Experimental Evaluation of the Impact of Gadolinium Orthovanadate GdVO4:Eu3+ Nanoparticles on the Carrageenan-Induced Intestinal Inflammation

A B S T R AC T Aim: To evaluate the effects of orally administered gadolinium orthovanadate GdVO4:Eu nanoparticles (VNPs) on the course of chronic carrageenan-induced intestinal inflammation. Methods: Samples of small intestinal tissue were collected from four groups of rats (intact, after administration of VNPs, with carrageenaninduced intestinal inflammation, with carrageenan-induced intestinal inflammation orally exposed to VNPs) to assess the intestinal morphology and HSP90α expression. Levels of seromucoid, C-reactive protein, TNF-α, IL-1β and IL-10 were determined in blood serum. Results: Oral exposure to VNPs was associated with neither elevation of inflammation markers in blood serum nor HSP90α overexpression in the small intestine, i.e. no toxic effects of VNPs were observed. Carrageenan-induced intestinal inflammation was accompanied by higher levels of TNF-α and IL-1β, as well as HSP90α upregulation in the intestinal mucosa, compared with controls. Administration of VNPs to rats with enteritis did not lead to statistically significant changes in concentrations of circulating pro-inflammatory cytokines with the trend towards their increase. Conclusion: No adverse effects were observed in rats orally exposed to VNPs at a dose of 20 μg/kg during two weeks. Using the experimental model of carrageenan-induced enteritis, it was demonstrated that VNPs at the dose used in our study did not affect the course of intestinal inflammation.


INTRODUCTION
Inflammatory bowel disease (IBD) is characterized by a chronic intestinal inflammation and includes two subtypes: Crohn's disease (CD) and ulcerative colitis (UC) (1). In CD, the inflammation is transmural and may af fect the entire gut, whereas in UC it is mainly limited to the mucosal layer of the large intestine (2). IBD is a multifacet ed disease whose development is associated with complex interactions between genetic and environmental factors, including features of intestinal microbiota, abnormalities of the innate immune system, dietary habits, etc. (3). Its conventional treatment includes 5aminosalicylate, gluco corticoid drugs, methotraxate, azathioprine, and antitu mor necrosis factor (TNF) agents such as infliximab, adali mumab, golimumab. Furthermore, IL12/23 antagonists (ustekinumab), inhibitors of intestinal lymphocyte traf ficking (vedolizumab, a monoclonal antibody to the α4β7 integrin), and small molecule inhibitors of Janus kinases, including tofacitinib, are currently available in the market (4)(5)(6)(7)(8). Nevertheless, the development of novel therapeutic agents for the treatment of IBD remains of huge impor tance, since the current firstline antiTNF treatment may be ineffective or the intolerance to antiTNF agents can emerge in patients (9).
There is strong evidence that IBD is accompanied by the excessive generation of reactive oxygen species (ROS) and subsequent development of oxidative stress (10,11). Overproduction of ROS in IBD results in oxidative dam age to macromolecules (lipid peroxidation and protein oxidative modification), loss of cell membrane integrity, low ATP production by mitochondria, apoptosis, etc. (10,12). This fact substantiates the search for novel, effective antioxidantbased agents for the treatment of IBD (13). In particular, the therapeutic potential of nanoparticles with antioxidant properties has been studied for decades. Converging lines of evidence indicate that they act as ROS scavengers (14,15). It has been reported that gadolinium orthovanadate GdVO 4 :Eu 3+ nanoparticles (VNPs) can scav enge free radicals in vitro (16). However, little is known about the therapeutic action of VNPs in vivo. To assess the therapeutic potential of VNPs, we have chosen the already characterized experimental model of carrageenanin duced intestinal inflammation (17)(18)(19).
The aim of our research was to study the impact of oral ly administered VNPs on the course of carrageenanin duced intestinal inflammation.

DESIGN OF THE STUDY, CHARACTERISTICS OF ANIMALS AND GROUPS
Fifty female WAG rats weighing 160-190 g were provided by the vivarium of Kharkiv National Medical University. They were randomly subdivided into five equal groups (n = 10). Carrageenaninduced intestinal inflammation was induced in the rats from groups A and B. The animals from group B were orally administered a water solution of VNPs at a dose of 20 μg/kg of weight against the back ground of intestinal inflammation. Group C included the intact animals fed on a standard diet and obtained a solu tion of VNPs at a dose of 20 μg/kg of weight. Groups D 1 and D 2 served as controls and consisted of intact rats. The rats were housed in cages. They were maintained in standard laboratory conditions at room temperature (24 ± 2 °C). Access to food was free. All the animals were sacrificed. Blood samples were collected to prepare serum for eval uating the systemic levels of inflammation markers. Fur thermore, fragments of small intestine were sampled for immunohistochemical studies.

