonstrated very diferent in diferent regions of gastrointestinal tract that the neurotoxic efects caused by administering of [10, 51], and there may be a bias to determine the micro- ­SiO2NPs, specifcally occurred by the elevation in oxida- bial composition in the intestine. tive stress and activation of microglial functions, leading Beyond the gut, it is well established that the gut micro- to highly negative impacts on the neurons [24, 25, 28, 55]. biota can modulate the function of other organs, such as However, to date, whether SiO­ 2NPs is capable of trans- brain, liver, and lung etc. [29, 53, 54]. Te bidirectional locating in the brain after oral administration has not communication that is enabled by many bacterial metab- been established, and the detailed mechanisms underly- olites, gut-derived chemicals, peptides. Tese gut-orig- ing how ­SiO2NPs exposure induces cognitive dysfunc- inated substances can signifcantly afect distant organs tions remain substantially unclear. Herein, the potential either directly through systemic circulation or indirectly limitation is that we can not exclude the possibility that by signaling via vagus nerve or chemicals from the gut accumulated ­SiO2NPs in the brain directly triggers the [29, 46, 48, 53, 54]. In this study, the obtained results neuronal damage. Notwithstanding, the current fndings revealed that, the enteric neurons were excited after presented in this study will still provide valuable informa- ­SiO2NPs administration, and the mRNA expressions of tion in the understanding of gut–brain axis involved in Vipr1 and Sstr2 were signifcantly increased by SiO­ 2NPs, ­SiO2NPs-induced neurobehavioral impairments. accompanied with pathological changes and neurobe- In this study, oral exposure to ­SiO2NPs for 28 days havioral impairments. Tese data may together support in young mice was shown to damage the intestinal and the hypothesis that exposure to SiO­ 2NPs by oral treat- cerebral cortex tissues. Similar trends were reported in ment causes the spatial learning and memory defcits and the previous investigations conducted by You et al. [26] locomotor inhibition by disrupting the gut–brain axis. and Salem et al. [56], in both of which neuropathologi- On one hand, the disrupted microbiota may infuence cal changes were presented in ­SiO2NPs-exposed ani- the physiological processes of specifc gut-originated mals. Interestingly, we did not fnd that oral exposure substances through their efects on the central nervous of ­SiO2NPs directly triggered infammatory response in system. On the other, the damaged brain also plays a crit- these two regions. Conversely, increased levels of neuro- ical part in reshaping the microbiota that may be harm- infammation in the cerebral cortex and hippocampus, ful for its metabolic activities. Intriguingly, our fndings manifested by the enhancement of Il-6 and Il-1β mRNA illustrated that treatment of ­SiO2NPs did not show any expressions, were observed in mice intranasal admin- impacts on the gut–liver and gut–lung axis, manifested istrated with ­SiO2NPs [26]. Such inconsistency on the by the unchanged indicators determined in either lung or induction of infammation may due to the diferent routes liver tissues in both two groups. Tese fndings will rep- of exposure as well as the dosage used in the animal resent the basis for better understanding how exposure model. Furthermore, the results also clearly elucidated to ­SiO2NPs distinctively shape the interactions between that the number of toluidine blue-positive mast cells was microbiome and neuronal functions. elevated markedly after ­SiO2NPs exposure. Because of Notably, previous reports have proved the capacity of the reciprocal interactions between mast cells and neu- ­SiO2NPs to translocate into the brain after intranasal roendocrine immune network, the decreased number of instillation or intravenous injection [19, 26]. Tey can mast cells may partially contribute to the alterations of gradually accumulate in the diferent regions of brain gut–brain signaling chemicals, such as Vipr1 and Sstr2. due to the limited excretion of SiO­ 2NPs in the body, and In addition, mast cells are also reported to be involved in thereby resulting in injuries to neuronal cells. Based on the modulation of digestive secretions by release of his- this observation, recent investigations have focused on tamine via gut–brain axis [57]. In this exposure model the role of ­SiO2NPs on brain functions. Either injected of ­SiO2NPs, our results precisely demonstrated that the intravenously, or intranasal instillation of ­SiO2NPs, leads activities of sucrase were increased signifcantly. Given to the accumulation of nanoparticles in the brain, and that the activity of sucrase is usually elevated as the

(See fgure on next page.)

