Preparation of Hydrophilic Iron Oxide Nanoparticles
It is well-established that surface modifications of IONPs could not only effectively enhance the stability of nanoparticles but also potentially affect the biocompatibility and pharmacokinetics of nanoparticles in vivo [30]. Some studies have found that positively-charged IONPs possessed a higher affinity to attach to the cell membranes, and they were more likely to be internalized in much larger amounts compared with negatively-charged IONPs [31]. Moreover, greater bioaccumulation, more protein adsorption and significant toxicity were also observed in positively-charged IONPs [31, 32]. Hypersensitivity reaction reported in most iron-based agents has recently attracted close attention [33, 34]. Attractively, lower probability of hypersensitivity was benefited from the high carboxyl group coating density of negatively-charged IONPs [35]. According to the aforementioned discoveries, a kind of low molecular weight poly (acrylic acid) (PAA) was selected to decorate and manufacture the monodisperse ultrasmall iron oxide nanoparticles via a modified protocol. The as-prepared 6 nm-diameter oleic acid-capped iron oxide nanoparticles initially dispersed in chloroform (Additional file 1: Fig. S1) [28]. After ligand exchange on nanoparticle surfaces by PAA, the hydrophilic PAA-capped iron oxide nanoparticles (IONPs) were obtained [29]. IONPs dispersed into aqueous solution were then examined by transmission electron microscopy (TEM) and dynamic light scattering (DLS) to demonstrate uniformity of size and distribution. TEM revealed the particle size of IONPs was ~6 nm (Fig. 2a), and the hydrodynamic diameter in aqueous solution measured by DLS was 8.62 ± 2.18 nm (Fig. 1b), validating its excellent monodispersion and stability. The overall zeta potential of IONPs (-22 mV) was measured using zeta seizer, indicating successful surface functionalization of the as-prepared hydrophilic nanoparticles. The FT-IR spectra further confirmed the surface modification of IONPs by PAA (Additional file 2: Fig. S2). The adsorption bands at 2923 and 2853 cm−1 were attributed to the -CH2- groups [36]. The peak at 1718 cm−1 was assigned to the carbonyl group (C=O) [29]. The appearance of two peaks at 3440 and 1631 cm−1 were ascribed to the -OH bending vibration, which was consistent with previously reported PAA-coated iron oxide nanoparticles [29, 36, 37]. The X-ray diffraction (XRD) pattern of IONPs was illustrated in Fig. 2c, which corresponded to magnetite (JCPDS 19-0629). IONPs dispersed in aqueous solution could form a stable suspension, and it showed a good response to the magnet (inset of Fig. 2d). A SQUID test was carried out to investigate the magnetic property of the acquired IONPs. As shown in Fig. 2d, IONPs had a saturation magnetization (Ms) of 65.95 emu/g Fe at room temperature.
Relaxometric Property of IONPs
To ascertain the MRI relaxation property of IONPs, the transverse relaxation time of IONPs with different Fe concentrations was measured by a clinical 3.0 T MRI scanner. An enhancement in the T2-weighted MR signals was observed as the concentration of IONPs increased (from 0.02 to 0.08 mM of Fe), leading to a raise in the signal of the corresponding MR images (Fig. 2e). The corresponding transverse relaxivity (r2) of IONPs was calculated to be 231.49 mM−1s−1, indicating a good MR imaging behavior (Fig. 2f).