CHARACTERISTICS OF NANOPARTICLES
The synthesis of GdVO 4 :Eu 3+ nanoparticle water colloidal solution was carried out in accordance with the meth od reported earlier (21). Briefly, 10 mL of aqueous solu tion of rareearth chlorides (0.01 mol/L) was mixed with 8 mL of ethylenediaminetetraacetic acid disodium salt (EDTA 2 Na) solution (0.01 mol/L). It was followed by the addition of 8 mL Na 3 VO 4 (0.01 mol/L) to the solution ob tained dropwise (рН = 13). The mixture was intensively stirred by a magnetic stirrer until yellowish transparent solution was formed.
The colorless transparent solution obtained as a result scattered light under the side illumination (Tindal cone). The solution was cooled and dialyzed against water for 24 h to remove the excess of ions. A dialysis membrane with a molecular weight cutoff of 12 kDa with a pore size of approximately 2.5 nm was used. The composition of spindlelike nanoparticles -Gd (0,9) Eu (0,1) VO 4 -with average size of 8 × 25 nm was formed ( Figure 1).

CARRAGEENAN-INDUCED INTESTINAL INFLAMMATION MODEL
Intestinal inflammation in the rats from group A and group B was induced by the daily oral administration of kcar rageenancontaining 1% processed Eucheuma seaweed (PES) in drinking water (140 mg per kg of weight) during 4 months. In addition to carrageenan, PES contained less than 15% of algal cellulose. The solution was prepared at least 24 h prior to its administration and stored at low tem perature (2 °C).
Development of intestinal inflammation was con firmed in each animal from groups A and B using routine histological staining techniques (hematoxylin and eosin staining, PAS reaction, and hallocyaninechrome alum Einarsson's stain).

DETERMINATION OF SYSTEMIC LEVELS OF INFLAMMATORY AND ANTI-INFLAMMATORY BIOMARKERS
Systemic levels of proinflammatory cytokines TNFα and IL1β were assessed by commercially available ELISA kits purchased from eBioScience (Austria). The procedures were done strictly in accordance with manufacturers' instruc tions. Concentrations of TNFα, IL1β and IL10 in blood serum were expressed in pg/ml. ELISA method was also used to assess the levels of antiinflammatory cytokine IL10 in blood serum of animals (eBioScience ELISA kit).
Furthermore, the levels of inflammatory markers such as seromucoid and Creactive protein were determined in blood serum of rats from groups C and D 2 by routine tech niques. Seromucoid and Creactive protein levels were assessed using commercially available kits manufactured by Filicit-Diagnostika (Ukraine). Seromucoid levels were expressed in units of the ShankHoagland scale (SH units), whereas the content of Creactive protein in blood serum was represented in mg/L. In addition, the content of mid dle molecules was determined in blood serum of animals from groups C and D 2 by the Gabrielyan's method to evalu ate the severity of endogenous intoxication (22). Tricho loacetic acid was added to serum. Then the mixture was centrifuged during 20 minutes at 3000 rpm. After cen trifugation the samples were 10fold diluted with distilled water. After stirring, the measurement was performed at λ = 254 nm and at λ = 280 nm. The 280 nm / 254 nm absorbance ratio was calculated. Concentrations of middle molecules were expressed in standard units.

IMMUNOHISTOCHEMICAL EVALUATION OF HSP90α EXPRESSION IN THE SMALL INTESTINE
Tissue samples of small intestine were fixed in a 10% for malin solution. Then paraffinembedded tissues were used to obtain 4μmthick sections, which were immunostained using commercially available mouse monoclonal antibod ies to HSP90α purchased from Thermo Fischer Scientific (USA). After incubation with the primary antibodies, the microslides were treated with an anti(mouse IgG)-horse radish peroxidase conjugate. Visualization was carried out using 3,3´diaminobenzidine (DAB) staining. The presence of brown coloration indicated the positive reaction.

BIOETHICS
All the experimental procedures were performed fol lowing the guidelines of EU Directive 2010/63/EU on the protection of animals used for scientific purposes, which is based on the Council of Europe Convection for the Pro tection of Vertebrate Animals used for Experimental and other Scientific Purposes (ETS123).

STATISTICAL ANALYSIS
Numerical data were analyzed using the KruskalWallis ANOVA test if three independent parameters were com pared. It was followed by the Dunn's multiple comparison posthoc test. Two independent groups of variables were compared by a nonparametric MannWhitney U test. It was selected based on the outcome of the ShapiroWilk and KolmogorovSmirnov normality tests. Differences were considered statistically significant at p < 0.05. The data obtained in our research were analyzed with Graph Pad Prism 5.0 (GraphPad software, USA).

RESULTS
To assess the toxicity and proinflammatory potential of VNPs, we determined the levels IL1β, middle molecules, Creactive protein, and seromucoid in blood serum of rats from group C and compared them with the correspond ing parameters of animals from the control group D 2 . The concentrations of proinflammatory IL1β, middle molecules and acute phase proteins (seromucoid and Creactive protein) in blood serum of rats orally exposed to the solution of VNPs were statistically insignificant ly (p > 0.05) higher than in the animals from the control group D 2 (Table 1).
Morphological studies of small intestine in rats from both control groups demonstrated that the epithelial layer of villi was intact. Epithelial cells at the top of intestinal villi regenerated well. No significant leukocyte infiltrate was found (Figures 2, 3).
Immunolabelling allowed us to find out that HSP90α was primarily expressed in the cytosol. HSP90αpositive cells were detected both in the epithelial lining and intesti nal glands. Immunostaining was also observed in the lami Tab. 1 Evaluation of orally administered GdVO 4 nanoparticle toxicity (Me (IQR)).

Groups of animals
Group D 2 (intact animals, n = 10) Group C (rats orally exposed to VNPs, n = 10)  (Figures 2, 3). Evaluation of the impact of VNPs on small intestine morphology and HSP90α expression showed that the oral exposure of rats to nanoparticles affected neither histo logical features of the small intestine nor the chaperone expression. No signs of intestinal inflammation were re vealed in rats from group C. Epithelium and villi were not damaged. The leukocyte infiltrate was as nonabundant as in both control groups D 1 and D 2 (Figure 2).
In this study, we observed almost the same pattern of HSP90α expression in rats orally exposed to VNPs as in groups D 1 and D 2 . Qualitative analysis indicated that HSP90α was moderately expressed in the lamina propria, epithelial cells, and glands ( Figure 2).
We observed the statistically significant (p < 0.0001 and p = 0.005, respectively) 4.1fold and 1.8fold increase in the concentrations of circulating proinflammatory cytokines TNFα and IL1β in rats from group A compared with the control group D 1 ( Table 2). The content of antiinflamma tory IL10 did not differ from controls (p > 0.05). Levels of TNFα and IL1β in rats with carrageenaninduced in flammation treated with VNPs was higher than in group A. However, the difference was found to be statistically insig nificant. Circulating IL10 levels in animals from group B were statistically insignificantly (p > 0.05) higher than in both group D 1 and group A ( Table 2).

Blood serum parameters (units)
TNF-α (pg/ml) IL-1β (pg/ml) IL-10 (pg/ml) Note: Differences were considered statistically significant at p < 0.05 (* indicates the statistical significance of differences between two independent variables). p 1 is the difference between groups D 1 and A, while p 2 is the difference between groups A and B.  We demonstrated that the oral intake of carrageen ancontaining solution by animals from group A resulted in the development of intestinal inflammation, evidenced by the damage to the intestinal villi, especially at their top. Furthermore, the damaged intestinal villi lacked epithelial cells in some regions. The lamina propria both in the villi and at the level of glands was significantly infiltrated with macrophages ( Figure 3).

Groups of animals
Analysis of HSP90α immunostaining showed that the epithelial cells of villi were strongly labeled. Moreover, the significant HSP90α upregulation was detected in glandu lar epithelial cells, not only at the base of intestinal glands but also above. Both the number of HSP90αlabeled epi thelial cells and the intensity of immunostaining were higher in group A compared with controls ( Figure 3).
Administration of VNPs against the background of car rageenaninduced inflammation by rats from group B was associated with the leukocyte infiltration with the pre dominance of macrophages. The infiltration abundance in rats from group B did not differ significantly from group A. In addition, the villi with the undamaged epithelial lining were found. It is interesting to note that some regions of the intestinal wall contained villi with the destroyed tops, while the lower portions of villi were well epithelialized, indicating the rapid regeneration ( Figure 4).
Strong HSP90α staining was primarily observed at the top of villi. However, some villi were either not or weakly immunostained. In some regions, the moderate HSP90α expression was revealed.