Fig. 8 Oral exposure to ­SiO2NPs did not afect the gut-lung axis and gut liver axis. A Efects of SiO­ 2NPs on the cell count in bronchoalveolar lavage fuid. B The levels of sIgA in the lung tissues of two groups were measured by ELISA assay. Efects of SiO­ 2NPs on the activities of SOD and contents of MDA in the lung and liver tissues were displayed in (C and D). Efects of SiO­ 2NPs on the mRNA expressions of Il-6, Tnf Col1a2, Tgfbr2, Ccl2 and Serpine1 in the lung and liver tissues were shown in (E and F). Data were reported as mean S.E.M. Statistical analysis was conducted by using ± independent student-t test. Asterisk * indicated P < 0.05 Diao et al. J Nanobiotechnol (2021) 19:174 Page 17 of 20

A B

BALF sIgA (Lung) ) ) 5 5 N.S.

0 9000 N.S. 4 8000

3 7000

2 6000

5000

1 concentration (ng/ml

4000 sIgA Cell count in BALF (×1 0

Vehicle SiO2NPs Vehicle SiO2NPs CD

SOD activity MDA content Vehicle Vehicle

SiO NPs SiO2NPs 1.5 2 1.5 N.S. N.S. N.S. l l N.S.

1.0 1.0

0.5 0.5 Fold change to contro Fold change to contro

0.0 0.0 Lung Liver Lung Liver

EF

Lung Liver 4 3 N.S. Vehicle N.S. Vehicle SiO2NPs SiO2NPs 3 N.S. 2 N.S. N.S. N.S. N.S. N.S. N.S. N.S. 2 N.S. 1 1 Relative mRNA expression Relative mRNA expression

0 0 Il-6 Tnf Tgfbr2 Serpine1 Col1a2 Il-6 Tnf Ccl2 Serpine1 Col1a2 Tgfbr2 Diao et al. J Nanobiotechnol (2021) 19:174 Page 18 of 20

