Star-block copolymers were constituted with PCL core and ɣ-tert butyl ester substituted PCL segments at the periphery as shown in Fig. 2a employing an in-house built melt reactor (Supplementary Figure S1a) to perform solvent-free ring opening polymerization (ROP). Three star-polymers were synthesized in a sequential ROP process in which initial polymerization of ε-caprolactone (CL) yielded a 6-arm PCL macroinitiator (MI) having statistically 5, 10 and 20 units per arm. These PCL MI were subsequently employed for the ROP of t-butyl ester substituted caprolactone monomer49 (t-BECL, Supplementary Figure S1b) (Fig. 2a and Supplementary Figure S1c) to yield three star-di-block polymers having 35/35, 60/60 and 120/120 units. For instance, the 60/60 diblock has 10 PCL units and 10 t-butyl ester PCL units per arm. The star block copolymers (SB) are referred to as SB-35, SB-60 and SB-100 where the number represents the total content of carboxylic ester substituted PCL segments at the periphery as shown schematically in Supplementary Figure S1d. To determine the structure and degree of polymerization (number of units), the peak intensities in 1H-NMR were analysed in detail (Supplementary Figures S2 and S3). For feed [M]/[PCL MI] = 60, the actual incorporation was determined to be 64 ± 3 repeating units confirming the statistical distribution of 10 units per arm in the second block (Supplementary Figure S2a). Deprotection of t-butyl ester in these star block copolymers yielded their carboxylic acid substituted PCL block copolymers (1H-NMR, Supplementary Figures S2 and S3). Size exclusion chromatography (SEC) was employed to determine the Mn, Mw and polydispersity index (PDI) and the values are summarized in table in Fig. 2b (SEC plots, Supplementary Figure S3b). Star-block copolymers were produced in very high molecular weights of 40 to 60 kDa which are sustainably enough to fold or self-assemble into a single polymer entity. Thermal properties of these new star polymers are in described in Supplementary Figure S3c.
Hydrophilic carboxylic PCL units in the periphery and hydrophobic PCL units in the core provided perfect molecular geometry for the star polymers to acquire required amphiphilicity. Star block copolymers were self-assembled by dialysis method (details in the experimental section). Dynamic light scattering (DLS) analysis determined the hydrodynamic diameter (Dh) of the star-block copolymer SB-60 to be Dh = 25 ± 7 nm (Fig. 2c). In Fig. 2c, FESEM and AFM images revealed the formation of spherical nanoparticles with average sizes of 23 ± 3 nm and 22 ± 5 nm, respectively. HR-TEM images were in coherence and histogram generated (not shown) from ~ 50 particles gave an average value of 24 ± 4 nm. From this data, the hydrodynamic radius (Rh) of the SB-60 NP was estimated to be 12 ± 3 nm. Small-angle X-ray scattering (SAXS) was employed to determine the radius of gyration (Rg) of SB-60, and based on the Guinier approximation, the Rg was estimated to be 7.1 ± 2.3 nm (Supplementary Figure S4a). The ratio of Rg/Rh was estimated to be 0.77 with respect to the existence of unimolecular micellar formulation.54 Pyrene encapsulation study showed no change in the ratio of I1/I3 at different polymer concentrations depicting the unimolecular micelle self-assembly by the star-block copolymer SB-60 and SB-100 (Supplementary Figure S4b). SB-35 was not readily dispersable in water and produced turbid solution rendering it unusable. The above analysis is evident for the existence of core-shell < 30 nm sized star-UMMs by both SB-60 and SB-100.
Encapsulation capabilities of star-UMM NPs were studied for doxorubicin (DOX) and NIR dye IR780 by the dialysis method (experimental section). The DLC for SB-60, SB-100, SB-35 was obtained as 14.2%, 13.1% and < 1% for DOX, respectively (Fig. 2d). This suggested that SB-60 exhibited the most optimized core-shell structure to attain the highest DLC for star-UMM. The SB-35 does not have sufficient compartmentalization for DOX loading. The core-shell geometry probably attained maximum packing at SB-60; thus, no significant DLC increase was observed in SB-100. Encapsulation of NIR dye IR780 in SB-60 showed a very good DLC of 5% which is excellent for deep-tissue bioimaging analysis. The sizes of nascent and DOX-loaded star block copolymer formulations have been tabulated in Fig. 2b (DLS plots in Supplementary Figure S4c). All the nano-formulations showed monomodal size distribution with narrow PDI values. The IR780-loaded star block copolymer SB-60 analogue exhibited size in the range of 120 ± 10 nm. The zeta potential for nascent and DOX-loaded micelles was determined in the range of -3 mV to -26 mV (Fig. 2b), which implies the formation of neutral to slightly negatively charged nano-formulations.
To rationalize the role of topology of the star-block copolymer architectures towards its ability to self-assemble into unimolecular micelles in aqueous medium; two controlled molecules were made having linear di-block (LB-60) and star-random copolymer (SR-60, no segregation of core and shell) architectures. The chemical compositions and molecular weights of the SB-60, LB-60 and SR-60 are identical and they differ only by the arrangements of repeating units (Supplementary Figure S5a). The SR-60 self-assembled as Dh = 30 ± 5 nm sized NP, similar to that of SB-60 (Supplementary Figure S5b). However, the linear di-block LB-60 exhibited the Dh to be 170 ± 10 nm with respect to the formation of large-sized aggregated micelle (FESEM and HR-TEM images in Supplementary Figure S5c). The pyrene encapsulation experiment in LB-60 showed a breakpoint with respect to critical micellar concentration at 1 µg/mL (Supplementary Figure S5c) as typically reported for aggregated micelles. The SR-60 and LB-60 exhibited DLC = 3% and 2% for DOX encapsulation which is almost 7-fold lower than that of SB-60 star-UMM (Fig. 2d). This reiterates the importance of polymer topology for producing UMM with high degree of drug loading, and star block copolymers seems to be the way forward.
Aliphatic polyester backbone in star-UMM makes them fully lysosomal enzymatic-biodegradable49. In the presence of horse-liver esterase enzyme (10 U), > 95% drug release was observed within 24 h (Fig. 3a). The enzymatic cleavage by α-Chymotrypsin (8 U) resulted in release of only 60% DOX molecules. In the control, percentage release was substantially lower at a value of ~ 15%, indicating that the degradation of the UMM only occurred in the presence of lysosomal enzymes. The horse-liver esterase enzyme seemed to be the most suitable for complete degradation of the nano-assemblies. To determine the cyto-compatibility of these nanocarriers, MTT assay was employed in WT-MEF cell lines. Various concentrations of the nascent star block copolymer scaffold SB-60 were incubated with WT-MEF cells for 72 h (Supplementary Figure S6a). As is evident from the histogram, the polymer scaffold displayed 100% biocompatibility up to 100 µg/mL, and about 70–80% of cells were viable up to 200–500 µg/mL. Further, the cytotoxicity of the DOX-loaded scaffold SB-60 + DOX showed that the free DOX was more toxic to cells compared to their delivery from the polymer platform (Fig. 3b). CLSM images in Fig. 3c captured for cells incubated with free DOX and SB-60 + DOX unimolecular micelle are shown for the 180 min time point (live cell). The lysotracker staining helped to visualize the uptake of nanoparticles by the cells and co-localization of the DOX signal at lysosomal compartments. The signals from lysotracker (green) and DOX (red) can be seen as yellow in the merged image. This trend attributes that the star-nanocarrier is readily taken up by the cells via endocytosis and internalized at the lysosomal compartment for biodegradation. A control experiment with free DOX (in Fig. 3c and other time points in Supplementary Figure S6b) exhibited no colocalization with lysotracker, indicating that DOX being a small molecule, is taken up by the cells via diffusion.
NIR dye IR780, having excitation in the near-infrared region, overcomes the limitation of tissue auto-fluorescence whilst offering the advantage of deeper tissue penetration.55 Two groups of mice (n = 3) were used; mice injected with free IR780 iodide dye (group 1) and SB-60 + IR780 (group 2) were utilized along with a control mouse (injected with 1X PBS) for in vivo biodistribution using the In Vivo Imaging System (IVIS, PerkinElmer). The IVIS imaging for tracking the biodistribution was carried out at 24 h, 48 h and 72 h, as can be seen in Fig. 3d (dorsal view). Using the dorsal view images of the mice, an attempt was made to quantify the IR780 distribution in the most rostral part of body with time across the two groups by selecting a region of interest (ROI). The plot of total radiant efficiency vs. time in Fig. 3e clearly demonstrated that the UMM SB-60 + IR780 exhibited superior potential to penetrate the BBB as opposed to the free dye and this distribution did not change with time up to 72 h. The plasma samples were also subjected to IVIS imaging to quantify the amount of IR780, the representative photographic image of the wells can be seen in Fig. 3f. The SB60 + IR780 nano-scaffolds demonstrated their ascendancy in their ability to be in circulation for longer duration than the free dye, (see plot of total flux vs. time in Fig. 3g). The whole-organ representative IVIS image for the brain captured at the 72 h time point can be seen in Fig. 3h. As can be seen from the average radiant efficiency plot, the SB-60 + IR780 unimolecular micelle depicted higher uptake in the brain compared to free IR780 dye. The whole-organ imaging reiterated the ability of the SB-60 + IR780 star UMM to breach the BBB.
Having observed the brain specific uptake, the biodistribution and biochemical analysis of the DOX loaded star-UMM was investigated. A total of 19 mice of 8–12 weeks old female balb/c strain (~ 25 g in weight) for two experimental groups, ‘free DOX’ (group 1) and DOX loaded star-block copolymer ‘SB-60 + DOX’ (group 2) were utilized. Each group comprised of n = 6 mice; half of the mice (n = 3) were used for confocal imaging analysis of drug uptake by organs and histological analysis, while the other half (n = 3) were utilized for biochemical analysis to quantify the amount of DOX in each organ. To understand the biocompatibility of the polymers alone, an additional group (n = 6) was constituted as the control group, namely SB-60 (group 3), to confirm that the nascent polymer itself won’t alter the physiology of the animal. As can be seen from Fig. 4a, for each mouse, the organs collected 24 h-post-injection were brain, heart, kidneys, liver, and spleen (see Supplementary Table ST1 for the organ weights). The DOX concentrations were determined and summarized in the bar diagram plot, as shown in Fig. 4b. An important observation was the significantly higher uptake of the SB-60 + DOX UMM in the brain tissue compared to free DOX. This was further closely investigated by measuring DOX biodistribution across brain regions under different conditions of delivering DOX. The % ID in the heart was the highest in case of free DOX, whereas the value for SB-60 + DOX was substantially lower. This is one of the highlights of our study considering the fact that the major side effect of chemotherapy with DOX is cardiotoxicity, and hence, reducing uptake of DOX in the heart by means of the nano-formulations would greatly overcome the toxic effects of the drug. The uptake by the liver, kidney and spleen tissues were almost similar across the two groups, with no significant difference. The DOX concentration was also determined in the plasma samples collected via retro-orbital plexus puncture at defined time points, and these values are shown in the Pharmacokinetics plot (Fig. 4c). The DOX concentration in plasma for the free DOX group was already less than half of the injected dose at the initial time point, thus their half-life could not be determined. The t1/2 for the SB-60 + DOX group was determined to be 3.5 h using the non-compartmental analysis in the PK Solver software employing the log-linear trapezoidal method.
To study the immune response in the mice, the plasma samples collected from all the 19 mice (3 groups plus 1 mouse injected with 1X PBS as control) at 24 h time point were employed to determine the concentration of five cytokines, i.e., IL-2, IL-4, IL-17A, IFN-ɣ and TNF-α using the cytokine standard values. The mouse injected with PBS acts as the negative control, wherein no immune response is expected in the mouse since PBS is non-immunogenic. Across both groups, the cytokine expressions were very similar to that observed in the case of PBS, and the values were < 5 pg/mL (Supplementary Figure S6c), thereby suggesting that the nano-formulations were not immunogenic. Further hemolysis assay clearly exhibited the polymer biocompatibility with almost negligible hemolysis values at concentrations as high as 1000 µg/mL (Supplementary Figure S6d). Histological images of various organs upon SB-60 + DOX uptake via H&E staining are shown in Fig. 4d. The high-magnification images (40 X) obtained using a bright field microscope (Carl Zeiss) showed no signs of necrosis, blood clotting, or morphological alterations in any of the tissue samples for the SB-60 + DOX group. However, in the panel corresponding to free DOX group some blood clotting was observed in the heart tissue, and according to the literature,56 that is a sign of cardiotoxicity augmented by free DOX (see Supplementary Figure S7). Hence, the star block copolymer DOX loaded assemblies did not result in any damage to the tissues, unlike the free DOX.
The DAPI stained 50 µm sections were imaged to measure DOX uptake in brain and heart tissues across different groups, as shown in Figs. 4e (63 X) and Supplementary Figure S8 (63 X). Images from the kidney, liver, spleen, brain and heart (10 X) are shown in Supplementary Figure S8. In Fig. 4e, in the brain images, only SB-60 + DOX nano-formulation exhibited a strong red fluorescence signal (DOX) whereas, substantially low signal was observed in the free DOX group. This is attributed to the SB-60 + DOX UMM being 18–30 nm making them ideal nano-formulations to cross the BBB. These SB-60 + DOX micelles were able to passive selectively cross the BBB, accumulate in the brain tissue and these being stable at infinite dilution would not result in a premature drug release. Another brownie point for these SB-60 + DOX micelles was their ability to reduce the uptake in heart tissue, as can be seen from Supplementary Figure S8 of the heart tissue images. The signal of the SB-60 + DOX micelles was extremely weak compared to the strong red signal corresponding to free DOX. Thus, the current design of SB-60 UMM can overcome the aforementioned limitation and facilitates the use of even higher doses of DOX for chemotherapy in long term. The DOX uptake from the confocal images was quantified by determining the normalized mean gray values in Fig. 4f. These values for the SB-60 + DOX micelle exhibited enhanced uptake in the brain and drastically reduced uptake by the heart compared to an opposing trend in free DOX group. The renal clearance was higher in case of free DOX compared to the other group. The SB-60 + DOX exhibited reduced RES uptake, as can be seen from the low mean gray values in the liver and spleen tissues.
Quantifying the DOX intensities from the confocal images in Supplementary Figure S9 led to the interrogation of whether the nano-formulations were taken up equally by different parts of the brain tissue. Five different areas of the brain, namely cerebellum, amygdala, hippocampus, cortex, and olfactory bulb were used (posterior to anterior, see the labelling on the sagittal section of the mouse brain in Fig. 5a, adapted from the Gene Expression Nervous System Atlas).57 The representative confocal images are shown in Supplementary Figure S9, and the DOX intensities visibly decreased upon going from the posterior part (cerebellum) to the anterior-most part (olfactory bulb). This was corroborated by the quantification of DOX intensities as normalized mean gray values plotted against the corresponding region of the brain (Fig. 5b). This trend can be attributed to the variation in the BBB heterogeneity and permeability across different brain regions, which depends on factors such as differential astrocyte and pericyte coverage, differences in tight-junction proteins like Zonula occludens (ZO)-1 and ZO-2, variation in cellular interactions between white and gray matter in different regions, changes in the vascular density, so on and so forth.58 In order to examine the intracellular uptake of these micelles, immunostaining was carried out marking different brain tissue cell types, i.e., MAP 2 for staining the mature neurons, NeuN as the nuclei marker and GFAP for labelling the glial cells and astrocytes. The intent was to investigate the specificity of DOX loaded unimolecular micelles across the brain tissue, if any, and the results can be seen in Fig. 5c. The emission of DOX coming from within the mature neurons appeared to co-localize with the markers, as can be seen from the merged images (both 10 X-first column and 63 X-second column) in Fig. 5c. This gave an affirmation of the intracellular uptake of the cargo and that uptake was similar across the brain tissue with no specificity for any cell type.
Mechanism illustrating the ability of the nano-scaffold to penetrate the BBB over the free drug has been outlined in Fig. 5d. The transcytosis across the barrier is dictated by factors such as lipophilicity, size, charge, molecular weight etc., and even with the requisite parameters the molecules are effluxed out into the blood stream via the P-glycoprotein (Pgp) pump.59–61 The crucial parameters for nanoparticles intended to penetrate the BBB are their size < 50 nm, appropriate lipophilicity and surface charge is preferred to be near neutral.59–61 In the current investigation, the plausible mechanism of BBB breach can be attributed to the caveolae-mediated transcytosis, that is known to transport albumin like macromolecules from the luminal side to brain parenchyma.59–61 The crossing of the BBB by the star block copolymer UMM in the present study could be attributed to the following factors: i) prolonged blood circulation making it more bioavailable; ii) appropriate lipophilicity favourable for transcytosis and, iii) the presence of carboxylate groups on the periphery that circumvent the Pgp efflux pumps. All these factors combined with the sub-30 nm size range of the UMM could be the aid in breaching the BBB. To validate the ability of star block copolymer UMM to enter the neuronal cells upon crossing the BBB, an in vitro time-dependent experiment was envisaged. Herein, the SB-60 + DOX UMM were incubated with olfactory bulb (OB) and mixed cortical primary neuronal culture for 1 h and 4 h followed by immunocytochemistry and the panels are shown in Fig. 5e. DAPI and MAP 2 antibody staining was employed for imaging the differentiated mature neurons; represented via the blue (λexc = 405 nm) and green channels (λexc = 633 nm), respectively. The DOX emission (red channel, λexc = 488 nm) from the UMMs was observed to have substantial co-localization with the neuronal markers. The SB-60 + DOX uptake in both OB and cortical neurons was quantified and the plot of uptake (%) vs time depicted significantly higher uptake of SB-60 + DOX UMM in the cortical neurons as compared to the OB neurons across both the time points (Fig. 5e). The in vitro neuronal culture data was in coherence with the in vivo brain biodistribution data, wherein the UMM uptake was higher in the cortex as opposed to the OB (Fig. 5b). Taking cognizance of the in vivo and in vitro analysis, it can be stated that star-UMM demonstrated exceeding capability to breach the most tightly regulated biological barrier, i.e., the blood-brain barrier. Recent reports provide evidence for the role of synaptic signalling in the progression of brain tumours.62,63 As expected from these results, sensory experience has been shown to modulate gliomagenesis, proven by regulating olfactory information processing in a mouse model.64 These recent findings call for further experiments employing potential delivery systems combined with targeted delivery65 and precise behavioural paradigms controlling sensory experiences.66,67 Our results thus provide a potential drug delivery method for brain tumors adding significantly to the emerging field of “cancer neuroscience” research.68,69