In vivo comparison of the biodistribution and toxicity of InP/ZnS quantum dots with different surface functional groups

: Background: Indium phosphide (InP) quantum dots (QDs) have shown a broad application prospect in the fields of biophotonics and nanomedicine. However, the potential toxicity of InP QDs has not been systematically evaluated. In particular, the effects of different surface modifications on the biodistribution and toxicity of InP QDs is still unknown, which hinders their further developments. The present study aims to investigate the biodistribution and in vivo toxicity of three types of InP/ZnS QDs with different surface modifications, hQDs (QDs-OH), aQDs (QDs-NH 2 ), and cQDs (QDs-COOH) respectively. These InP/ZnS QDs with 2.5 mg/kg BW or 25 mg/kg BW were intravenously injected into BALB/c mice. Biodistribution of three QDs was determined through cryosection fluorescence microscopy and ICP-MS analysis. The subsequent effects of InP/ZnS QDs on histopathology, hematology and blood biochemistry were evaluated at 1, 3, 7, 14 and 28 days post-injection. Results: The results showed these types of InP/ZnS QDs QDs were rapidly distributed in the major organs of mice, mainly in the liver and spleen, and lasted for 28 days. No abnormal behavior, weight change or organ index were observed during the whole observation period, except that 2 mice died on Day 1 after 25 mg/kg BW hQDs treatment. The results of H&E staining showed that no obvious histopathological abnormalities were observed in the main organs (including heart, liver, spleen, lung, kidney, and brain ) of all mice injected with different surface functionalized QDs. Low concentration exposure of three QDs hardly caused obvious toxicity, while high concentration exposure of the three QDs could cause some changes in hematological parameters or biochemical parameters related to liver function or cardiac function. More attention need to be paid on cQDs as high dose exposure of cQDs induced death, acute inflammatory reaction and slight changes in liver function in mice. Conclusions: Generally speaking, the surface modification and exposure dose can influence the biological behavior and in vivo toxicity of QDs. The surface chemistry