Investigation of the Biocompatibility of IONPs in vivo
Intravenous injection is the most routinely used approach for administration of IONPs as MRI contrast agents, and intravenously injected IONPs could also circulate through the bloodstream to multiple organs. In general, IONPs are selectively taken up by the liver and spleen [17, 18, 38], while a few reports also documented the retention or elimination of IONPs in the lung or kidney [39, 40], subsequently raising the toxic concerns in major organs. As IONPs were injected intravenously, the hemolysis rate of IONPs was firstly evaluated, and the rate was calculated to be almost zero, indicating the excellent haemocompatibility of IONPs and promising intravenous administration approach (Additional file 3: Fig. S3). Secondly, the potential toxicity of IONPs to major organs has been evaluated. Body weight (BW) is one of the simple, intuitive and effective indexes to reflect systemic toxicity, as shown in Fig. 3a, compared with the PBS group, there was no BW loss in the IONPs group treated with different doses. The BW of mice injected with high dose of IONPs (26.97 ± 0.34 g) on the first day were equivalent to the control group (26.87 ± 0.43 g), and on the 28th day, the BW of mice administrated with high-dose IONPs (38.57 ± 1.22 g) were even slightly heavier than the control group (36.29 ± 0.54 g), which exhibited their negligible systemic toxicity in vivo. To lend more direct evidence regarding IONPs on specific major organs, the accumulation profile of IONPs in heart, liver, spleen, lung, kidney and brain was depicted inductively coupled plasma atomic emission spectrometer (ICP-AES) on 1, 3 and 7 days after intravenous injection of IONPs. As a result, the average amounts of Fe ions retained in organs were 22.63 % ID/g (liver), 8.98 % ID/g (spleen), 8.22 % ID/g (lung), 3.92 % ID/g (heart), 1.63 % ID/g (kidney) and 0.56% ID/g (brain) in descending order 24 h after initial injection (Fig. 3b). Obviously, IONPs were mainly accumulated in liver, spleen and lung, while negligible amounts were distributed in brain, which process may be hampered by blood-brain barrier (BBB). With the increase of time, the concentration of Fe ions in the tissues decreased rapidly. For example, the amount of Fe ions in liver reduced from 22.63 % ID/g of the first day to 4.07 % ID/g of the seventh day, and the decrease trend was also observed in the other organs. Later on, the histopathological examination (hematoxylin-eosin staining, H&E staining) was utilized to support the biosafety profile of IONPs in major organs (Fig. 3c and Additional file 4-7: Fig. S4-7), in which no abnormal changes in the pathological cellular structures or hyperemia, edema, cell death and other obvious injuries were observed, indicating no significant risk of the IONPs to ICR mice following intravenous injection. In details, normal lung parenchyma was observed, and the alveoli presented a vacuolated thin-walled structure. In the hepatic tissues, hepatocytes arranged in radial lines around the central vein. Meantime, the red pulp and follicles of the white pulp were well organized, indicating physicological spleen tissues. For the kidney, the mesangium appeared normal without obvious damage and the glomerular capillary or Bowman's capsule was observed in a typical nephron. Lastly, the myocardium structure was complete and myocardial cell morphology was normal. In addition, the most important serum biochemical parameters, including blood urine nitrogen (BUN), creatinine (CRE), aspartate aminotransferase (AST) and alanine aminotransferase (ALT), were used to present the function of main organs after IONPs administration (Fig. 3d-g). The results showed that these laboratory parameters in IONPs-treated mice with doses from 2.5 mg/kg to 20 mg/kg at each timepoints (up to the 28th day) were within the normal range, in other words, they have not significantly altered compared with the control group. In addition, there were no significant differences among each groups treated with low-, middle- and high-dose IONPs, respectively. For instance, no obvious increase in ALT value was documented on the first day from low-dose group (17.18 ± 2.17 U/L) to medium-dose group (17.58 ± 1.31 U/L) or high-dose group (16.29 ± 2.07 U/L). Collectively, all the above results indicated that there were no obvious toxicities to major organs of IONPs, which was in consistent with the previous reports regarding systematic toxicity evaluation of IONPs [6–8].