concentration of sucrose increases in the intestine, the Supplementary Information The online version contains supplementary material available at https://​doi.​ enhancement of its activity indicate that SiO­ 2NPs may adaptively increase the level of sucrose, which is of capac- org/​10.​1186/​s12951-​021-​00916-2. ity to lower gut microbial diversity. Another cause for the improvement of digestive enzyme activity perhaps is that Additional fle 1: Figure S1. (A) The rarefaction curves of two groups were displayed. (B) The Kyoto Encyclopedia of Genes and Genomes ­SiO2NPs modifed the secretion of bacterial enzyme. (KEGG) function analysis obtained top 3 elevated functional abundances Intestinal mucosal barrier is the frst line of physical and top 3 reduced functional abundances in top 50 highest abundances defense against external substances. Tis barrier is a het- of KEGG orthology (KO). (C) Functional abundances were measured on the KEGG pathway level 2. erogeneous entity mainly composed of mechanical, bio- Additional fle 2: Table S1. Primer sequences of target genes. chemical, and immune barriers [58]. Te obtained results explicated that the mRNA expressions of Tjp1 and Ocln, both of which were tight junction proteins, were sharply Acknowledgements We thank Professor Yan Tang (Southwest Medical University, School of Public reduced by ­SiO2NPs. In H&E staining assay, pathological Health) for her help on doing the neurobehavioral tests on animals. changes were also observed in ­SiO2NPs-treated animals. Tese fndings together indicate the mechanical barrier Authors’ contributions JD: Methodology, investigation, software; YX: methodology, investigation, for- of gut is signifcantly damaged by ­SiO2NPs, which may mal analysis; SC: conceptualization, supervision, data curation, formal analysis; further lead to the disturbance of gut microenvironment XJ: writing-original draft, data curation, formal analysis, funding acquisition; and increase the sensitivity of gut to exogenous stimuli. ZJ: conceptualization, supervision; JQ: conceptualization, supervision, funding acquisition; XQ: resources, validation, funding acquisition; MG, JS, XD, JF: meth- Meanwhile, the sharply reduced contents of sIgA, Paneth odology, validation; ZZ: project administration, conceptualization, writing- cell compartment and overall granule also signify that the review and editing; CC: project administration, conceptualization, writing- capacity of immune host defenses in the gut is weaken review and editing. All authors read and approved the fnal manuscript. after oral exposure of ­SiO2NPs. Notably, the disrupted Funding barrier function may change the intestinal permeability This research was supported by National Natural Science Foundation of China that harmful substances or pathogens are easily to pass (81703187, 81602820, 81903358); Natural Science Foundation of Chongqing (cstc2020jcyj-msxmX0192 and cstc2020jcyj-msxmX0155); Science and Tech- through epithelial barrier, thereby causing the adverse nology Research Program of Chongqing Municipal Education Commission efects in diferent distant organs. (KJQN201800434, KJQN201900419, KJQN 201900421 and KJCXZD2020020); Te limitations of this study should be considered. At Research Program of Basic Research and Frontier Technology of Chongqing Yuzhong district (20200105); Chongqing Talent Project (CQYC2020058650). frst, in animal studies, the dosage and duration of oral Z.Z. and C.C. were both the Young talent of Bayu scholar. administration of SiO­ 2NPs are important factors for mimicking exposure conditions to which humans may be Availability of data and materials The datasets used and/or analyzed during the current study are available from exposed through the daily dietary intake [10]. Tus, we the corresponding author on reasonable request. herein calculated the dosage of SiO­ 2NPs according to the upper limit of ­SiO2 added in the food additive. Although Declarations this dosage seems unusually high, it is needed for health risk assessment of foodborne ­SiO NPs. Secondly, due Ethics approval and consent to participate 2 This work was approved by the ethics committee of Chongqing Medical to the lack of cogent evidence obtained from germ-free Univeristy. animals treated with SiO­ 2NPs, the causal efects of gut microbiota on the brains and behaviors should be verifed Competing interests The authors deny any competing interests related to this study. in the further studies. Author details Conclusion 1 Department of Occupational and Environmental Health, School of Public Health and Management, Chongqing Medical University, Chongqing 400016, To the best of our knowledge, this is the frst work that People’s Republic of China. 2 Center of Experimental Teaching for Public elucidates that oral exposure to SiO­ 2NPs results in the Health, Experimental Teaching and Management Center, Chongqing Medical 3 spatial learning and memory impairments and locomotor University, Chongqing 400016, People’s Republic of China. Department of Health Laboratory Technology, School of Public Health and Management, inhibition via distinctive gut–brain axis. Te implications Chongqing Medical University, Chongqing 400016, People’s Republic of China. of this study include the novel idea that proper mainte- 4 Department of Pharmacy, The First Afliated Hospital of Chongqing Medical 5 nance and coordination of gut functions may be critical University, Chongqing 400016, People’s Republic of China. Molecular Biology Laboratory of Respiratory Disease, Institute of Life Sciences, Chongqing Medi- for protection against the neurotoxicity of foodborne cal University, Chongqing 400016, People’s Republic of China. 6 Dongsheng ­SiO2NPs because of the multifaceted system of gut– Lung‑Brain Disease Joint Lab, Chongqing Medical University, No. 1, Yixueyuan brain bidirectional communication. Tis study may also Road, Yuzhong District, Chongqing 400016, People’s Republic of China. provide valuable scientifc evidence for policy makers to Received: 9 March 2021 Accepted: 28 May 2021 evaluate the safety of ­SiO2NPs in infant food applications. Diao et al. J Nanobiotechnol (2021) 19:174 Page 19 of 20