Background
In the past few years, nanomaterials have attracted widespread interest owing to their unique magnetic, optical, thermal or conductive properties. Quantum dots (QDs), as a kind of interesting semiconductor nanomaterials, are composed of group Ⅱ -Ⅳ or Ⅲ -Ⅴ elements with the diameter of 1-10 nm. They are increasingly used in medical and pharmaceutical science due to their properties, such as broad absorption spectra, narrow tunable photoluminescence (PL) spectra, high quantum yield, light bleaching resistance, and large specific surface area [1; 2]. These advantages enable QDs to be used in the field of biological imaging, drug deliver and diagnostics after they are coupled with different biomolecules [3]. Despite the numerous benefits provided by QDs, people still have doubts about their potential harmful health effects, which are related to the heavy metals such as Cd, Pb, As, and Te in these materials [4; 5].
Considering the inherent toxicity of Cd-based QDs, it is critical to find other kinds of safer QDs.
Among the Cd-free alternatives, indium phosphide (InP) QDs have shown great potential as a replacement for Cd-based QDs. Because InP QDs do not contain heavy metal elements, show stable quantum yield and size-tunable PL emission from visible to near infrared (NIR) range, InP QDs are beneficiary in biological applications [6; 7].
For example, Zhang et al carried out in vivo imaging of tumor bearing nude mice with silica medium composite probe encapsulated InP/ZnS QDs, and found that the composite probe had excellent tumor targeting and fluorescence imaging capabilities [8]. More importantly, different from the ionic bond in CdSe QDs, the covalent bond in InP QDs is stronger and more robust, which make them less toxicity [9; 10]. Some studies have proved the biosafety of InP QDs. For instance,. Yaghini et al. found that InP QDs mainly accumulated in the liver and spleen of rats, with no abvious organ damage, histopathological lesions or serum biochemical changes when the rats were injected intravenously with InP QDs at the dose of 12.5mg/kg body weight (BW) or 50mg/kg BW [11]. Brunetti et al. found that the toxicity of InP/ZnS QDs was much lower than that of CdSe/ZnS QDs by comparing their toxicity in vitro and in vivo (animal model Drosophila), and they considered InP/ZnS QDs were safer alternatives to CdSe/ZnS QDs [12]. However, some published studies have provided evidence of biological damage caused by InP QDs. For example, Yamazaki et al. investigated the survival of the Syrian golden hamster for two years, during which they were given 3 mg/kg InP particles intratracheally twice a week for 8 weeks. Severe pulmonary inflammation and localized bronchioloalveolar cell proliferation was observed after the last administration [13]. Chen et al. reported InP/ZnS QDs could cause the deformation and death during the development of Chinese rare minnow embryos, although the effects were weaker than Cd-based QDs and CuInS/ZnS QDs [14].
As we all know, QDs are possibly exposed to organisms through many routes, including lung inhalation, ingestion, skin contact, and even intravenous injection during the process of production, application and wasting [15]. Then they may circulate, metabolize, excrete or accumulate in the body, and produce varing levels of biological toxic effects on organisms. The in vivo biological features of QDs mainly depend on the physicochemical characteristics such as elemental composition, particle size, surface charge, surface chemical properties [16; 17; 18]. Du et al. reported that PEG functionalization could reduce the accumulation of CdTe QDs in liver and kidney, and also decrease the oxidative stress variation [19]. Since the majority of QDs are composed of toxic substances, shells or macromolecules are often coated on the surface in order to reduce the biotoxicity [20; 21]. In addition, researchers hope the QDs could be eliminated from the body over time, rather than being broken down or accumulated in organs for a long time. Thus, the in vivo distribution and toxicity of QDs with different physicochemical properties need to be deeply addressed before wide-scale biological application of QDs. However, there are limited reports on the toxicity of InP QDs, and most of them focus on toxicity studies using in vitro cell model or simple model organisms, which may hinder the biomedical application of InP QDs.
In the present study, in order to evaluate the in vivo distribution and biotoxicity of InP/ZnS QDs with different surface modifications, three kinds of commercially available InP/ZnS QDs, including hydroxylated QDs (hQDs), amino QDs (aQDs), and carboxylic QDs (cQDs) were injected intravenously into mice, respectively. Mice were sacrificed at 1, 3, 7, 14 and 28 days post-exposure. By comparing the behavior, weight, organ coefficient, hematological parameters, serum biochemical parameters, biodistribution and organ histopathology of the animal models in different groups, the effects of different surface functional groups on the toxicity of InP/ZnS QDs were demonstrated. Our study may provide a better understanding of the effects of surface chemistry on QDs toxicity at the animal level and also facilitate the synthesis of safer QDs in biomedical applications.

Characterization of InP/ZnS QDs
TEM was used to evaluate the shape, size and morphology of the three types of InP/ZnS QDs with different surface functional groups (-OH, -NH2 and -COOH). The representative images of hQDs, aQDs and cQDs were shown in Figure 1(A-C), respectively, suggesting that the three kinds of InP/ZnS QDs were spherical or ellipsoidal, with uniform particle size of approximately 3-7 nm. The diameter of aQDs was slightly smaller than that of hQDs and cQDs. The hydrodynamic diameter and Zeta potential of these three water-soluble InP/ZnS QDs were investigated by zeta potential and particle size analyzer. The results were shown in Figure 1(D-F) and Optical properties of QDs were also evaluated and the results were shown in Figure   1(G-I). Evaluation of the absorbance spectra indicated the first exciton peak of the three InP/ZnS QDs was around at 350 nm. From the results of fluorescence spectra, the emission spectra of the three water-soluble QDs were narrow and symmetrical, and the emission peak was about at (625 ± 10) nm with the excitation at 380 nm.