DISCUSSION
Nanotechnology has already shown its significant poten tial in the field of medicine. Biomedical application of nanoparticles seems to be promising therapeutic agents due to their relatively small size and unique characteris tics (23,24). Nevertheless, the possibility of administer ing nanoparticles as drugs raises concerns regarding their adverse effects and probable toxicity. Thus, we evaluated safety and oral exposure risks of VNPs. Our findings indi cate that the oral consumption of VNPs during two weeks is not associated with the statistically significant changes in the content of circulating inflammatory markers such as IL1β, seromucoid, and Creactive proteins. Biochemical data are supported by the results of morphological stud ies. No morphological signs of intestinal inflammation were found in animals exposed to VNPs. Furthermore, the development of intoxication in response to VNPs oral consumption was not found, evidenced by the absence of middle molecules elevation in blood serum.
We also assessed expression of HSP90α, which is a molecular chaperone involved in the regulation of cellu lar proteostasis promoting protein folding and refolding in response to stress factors (25). It is worth mentioning that HSP90α is an isoform of the chaperone upregulated in stress conditions, while its β form is expressed consti tutively (26). It has been reported that HSP90α is upregu lated during inflammation (including the intestinal one) and in response to oxidative stress (27). No changes in its expression confirm the data of biochemical studies and indicate the absence of inflammation in the intestine after the oral consumption of VNPs.
Our biochemical and histological findings suggest that VNPs have no toxic effects when exposed orally at a dose of 20 μg/kg of weight during two weeks. Based on our findings, VNPs cannot be considered proinflammatory agents. Such conclusion is consistent with data of studies focused on elucidation of VNP properties and biological effects (28)(29)(30)(31).
The next task of our research was to assess the thera peutic potential of VNPs in intestinal inflammation caused by oral consumption of a carrageenancontaining solu tion. Carrageenans are sulfated hydrocolloids of polysac charide nature extracted from microalgae and used in food industry as thickeners, stabilizers, and emulsifiers (32). In addition, this food additive can trigger the development of intestinal inflammation as a result of its oral consumption by rats (17)(18)(19)(20). The development of inflammation in the rats from group A was confirmed in this study histologi cally and biochemically. Changes in the blood serum cy tokine profile observed in our study indicated the active inflammatory process in the intestine. We believe that ele vation of circulating proinflammatory TNFα and IL1β is mediated, at least partially by ROS, whose overexpression is known to be stimulated by carrageenan (33). In our pre vious study, we linked HSP90α intestinal overexpression revealed in this study with the development of oxidative stress in carrageenaninduced enteritis as well (17). This overexpression seems to be protective and aim at provid ing refolding of damaged protein to promote survival of enterocytes. . However, the epithelial layer below is preserved. Macrophage infiltration can be seen. Immunohistochemical reaction with antibodies to HSP90α. ×100. B) Very significant HSP90α labeling is found in the intestinal villi against the background of leukocyte infiltration. HSP90αpositive cells are marked with red arrows. Immunohistochemical reaction with antibodies to HSP90α. ×400. C). Strong HSP90α staining is revealed in the intestinal glands (marked with red arrows). Immunohistochemical reaction with antibodies to HSP90α. ×400. D) Fragments of the destroyed villi with strong HSP90α immunostaining are seen in the small intestinal lumen (marked with red arrows). Furthermore, the strongest HSP90α staining was found at the top of villi (marked with black arrows). Immunohistochemical reaction with antibodies to HSP90α. ×100.
VNPs did not stimulate the synthesis of antiinflamma tory IL10 and even worsened the imbalance between cir culating proinflammatory and antiinflammatory cyto kines, albeit the difference was statistically insignificant. Thus, we believe that VNPs at the dose used in our study does not affect the course of inflammation. It is worth not ing that that their oral consumption does not lead to the intensification of inflammatory response. Furthermore, the strongest HSP90α immunostaining in animals from group B is observed at the top of intestinal villi and seem to be compensatory. However, this was not sufficient to pro vide the cell survival and resulted in the reduced viability of cells and activation of cell death. In Figure 4 (D) we can notice such damaged villi alienated from the mucosa in the lumen of small intestine with strong HSP90α expression. In response to cell death, the regeneration should be acti vated. And we have managed to find the areas of extensive regeneration of enterocytes at the bottom of villi. We be lieve that such regeneration may be protective and can be associated with the action of VNPs. Such regions with so intense regeneration were not found in nontreated rats.

CONCLUSION
Oral exposure to VNPs at a dose of 20 μg/kg of weight by rats during two weeks showed no adverse effects. VNPs neither affect the level of circulating inflammatory mark ers nor influence the small intestinal morphology. Fur thermore, their oral intake was not associated with over expression of ROSinducable chaperone HSP90α in the intestinal mucosa. Evaluation of VNP therapeutic poten tial using an experimental model of carrageenaninduced enteritis demonstrated no significant effects on the course of inflammation. However, HSP90α overexpression in rats with carrageenaninduced intestinal inflammation treat ed with VNPs prevailed at the top of villi in a combination with the active proliferation at the bottom.

FUNDING
The study was not funded in any way.