References 22. Ahamed M, Akhtar M, Alhadlaq HJ. Infuence of silica nanoparticles on 1. Abdelhamid HN, Wu B-S, Wu H-F. Graphene coated silica applied for high cadmium-induced cytotoxicity, oxidative stress, and apoptosis in human ionization matrix assisted laser desorption/ionization mass spectrometry: liver HepG2 cells. Environ Toxicol. 2020;35:599–608. a novel approach for environmental and biomolecule analysis. Talanta. 23. Liu J, Li X, Zhou G, Sang Y, Zhang Y, Zhao Y, Ge W, Sun Z, Zhou XJ. Silica 2014;126:27–37. nanoparticles induce spermatogenesis disorders via L3MBTL2-DNA 2. Bhaisare ML, Abdelhamid HN, Wu B-S, Wu H-F. Rapid and direct MALDI- damage-p53 apoptosis and RNF8-ubH2A/ubH2B pathway in mice. MS identifcation of pathogenic bacteria from blood using ionic liquid- Environ Pollut. 2020;265:114974. modifed magnetic nanoparticles (Fe 3 O 4@ SiO 2). J Mater Chem B. 24. Li X, Ji X, Wang R, Zhao J, Dang J, Gao Y, Jin MJ. Zebrafsh behavioral 2014;2:4671–83. phenomics employed for characterizing behavioral neurotoxicity caused 3. Additives EPoF, Food NSat, Younes M, Aggett P, Aguilar F, Crebelli R, Duse- by silica nanoparticles. Chemosphere. 2020;240:124937. mund B, Filipič M, Frutos MJ, Galtier P, Gott DJ. Re-evaluation of silicon 25. Wei J, Liu J, Liang S, Sun M, Duan JJ. Low-dose exposure of silica dioxide (E 551) as a food additive. EFSA J. 2018;16:e05088. nanoparticles induces neurotoxicity via neuroactive ligand-receptor 4. Fruijtier-Pölloth C. The safety of nanostructured synthetic amorphous interaction signaling pathway in zebrafsh embryos. Int J Nanomed. silica (SAS) as a food additive (E 551). Arch Toxicol. 2016;90:2885–916. 2020;15:4407–15. 5. Al-Rawi M, Diabaté S, Weiss C. Uptake and intracellular localization of 26. You R, Ho Y, Hung C, Liu Y, Huang C, Chan H, Ho S, Lui S, Li H, Chang RJP. submicron and nano-sized SiO 2 particles in HeLa cells. Arch Toxicol. Silica nanoparticles induce neurodegeneration-like changes in behavior, 2011;85:813–26. neuropathology, and afect synapse through MAPK activation. Part Fibre 6. Yang Y, Faust JJ, Schoepf J, Hristovski K, Capco DG, Herckes P, Wester- Toxicol. 2018;15:28. hof PJ. Survey of food-grade silica dioxide nanomaterial occurrence, 27. Roshanfekrnahzomi Z, Badpa P, Esfandiari B, Taheri S, Nouri M, Akhtari K, characterization, human gut impacts and fate across its lifecycle. Sci Total Shahpasand K, Falahati MJ. Silica nanoparticles induce conformational Environ. 2016;565:902–12. changes of tau protein and oxidative stress and apoptosis in neuroblas- 7. Athinarayanan J, Periasamy VS, Alsaif MA, Al-Warthan AA, Alshatwi AA. toma cell line. Int J Biol Macromol. 2019;124:1312–20. Toxicology: presence of nanosilica (E551) in commercial food products: 28. Yang X, He C, Li J, Chen H, Ma Q, Sui X, Tian S, Ying M, Zhang Q, Luo Y, TNF-mediated oxidative stress and altered cell cycle progression in et al. Uptake of silica nanoparticles: neurotoxicity and Alzheimer-like human lung fbroblast cells. Cell Biol Toxicol. 2014;30:89–100. pathology in human SK-N-SH and mouse neuro2a neuroblastoma cells. 8. Luyckx M, Hausman J-F, Lutts S, Guerriero GJFiPS. Silicon and plants: Toxicol Lett. 2014;229:240–9. current knowledge and technological perspectives. Front Plant Sci. 29. Jameson K, Olson C, Kazmi S, Hsiao EJ. Toward understanding microbi- 2017;8:411. ome-neuronal signaling. Mol Cell. 2020;78:577–83. 9. Dekkers S, Krystek P, Peters R, Lankveld D, Bokkers B, van Hoeven- 30. De-Paula VDJR, Forlenza A, Forlenza O. Relevance of gutmicrobiota Arentzen P, Bouwmeester H, Oomen AJN. Presence and risks of nanosilica in cognition, behaviour and Alzheimer’s disease. J Pharmacol Res. in food products. Nanotoxicology. 2011;5:393–405. 2018;136:29–34. 10. Lamas B, Martins Breyner N, Houdeau EJP. Impacts of foodborne 31. Sun M, Shen YJ. Dysbiosis of gut microbiota and microbial metabolites in inorganic nanoparticles on the gut microbiota-immune axis: potential Parkinson’s disease. Ageing Res Rev. 2018;45:53–61. consequences for host health. Part Fibre Toxicol. 2020;17:19. 32. Breznan D, Das DD, O’Brien JS, MacKinnon-Roy C, Nimesh S, Vuong NQ, 11. Murugadoss S, Lison D, Godderis L, Van Den Brule S, Mast J, Brassinne Bernatchez S, DeSilva N, Hill M, Kumarathasan P, Vincent R. Diferential F, Sebaihi N, Hoet PJ. Toxicology of silica nanoparticles: an update. Arch cytotoxic and infammatory potency of amorphous silicon dioxide Toxicol. 