Differences in Distribution and Biological Effects of F3O4@PEG Nanoparticles in Normotensive and Hypertensive Rats—Focus on Vascular Function and Liver

We investigate the distribution and biological effects of polyethylene glycol (PEG)-coated magnetite (Fe3O4@PEG) nanoparticles (~30 nm core size, ~51 nm hydrodynamic size, 2 mg Fe/kg/day, intravenously, for two days) in the aorta and liver of Wistar–Kyoto (WKY) and spontaneously hypertensive rats (SHR). Fe3O4@PEG had no effect on open-field behaviour but reduced the blood pressure (BP) of Fe3O4@PEG-treated SHR (SHRu) significantly, compared to both Fe3O4@PEG-treated WKY (WKYu) and saline-treated control SHR (SHRc). The Fe3O4@PEG content was significantly elevated in the aorta and liver of SHRu vs. WKYu. Nitric oxide synthase (NOS) activity was unaltered in the aorta, but significantly increased in the liver of SHRu vs. SHRc. In the aorta, Fe3O4@PEG treatment increased eNOS, iNOS, NRF2, and DMT1 gene expression (considered main effects). In the liver, Fe3O4@PEG significantly elevated eNOS and iNOS gene expression in SHRu vs. SHRc, as well as DMT1 and FTH1 gene expression (considered main effects). Noradrenaline-induced contractions of the femoral arteries were elevated, while endothelium-dependent contractions were reduced in SHRu vs. SHRc. No differences were found in these parameters in WKY rats. In conclusion, the results indicated that the altered haemodynamics in SHR affect the tissue distribution and selected biological effects of Fe3O4@PEG in the vasculature and liver, suggesting that caution should be taken when using iron oxide nanoparticles in hypertensive subjects.


Introduction
Metal nanoparticles (NPs), including ultra-small superparamagnetic iron oxide nanoparticles (USPIONs), possess strong potential for various biomedical applications [1][2][3]. It is well-known that the properties and biological effects of USPIONs are considerably dependent on their size, physicochemical factors, and surface modification.
Amongst metal NPs, magnetite (Fe 3 O 4 ) is used mainly due to its higher magnetization [4], while neutral biocompatible polyethylene glycol (PEG) is used as it has been approved by the Food and Drug Administration [5]. PEGylation has been shown to reduce the toxic effects of uncoated iron oxide NPs [6,7], including oxidative damage to DNA, proteins, and membrane lipids, as well as interference with the innate iron metabolism in mammals [2].

Experimental Design
In this study, we used normotensive WKY (n = 14) and spontaneously hypertensive (n = 16) male rats at the age of 14-16 weeks. All rats were born in the certified animal facility of the Institute of Normal and Pathological Physiology, Centre of Experimental Medicine, Slovak Academy of Sciences, in order to maintain a standardized environmental background for all animals. Rats were fed with pelleted chow for young rats deficient in phytoestrogens (Altromin 1314, variant P; Altromin Spezialfutter, Lage, Germany) until their ninth week of age. Afterward, rats were fed with Altromin 1324 chow (for adult rats). The iron content in both chows was 192.51 mg/kg. The pellet food and water were available ad libitum. Rats were housed under standard conditions at 22-24 • C, humidity 45-65%, and with a 12 h light/dark cycle.
Before the experiment, rats had a catheter implanted in the jugular vein, for infusion of USPIONs or saline (in the controls). The catheters were implanted under 2.5-3.5% isoflurane anaesthesia. The procedure for catheter implantation has been described in detail by Liskova et al. [19]. Rats were randomly divided into control WKY rats (WKYc), USPION-treated WKY rats (WKYu), control SHR rats (SHRc), and USPION-treated SHR rats (SHRu), with n = 7-8 per group. During the 10 min infusion, rats were placed in a plastic box with dark walls and a transparent lid (27 cm × 14 cm × 9 cm in size), which allowed for their free movement. USPIONs were administered intravenously on two consecutive days, at a dose of 2 mg Fe per kg of body weight per day. Rats were killed by decapitation after brief exposure to CO 2 (until the loss of consciousness) approximately 26 h after the second infusion, and trunk blood, tissues, urine, and faeces were collected for further analyses. Faeces were collected randomly from those excreted during the open-field test, keeping approximately the same time of excretion after USPION infusion. Serum was collected from trunk blood. Urine was collected from the bladder after decapitation. The design of the experiment is provided in Table 1. We used commercially available PEG-coated magnetite USPIONs. The USPIONs were purchased from Sigma-Aldrich (Bratislava, Slovakia, cat. No. 747408, PubChem SID 329765832, accessed on 18 March 2021). The iron content was 1 mg Fe/mL, and the concentration of NPs dispersed in water was 0.034 nmol/mL. The size of the USPIONs core (determined by transmission electron microscopy) was 28-32 nm, the zeta potential was -12 mV, the polydispersity index was 0.1, and the hydrodynamic size was about 45 nm (parameters declared by the manufacturer). USPIONs were autoclaved at 121 • C for 30 min. Before their use, the exact physico-chemical properties of the USPIONs were further determined, as published in detail in our earlier publication [20]. USPIONs were then dispersed with sterile saline to reach a final dose of 2 mg of Fe/kg of body weight in a final volume of 1 mL and infused intravenously into the jugular vein during the 10 min infusion using an infusion pump. Sterile saline (1 mL) was infused into control rats.
All of the chemicals used in this study were purchased from Sigma-Aldrich (Bratislava, Slovakia) and Merck Chemicals (Bratislava, Slovakia), unless stated otherwise.