Body weight and organ weight/BW coefficients
The behavior, mental status, food intake, urine and feces of mice were observed daily after intravenous injection of QDs. Two mice from cQDs-treated group died on the first day after treatment. Besides, no unusual changes were observed in food intake, fur, behavior and mental status after exposure of QDs. The body weight of mice was continuously recorded for 28 days and the data was shown in Figure 2A. The body weight of QDs-exposed groups and the control group showed comparable increasing trends through the study. Main organs of mice were removed and weighted carefully when the mice were sacrificed at various post-injection times. Organ index of the main organs was calculated as organ weight (mg)/BW (g). The organ index of heart, liver, spleen, lung, kidney and brain of mice on Day 28 were shown in Figure 2B, there was no significant statistical difference in all organ index between QDs treated groups and control group. The above results suggested that these InP/ZnS QDs did not interfere with the growth of mice.

In vivo distribution of QDs
After QDs enter the body through intravenous injection, they will be transported characteristics at least 28 days. Though very little fluorescence was observed in brain tissue sections after mice were exposed to QDs, it indicated that all three QDs could pass through the blood-brain barrier and distribute in the brain.
In order to further quantitate the accumulation of three kinds of QDs in liver, kidney and spleen, the In element concentration in tissues at different sampling times was measured by ICP-MS. The results were shown in Figure 6. The three QDs were mainly distributed in liver and spleen, and this result was consistent with that of cryosection fluorescence microscopy. The concentration of In in kidney was about one tenth of that in liver and spleen. The accumulation of three QDs in the liver reached the peak on 3 days post-injection (hQDs: 7.95 μg/g In, aQDs: 9.06μg/g In, and cQDs: 6.92 μg/g In). The distribution of aQDs and cQDs in spleen also reached the peak on Day 3 after administration, but In concentration of hQDs peaked in spleen at 7 days postinjection. After that, the In concentration in liver, spleen and kidney decreased gradually.
However, In could be still detected in liver, spleen and kidney tissues from mice treated with high-dose QDs on Day 28. Since In was not detected in all tissues of the control group, it could be considered that all detected In element in liver, spleen and kidney came from the residues of QDs in vivo. The above results showed that the three QDs (hQDs, aQDs, and cQDs) were mainly distributed in the liver and spleen when they were injected intravenously. Although QDs may be removed or broken down in vivo, this process will take quite a long time.

Histopathological detection results
Histological assessment was performed to evaluate the tissue damage and inflammation caused by QDs exposure. Main organs of all mice were sliced into 5 μm sections and stained with H&E. The representative histological results were shown in pathological changes were observed in heart, liver, spleen, lung, kidney, and brain of all mice at different sampling times when compared with the control group. Therefore, although all the three kinds of QDs could remain in the organs of mice, especially in liver and spleen for a long time, no obvious histopathological abnormalities were observed, which suggested InP/ZnS QDs with different surface functional groups (-OH, -NH2, and -COOH)caused low toxicity to these organs. Figure 7. Representative histological images of major organs including heart, liver, spleen, lung, kidney and brain collected from the control group mice and different surface functionalized QDs treated mice following intravenous injection at dose of 25 mg/kg at 1 day and 28 days post-injection (scale bar: 50 μm).