Evaluation of the Sperm Quantity and Quality in vivo
Reproductive diseases have raised growing concerns worldwide and the male factor accounts for a considerable proportion of the problems. Sperm cell is responsible for carrying the paternal genetic complement to the oocyte and forming an euploid zygote, which holds great importance to maintain normal fertility [41]. However, compared with somatic cells, spermatozoa are relatively vulnerable to oxidative stress, external stimuli and many other factors, owing to their limited capability of antioxidant protection and DNA repair mechanisms [42]. Therefore, the impact of IONPs on male reproductive system needs to be carefully evaluated, in addition to the systematic toxicity assessment of the aforementioned major organs. Accordingly, the semen analysis profiles on male ICR mice intravenously (i.v.) injected with IONPs at low-, middle- or high-dose (2.5 mg/kg, 10 mg/kg and 20 mg/kg, respectively) were closely monitored for 28 days. As a result, the amount of normal sperm in the epididymis significantly reduced in a dose-dependent manner. On the first day after injection, the number of sperm dropped precipitously, we could only observe the presence of a small amount of sperm (approximately 21.7 % of the normal value) in low-dose group and even hardly counted the sperm (less than 8 % of the normal value) in high-dose group (Fig. 4a). This dose-dependent reduction of sperm quantities also appeared on the 3rd and 7th days after initial injection, although the sperm counts have gradually increased, as compared with the first day post-injection. For instance, the sperm counts in epididymal tissue suspensions were approximately 7.1 %, 14.4 % and 25.6 % of the normal value on the first, third, and seventh day after high-dose injection, respectively (Fig. 4b). Notably, the sperm quantity has returned to the physiological range on the 14th day, and kept normal until the observed duration of 28 days (Fig. 4a, b). As for male reproductive function, sperm quality possesses equal or even more importance than sperm quantity [43]. Thus, sperm qualities including viability and motility were evaluated following IONPs injection. Consequently, in consistent with the alternation in sperm quantities, IONPs impaired sperm qualities in a dose-dependent manner at each checkpoints of the first, third, seventh day post-treatment, and the harmful effect achieved the peak on the first day after low-, middle- or high-dose injection. Concretely, on the first day after injection of IONPs, the sperm viability rate of the low-dose group (28.56 ± 1.14 %) was significantly lower than that of the control group (49.79 ± 0.57 %), let alone the middle-dose group (17.56 ± 0.81 %) or the high-dose group (8.75 ± 0.32 %) (Fig. 4c). At the meantime, sperm motility showed almost identical changing pattern with sperm viability, no matter which doses of IONPs were administrated in each timepoints. For instance, the motility rate was extremely lower in the high-dose group (2.74 ± 0.38 %), compared to that in the control normal group (39.43 ± 0.77 %) (Fig. 4d). Notably, in accordance with that the sperm counts gradually increased and returned to the normal range on the 14th day, the sperm viability and motility also recovered to the normal range in the same duration, suggesting that the toxicities of IONPs to sperm quantity and quality were temporary and could be reversible within 2 weeks. In consideration that the sperm were released from the epididymal fragment to the buffer followed by measurement [44], sperm with normal viability and motility in the control group could almost entirely release from the epididymis into the buffer, while in IONPs-injected groups, the total counted number of sperm might be lower than it actually was, demonstrating the proportion of functional sperms should be further lower than currently estimated. Collectively, these findings raised an important clinical implication that there would be a temporary but recoverable loss of sperm quantity and quality after intravenous injection of IONPs. If a male patient has received MRI examination with IONPs as contrast agents, the fertilized behavior should be delayed until the semen parameters are recovered, which may need some duration.