2017;91:2967–3010. nanoparticles of similar size in multiple cell lines. Nanotoxicology. 12. Napierska D, Thomassen L, Lison D, Martens J, Hoet PJP. The nanosilica 2017;11:223–35. hazard: another variable entity. Part Fibre Toxicol. 2010;7:39. 33. Rubio L, Pyrgiotakis G, Beltran-Huarac J, Zhang Y, Gaurav J, Deloid G, 13. Kusaczuk M, Krętowski R, Naumowicz M, Stypułkowska A, Cechowska- Spyrogianni A, Sarosiek KA, Bello D, Demokritou P. Safer-by-design Pasko MJ. Silica nanoparticle-induced oxidative stress and mitochondrial fame-sprayed silicon dioxide nanoparticles: the role of silanol content damage is followed by activation of intrinsic apoptosis pathway in on ROS generation, surface activity and cytotoxicity. Part Fibre Toxicol. glioblastoma cells. Int J Nanomed. 2018;13:2279–94. 2019;16:1–15. 14. Schmitz C, Welck J, Tavernaro I, Grinberg M, Rahnenführer J, Kiemer A, van 34. Reagan-Shaw S, Nihal M, Ahmad N. Dose translation from animal to Thriel C, Hengstler J, Kraegeloh AJN. Mechanical strain mimicking breath- human studies revisited. FASEB J. 2008;22:659–61. ing amplifes alterations in gene expression induced by SiO NPs in lung 35. Zhang S, Cheng S, Jiang X, Zhang J, Bai L, Qin X, Zou Z, Chen CJF, epithelial cells. Nanotoxicology. 2019;13:1227–43. Association ctaijpftBIBR. Gut-brain communication in hyperfunction of 15. Liu Y, Wei H, Tang J, Yuan J, Wu M, Yao C, Hosoi K, Yu S, Zhao X, Han Y, 5-hydroxytryptamine induced by oral zinc oxide nanoparticles exposure Chen GJC. Dysfunction of pulmonary epithelial tight junction induced in young mice. Food Chem Toxicol. 2020;135:110906. by silicon dioxide nanoparticles via the ROS/ERK pathway and protein 36. Zhang T, Jiang J, Liu J, Xu L, Duan S, Sun L, Zhao W, Qian FJ. MK2 is degradation. Chemosphere. 2020;255:126954. required for neutrophil-derived ROS production and infammatory bowel 16. Almansour M, Alarif S, Jarrar BJ. In vivo investigation on the chronic disease. Front Med. 2020;7:207. hepatotoxicity induced by intraperitoneal administration of 10-nm silicon 37. Liang W, Zhang G, Luo H, Liang L, Huang D, Zhang FJG. Tryptase and dioxide nanoparticles. Int J Nanomed. 2018;13:2685–96. protease-activated receptor 2 expression levels in irritable bowel syn- 17. Zhao X, Wei S, Li Z, Lin C, Zhu Z, Sun D, Bai R, Qian J, Gao X, Chen G, et al. drome. Gut Liver. 2016;10:382–90. Autophagic fux blockage in alveolar epithelial cells is essential in silica 38. Tang Q, Bai L, Zou Z, Meng P, Xia Y, Cheng S, Mu S, Zhou J, Wang X, Qin nanoparticle-induced pulmonary fbrosis. Cell Death Dis. 2019;10:127. X, et al. Ferroptosis is newly characterized form of neuronal cell death in 18. Dekkers S, Bouwmeester H, Bos P, Peters R, Rietveld A, Oomen AJN. response to arsenite exposure. Neurotoxicology. 2018;67:27–36. Knowledge gaps in risk assessment of nanosilica in food: evaluation of 39. Zhang S, Jiang X, Cheng S, Fan J, Qin X, Wang T, Zhang Y, Zhang J, Qiu the dissolution and toxicity of diferent forms of silica. Nanotoxicology. Y, Qiu J, et al. Titanium dioxide nanoparticles via oral exposure leads to 2013;7:367–77. adverse disturbance of gut microecology and locomotor activity in adult 19. Ivanov S, Zhuravsky S, Yukina G, Tomson V, Korolev D, Galagudza MJM. mice. Arch Toxicol. 2020;94:1173–90. In vivo toxicity of intravenously administered silica and silicon nanoparti- 40. Meng P, Zhang S, Jiang X, Cheng S, Zhang J, Cao X, Qin X, Zou Z, Chen cles. Materials. 2012;5:1873–89. CJE. Arsenite induces testicular oxidative stress in vivo and in vitro lead- 20. Leung CC, Yu IT, Chen W. Silicosis. Lancet. 2012;379:2008–18. ing to ferroptosis. Ecotoxicol Environ Saf. 2020;194:110360. 21. Hu H, Fan X, Guo Q, Wei X, Yang D, Zhang B, Liu J, Wu Q, Oh Y, Feng Y, et al. 41. Zhang Y, Jiang X, Zhang J, Xia Y, Qiu J, Wang T, Qiu Y, Qin X, Wang B, Zou Z, Silicon dioxide nanoparticles induce insulin resistance through endoplas- Chen CJ. Heterozygous disruption of beclin 1 mitigates arsenite-induced mic reticulum stress and generation of reactive oxygen species. Part Fibre neurobehavioral defcits via reshaping gut microbiota-brain axis. J Hazard Toxicol. 2019;16:41. Mater. 2020;398:122748. Diao et al. J Nanobiotechnol (2021) 19:174 Page 20 of 20