Open-Field Test
Locomotor activity and anxiety-like behaviour were measured using an open-field test (OF) between 07:00-07:30 h, using the Any-maze (Stoelting Europe, Dublin, Ireland) video tracking system in 10 min trials. The testing conditions have been previously described in detail by Kluknavsky et al. [24]. The total distance travelled in the OF and immobility time were determined as the parameters of locomotor activity. As the markers of anxiety-like behaviour, the relative distance travelled in the central zone (calculated as the percentage of distance travelled in the central zone with respect to total distance travelled), time spent in the central zone (central time) and counts of entries into the central zone (central entries) were determined.

Systolic Blood Pressure and Heart Rate
The systolic BP and HR were measured in pre-conditioned conscious rats using the noninvasive tail-cuff plethysmography method, as previously described by Puzserova et al. [25]. Each value was calculated as the average of five measurements. The BP and HR values were measured at the beginning of the experiment (basal) and 24 h after the second infusion (end).

Nitric Oxide Synthase Activity
NOS activity was determined on the basis of conversion of [ 3 H]-L-arginine (specific activity 5 GBq/mmol,~100,000 dpm; ARC, St. Louis, MO, USA) to [ 3 H]-L-citrulline, using 20% tissue homogenates (w:v) of the brainstem, aorta, and liver through the method described in detail previously [19]. The enzyme activity was expressed as pkat/g of protein.
The protein concentration was determined using the Lowry method.
The total RNA of the samples was isolated by using PureZOL™ RNA Isolation Reagent (Bio-Rad, Hercules, CA, USA), according to the manufacturer's protocols. The amount of total isolated RNA was spectrophotometrically quantified at 260/280 nm and 260/230 nm, using a NanoDrop spectrophotometer (Thermo Scientific, Waltham, MA, USA). In the next step, the isolated RNA was reverse transcribed into cDNA using Eppendorf Mastercycler (Eppendorf AG, Elbmarsch, Germany) and the iScript™ cDNA Synthesis Kit (Bio-Rad, Hercules, CA, USA) reaction mixture, according to the manufacturer's instructions. Gene amplification was performed using qPCR on a CFX96 Real-Time PCR detection system (Bio-Rad, Hercules, CA, USA). SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) was used for gene amplification. Primer pairs used to amplify selected genes are listed in the Table 2. The inner arterial diameter of the femoral arteries was set to 90% of the diameter predicted for the pressure at 100 mmHg in the wire myograph. After 30 min of stabilisation, the arteries achieved their basal tones. To test the viability of isolated arteries, the arteries were incubated in a depolarising solution; that is, modified Krebs-Henseleit solution in which NaCl was exchanged for an equimolar (125 mmol/L K + ; KPSS) concentration of KCl for 2 min. After being washed out, the experimental protocol was carried out. After the washing and stabilisation of the basal tone, cumulative concentrations of 5-hydroxytryptamine (5-HT, 10 −8 -10 −5 mol/L) or noradrenaline (NA, 10 −8 -10 −4 mol/L) were used to determine concentration-response curves. This was followed by application of acetylcholine (ACh; 3 × 10 −8 and 10 −6 mol/L), in order to induce endothelium-dependent relaxations. After maximal ACh-induced relaxation was achieved, the endothelium-dependent contraction developed within the next 3 min, as shown in Figure 1a,b. EC 50 , E max , and slope were calculated using Hill's equation, and endothelium-dependent ACh-induced relaxations were calculated as the % of maximal 5-HT contraction. Endothelium-dependent ACh-induced contractions were calculated as the % of maximal ACh-induced relaxation for each ACh concentration used (3 × 10 −8 and 10 −6 mol/L).

