The Effect of Particle Size on the Absorption of Cyclosporin A Nanosuspension Through the Gastrointestinal Barrier

Background: The particle size is one of great important properties of nanoparticles which affects the dissolution rate in vitro and pharmacokinetics in vivo. This study aimed to design an oral cyclosporin A nanosuspension (CsA-NS) and investigate the effect of particle size of cyclosporin A nanosuspension (CsA-NS) on absorption through the gastrointestinal barrier. Results: CsA-NSs with different particle sizes were prepared. Dissolution rate in vitro, transmembrane permeation, gastrointestinal transport properties and the oral absorption of CsA-NSs were promoted by reducing size, except cellular uptake. Specially the particle size of CsA-NSs was nanoscale, their bioavailability was bioequivalent with marked soft capsules (Sandimmun Neoral®) which is self-microemulsion. Conclusions: This study proposed the potential of developing CsA oral multi dosage form, taken the advantage of nanosuspensions.

(EMD Millipore, Billerica, MA, USA). HPLC-grade acetonitrile and methanol were obtained from Thermo Fisher Scienti c (Waltham, MA, USA). All other reagents were analytical grade.

Preparation of the CsA-NSs with Different Particle Sizes
CsA-NSs with different particle sizes were prepared by the wet bead milling method. HPC (7.5 g), TPGS (1.0 g) and SDS (0.1 g) were dissolved in deionized water (100 mL) to form a surfactant solution. CsA (15 mg) was added to the surfactant solution and stirred at 300 rpm with a magnetic agitator. The CsA suspensions and surfactant were processed with the high-pressure homogenization (HPH) technique by Ultra Turrax (T25; IKA, Staufen, Germany) at 10,000 rpm for 5 min, a uniform suspension was obtained. The uniform suspensions were then milled using the grinding machine (Dyno®-Mill Research Lab, WAB, Switzerland) at 1500, 2000, and 2500 rpm for 5 min, followed by 3000 rpm for 2.0 h to prepare CsA-NSs with the different particle sizes. The 0.3 mm yttrium-stabilized zirconium beads were used as milling media in this study.

Particle characterization
The mean particle size (MPS) and polydispersity index (PDI) of CsA-NSs with smaller particle sizes were detected by photon correlation spectroscopy (PCS), using a Malvern Zetasizer (ZS-90; Malvern Instruments, Malvern, UK). Potential larger particles or aggregates which cannot be detected by PCS. So, the median diameter (D 50 ) of CsA-NSs with larger particle sizes were analyzed by laser diffractrometry (LD) using a Mastersizer (2000; Malvern Instruments, Malvern, UK). Samples were diluted in water to a suitable concentration. The optical parameters of cyclosporin A were: real refractive index (RI) 1.49 and imaginary refractive index (IRI) 0.03 20 . D50 will also be referred to as MPS for ease of description in the following article.
Characterization of CsA-NSs with different particle sizes The morphologies of CsA-NSs (280 nm), CsA-NSs (522 nm) and CsA-NSs (2967 nm) were determined by scanning electron microscopy (SEM) (JSM-7900F, JEOL, Japan) was used to determine morphologies of CsA-NSs. Samples were a xed to aluminum stubs using a double-sided carbon tape and sputter-coated with gold under an argon atmosphere. Differential scanning calorimetry (DSC) was performed with a DSC 214 differential scanning calorimeter (NETZSCH, Germany). The thermal properties of CsA, the excipients (HPC, TPGS and SDS), their physical mixture (raw CsA and stabilizers), and the CsA-NSs with different particle sizes were analyzed. Accurately weighted samples of 3 mg were placed in open aluminium pans, and DSC scans were recorded at a heating rate of 10 K/min from 20 ℃ to 200 ℃ under nitrogen purge gas ow (20 mL/min). An empty pan was used as reference. X-ray diffraction (XRD) was performed using a diffractometer (D8-Advance, Bruker, Germany) equipped with an Apex II CCD detector. The crystallinity state of CsA in raw CsA powder and CsA-NSs were analyzed. The X-ray source was Kα radiation from a copper target with a graphite monochromator at a wavelength of 1.54 Å. Standard runs using a 40 kV voltage, a 40-mA current, and a scanning rate of Particle size stability studies This study aimed to investigate the effects of particle size on delivery of trans-intestinal epithelium transportation of CsA-NSs. So, it is necessary to investigate the stability of particle size in experimental process. The particle size of all the nanosuspensions were studied at 4 ℃ and room temperature. The particle size of CsA-NSs were also investigated in SGF, SIF, RPMI 1640 medium and HBSS. MPS and PDI of all samples were analyzed in triplicate and reported as the standard deviation.
The release of CsA-NS in vitro A release study in vitro was carried out using the paddle method Chinese Pharmacopoeia speci cations using a dissolution apparatus (RC1207DP, Tianda Tianfa Technology Co. Ltd., Tianjin, China). The CsA-NSs with different particle sizes containing an equivalent of 10 mg CsA were dropped into 900 mL of water, which was maintained at 37 ± 0.5 ℃ and stirred at 100 rpm. Samples (5 mL) were withdrawn and replaced with the equal volume of fresh medium at predetermined time intervals (5,10,20,30,40,50, 60 min). The dissolution samples were ltered through a membrane lter of 0.22 µm pore size (Tianjin Jinteng Experimental Equipment, Co. Ltd., Tianjin, China). Apart from water, hydrochloric acid (pH 1.2), phosphate buffer (pH 4.5) and phosphate buffer (pH 6.8) were also respectively used as release mediums to investigate the release in vitro of CsA-NSs.
The content of CsA was assayed by a high-performance liquid chromatography (1640, Agilent, USA). The absorbance wavelength was set at 214 nm. The mobile phase was a mixture of phosphate acid (pH 1.2) solution and acetonitrile at a 10:90 v/v. A CAPCELL PAK C18 column (5 µm, 4.6 mm × 250 mm, Shiseido, Japan) was used with a ow rate of 1 mL/min and the column temperature was maintained at 70 ℃ using a column heater. The injection volume was 20 µL.
In vitro cellular uptake and Caco-2 monolayerpermeation Cell culture The Caco-2 cells were cultured by following the regular procedures in 1640 medium at 37 ℃, 90% RH and 5% CO 2 in T-25 asks. The culture medium was changed every other day. The cells were passaged every 4-6 days after dissociation with 0.25% trypsin/0.02% EDTA solution when the cell fusion rate reached 85%.

In vitrocytotoxicity
Caco-2 cells were seeded in 96-well plates at a density of 5 × 10 4 cells in 200 μL medium per well and incubated for 48 h. The culture medium was removed, then culture medium containing various concentrations (from 200 to 400 µg/mL) of CsA-NSs with different particle sizes of 280 nm, 522 nm and 2967 nm were added to the cells. After 4 h, the medium was replaced with fresh medium containing 1 mg/mL of MTT, and the cells were further incubated for 4 h. Then the supernatant was removed, and the MTT formazan crystals were dissolved in 100 µL DMSO under gentle shaking for 30 min at room temperature. The absorbance was measured using a Spectrophotometer (3020, Thermo Fisher Scienti c Oy, USA) at 490 nm. Cell viability was calculated by measuring the absorbance. Bedford, MA, USA) in a density of 1 × 10 5 cells/cm 2 , and cultured for 21 d under 5% CO 2 , 90% relative humidity, and 37 ℃. The apical (AP) and basolateral (BL) compartments contained 0.5 and 1.5 mL of culture medium, respectively. The culture medium was replaced every other day for the rst week and daily thereafter. The trans-epithelial electrical resistance (TEER) was measured, and a threshold value of 500 Ω/cm 2 was set for transmembrane studies. Prior to the experiments, the culture medium was replaced with warm HBSS (37 ℃). The cell monolayer was equilibrated at 37 ℃ for 30 min before conducting the transport studies. HBSS was removed and 200 µL HBSS containing CsA-NSs (40 µg/mL CsA) were added to the AP compartments, while 1.0 mL HBSS was lled into the BL side. At predetermined time points (0.5, 1.0, 1.5, 2, 2.5 and 3 h), aliquots (100 µL) were withdrawn from the BL side, and an equivalent volume of HBSS was added to maintain a constant volume. The CsA concentration in the samples was determined by LC-MS/MS. The preparation of the samples and LC-MS/MS conditions are detailed in "In vitro cellular uptake of CsA-NSs in Caco-2 cells". The apparent permeability coe cient (P app , cm/s). P app was calculated using the following equation 10 : where dQ/dt is the transport rate (µg/min), C 0 is the initial drug concentration on the apical side (µg/mL), and A is the surface area of the membrane lter (0.3 cm -2 ).

Situ Single-Pass Intestinal Perfusion Experiments
Transport of CsA-NSs in the gastrointestinal (GI) tract was monitored by in situ single-pass intestinal perfusion experiments 21 (Fig.2). 15 male SD rats were randomly and equally divided into 3 groups: 280 nm-CsA-NS, 500 nm-CsA-NS and 2967 nm-CsA-NS. Cannulations were made at duodenum (1 to 11 cm downward from the pylorus), jejunum (15 to 25 cm downward from the pylorus), ileum (0 to 10 cm upward from the caecum) and colon (0 to 10 cm downward from the caecum) of anesthetized SD rats. The intestinal content was removed with physiological saline until the outlet solution appeared clear. CsA-NSs dilution (40 µg/mL CsA) were perfused along the bowel through the cannula at a ow rate of 0.20 mL/min to the intestinal segment. After a stabilization period of 30 min, perfusion uid was collected into pre-weighed 5.0 mL vials every 20 min for 6 times. Samples were weighed and assayed by HPLC. The length of the intestinal segments was measured at the end of the experiment, and nally mercy killing of the animals were exercised. 200 µL acetonitrile was added into 50 µL of the sample followed by 1 min vortex mixing, then the sample was centrifuged at 14,500 × g for 10 min to obtain supernatant. At last, 20 µL supernatant was injected into HPLC to detect the concentration of CsA. HPLC conditions are detailed in "In vitrorelease".
The absorption rate constants (K α ) and effective permeability coe cients (P eff ) of CsA-NSs across rat intestine were calculated based on the disappearance of the drug in perfusate using the following where Q is the ow rate of the drug through the intestine (0.2 mL/min), r is the radius of the rat intestine, and L is the length of intestinal segment perfused after completion of the perfusion experiment. C in , C out , V in and V out are the drug concentration (µg/mL) and volume (mL) in the inlet of the perfusate entering the intestinal segment and the exiting solution, respectively.

Pharmacokinetic Studies
The in vivo pharmacokinetics of Neoral® microemulsion, CsA-NS (280 nm), CsA-NS (522 nm) and CsA-NS (2967 nm) were investigated. 20 healthy male SD rats were randomly divided into 4 groups (n = 5) and then administered with Neoral® microemulsion or CsA-NSs at an equivalent CsA dosage of 25 mg/kg via oral gavages. Orbital blood samples were collected at 0. 25  10 µL CsD (200 ng/mL) solution was added into 50 µL blood and then was vortex mixed for 30 s. 200 µL acetonitrile were added into the mixture and vortex-mixed for 1 min. Samples were then centrifuged at 14,000 × g for 10 min to obtain supernatant. Finally, the supernatant 5 µL were injected to determine the CsA concentration in the whole blood using LC-MS/MS system. LC-MS/MS conditions are detailed in "In vitro cellular uptake of CsA-NSs in Caco-2 cells".

Data analysis
All results were expressed as mean ± SD. Statistical signi cance of the results was analyzed using oneway ANOVA. Values of p < 0.05 was considered statistically signi cant for all tests.

Characterization of CsA-NS Particle Size and Size Distribution
Particle Size of CsA-NSs were measured by DLS. As Figure 3 showed that the MPS becomes smaller and PDI becomes more uniform with prolongation of grinding time. Within 5 min, at a speed of 1500 rpm, MPS reduced to approx 3000 nm. With continuation of the milling, CsA-NSs with 500 nm and 280 nm were obtained after 30 min and 105 min at a speed of 3000 rpm, respectively. PDI ranged from 0.541 to 0.111. It was observed that the milling time more than 105 min could no longer decrease particle size.

Surface Morphology of CsA-NS
The morphology of CsA-NSs were observed under SEM. The scanning electron microscopy images are shown in Figure 4. CsA-NSs with particle size of 280 nm and 522 nm have near spheroid particle structures (A and B). Figure 2 (C) shows that CsA-NS with particle size of 2967 nm is irregular in shape.
Physical Status of CsA in the CsA -NS Beside the size and distribution, the physical status of CsA in CsA-NSs is a very critical parameter. By using X-ray diffraction, we can analyze whether the crystalline form has changed in the preparation process. As shown in Figure 5, the original CsA has obvious diffraction peaks between 2θ values of 5° and 45°. This indicates that CsA exists in a crystalline state in raw CsA. The peaks of CsA in the physical mixture (raw CsA and stabilizers) became lower because it was probably bounded by the surfactant molecules. CsA-NSs show no diffraction peaks, due to amorphous structure caused by milling process and absolutely bounded by the surfactant molecules in the samples 23 . The XRD results con rm that CsA experienced crystalline transformation during the media milling process.
DSC thermograms of stabilizers, raw CsA, physical mixture (raw CsA and stabilizers), and powder of CsA-NSs are shown in Figure 6. The DSC thermogram of raw CsA shows a weak endothermic peak at approximately 107-132 ℃ 24 . It means there is some CsA of crystalline state in raw CsA. The peak of pure CsA was not observed in the thermogram of the physical mixture, which shew obvious in uence from stabilizers. In the CsA-NSs with different particle size of 280 nm, 522 nm and 2967 nm, this peak was also not present, but the DSC thermograms showed a gentle exothermic peak at approximately 80 ℃, which presents crystallization process and indicates that CsA occurs in an amorphous form in CsA-NSs. The DSC thermograms of stabilizers and physical mixture show sharp transitions at 37 ℃, which correspond to the melting points of TPGS (37-41 ℃) 25 ,but this peak is absent in the freezedried CsA-NSs, it can been concluded that TPGS was absorbed by the surface of nanoparticles in molecular form.

Particle size stability studies
It is important that the CsA-NS remains stable in different condition. Especialy, for nanoparticle, particle size is one of most important indexes. As shown in Figure 7, no signi cant changes in MPS of the CsA-NSs could be observed within three months 4 ℃ and room temperature (RT), indicating that the particle size of CsA-NSs is su ciently stable in storage. The particle size of the CsA-NSs increases only slightly after the CsA-NSs were mixed with HBSS, 1640 culture, SGF and SIF at 37 ℃ for 4 h, 4 h, 2 h and 12 h, respectively, which suggests that they could remain stable in the whole experiments cycle.

In vitro dissolution of CsA-NS
The release pro les of CsA-NSs with different particle size were evaluated in four kinds of medium: water, hydrochloric acid pH 1.2, phosphate buffer pH 4.5 and phosphate buffer pH 6.8. As shown in Figure 8A, the CsA-NSs with smaller particle size showed higher dissolution rate in comparison to CsA-NSs with larger particle size in water. CsA-NSs (280 nm) released more than 90% of CsA within 20 min. The release percentage of CsA from the CsA-NS (522 nm) and CsA-NS (2967 nm) were approximately 85% and 70% within 40 min, respectively. The same release trend also occurs in other medium, hydrochloric acid pH 1.2 ( Figure 8B), phosphate buffer pH 4.5 ( Figure 8C) and phosphate buffer pH 6.8 ( Figure 8D).

Cytotoxicity
The cytotoxic effects of CsA-NSs with different particle sizes were evaluated in the Caco-2 cell line. As shown in Figure 9, there was no signi cant differences in cell viability compared with the control, when the CsA concentrations in cell medium were lower than 200 µg/mL. When the CsA concentration of CsA-NS (2967 nm) or CsA-NS (522 nm) in cell medium was 400 µg/mL or 600 µg/mL respectively, the cell viability was signi cantly different from the control group. This suggested that CsA-NSs with a larger particle size (522 nm and 2967 nm) were more toxic to Caco-2 cell.

Cellular Uptake Studies
To study the bioavailability potential of the CsA-NSs for oral delivery, it is important to understand the effect of particle size on the cellular uptake of Caco-2 cells. As shown in Figure 10, the cellular uptake of all CsA-NSs increased during the whole incubation stage from 0.5 to 2 h. The cellular uptake of CsA-NSs in Caco-2 cells reduced with particle size increasing, this was consistent with the result of the previous toxicity test. After 2 h incubation with CsA-NS (2967 nm), the cells contained 4409 ng CsA per mg of protein, while 4202 and 3846 ng CsA per mg of protein for CsA-NS (522 nm) and CsA-NS (280 nm) respectively.

Transport of CsA-NSs across the Caco-2 cell monolayers
To evaluate the transmembrane capacity of CsA-NSs with different particle size, we determined the amount of CsA transporting across Caco-2 cell monolayers from the AP side to the BL side. In all the experiments, the TEER values showed no changes,which indicates there is no cellular monolayers damage during experiment. Relative cumulative transport of CsA in CsA-NSs across Caco-2 cell monolayers from the apical to the basolateral side after 3 h of incubation at 37 ℃ were shown in Figure  11A. The amount of CsA in the BL side of monolayers treated with CsA-NS (280 nm) is highest among the three kinds of CsA-NSs with different particle size at each time point. After 3h incubation, the relative cumulants of CsA in the BL side was 4.18%, 3.88% and 2.98% for NS (280 nm), NS (522 nm) and NS (2967 nm) respectively. These results indicate that the relative cumulative transport of CsA increased with the decrease of particle size and the extension of time in the BL side.
The P app of CsA from the apical side to the basolateral side was calculated after incubating Caco-2 cells with CsA-NSs for 0 to 3.0 h. As shown in Figure 11B

In situ perfusion
In the intestinal absorption studies, the absorption characteristics of CsA-NSs with different size were assessed in four segments of small intestine from alive SD rats: duodenum, jejunum, ileum and colon. The absorption rate constants (K a ) and effective permeability coe cients (P eff ) obtained in the singlepass intestinal perfusion (SPIP) models are presented in Table 1

Pharmacokinetics Studies
In pharmacokinetic studies of CsA-NSs after oral administration, SD rats were selected as model animals. The blood concentration-time pro les of CsA are shown in Figure 12, and the pharmacokinetic parameters are shown in Table 2. The results revealed that, CsA-NS (280 nm) showed the highest C max The AUC 0-48h of 280 nm, 522 nm and 2967 nm CsA-NSs were compared with the reference (Neoral®) values, resulting in a relative bioavailability of 90.20%, 80.18% and 59.61%, respectively. It was evident that, the AUC 0-48h and C max values of CsA-NSs formulations increased in descending order: 280 nm > 522 nm > 2967 nm, while the T max values followed the reverse rank order for the different sized samples. Abbreviations: C max , maximum whole blood concentration; T max , time to reach maximum whole blood concentration; AUC, area under the whole blood concentration, Frel, relative oral bioavailability.

Discussion
In the present study, CsA-NSs with different particle sizes were prepared. Extension of the milling time can obviously reduce the particle diameter within 105 min. And CsA-NS with the smallest particle size 280 nm was obtained. But in this study, if the milling time is over 105 min, the particle size could no longer decrease, conversely, the particle size of CsA-NSs increased. A similar phenomenon as above has been seen in other studies 26 . Thus, for CsA-NSs in this study, the diameter of 280 nm is probably minimum lower limit. It's reported that the upper size limit of nanoparticle internalized into non-phagocytotic cells by means of nonspeci c endocytosis was 3000 nm 27 , and the nanoparticle of 3000 nm also can be adsorbed and immobile within the submucosal layer of the thicker mucosa and the Peyer's patches 28 .
For 500 nm particles, many researches con rmed that it can be absorbed by Peyer's patches in intestine and found absorption maximum among particles between 100 nm and 3 µm 11,28 . So, the CsA-NSs with size of 280 nm, 522 nm and 2967 nm were applied to the following research process.
This study aimed to investigate the effects of particle size on trans-intestinal epithelium transportation of CsA-NSs. It is necessary to investigate the particle size stability of CsA-NSs in different conditions involved in the experiments. As shown in Figure 7, the particle size of CsA-NSs has no signi cantly change in storage and remain stable in the whole experiments. In dissolution study, the release rate of CsA-NS (280 nm) was the fastest, and the cumulative drug release was the highest. These results indicated that the drug release or dissolution rate from CsA-NSs followed a particle size dependent dissolution trend. The increased rate of drug release is likely due to the small size and large surface area of CsA-NSs, as predicted by the Noyes-Whitney equation.
Caco-2 cells derived from a colon carcinoma and can mimic successfully a biological barrier, which also possess high levels of p-glycoprotein. Given gastrointestinal condition, especially a high expression of pglycoprotein in the lower GI-tract and colon, Caco-2 cells were used to evaluate the cellular uptake, in this study. The cellular uptake gradually increased during the whole incubation stage from 0.5 to 2 h. Furthermore, the increase range of cellular uptake for CsA-NSs with a larger particle size was higher (Fig 10.), which is consistent with the toxicity test (Fig 9.). These results con ict with other reported research 29,30 , in which the cell uptake of microparticle is dependent upon the particle size, the smaller particles are, the greater uptake they have 31 . We analyzed the reasons of the phenomenon appealing from the following aspects (Fig 13.). Firstly, as a classical transported substrate of pglycoprotein, CsA can be secreted from the Caco-2 cells by p-glycoprotein 32 . Secondly, the smaller size microparticles < 200 nm were mostly localized in the lysosomals 14,33 , where CsA can be metabolized partly. Thirdly, smaller nanoparticles can penetrate cell easier by transcellular transportation including the endocytosis and exocytosis of nanoparticles [34][35][36][37] . Although the internalization of 2967 nm CsA-NS might be less, as compared with 280 nm and 522 nm, actual CsA in whole cell might well be higher due to lower e uent, degradation and exocytosis. In this work, we nd that the e ciency of cellular uptake depends on not only particle size but also drug properties. For other drug and cell types, particle size-dependent endocytosis may not be the rate limiting step in cellular uptake e ciency, and more parameters should be considered for in cell studies. These results are limited to cellular uptake study, and the ultimate absorption through the gastrointestinal barrier needs been characterized by transmembrane transport and in vivo pharmacokinetics.
After 20-21 days culture the Caco-2 cells formed enterocyte-like cell monolayer and expressed tight junction, micro villi and brush border 22 . The Caco-2 cell monolayers, a good model for intestinal epithelium, could be used prior to in vivo studies for a rapid assessment of the factors. No signi cant differences in TEER values of cell monolayers was observed before and after experiments, which demonstrated the unchanged integrity of cell monolayers 38 . Cumulative transport of CsA-NSs across Caco-2 cell monolayers from the apical to the basolateral side after 3 h incubation at 37°C is shown in Figure 8. Our results show that the concentration of CsA in the basolateral side increased with the decrease of particle size and the extension of time. The increase of dissolution rate and saturation solubility caused by the decrease of particle size are responsible for improving transport of CsA through caco-2 cell monolayer. Meanwhile, nano-particles with smaller size also appeared to across from apical to basolateral monolayer more than the larger ones 35 . Though p-glycoprotein presenting in Caco-2 cell inhibits the overall permeation of CsA, the transport system back into the apical lumen becomes saturated and the diffusion rate through the cells monolayer is the overall limiting factor at higher concentrations 32 . Therefore, reducing particle size and increasing solubility is one of the effective means to increase cells monolayer transport.
Compared with Caco-2 models, the in-situ perfusion models address the complexity of intestinal processes, which eventually determine in vivo intestinal absorption. These complexity not only includes normal expression levels of P450 enzymes and existence of a protective mucus layer 39 , but also remains intact blood vessels and nerves 40 . The obtained assessment is based on the disappearance of the drug in the lumen. The results show the K a and P eff appeared to be higher with the particle size of CsA-NSs decrease. This is because that the transport of CsA-NSs across intestine could be stimulated by improving dissolution rate and saturation solubility, prolonging the time of mucoadhesion to GI 41 , as well as adsorptive endocytosis caused by the decrease of particle. In addition, the results also show that CsA-NSs in duodenal absorption is the best, followed by it is ileum and jejunum, nally colon. The reasons for this result are complex (Fig 14.). characteristics of the duodenum leading to high metabolism, low effusion, high solubility, and e cient particle endocytosis, the disappearance of CsA in the CsA-NSs perfusion uids with different particle sizes was the largest in the duodenum. It's reported that ileum carries more Peyer's patches and M cells than jejunum, in which particles above 1 µm can be trapped in Peyer's patches 35,44 . This may be the reason why more CsA-NSs is absorbed in ileum than jejunum.
The pharmacokinetics of CsA-NSs with different particle sizes following a dose oral administration of 25 mg/kg was investigated in SD rats using the marketed microemulsion (Neoral®) as reference. Mean drug plasma concentration-time pro les and pharmacokinetic parameters of CsA-NSs are showed in Figure 11 and

Conclusion
In the present study, CsA-NSs with particle sizes of 280 nm, 522 nm and 2967 nm were prepared by wet bead milling method. The particle size had a signi cant in uence on the dissolution behavior, cytotoxicity and cellular uptake in Caco-2 cells, transmembrane permeation across Caco-2 cell monolayers, transport properties in the small intestine and pharmacokinetics properties in SD rats of CsA-NSs. The potential of CsA-NSs with smaller particle size were appealing. This study will provide guidance for novel dosage forms of CsA. Availability of data and materials

Abbreviations
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.  Structure scheme of CsA-NS.