Evaluation of Hormones and Testicular Indexes after IONPs Accumulation
Sperms are produced in seminiferous tubules of the testis, followed by being transported and stored in the epididymis for leaving the male body (Fig. 4e). Therefore, the damage to testis or/and epididymis may be potential causes of the decline of sperm quantity and quality after injection with IONPs. Normal testicular function, being dependent upon hormones acting through endocrine or paracrine manners, is essential for germ cell homeostasis in vivo [45]. Among these hormones, follicle-stimulating hormone (FSH), luteinizing hormone (LH) and testosterone (T) are particularly important to male germ cell fate, as the disorder of them would induce aberrant sperm parameters and even germ cell apoptosis [46]. To explore whether the reduction of sperm quantity and quality was associated with the abnormal levels of hormones in the mice injected with IONPs, we measured FSH, LH and testosterone, respectively. The results revealed the three hormones in mice treated with IONPs fluctuated within the normal ranges (Fig. 5a-c). Specifically, the levels of FSH (29.99 ± 0.99 ng/mL) and LH (5.93 ± 0.08 pg/mL) in mice treated with 20 mg/kg IONPs at the first day were almost equivalent to the control group (31.40 ± 0.53 ng/mL and 5.87 ± 0.04 pg/mL). In addition to these two hormones that secreted by pituitary gland, testosterone also presented the negligible alternation between the IONPs-injected group (2045.09 ± 13.14 pg/mL) and the control group (2084.35 ± 14.48 pg/mL) at the first day. Therefore, the normal hormone levels indicated that the damage to the testis of IONPs was extremely limited or even absent. Later on, to further explore the cause of the sperm loss and disability, we observed the testicular size, weight and shape in each groups. The results showed that there were no significant changes in testicular size and shape of the mice injected with IONPs with the comparable control group (Fig. 5d). Concretely, testicular weight increased from 105.80 ± 2.15 mg of the first day to 142.20 ± 2.87 mg of the 28th day after 20 mg/kg IONPs injection, which tendency was also observed in the other groups (Additional file 8: Fig. S8). To exclude the effect of age on testicular weight and more intuitively understand testicular changes, testicular index was introduced to further evaluation, which also verified the negligible differences among each groups (Fig. 5e), indicating that the macro-damage of IONPs to the testis was ignorable.
Unlike most other organs, testis possesses a unique structure called blood-testis barrier (BTB) that is composed of connective tissue, capillary endothelium, basal membrane of spermatogenic epithelium and Sertoli cell tight junction, which could not only preventing sperm antigens from escaping outside the spermatogenic tubules but also avoiding harmful substances from entering the testis to maintain environmental homeostasis for sperm development [47]. Therefore, to identify whether IONPs could enter and accumulate in the testis, ICP-AES was used to detect the abundance of Fe ions. Compared with the control group, there was no additional accumulation of Fe ions in the testis of mice injected with IONPs at each timepoints (Fig. 5f). Even at the first day after injection of IONPs that the sperm quantity and quality decreased most obviously, the concentration of Fe ions in the control group (32.75 ± 0.71 µg/g) was not significantly higher than that in the low-, middle-, high-dose groups administrated with IONPs (31.60 ± 2.20 µg/g, 29.73 ± 1.20 µg/g and 32.86 ± 1.26 µg/g, respectively). These results confirmed that IONPs have not entered into the testis.
Observation of the Histological Changes in Testis after IONPs Accumulation
The index to reflect the spermiogenesis is an important reference for judging testicular tissue injury, but not the only criterion. In addition to the detection of the objective laboratory indicators, slight injury which may cause the subtle changes of the testis, should also be carefully assessed. Accordingly, the H&E staining was used to display the structures of the testis. As a result, the structures have not shown significant differences between the control group with low- and middle-dose IONPs injection groups, while a few scattered abnormal cells were observed in the testis of mice with high-dose injection of IONPs on the first day (Fig. 5g), suggesting high-dose IONPs might cause the apoptotic cell death in testis. Johnsen score is routinely used for histologically grading in testis H&E staining, which is an effective and quantitative method to evaluate histological features and morphological changes to various pathological factors affecting testicular cells, mostly when there is no immediate or only minimal clinical alteration [48]. In this study, we used the standard Johnsen scoring system to evaluate the pathological changes of testis. As shown in Additional file 9: Fig. S9, at the first day after injection with low- and middle-dose IONPs, the Johnsen scores of the testis were both 9.67 ± 0.21, almost the same with that of the control group, while after the high-dose IONPs injection (20 mg/kg), the Johnsen score of the testis (9.17 ± 0.31) was significantly lower than the control group (9.83 ± 0.17), suggesting slightly impaired spermatogenesis, which was characterized by increased late spermatids and disorganized epithelium in testis [48]. Consistent with the observation of sperm parameters and other indicators, the most severe damage occurred at the first day after IONPs injection and gradually recovered within 2 weeks (Additional file 9: Fig. S9). However, as the presence of BTB prevented the entry of IONPs into the testis, it was quite interesting to explore why did the apoptotic cells appear after high-dose IONPs injection. In testis, there is a population of cells named Sertoli cell, which has receptors for the hormonal regulators of spermatogenesis such as FSH and testosterone, and also provides paracrine factors, nutrients, cytokines and biologically active peptides that essential for supporting spermatogenesis and fertility [49]. Notably, BTB is created by adjacent Sertoli cells near the basement membrane to divide the seminiferous epithelium into the basal and adluminal (apical) compartments, and subsequently serves for successful postmeiotic spermatid development in the apical compartment [50]. Sertoli cells on the front lines might be damaged by those toxic substances and lead to functional impairment or even cell death in the testis. As no morphological abnormalities in the Sertoli cells were observed from H&E staining of testis, we hypothesized that the extremely low-amount IONPs have entered into the Sertoli cells and impaired their function, consequently leading to the apoptosis of a few spermatogonia. To verify this hypothesis, we have detected the markers near the basal membrane to evaluate the function of Sertoli cells. Among these markers, Occludin is a tight junction protein, which participates in forming the paracellular barrier that mediates the communications of substances in the intercellular compartments between Sertoli cells [51], and N-cadherin is responsible for cell adhesion and recognition [52]. The immunohistochemical (IHC) images revealed that IONPs injection decreased the local expression and density of the markers (Fig. 5h, i). In addition, the tight junction supported by Occludin became discontinuous, which further indicated that the function of the Sertoli cells was partially impaired (Fig. 5i). Collectively, Sertoli cells internalized a tiny amount of IONPs when preventing them from entering the testis, which decreased the ability to nourish the spermatogonia and resulted in a few apoptotic spermatogonia. Notably, although low- and middle-dose administration of IONPs in mice elicited significant sperm quantity and quality decline within seven days (Fig. 4a), extremely limited or no impairment of Sertoli cells has been observed in those mice, suggesting the abnormal sperm parameters elicited by IONPs, to a large extent, were not attributed to the limited Sertoli cells impairment.
Investigation of the Toxicity of IONPs to Epididymis in vivo
As mentioned above, the damage to testis or/and epididymides may be potential causes of the abnormal sperm parameters in mice injected with IONPs. Since testis has not been influenced by IONPs obviously, it makes sense to explore whether epididymidis has been impaired after IONPs injection and subsequently caused sperm quantity and quality reduction. Immature sperms from the testis will be transported into the epididymidis to form mature sperms followed by storage during asexual activity. Thus, the significant reduction of sperms in epididymidis of mice injected with IONPs at the first day was probably not correlated with the decreased production of immature sperms or the presence of very few apoptotic cells in the testis. In contrast, we speculated that the reduction of sperm count in epididymidis might be attributed to the direct toxicity of IONPs, as the deterioration of the epididymal environment could also lead to the death of mature sperms. Herein, epididymis index was firstly measured to evaluate epididymal changes. At the first day after injection of IONPs, there were no statistical differences of epididymis index between the control group (mean value, 0.12 %) and even the high-dose IONPs injection group (mean value, 0.10 %) (P > 0.05). With the growth of mice, on the 28th day after injection, the epididymis index in all mice increased, but there were still no statistical differences between the control group with each IONPs-injected groups (P > 0.05) (Fig. 6a). In addition, we have assessed the weight change of epididymidis and recorded the remarkably similar values between the control group with each administration groups (P > 0.05) (Additional file 10: Fig. S10). Moreover, due to the absence of BTB, we speculated that IONPs could directly enter the epididymidis and subsequently accumulated, leading to the death of sperms. This hypothesis was confirmed via the ICP-AES analysis, demonstrated by that the concentration of Fe ions rapidly accumulated in the epididymidis to a maximum value of 44.15 ± 5.20 % ID/g after 1 h, while rapidly dropping to an extremely low level within 24 h (1.99 ± 0.35 % ID/g) (Fig. 6b). The tendency was coincided with the decline and recovery time of sperm count and suggested that the decrease of sperms in the epididymidis was tightly correlated with the accumulation of IONPs, characterized by Fe ions in ICP-AES analysis. In addition to the evaluation of the quantity and quality by analyzing sperms in the suspension, H&E staining was also applied to directly observe sperms in the epididymis. The presence of sperms in the epididymal lumen was hardly observed at the first day after injection with IONPs. As time increased, the number of sperms continued to increase until it returned to the normal range at the 14th day after administration (Fig. 6c). These results were consistent with the parameters revealed by a computer-assisted semen analysis system (Fig. 4a-d). Notably, previous studies have shown that the accumulation of intracellular Fe ions could augment the Fenton reaction to produce a large number of harmful oxidative products [53]. Therefore, oxidative stress may be the underlying mechanism of sperm death. Malondialdehyde (MDA) is well-known as one of the final products of polyunsaturated fatty acids peroxidation and the increased oxidative stress would cause the accumulation of MDA [54]. However, superoxide dismutase (SOD) is commonly regarded as a natural scavenger of free radicals and main antioxidant enzyme to fight against oxidative stress in the body [55]. To assess the the extent of oxidative stress in epididymidis, the levels of MDA and the activities of SOD were measured (Fig. 6d, e). In the group injected with high-dose (20 mg/mL) IONPs, the MDA level was 18.26 ± 0.57 mmol/mgprot, which was significantly higher than the value of the control group (7.93 ± 0.78 mmol/mgprot) (Fig. 6d). Meanwhile, the SOD activity of 43.05 ± 3.10 U/mgprot in the control group decreased to 26.37 ± 0.82 U/mgprot in the high-dose IONPs injection group (Fig. 6e). These results together suggested that IONPs elevated the oxidative stress in the epididymidis. To further evaluate the damage to epididymidis after IONPs injection, TUNEL staining was utilized to explore the underlying details. As shown in Fig. 6f, a considerable amount of apoptotic cells were observed in the epididymal lumen after the high-dose IONPs administration. The shape and size of apoptotic cells were similar with the abnormal cells indicated by H&E staining of testis, which suggested that the apoptotic cells observed by TUNEL staining may originate from the testis. However, for the epididymal tissues themselves, such as epididymal epithelium, mesenchymal cells, no apoptotic cells were observed. This phenomenon further confirmed that sperms were more sensitive to harmful substances than the other somatic cells. Collectively, we have identified that the elevated oxidative stress levels caused by the accumulation of IONPs in the epididymidis might be one of the underlying mechanisms for the rapid decline of sperm parameters. Given that the sperm damage was temporary and reversible, it may provide some important implications in clinical practice. It is well established that the spermatogenic cycle is critical for continuous sperm production, and the spermatogenic cycle of mice is eight weeks [56]. In consideration of that the reduction of sperm quantity and quality caused by the IONPs injection was able to completely returned to the normal range within two weeks, we speculated that the sperm parameters would be free of abnormalities within a quarter of spermatogenic cycle in male patients who have received MRI examination with IONPs injection, in other words, the fertilized behavior of them was strongly recommended after about 2.5 weeks, which is a quarter of spermatogenic cycle of human beings (10 weeks) [57].