Determination of the USPIONs Content
Determination of USPION-originated iron content in the serum, urine, faeces, aorta, and liver was performed by measuring their magnetic properties, as described previously by Skratek et al. [20]. USPION content in blood was calculated as a sum of serum content plus USPION content found previously in erythrocytes [26]. Briefly, a Quantum Design (San Diego, CA, USA) SQUID magnetometer MPMS-XL 7AC was used. Magnetic characterization of USPIONs, saline-treated control (both SHRc and WKYc), and Fe 3 O 4 @PEG-treated (SHRu and WKYu) was carried out by measuring the field dependence of the mass magnetization (M) at a temperature of 300 K, using an applied magnetic field up to 1 T. The USPION-originated iron content in the USPION-treated rats was calculated as M' = M SAMPLEu − M c (where M SAMPLEu is magnetization of the tissue sample of USPION-treated rat, and M c is the average magnetization of the given tissue from control rats) and determined by comparison to M USPION , which is magnetization of USPIONs with a declared concentration of iron [20]. This method distinguishes USPION-originated iron from iron naturally present in the tissues, on the basis of their different magnetic properties under the given experimental conditions. Tissues for magnetic measurements were dissected using ceramic scissors and forceps, cleaned of connective tissues, and immediately frozen in liquid nitrogen. All samples were kept at −80 • C until magnetometric measurements were performed.

Transmission Electron Microscopy of the Tissues
Aortae of WKYu and SHRu rats were cleaned of connective tissue (n = 3 for each group) and cut into 3 mm long rings. Liver tissue was cut into small pieces, about 2 mm 3 in size. Tissues were fixed by immersion in 2.5% glutaraldehyde in 0.1 mol/L cacodylate buffer (Serva, Heidelberg, Germany) at a pH of 7.4, for 3 h at 40 • C. After washing in the buffer, the samples were post-fixed in 1% OsO 4 for 30 min, dehydrated in an alcohol series, infiltrated in propylene oxide, and embedded in Epon 812 (Serva, Heidelberg, Germany). Ultrathin sections of the aortae were cut using a Leica ultramicrotome (Leica Microsysteme EM UC7, Wien, Austria), and then mounted on nickel grids [27]. Unstained sections were examined using Tesla BS 500 transmission electron microscope (Brno, Czech Republic), in order to detect the distribution of nanoparticles. Semi-quantitative evaluation was used to determine the differences in amount and distribution of USPIONs in the aorta and liver of WKYu and SHRu. Due to unaltered endothelium-dependent relaxation and increased NA-induced contractions, we focused on the presence of USPIONs in the tunica media of the aorta. For both tissues investigated, the score was defined from minus (−), indicating the absence of USPIONs, to (+++), as their highest amount.

Statistical Analysis
Statistical analyses of blood pressure (absolute values), heart rate (absolute values), and open-field behaviour were performed using three-way ANOVA for repeated measures, with measurement (basal and end) as a repeated factor. Treatment (saline in controls or USPIONs) and rat strain (WKY and SHR) were the independent factors. Two-way ANOVA (strain and treatment) was used for all other analyses (except for magnetometric measurements). All ANOVA analyses were followed with Bonferroni's post hoc test. Normality of the data distribution was tested using the Kolmogorov-Smirnov test. Magnetometric measurements were analysed by two-tailed Student's t-test. Correlations between variables were analysed using Pearson's correlation coefficient (r). The values were considered to differ significantly when p < 0.05. The results are presented as mean ± standard error of means (SEM). The GraphPad Prism v7.02 software (GraphPad Software, Inc., San Diego, CA, USA) and Statistica v13.5 (StatSoft Europe, Hamburg, Germany) was used for the statistical analyses.

Open-Field Behaviour
ANOVA revealed significant strain-dependent differences for total distance travelled  Figure 2d) in both strains; however, treatment and interaction of the factors were insignificant, and no differences in the above-mentioned parameters were found among the groups. Importantly, USPION-treatment did not alter the above-mentioned variables, compared to the control groups, in the respective rat strains.

Blood Pressure and Heart Rate
ANOVA revealed significant strain-dependent differences in BP of rats (F(1,26) = 32.22, p < 0.001) with significantly higher BP in SHR (Figure 3a). The main effect of treatment was not statistically significant. There was a significant effect of repeated BP measurement (F(1,26) = 28.71, p < 0.001). ANOVA also found a significant effect of interaction of strain, measurement, and treatment (F(1,26) = 7.29, p < 0.02). The BP of USPION-treated SHR was significantly reduced, compared to their basal levels, which was not observed in the SHRc (Figure 3a). Calculation of the BP changes relative to their basal BP levels (∆ BP in percentage of basal levels) showed a significant decrease in BP in the SHRu group vs. both WKYu and SHRc groups (Figure 3b). The HR of SHR was also significantly higher than in WKY (F(1,26) = 265, p < 0.0001; main effect of strain). The effects of measurement and treatment, as well as the interactions of all factors, were insignificant (Figure 3c). The relative changes of HR related to the basal levels (∆ HR in percentage of basal levels) did not differ among the groups (Figure 3d).

NO Synthase Activity
NOS activity was determined in the brainstem (due to its function in cardiovascular and motor control), liver (due to its function in the excretion/metabolism of USPIONs), and aorta (due to its function in cardiovascular regulation). ANOVA did not reveal significant differences in NOS activity in the brainstem ( Figure 4) and aorta (Figure 5a). In the liver, ANOVA revealed a significant effect of strain (F(1,24) = 50.64, p < 0.0001), with higher levels found in SHR. In addition, there was significant effect of interaction of strain and treatment (F(1,24) = 5.79, p < 0.03) with NOS activity, which was significantly elevated in SHRu vs. both SHRc and WKYu (Figure 6a).

Gene Expression
Two-way ANOVA revealed a significant effect of strain for eNOS gene expression in the aorta, with reduced mRNA levels found in SHR vs. WKY (F(1,23) = 6.73, p < 0.02; Figure 5b). In addition, eNOS and iNOS expressions were elevated in the aorta of USPIONtreated rats (eNOS: F(1,23) = 8.31, p < 0.01; iNOS: F(1,23) = 7.76, p < 0.02; Figure 5b,c). There was a significant positive correlation between eNOS and iNOS expression ( Figure 5d); however, eNOS expression correlated negatively with total NOS activity ( Figure 5e) and a trend of negative correlation was found between iNOS and total NOS activity in the aorta (Figure 5f).

5-HT
ACh-induced relaxations were investigated in the arterial segments after 5-HT-induced contraction. There were obvious strain-dependent differences in ACh-induced relaxations, which were reduced in SHR vs. WKY (F(1,23) = 14.58, p < 0.001; main effect of strain, Figure 10a). Higher ACh concentration resulted in higher relaxations (F(1,21) = 389.74, p < 0.0001; main effect of concentration) in both rat strains (Figure 10a).  As high ACh concentration is known to induce the release of endothelium-derived contracting factors [21,28], endothelium-dependent contractions of the arterial segments induced by high ACh dose were investigated (Figure 10b). There was a significant effect of strain (F(1,23) = 93.03, p < 0.0001; with higher contractions in SHR vs. WKY), and the effect of treatment (F(1,23) = 4.37, p < 0.05) with lower contractions in USPION-treated rats vs. controls. The effect of the interaction between strain and treatment was also significant (F(1,23) = 4.52, p < 0.05), and USPIONs reduced the endothelium-dependent contractions after both applications of ACh in SHRu vs. SHRc (p < 0.02), while no differences were found in WKY (Figure 10b).

Electron Microscopy
A semi-quantitative analysis of USPION content in the aorta and liver demonstrated a higher number of nanoparticles in both tissues of SHRu than in WKYu (Table 4). Electron microscopy revealed heterogeneous distributions of USPIONs in the aortic wall and liver of both strains. In WKY rats, USPIONs were observed in the elastic layers of the aorta (Figure 12a,b), while in the aorta of SHR, the USPIONs were localized in both the elastic layers and smooth muscle cells (Figure 12c,d). In the liver, USPIONs were present in the hepatocytes of both strains (Figure 9e,f).

Discussion
In this work, we evaluated the tissue distribution and selected biological effects of Fe 3 O 4 @PEG nanoparticles in normotensive WKY and spontaneously hypertensive rats. The main findings of our study were as follows: (1) USPIONs did not alter the spontaneous open-field behaviour in both rat strains; (2) hypertension was associated with increased deposition of USPIONs into the aortic vascular wall and liver; and (3) USPIONs elevated NA-induced contractions in isolated femoral artery, but reduced endothelium-dependent contractions and blood pressure in hypertensive rats. In addition, we found elevated NOS activity in the liver of SHRu, associated with elevated gene expression of eNOS, iNOS, and DMT1 in this organ, while no changes in NOS activities were found in the aorta and brainstem of any rat strain.
As first, we evaluated the possible USPION-induced behavioural changes in an openfield test. We found no changes in the parameters associated with locomotor activity and anxiety-like behaviour in WKY. Similarly, Askri et al. [29] have shown that sub-acute intranasal exposure to IONs did not affect the emotional behaviour, anxiety index, and learning/memory capacities of Wistar rats. In addition, low-dose IONs failed to alter behaviour in open-field and forced swim tests in Wistar rats [30]. On the other hand, high doses of uncoated IONs administered to mice i.p. once a week for 4 weeks led to damage to the blood-brain barrier, altering locomotor behaviour and spatial memory [31]. To our knowledge, this study is the first dealing with the behavioural effects of Fe 3 O 4 @PEG in SHR. Despite significant locomotor hyperactivity and damage to the blood-brain barrier, both of which are well-known in SHR [32][33][34], repeated i.v. administration of USPIONs did not alter the open-field behaviour of rats in terms of locomotion or parameters associated with anxiety-like behaviour. In addition, no changes in NO production were found in the brainstem, which was selected due to its roles in regulating both motoric functions and BP.
Regarding BP, USPIONs significantly reduced BP in SHR but not in WKY. In our previous studies, we found a delayed decrease in BP in acute stress-exposed USPION-treated WKY rats, but not in rats treated with USPIONs under control conditions [19]. Using polyacrylic acid-coated γ-Fe 2 O 3 NPs, Iversen et al. have reported a transient decrease in BP when determined 12-24 h after administration in mice [35]. In human studies, hypotension has been observed after administration of IONs to improve magnetic resonance imaging contrast [21]. On the other hand, acute intravenous administration of Fe 3 O 4 @PEG-Alendronate NPs to SHR rats did not produce changes in BP when investigated 100 min after acute i.v. infusion [36]. In this study, a decrease in BP occurred only in SHR 24 h post-treatment, suggesting possible changes in vascular functions.
For the investigation of vascular function, we used the femoral artery, the tone of which is regulated by both the endothelium-derived relaxing and contracting factors [19], making it ideal for the study of vascular responses in various experimental approaches. Previously, we found that an acutely infused lower dose of the USPIONs did not alter 5-HT-induced contractions and overall endothelium-dependent ACh-induced relaxations but elevated its NO-dependent component in WKY. Similarly, Fe 3 O 4 @PEG-alendronatecoated NPs did not damage the vascular function of the femoral artery in SHR after a single infusion [36]. In this study, 5-HT-induced contractions were unaltered by repeated USPION administration, while NA-induced contractions were significantly increased, compared to the control SHR, suggesting elevated sympathetic nervous system tone. On the other hand, ACh-induced endothelium-dependent contractions were reduced in USPION-treated SHR, possibly indicating a compensatory mechanism. Regarding USPION localisation, we showed that USPIONs were present only in a smaller number, and dominantly in the elastic layers, in USPION-treated WKY rats. However, in SHR, USPIONs were present in a greater amount in both the elastic layers and vascular smooth muscle cells. The elevated amount of USPION-originated iron in the aorta of SHRu, compared to WKYu was also confirmed using a magnetometric method. This may result from hypertension, which can alter or damage cell-to-cell connections [37], thus allowing for increased incorporation of USPIONs into the vascular wall. In addition, the permeability of vascular smooth muscle cell membranes may also be increased under hypertensive conditions, which may allow USPIONs to enter the vascular smooth muscle cells. In agreement with these findings, we also found elevated gene expression of DMT1 (involved in the transport of divalent metals, including Fe 2+ ), as well as nuclear factors NRF2 and PPARγ-which are involved in the regulation of various genes, including those involved in antioxidant defence and iron metabolism [14,38,39]-when calculated as the main effect of treatment, in both rat strains. Such changes were associated with elevated intracellular levels of iron. We also observed a significant main effect of USPIONs on eNOS and iNOS gene expression in the aorta, both of which were negatively correlated with total NOS activity (which remained unchanged). As the expression of SOD1 and SOD2 were also unchanged, USPIONs supposedly did not induce elevated superoxide production, which could react with NO. Thus, NO bioavailability should remain unchanged, which was supported by the finding of unchanged ACh-induced relaxation in the femoral artery. Collectively, despite higher incorporation of USPIONs into the vascular wall and smooth muscle cells in SHR, there were no changes in NO bioavailability in the aorta and ACh-induced relaxations in the femoral artery. Thus, the above-mentioned alterations in the arteries cannot explain the BP decrease in USPION-treated SHR, without the contribution of other factors.
Such an important factor could be alterations in the liver vascular bed and renal and hepatic function. The liver is well-known as the main organ for excretion of USPIONs of the size used in this study. However, the decomposition of USPIONs can result in renal excretion, which we observed in both rat strains. This, however, was not associated with significant kidney damage as we have published previously [26]. Regarding the liver, we found that NO synthase activity was significantly elevated in the USPION-treated SHR, which was correlated positively with elevated iNOS and eNOS gene expression. Thus, a factor considerably contributing to the decrease in BP in USPION-treated rats may be elevated relaxation in the hepatic vasculature resulting from elevated NO production by both eNOS and iNOS [23], further reducing peripheral vascular resistance. In addition, elevated expression of PPARγ may suggest early stages of liver damage [17,18].
We also found that USPIONs increased the gene expression of DMT1 in SHRu and elevated gene expression of FTH1 (main effect of treatment) in the liver, which may suggest elevated iron levels in this organ. Interestingly, the gene expression of HAMP gene, encoding hepcidin, a main hormone produced in the liver that regulates iron absorption in enterocytes and other tissues, was unaltered in both rat strains. On the basis of these findings, we can assume that USPIONs did not modify the signal pathways involved in hepcidin release in the liver at this stage of their action, as iron from (at least partially) decomposed USPIONs was successfully trapped in ferritin cores and/or excreted. The unchanged hepcidin levels were also in agreement with our finding of unaltered blood content of USPIONs in SHRu. Importantly, our previous studies showed a reduced amount of USPION-originated iron in erythrocytes in SHRu compared to WKYu [26]. This is in agreement with the trend of elevated USPION levels in sera of SHRu in this study, suggesting differences in erythrocyte membrane properties resulting in higher binding of USPIONs on erythrocytes of WKY rats, compared to SHR. Higher amount of USPIONs unbound to erythrocytes can, then, lead to elevated USPION incorporation into the tissues. The higher amount of USPIONs in the bloodstream of SHR may also lead to partially elevated renal and hepatic excretion of USPIONs in SHRu. However, the absolute amounts of USPION-originated iron in renal and hepatic excretions from both rat strains remain unknown, as we did not collect all urine and faeces during the experiment for calculation of these parameters.

Conclusions
In conclusion, our results showed an altered distribution of USPIONs in SHRu, compared to WKYu; namely, elevated USPION incorporation into the vascular wall and the liver, supposedly as a result of a higher amount of erythrocyte-unbound NPs in the bloodstream of SHR. This could lead to increase in NO production in the liver, due to the activation of iNOS and eNOS gene expression, resulting in lower peripheral vascular resistance associated with BP decrease only in SHR. Thus, our findings suggest caution when using Fe 3 O 4 @PEG-coated NPs in hypertensive subjects, as they may produce a BP decrease.