Hematology analysis
Since the three QDs will remain in vivo for a long time, the changes of blood cells can reflect some pathological reactions induced by QDs. Routine Blood analysis was performed and the results were shown in Figure 8. Most of hematological parameters obtained from QDs-treated mice were comparable to those in control group. On the first day after treatment, WBC counts in aQDs and cQDs group were obviously higher than that in control group (P < 0.05), which indicated that these two QDs may cause an acute inflammatory response after entering the body. The percent of Nue and Lym on Day 3 and Day 28 in aQDs groups was significantly higher than that in control group (P < 0.05). The levels of RDW-CV on Day 3, PCT and MPV on Day 7 in hQDs group were remarkabley higher than those in control group (P < 0.05). The levels of PLT, PCT and MPV in cQDs groups were obviously changed compared to those of the control group (P < 0.05). No significant differences were found in the other parameters. The above results showed that a high dose of the three QDs caused changes in hematological indexes. In particular, aQDs and cQDs could cause acute inflammation in the body, hQDs mainly had adverse effects on red blood cells and platelets. higher than that in control group (P < 0.05). In high dose cQDs treated groups, LDL-C on Day 1, A/G on Day 3 and GLU on Day 28 were obviously higher than that in control group (P < 0.05). Almost all the low dose QD exposed groups showed no significant changes in the above biochemical indexes. The results showed that the three surface functionalized InP/ZnS QDs could affect the liver function of mice when exposed at high concentration, the effect of hQDs was the most serious, and the effect of aQDs and cQDs was relatively mild.
Heart function indexes including LHD, CK, CK-MB and α-HBDH of all mice were measured and the results were shown in Figure 9(O-R). Levels of LDH on Day 3 and Day 28, α-HBDH on Day 3 in high dose hQDs treated group were significantly higher than that in control group (P < 0.05). Levels of α-HBDH on Day 7 and LDH on Day 14 in high dose aQDs treated group were obviously higher than that in control group (P < 0.05). There were no significant differences in CK and CK MB levels between QDs treated groups and control group. The above results showed that the hQDs and aQDs could cause slight effects on cardiac function, while cQDs had almost no effect on cardiac function. Furthermore, renal function indexes including UA, Urea and CREA of all mice were also detected, and the results were shown in Figure9 (S-U).
There were no significant changes in UA, Urea and CREA levels of the three QDs treated mice when compared with the control groups at the same sampling time. It illustrated that the three QDs with different surface functional groups had no toxicity to the kidney. which are carried out by in vitro cell lines [22] or simple model organisms (Hydra vulgaris [23], rare minnow embryos [24] It is very important to comprehensively evaluate the disposition of nanoparticles in vivo, so as to understand and predict their effectiveness and side effects [25]. In this study, In this study, the fluorescence of QDs appeared to be weakening over time, but it could still be observed on 28 days post-injection, indicating that there were still intact QDs in tissues. There are two possible reasons for the weakening of fluorescence.
Firstly, the integrity of QDs was broken after they entered the body because of the corrosive internal environment [35]. Once the QDs were destroyed, the surface trap states produced on the particles and affected the electron hole recombination process, and eventually lead to the decrease or disappearance of fluorescence intensity. Although the structural stability of InP QDs is better than Cd-based QDs, a recent study by Veronesi has reported that InP core was rapidly degraded in Hydra tissues without evident toxicity [36]. Another reason is that QDs were excreted through kidneys, bile ducts, lungs, secretory glands or other organs. It is generally believed that nanoparticles with a diameter less than 3 nm can extravasate into tissues nonspecifically.
Nanoparticles with a diameter less than 5.5 nm can be excreted via renal clearance rapidly and efficiently, while the nanoparticles with a diameter more than 15 nm are hardly cleared via the urinary excretion [37; 38]. or -COOH groups respectively in mice following intravenous tail vein injection.
Following administration, the three kinds of InP/ZnS QDs could be rapidly distributed into the main organs of mice, especiallyin the liver and spleen. The three QDs were excreted from the body gradually, but the In element from QDs still could be detected in liver, spleen and kidney over 28 days period. Low concentration exposure of three QDs hardly caused obvious toxicity, while high concentration exposure of the three QDs could cause some changes in hematological parameters or biochemical parameters, and had no effect on histopathological changes. It should be noted that high dose exposure of cQDs could lead to death, acute inflammatory reaction and slight changes in liver function in mice, which was considered to be more toxic compared with the other two QDs. In conclusion, InP/ZnS QDs could be distributed to the main organs in vivo for a long time. Different surface modifications are crucial to the in vivo toxicity of QDs, which need to be taken into consideration in the synthesis and application of QDs in the future. Even though it will take quite a long time to realize the clinical transformation of QDs, especially in human body, we still hope that QDs can give full play to their excellent optical properties and drug delivery capacity in biological application with the rapid development of chemical synthesis technology, nano application and toxicology technology.

Characterization of InP/ZnS QDs
InP/ZnS QDs used in this study were prepared by Najing tech Company, China.
Prior to be used in our experiments, the surface of QDs was modified with hydroxyl, amino and carboxyl groups, respectively. Finally, three water-soluble InP/ZnS QDs, including hydroxylated QDs (hQDs), amino QDs (aQDs), and carboxylic QDs (cQDs) were obtained. The size and morphology of the three QDs were characterized by transmission electron microscopy (TEM) (HT7700, HITACHI, Japan). Zeta potential and hydrodynamic size distribution of the three QDs was characterized by zeta potential

Animal treatments and sample collection
Mice were randomly divided into 7 groups according to the kinds and doses of QDs exposed, as follows: (1)

Cryosection fluorescence microscopy
Mice were administered intravenously with QDs with different surface functional groups at 25 mg/kg BW or 2.5 mg/kg BW and were sacrificed at various post-injection time points. The heart, liver, spleen, lung, kidney and brain was harvested, embedded into OCT compound and frozen at -80°C. The tissue frozen section was cut into 5 μm thick by freezing microtome (CM3050S, Leica, Germany). Fluorescence microscopic imaging of QDs was observed by a fluorescence microscope (Axio Observer, ZEISS, Germany).

Quantification of uptake in organs
Liver, spleen and kidney of each mouse were digested in the microwave digestion instrument by adding 4 mL 65% nitric acid (HNO3) and 1 mL 30% hydrogen peroxide

Histopathological examination
Mice were sacrificed on Day1, Day 3, Day 7, Day 14 and Day 28 after the injection.
Major organs including heart, liver, spleen, lung, kidney and brain were removed and fixed in tissue fixative. After gradient dehydration with different concentrations of alcohol in an automatic tissue dehydrator (APS300S, Leica, Germany), tissues were embedded in paraffin blocks by paraffin embedding station (Leica, Germany). Then the tissues were cut into 5 μm thin slices by an ultra-thin semiautomatic microtome (RM2236, Leica, Germany) and adhered to the slides. After the slides were stained with hematoxylin and eosin (H&E), histopathological morphology was evaluated under the microscope (Axio Observer, ZEISS, Germany) by an independent pathologist unaware of the treatment.

Hematology analysis
After the mice were anaesthetized, blood samples were harvested from the posterior orbital venous plexus of mice. About 50 μL of blood were collected in tubes containing heparin sodium and routine blood analysis was determined by fully

Serum biochemical analysis
The whole blood of mice was collected in disposable venous blood vessels containing separating gel and the serum was obtained by centrifugation for 15 mins at

Statistical analysis
All statistical analysis was performed by SPSS 22.0 statistical software packages and figures were drawn with GraphPad Prism software package. Data was expressed as mean ± standard deviation (SD). The difference among the different groups was compared by one-way ANOVA. Results were considered significant if P < 0.05.  inductively coupled plasma mass spectrometry.

Availability of data and materials
All data generated or analyzed during this study are included in this published article.

Acknowledgements
Thanks to Instrumental Analysis Center of Shenzhen University.

Ethics approval and consent to participate
The study was conducted in accordance with the Guide for the Care and Use of Laboratory Animals Center of Shenzhen University and approved by the Experimental Animal Ethics Committee of Shenzhen University (Permit No.2017011).

Declaration of interest statement
No conflict of interest in the submission of this manuscript is declared.

Consent for publication
All the authors have approved the manuscript and agree with submission to your esteemed journal.