42. Bai L, Tang Q, Zou Z, Meng P, Tu B, Xia Y, Cheng S, Zhang L, Yang K, Mu S, 52. Chen H, Zhao R, Wang B, Cai C, Zheng L, Wang H, Wang M, Ouyang H, et al. m6A Demethylase FTO regulates dopaminergic neurotransmission Zhou X, Chai Z, et al. The efects of orally administered Ag, TiO2 and SiO2 defcits caused by arsenite. Toxicol Sci. 2018;165:431–46. nanoparticles on gut microbiota composition and colitis induction in 43. Vorhees C, Williams M. Morris water maze: procedures for assessing spa- mice. NanoImpact. 2017;8:80–8. tial and related forms of learning and memory. Nat Protoc. 2006;1:848–58. 53. Tripathi A, Debelius J, Brenner DA, Karin M, Loomba R, Schnabl B, Knight 44. Gould TD, Dao DT, Kovacsics CE. The open feld test. In: Mood and anxiety RJN. The gut–liver axis and the intersection with the microbiome. Nat Rev related phenotypes in mice. Springer; 2009:1–20. Gastroenterol Hepatol. 2018;15:397–411. 45. Kamada N, Seo S-U, Chen GY, Núñez GJ. Role of the gut microbiota in 54. Budden KF, Gellatly SL, Wood DL, Cooper MA, Morrison M, Hugenholtz immunity and infammatory disease. Nat Rev Immunol. 2013;13:321–35. P, Hansbro PM. Emerging pathogenic links between microbiota and the 46. Lavelle A, Sokol HJ. Gut microbiota-derived metabolites as key actors gut–lung axis. Nat Rev Microbiol. 2017;15:55–63. in infammatory bowel disease. Nat Rev Gastroenterol Hepatol. 55. Liu X, Sui B, Sun JJB. Blood–brain barrier dysfunction induced by silica NPs 2020;17:233–7. in vitro and in vivo: involvement of oxidative stress and Rho-kinase/JNK 47. Malo M, Alam SN, Mostafa G, Zeller S, Johnson P, Mohammad N, Chen K, signaling pathways. Biomaterials. 2017;121:64–82. Moss A, Ramasamy S, Faruqui AJG. Intestinal alkaline phosphatase pre- 56. Salem A, Oudhabechi A, Sakly M. Efect of nano-sized SiO(2) particles on serves the normal homeostasis of gut microbiota. Gut. 2010;59:1476–84. the cognitive function and biochemical response. Arch Environ Occup 48. Spencer SP, Fragiadakis GK, Sonnenburg JLJI. Pursuing human-relevant Health. 2019;74:140–6. gut microbiota-immune interactions. Immunity. 2019;51:225–39. 57. Buhner S, Schemann MJ. Mast cell–nerve axis with a focus on the human 49. Mantis NJ, Rol N, Corthésy BJ. Secretory IgA’s complex roles in immunity gut. Biochim Biophys Acta Mol Basis Dis. 2012;1822:85–92. and mucosal homeostasis in the gut. Mucosal Immunol. 2011;4:603–11. 58. Turner JR. Intestinal mucosal barrier function in health and disease. Nat 50. Sasabe J, Miyoshi Y, Rakof-Nahoum S, Zhang T, Mita M, Davis BM, Hamase Rev Immunol. 2009;9:799–809. K, Waldor MKJ. Interplay between microbial d-amino acids and host d-amino acid oxidase modifes murine mucosal defence and gut micro- biota. Nat Microbiol. 2016;1:1–7. Publisher’s Note 51. Bouwmeester H, van der Zande M, Jepson MJ. Efects of food-borne Springer Nature remains neutral with regard to jurisdictional claims in pub- nanomaterials on gastrointestinal tissues and microbiota. Wiley Interdis- lished maps and institutional afliations. cip Rev Nanomed Nanobiotechnol. 2018;10:e1481.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions