Development of Traceable Rituximab-Modified PEO-Polyester Micelles by Postinsertion of PEG-phospholipids for Targeting of B-cell Lymphoma.

The objective of this work was to develop rituximab (RTX)-modified polymeric micelles for targeting of B-cell lymphoma cells, through postinsertion of RTX-poly(ethylene glycol)-1,2-distearoyl-sn-glycero-3-phosphoethanolamine (RTX-PEG-DSPE) into methoxy poly(ethylene oxide)-poly(ε-caprolactone) (PEO-PCL) or methoxy poly(ethylene oxide)-poly(ε-benzylcarboxylate-ε-caprolactone) (PEO-PBCL) micelles. Mixed micelles were made traceable by introducing Cy5.5 to RTX and conjugating Cy3 to propargyl moiety, end-capped PCL or PBCL. Successful adaptation of the postinsertion method for the formation of immunomicelles was evidenced by measurement of RTX levels on the micellar surface, purified from free RTX by size exclusion chromatography, using microBSA assay. A change in the micellar diameter, from 50-70 nm for PEO-PCL and PEO-PBCL micelles and 20 nm for PEG-DSPE micelles, to 80-95 nm for the mixed micellar population as well as the critical micellar concentration of mixed micelles provided further proof for the success of the postinsertion method applied here. Mixed micelles containing PCL or PBCL with a degree of polymerization of 22 (PCL22 and PBCL22) were thermodynamically and kinetically more stable than those with PCL15. Accordingly, RTX micelles containing PCL22 or PBCL22 showed a higher percentage of Cy3+/Cy5.5+ cell population in CD20+ KG-15 cells, than those with PCL15. The percentage of Cy3+/Cy5.5+ cell population drastically reduced in the presence of competing RTX for micelles containing PCL22 or PBCL22 cores, indicating the superiority of these structures for active targeting of CD20+ cells. No significant difference in the cytotoxicity of paclitaxel in RTX-micelles versus plain ones was observed, reflecting the noninternalizing function of CD20. The results show that traceable mixed micelles prepared through postinsertion of RTX-PEG-DSPE to PEO-PCL22 or PEO-PBCL22 micelles can be used for targeting and/or imaging of CD20+ B cell lymphoma cells. The postinsertion method can be adopted to prepare other PEO-poly(ester)-based immunomicelles for active targeting of other diseased cells.

Introduction

Lymphoma can be defined as neoplasm of lymphoid origin, with more than 25 different histological subtypes. B-cell non-Hodgkin Lymphoma (NHL), which accounts for more than 90% of all adult cases of lymphoma, is CD20 positive.[1,2] Conventional chemotherapy regimen used in lymphoma usually consists of cyclophosphamide, doxorubicin (DOX), vincristine, prednisone (CHOP).[3] The survival of a patient with aggressive lymphoma is reported to be less than 50%, which urges the need for new therapies.[4]

The introduction of rituximab (RTX), a chimeric anti CD20 monoclonal antibody, has changed the treatment of B-cell malignancies. Rituximab was first tested as a single agent in the treatment of NHL in 1997, with limited success.[5] Clinical studies over the years have shown the huge impact of adding RTX to the conventional chemotherapy regime in NHL therapy.[5] For instance, a randomized study (399 patients) conducted by 86 centers in France, Belgium, and Switzerland showed a significantly higher rate of complete remission in patients receiving RTX along with CHOP (76 vs 63%; P = 0.005). The median following the survival rate for 2 years was also significantly improved (57 vs 38%; P = 0.001).[6] Another study showed a significant improvement in the event free survival (42% for RTX with CHOP vs 25% for CHOP alone; P < 0.0001).[7]

In search for the development of more effective and less toxic drugs for NHL, targeting of NHL cells using CD20 monoclonal antibody radio-immunoconjugates has been pursued.[8,9] This included ibritumomab tiuxetan, a yttrium 90–conjugated monoclonal antibody to CD20, which has shown an overall response rate of 80% compared with 56% for RTX in clinical studies (P = 0.002), perhaps by locating the radiopharmaceutical in close proximity of the malignant NHL cells.[10] Application of antibody-drug conjugates (ADC)s using monoclonal antibodies (mAb)s against CD20 has not been explored due to the noninternalizing nature of CD20, which will limit the success of this approach. Instead, internalized antigens such as CD19 and CD22 were targeted for this purpose.[11−13]

An alternative approach for targeted drug/radiophamaceutical delivery to NHL is the use of nanodelivery systems modified on their surface with mAbs against NHL antigens. In this approach, the nanocarrier can physically load several moles of drug or radiochemical inside, carry the drug toward the target cells, and then release it either in the vicinity of malignant cells or inside the cells following carrier internalization.[14] This approach has several advantages over the use of ADCs or radio-immunoconjugates: (a) it can take advantage of the physical barrier provided by the nanocarrier against drug distribution and toxicity in normal organs; (b) it can lead to enhanced anticancer effects for the incorporated drug even using noninternalizing antigens, including CD20, for drug targeting; (c) it can increase the ratio of the delivered drug per mAb in the system; and finally (d) it can be used for the delivery of drug combinations.

The development of RTX-modified liposomes and nanoparticles has been pursued in previous studies showing favorable results. For instance, Wu et al. have studied the effect of adriamycin-containing liposomes modified on their surface with a fab fragment of RTX in NHL xeno-transplant in SCID mice and showed a significant reduction in tumor burden in animals treated with this formulation compared with plain liposomes carrying adriamycin or free drug.[15] In another study, Zhou et al. prepared mesoporous silica nanoparticles decorated with RTX and loaded with doxorubicin (DOX). They have also observed significant inhibition of tumor growth for nanocarriers of DOX modified with RTX on their surface compared with plain nanoparticles and free DOX in a Raji lymphoma-bearing mice model.[16]

Polymeric micelles (PMs) are nanodelivery systems extensively explored for application in cancer therapy because of their unique and favorable properties in tumor targeting.[17−20] PMs consist of amphiphilic block copolymers that can self-assemble and form core/shell structures. In an aqueous environment, the hydrophobic core of PMs can solubilize lipophilic drugs. In this environment, the shell is hydrophilic, providing stealth properties, protecting the carrier from aggregation and early uptake by phagocytic cells. Development of antibody-modified polymeric micelles has been mostly conducted using poly(ethylene glycol)-phospholipid (PEG-PL) micelles, which are known to have suboptimal stability for tumor targeting.[21,22] Few studies have reported on the development of other classes of polymeric micelles, including poly(ethylene oxide)-poly(caprolactone) (PEO-PCL), modified on their surface with antibodies through maleimide functional groups on the PEO end.[23]

The objective of this study was to develop an easy method for the preparation of mAb-modified poly(ester)-based micelles of different structures. For this purpose, we explored postinsertion of RTX-PEG-PLs into PEO-poly(ester) micellar structures. In this context, RTX or its Cy5.5 conjugated counterpart were chemically linked to commercially available 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-N-[succinimidyl (polyethylene glycol)] (NHS-PEG-DSPE). The Cy5.5-labeled RTX-PEG-DSPE was then mixed with either Cy3-conjugated methoxy poly(ethylene oxide)-b-poly(ε-caprolactone)-b-poly(α-propargylcarboxylate-ε-caprolactone)(PEO-PCL-PPrCL-Cy3) with different PCL chain lengths (15 or 22 PCL degree of polymerization (DP)) or Cy3-conjugated methoxy poly(ethylene oxide)-b-poly(α-benzylcarboxylate-ε-caprolactone)-b-poly(α-propargylcarboxylate-ε-caprolactone) (PEO-PBCL-PPrCL-Cy3) (22 PBCL DP). The formation of mixed micelles, thermodynamic and kinetic stability of mixed micelles, as well as preferential uptake of micelles by CD20 positive and negative cells were then evaluated. Finally, the capacity of different mixed micelles for loading, release, and in vitro delivery of paclitaxel (PTX) to CD20 positive versus CD20 negative cells was assessed. The results showed the success of the postinsertion method in the formation of traceable mAb-modified PEO-poly(ester)s mixed micelles and shed light on the higher stability of mixed micelles containing PCL22 and particularly those with PBCL22 leading to improved targeting of CD20 positive cells by these structures.

Results

Characterization of Synthesized Block Copolymers and Their Micelles

The characteristics of prepared block copolymers and the associated micelles are summarized in Table . All micelles showed critical micelle concentration (CMC)s in the μg/mL range (Table ). The CMC of MM-PCL22 and MM-PBCL22 was higher than that of the micelles prepared from individual PEO114-PCL22-PPrCL4 or PEO114-PBCL22-PPrCL4 block copolymers. However, it was significantly lower than the CMC of micelles prepared from NHS-PEG-DSPE. Among the prepared mixed micelles, MM-PBCL22 micelles had the lowest CMC (12.3 μg/mL), which indicates that these mixed micelles are the most thermodynamically stable ones. On the other hand, the MM-PCL15 micelles showed a high CMC around 74.9 μg/mL.

Table 1

Characteristics of Synthesized Block Copolymers and Micelles Self-Assembled from Single Polymers or upon Incubation of PEO-Poly(ester)s with NHS-PEG-DSPE or RTX-PEG-DSPE

block copolymer(s) forming micellesa	abreviation used for micelles	Mn (g mol–1)b	average micellar sizec± SD (nm)	PDId ± SD	CMCe ± SD (μg/mL)
PEO114-PCL15-PPrCL4	PCL15	6500	57 ± 0.6	0.215 ±0.003	5.1 ± 0.6
PEO114-PCL22-PPrCL4	PCL22	7800	50 ± 3.4	0.263 ± 0.002	2.2 ± 0.1f
PEO114-PBCL22-PPrCL4	PBCL22	9960	71 ± 0.4f	0.127 ± 0.019	2.2 ± 0.4f
NHS-PEG-DSPE	DSPE	3400	23 ± 2.1	0.368 ± 0.100	45.1 ± 4.8
NHS-PEG-DSPE and PEO114-PCL15-PrPCL4	MM-PCL15		78 ± 0.6g	0.268 ± 0.003	74.9 ± 3.7g
NHS-PEG-DSPE and PEO114-PCL22-PPrCL4	MM-PCL22		84 ± 2.9g	0.399 ±0.0170	18.2 ± 2.5g
NHS-PEG-DSPE and PEO114-PBCL22-PPrCL4	MM-PBCL22		93 ± 3.6g	0.242 ± 0.020	12.3 ± 3.1g
Rituximab-NHS-PEG-DSPE and PEO114-PCL15-PPrCL4	RTX-MM-PCL15		93 ± 7.5	0.234 ± 0.070
Rituximab-NHS-PEG-DSPE and PEO114-PCL22-PPrCL4	RTX-MM-PCL22		95 ± 20.6	0.284 ± 0.068
Rituximab-NHS-PEG-DSPE and PEO114-PBCL22-PPrCL4	RTX-MM-PBCL22		110 ± 11.6	0.216 ± 0.010

The number shown as subscript indicates the DP of each block as determined by 1H NMR.

The average number molecular weight determined by 1H NMR.

Hydrodynamic diameter (Z average) estimated from dynamic light scattering (DLS).

Poly dispersity index measured by DLS.

Critical micellar concentration measured by DLS.

Significantly different from PCL15 micelles (unpaired Student t-test P < 0.05).

The data for mixed micelles are statistically different from their counterpart micelles prepared from single block copolymers (unpaired Student t-test P < 0.05).

The Z average diameter of the self-assembled structures was below 100 nm and they showed a relatively narrow polydispersity index. To confirm the successful formation of mixed micelles, the size of a micelle formed from individual block copolymers, i.e., PEO114-PCL15-PPrCL4, PEO114-PCL22-PPrCL4, PEO114-PBCL22-PPrCL4, or NHS-PEG-DSPE, was measured separately before mixing. After mixing, the size of PEO114-PCL15-PPrCL4/NHS-PEG-DSPE, PEO114-PCL22-PPrCL4/NHS-PEG-DSPE, or PEO114-PBCL22-PPrCL4/NHS-PEG-DSPE pairs was also measured at different incubation time intervals. For mixed micelle samples, at time zero, two peaks reflecting the size of micelles from each individual block copolymer appeared. As the incubation continued for 24 h, only one peak was observed. The average diameter of mixed micelles measured at this time point was shown to be significantly larger than the average diameter of micelles from individual polymers, as shown in Table . Moreover, the average diameter of RTX-modified mixed micelles was between 93 and 110 nm compared with average diameters of 78–93 nm for their counterparts without RTX modification.

Quantification of RTX on Micelles

The amount of RTX conjugated to the mixed micelles was quantified by Micro Bicinchoninic Acid (BCA) protein assay. This was done after separating different populations formed following mixing of PCL15, PCL22, or PBCL22 based block copolymers with PEG-DSPE ones, using the Sepharose column. All 32 collected fractions from the Sepharose column were primarily subjected to DLS measurements to identify the average size of particles in the corresponding fraction. Thereafter, the microBCA assay was used to identify the fractions that contained RTX. The BCA reading was also used to quantify RTX concentration and conjugation efficiency in the collected fractions. Combining the data from these measurements, we found that the first 5 fractions collected from the Sepharose column had neither micelles nor RTX (only phosphate-buffered saline (PBS)) (Table ). This was evidenced by no absorbance in the BCA assay or reading in the DLS. Fractions 6–11 were suggested to have RTX-conjugated mixed micelle, because DLS measurement showed the existence of colloidal particles at a 90–110 nm size range for these samples. Additionally, we had a high absorbance in the BCA assay in these fractions. Fractions 12 to 14 had only the plain micelles, which was implicated by the absence of BCA response and a reading in the DLS similar to the size identified for plain micelles (from 45 to 65 nm). Fractions 15–16, on the other hand, were responsive to BCA assay and showed a size similar to that identified for RTX-PEG-DSPE micelles (around 30 nm). Fractions 17–28 most likely included a mixture of the free antibody and RTX-PEG-DSPE/NHS-PEG-DSPE unimers. Fractions 28–32 showed no reading by DLS and no absorbance by the BCA assay. The high polydispersity index and average diameter recorded for fraction 17–22, may be attributed to the aggregation of free RTX under experimental conditions.

Table 2

Characterization of the Different Fractions Separated during the Purification of Mixed Micelles (RTX-MM-PCL15, RTX-MM-PCL22, and RTX-MM-PBCL22) by a Sephrose Columna

identified composition of fraction	fraction	diameter (nm) ±SD	PDI ± SD	RTX concn (μg/mL)	RTX recoveryb (%)
RTX −MM-PCL15	6–11	93.3 ± 7.5	0.234 ± 0.07	0.255	17.5
RTX-MM-PCL22		95.3 ± 20.6	0.284 ± 0.06	0.257	18
RTX-MM-PBCL22		110 ± 11.6	0.216 ± 0.01	0.276	19
PCL15 micelles	12–14	52.6 ± 5.2	0.198 ± 0.02	0.025	1.7
PCL22 micelles		48.86 ± 7.3	0.378 ± 0.026	0.043	2.9
PBCL22 micelles		65.4 ± 5.6	0.106 ± 0.007	0.01	0.6
RTX-PEG-DSPE micelle	15–16	30.2 ± 5.7	0.432 ± 0.01	0.136	9.3
free RTX	17–22	16.05 ± 4.4	0.568 ± 0.075	0.853	59
RTX-PEG-DSPE and NHS-PEG-DSPE unimers	23–28			0.214	14

The fractions were identified using the combination of data from size measurement by DLS and RTX concentration by BCA assay.

RTX recovery efficiency is calculated as (amount of detected RTX/amount of added RTX) × 100.

Kinetic Stability of Polymeric Micelles

The kinetic stability of micelles was studied by DLS after their incubation with sodium dodecyl-sulfate (SDS), which acts as a destabilizing agent. Figure A,B shows the % intensity of the micellar peak and its polydispersity, respectively, for PCL15, PCL22, PBCL22, NHS-PEG-DSPE, as well as MM-PCL15, MM-PCL22, and MM-PBCL22 (descriptions in Table ) over time in the presence of SDS. As shown in the figure, NHS-PEG-DSPE showed a dramatic drop in the signal intensity of the micellar peak immediately after incubation with SDS. After 2 h of incubation, the intensity of the micellar peak was less than 5%. In the case of PBCL22 micelles and MM-PBCL22, there were neither a change in the intensity of the micellar peak nor the polydispersity index (PDI) even after 24 h of incubation, which shows that these micelles remained intact throughout incubation with SDS. For PCL22 micelles and MM-PCL22, the signal intensity of the micellar peak dropped around 60% after 6 h of incubation with SDS. Their PDI also started to increase above 0.5, around the same time. The behavior of MM-PCL15, on the other hand, was different from that of micelles of PCL15. The PDI of MM-PCL15 reached 1 only after 2 h, implying micellar dissociation. For micelles of PCL15 from single block copolymer, the PDI remained intact for upto 4 h and then increased. Overall, among block copolymers under study, the PEO-PBCL-based micelles (from single block copolymer or mixed) seemed to be the most stable ones, kinetically. Individual and mixed micelles containing PCL22 were the next kinetically stable ones.

Figure 1

Assessment of kinetic stability for micelles from individual polymers vs associated mixed micelles. Micelles (1 mg/mL) were incubated with SDS (6.7 mg/mL in the final solution). (A) The time-dependent change in the micelle peak intensity. (B) Time-dependent change in the polydispersity index (PDI). Data represent average ± standard deviation (SD) (n = 3). The data related to NHS-PEG-DSPE micelles is repeated in all graphs for comparison with associated mixed micelles.

In Vitro Cellular Association

The RTX-modified mixed micelles were labeled with 2 fluorescent dyes; Cy5.5 (conjugated to RTX) and Cy3 (conjugated to the end of poly(ester) block in micelle core) (Figures and 3).

Figure 2

Chemical synthesis and models for the preparation of PEO-PCL-PrPCL block copolymers labeled with Cy3 at the PCL end. The PEO-PBCL-PrPCL-Cy3 micelles are prepared using the same procedure.

Figure 3

Models for the preparation of RTX-modified micelles double-labeled with Cy5.5/Cy3 through post insertion of Cy5.5-RTX-PEG-DSPE into PEO-PCL-PrPCL-Cy3 or PEO-PBCL-PrPCL-Cy3 micelles.

Figures and 5 summarize the results of flow cytometry measurements for the association of the RTX-MM-PCL15, RTX-MM-PCL22, and RTX-MM-PBCL22 with KG-15 (CD20+) or SUP-M2 cells (CD20–). In the dot plot (Figure S1), the bottom left section represents Cy3 and Cy5.5 negative cells, the bottom right section shows the cells positive for Cy3 only, the top left section is cells positive for Cy5.5 only, and the top right section shows cells double positive for Cy3 and Cy5.5.

Figure 4

Bar graph showing the percentage of Cy3/Cy5.5 positive cells after 4 h of incubation of KG-15 and SUP-M2 cells with different mixed micelles with or without pretreatment with free RTX. Each bar represents an average percentage of positive cells ± SD (n = 3). * denotes statistically significant (unpaired Student’s t-test, P <0.05). Dot plots are shown in Figure S1 in the Supporting Information.

Figure 5

Percentage of Cy3/Cy5.5 positive KG-15 and SUP-M2 cells after 4 h of incubation with RTX-conjugated MM-PCL15, MM-PCL22, and MM-PBCL22 micelles. Each bar represents the average percentage of Cy3/Cy5.5 positive cells ± SD. (n = 3)* denotes statistically significant (unpaired Student’s t-test P < 0.05).

As shown in Figure , cells treated with RTX- modified MM-PCL22 and MM-PBCL22 were mostly positive for both Cy3 and Cy5.5. This observation suggests that these micellar systems remained intact in the media when incubated with the cells. In addition, the association of the RTX modified MM-PCL22 and MM-PBCL22 with KG-15 (CD20+) was 3.7- and 4.2-fold enhanced compared with their association with SUP-M2 (CD20–) cells, respectively (unpaired Student t-test, P < 0.05) (Figure ). On the other hand, for RTX-modified MM-PCL15, the vast majority of the cells was only positive for Cy5.5 dye, which suggests that these mixed micelles dissociated in the media and the Cy5.5-labeled RTX-PEG-DSPE micelle is the population associating with the CD20+ cells (Figure S1).

To assess the effect of the receptor (CD20) mediating the cellular association of RTX-conjugated mixed micelles, a competition study with preincubation of cells with free RTX was conducted. In this experiment, we compared the percentage of cells that were positive for both Cy3/Cy5.5 without and with pretreatment with free RTX. The results showed a significant reduction of the Cy3/Cy5.5 double positive cells for RTX-MM-PCL15, RTX-MM-PCL22, and RTX-MM-PBCL22 on the KG-15 following pretreatment with free RTX (Figure ) (unpaired Student t-test, P < 0.05). In contrast, pretreating the SUP-M2 cells with free RTX did not change the association of the RTX-modified micelles with the SUP-M2 cells. We investigated the effect of RTX competition on the association of plain micelles with both KG-15 and SUP-M2 cells by looking into the difference in the percentage of Cy3 positive cells with and without pretreatment with free RTX. The results showed no significant difference in the association of the plain MM-PCL15, MM-PCL22, and MM-PBCL22 micelles upon competition with free mAb compared with those without RTX pretreatment in both cell lines (unpaired Student t-test, P > 0.05). Pretreatment with RTX also reduced the percentage of Cy5.5 positive KG-15 cells for RTX-PEG-DSPE micelles (P < 0.05, unpaired Students’ t teat) reflecting the competition of free RTX for association of RTX-PEG-DSPE micelles with CD20 cells. This difference was not seen for SUP-M2 cells (P > 0.05, unpaired Student’s test).

Characterization of PTX-Loaded Polymeric Micelles

Solubilization of PTX in MM-PCL15, MM-PCL22, MM-PBCL22, and its counterpart micelles from single block copolymers was examined.[24] Under identical loading conditions, no significant difference in the encapsulation efficacy between PCL15 versus MM-PCL15, PCL22 versus MM-PCL22, and PBCL22 versus MM-PBCL22 micelles was observed (unpaired Student’s t-test, P > 0.05). Overall, MM-PCL22 and MM-PBCL22 showed significantly higher encapsulation efficacy than MM-PCL15 (one way analysis of variance (ANOVA), P < 0.05) (Table ).

Table 3

Characteristics of the PTX-Loaded Polymeric Micelles under Study

micelle	average diameter ± SD (nm)	PDI	PTX loading ± SD (wt %)	encapsulation efficiency ± SD (%)	PTX release at 48 h (%)
PCL15			0.55 ± 0.01	14.8 ± 1.50	47.0 ± 5.3
PCl22			0.79 ± 0.08a	24.7 ± 2.50a	53.0 ± 4.6
PBCL22			0.60 ± 0.04	23.6 ± 2.10a	58.1 ± 8.2
MM-PCL15	80 ± 1.5	0.376 ± 0.01	0.37 ± 0.01c	15.0 ± 1.2	75.5 ± 5.1c
MM-PCL22	83 ± 0.7	0.217 ± 0.003	0.45 ± 0.03c,b	23.6 ± 0.5b	40.7 ± 4.5c,b
MM-PBCL22	78 ± 0.1	0.148 ± 0.01	0.39 ± 0.03c	20.7 ± 0.8b	60.75 ± 3.7b

Significantly different from PCL15.micelles.

MM micelles significantly different from MM-PCL15.

MM micelles significantly different from their counterpart single micelles.

PTX Release from Polymeric Micelles

The in vitro release profile of PTX from polymeric micelles under study is illustrated in Figure . A comparison of the release profile between MM-PCL15, MM-PCL22, MM-PBCL22 with their counterpart micelles prepared from single corresponding PEO-poly(ester) block copolymers was conducted by the f2 similarity factor. In addition to the f2 factor, the unpaired Student t-test was also conducted for the purpose of point-to-point analysis. Our results showed that MM-PCL22 significantly slowed down the release of PTX compared with PCL22 micelles at a 48 h time point (40.7% of PTX released from MM-PCL22 compared with 53.1% release from PCL22) (unpaired Student’s t-test, P < 0.05). On the other hand, when comparing the release behavior at an early time point, for instance, after 2 h, we notice that the release of PTX from MM-PCL22 (8.49%) was significantly higher than its release from PCL22 (2.2%) (Unpaired Student’s t-test, P < 0.05). However, when measuring the f2 factor, no significant difference between the release profile of MM-PCL22 and PCL22 micelles was observed (f2 = 52.3). In contrast, the release of PTX from MM-PCL15 after 48 h was significantly accelerated compered with its release from PCL15 (75.05 versus 46.86% release for MM-PCL15 and PCL15, respectively) (unpaired Student’s t-test, P < 0.05). The same pattern was also noticed after 2 h, where 1.2% of PTX was released from PCL15 whereas 23.5% was released from MM-PCL15 (Unpaired Student’s t-test, P > 0.05). When conducting the f2 factor analysis, a significant difference between the release profile of PTX from PCL15 and MM-PCL15 was noted (f2 = 10.9). In the case of MM-PBCL22, there was no significant difference between the release profile of PTX from and PBCL22 (58.1%) and MM-PBCL22 (60.75%) at 48 h (Unpaired Student’s t-test, P > 0.05). Interestingly, when looking at the release profile of PTX after 2 h, we see a significantly noticeable burst release of PTX from PBCL22 (30.5%) compared with only (6.5%) from MM-PBCL22 (Unpaired Student’s t-test, P < 0.05). Additionally, when conducting the f2 similarity factor analysis, a significant difference in the release profile between PBCL22 and MM-PBCL22 was shown (f2 = 38.4). Overall, by comparing the release profile of all prepared mixed micelles, MM-PCL22 appeared to have a significantly slower release profile, since only 14.7 and 40.7% of PTX were released after 6 and 48 h, respectively, from this formulation while 55.78 and 75.1% were released from MM-PCL15 after 6 and 48 h, respectively. For MM-PBCL22, 26.6 and 60.8% of PTX were released after 6 and 48 h, respectively, making this formulation between MM-PCL22 and MM-PCL15 in terms of PTX release.

Figure 6

In vitro release profile of physically loaded PTX from different micellar formulations at 10% fetal bovine serum (FBS) media. Data represent average ± SD (n = 3).

In Vitro Cytotoxicity

The results of in vitro cytotoxicity of PTX loaded in RTX-MMs versus plain MMs and free PTX against both KG-15 and SUP-M2 cells following 24 and 72 h of incubation is illustrated in Figure A,B, respectively. This study was conducted at a concentration equivalent to 400 ng/mL of PTX (close to IC50 of PTX in the cell lines under study). Overall, at 24 h of incubation time, irrespective of the cell line used and the core forming block in the mixed micellar formulations of PTX, no significant difference between the cytotoxicity of RTX-modified and plain micelles was observed (Unpaired Student’s t-test, P > 0.05). However, free PTX was significantly more cytotoxic than the micellar formulations in both cell lines (Unpaired Student’s t-test, P < 0.05). At 72 h, for the SUP-M2 cells (CD20–), there was no significant difference between the cytotoxicity of PTX loaded in RTX-conjugated mixed micelles and plain mixed micelles regardless of the core-forming structure (Unpaired Student t-test P > 0.05). However, free PTX was found to be more cytotoxic than micellar PTX formulations (both plain and RTX modified) (Unpaired Student t-test P < 0.05). In KG-15(CD20+) cells, free PTX showed similar cytotoxicity to that of encapsulated PTX in both plain and RTX-modified micelles (Unpaired Student t-test P > 0.05).

Figure 7

In vitro cytotoxicity of PTX encapsulated in RTX-modified MM-PCL15, RTX modified MM-PCL22, and RTX-modified MM-PBCL22, in comparison with PTX encapsulated in plain mixed micelles and free PTX against SUP-M2 and KG-15 cells after (A) 24 h and (B) 72 h. Data represent average ± SD (n = 3) *denotes a statistically significant difference (Unpaired Student’s t-test P < 0.5), while ns denotes a statistically nonsignificant difference (Unpaired Student’s t-test P > 0.5).

Discussion

The main objective of this research was to come up with simple methods for the development of traceable mAb-modified poly(ester)-based micelles for active cancer targeting so that the fate of nanocarriers can be easily followed in the biological systems. Our aim was to develop methods that are (a) easy and feasible, (b) use commercially available functional materials for mAb conjugation (c) ensure the maintenance of a fluorescent tag with the nanocarrier, and (d) do not jeopardize the thermodynamic and kinetic stability of developed nanocarriers upon introduction of the fluorescent tag. For this purpose, RTX was used as a representative clinically relevant mAb for active targeting of CD20 positive NHL cells. In effect, post insertion of Cy5.5 tagged RTX-PEG-DSPE into polymeric micelles composed of PEO-PCL-Cy3 or PEO-PBCL-Cy3 of different core lengths was pursued (Figures and 3). Since the Cy3 azide was chemically conjugated into the core of the micelles and the Cy5.5 NHS ester dye was conjugated to the RTX on the distal end of the shell of the micellar system (Figure ), no interference between the two dyes was expected or observed. In addition, this design will allow tracing of the components of the mixed micelles in the cells as well as in vivo following administration, providing clues on the micellar stability in biological systems.

Rituximab was covalently attached to the distal end of PEG-DSPE micelles through the reaction of the reactive NHS ester end of NHS-PEG-DSPE with the aliphatic free amine from the lysine of RTX to form a stable amide bond. This reaction was found to be appropriate for our purpose for several reasons. First, the reaction is simple and can be carried out in aqueous media at mild conditions. The amino group of the mAb is a good nucleophile at pH = 8, which can attack the carboxyl group of the NHS ester of the PEG-DSPE micelles to form a stable amide bond. Conveniently, the complementary determining region (CDR) of RTX will not be inactivated by this reaction, since, CDR doesn’t include any lysine.[25−27] Besides, NHS-PEG-DSPE is commercially available.

For the development of Cy3-tagged micelles, we prepared PEO-PCL-PPrCL and PEO-PBCL-PPrCL block copolymers, first. Cy3 was then conjugated to the PPrCL end of the block polymers. Mixed micelles were then prepared through co-incubation of RTX-modified or plain PEG-DSPE micelles with PEO-PCL/PEO-PBCL ones. It was assumed that the PEG-phospholipids can spontaneously insert themselves into the more stable PEO-poly(ester) micelles to form mixed immunomicelles.[28−30] The validity of this assumption was confirmed by following the changes in the size of particles upon co-incubation of two micellar populations that turned into one population over time (Table ). The level of success in the formation of stable traceable micelles by postinsertion of RTX-PEG-DSPE into PEO-PBCL or PEO-PCLs of different PCL lengths was further evaluated by following the kinetic and thermodynamic stability of mixed micelles versus micelles from their individual polymer/lipid counterparts (Figure ). The stability of mixed micelles was also evaluated investigating the proportion of Cy5.5/Cy3double positive CD20+ and CD20– cells versus the proportion of cells that were only positive for either Cy5.5 or Cy3 (Figure ).

Previous studies have shown a direct correlation between the hydrophobicity of block copolymers and the thermodynamic stability of polymeric micelles as reflected by a decrease in their CMC.[31,32] Increasing the hydrophobicity of the polymer, either through elongation of PCL chain or the introduction of more hydrophobic benzyl groups in the core-forming block has been shown to improve the thermodynamic and kinetic stability of micelles formed from individual block copolymers.[33,34] This has been the case here for micelles prepared from individual block copolymers and mixed micelles and reflected in the reduced CMC of micelles (an indication of thermodynamic stability) showing higher tendency of block copolymers for self-assembly at lower concentrations. Hydrophobicity of block copolymers (those having longer PCLs or those with PBCL structures) also contributed to the resistance of the micellar structure to dissociation in the presence of SDS, implying improved core rigidity due to hydrophobic interactions within the micellar core. In general, mixing with PEG-DSPE appeared to lower the thermodynamic stability (increase the CMC) of PEO-poly(ester)s under current study. This effect was more noticeable for PEO-PCL15 than PEO-PCL22 and PEO-PBCL22 micelles, however. MM-PBCL22 was the most thermodynamically stable system among the three mixed micelles (Table ), which is due to its higher hydrophobicity. Mixing of PEG-DSPE did not seem to affect the kinetic stability of PEO-PCL22 and particularly that of PEO-PBCL22 micelles, but loosened the micellar structure for those formed from PEO-PCL15 (Figure ).

We then investigated the formation of mixed micelles and their stability in cell media, by comparing the association of Cy5.5/Cy3 dye with CD20 positive versus CD20 negative cells for MM-PCL15, MM-PCL22, and MM-PBCL22.[21,35] In our work, RTX was reacted with NHS-PEG-DSPE at a very low ratio (1:100) so that the stealth properties provided by the PEG shell were not compromised following RTX modification of micellar shell (Table ).[28] Despite the low RTX density, the RTX-conjugated MM-PCL22 and MM-PBCL2 showed significantly increased association with CD20+ cells. This was evidenced by a significant increase in the Cy5.5+/Cy3+ population for KG-15 cells treated with RTX-micelles compared with those treated with plain MMs (Figure ). This observation reflects the high affinity and selectivity of RTX toward CD20 as well as the abundant and well-exposed countenance of CD20 on the surface of KG-15 cells.[28,36,37]

Among different RTX-MMs under study, RTX-MM-PCL15 showed the lowest population of Cy5.5/Cy3+ cells (Figures and 5), whereas a high population of Cy5.5+ positive cells was observed. The latter points to the instability of MM-PCL15 and its rapid dissociation to two separate micellar populations consisting of Cy5.5-labeled RTX-PEG-PSPE and Cy3-labeled PEO-PCL15 micelles and is in line with our earlier observation on the low kinetic stability of these micelles. In agreement with the above observation, pretreatment of CD20+ cells with free RTX was the most effective in lowering the proportion of Cy5.5+/Cy3+ cells for MM-PCL22 and MM-PBCL, again reflecting the higher stability of these micelles.

Upon PTX loading, among different mixed micellar formulations, MM-PCL22 showed the highest EE (Table ) and the slowest release profile (Figure ). These results were also in line with our previous findings.[24] With the exception of KG-15 cells at 72 h incubation time, micellar PTX formulations under study showed lower cytotoxicity compared with free PTX. This may be attributed to the slow release of PTX from the micellar formulations. No significant difference between the cytotoxicity of PTX loaded in RTX-conjugated mixed micelles following 24 or 48 h of incubation and that of plain mixed micelles was observed, irrespective of CD20 positivity of treated cells (Figure ). This may be a reflection of the noninternalizing nature of CD20, where RTX-modified micelles are expected to provide extracellular release of the loaded PTX in the vicinity of cancer cells. Nevertheless, the conditions of the in vitro study where plain and RTX-modified micelles are both in similar contact with cells irrespective of their CD20 level of expression may not be a good predictor of in vivo conditions, where more rapid wash off of plain (unmodified carriers) from the surface of the cells can take effect.[28,38] The loading level of PTX in the formulations under study was below 1%. Despite high potency of PTX, developed formulations would need a high level of excipients to achieve clinically relavant doses in animal models for in vivo investigations pointing to the necessity of further optimization studies to increase drug loading levels.

Conclusions

Traceable mAb-modified polymeric micelles without jeopardizing the functionality of the mAb and the stability of the nanocarrier can be developed through postinsertion of fluorescently tagged mAb-PEG-DSPE to PEO-poly(ester) micelles. A positive correlation between the hydrophobicity of the poly(ester) core and the stability of mixed micelles was established that led to an improved targeting efficiency for RTX-modified mixed micelles of higher hydrophobicity in the poly(ester) core.

Experimental Section

Materials

Methoxy poly(ethylene oxide) (PEO) (Mwt 5000 Da), sodium dodecyl sulfate, and L-Ascorbic acid (99%) were purchased from Sigma (St Louis, MO). ε-Caprolactone was purchased from Lancaster synthesis (U.K.). α-Benzylcarboxylate-ε-caprolactone (BCL) and α-propargyl-carboxylate-ε-caprolactone were synthesized by Alberta Research Chemicals Inc. (ARCI, Edmonton, AB, Canada) based on methods published previously by our group.[39] 1, 2-Distearoyl-sn-glycero-3-phosphoethanolamine-N-[succinimidyl (polyethylene glycol)] (DSPE-PEG3400-NHS) and methoxy poly(ethylene oxide) (Mwt 2000 Da), covalently linked to distearoyphosphadidyl ethanolamine (mPEO2000-DSPE), were purchased from Nanocs Inc. Paclitaxel (PTX) (purity> 99.5) was purchased from LC Laboratories (Woburn, MA). Rituximab (anti CD20 antibody) and Taxol were provided by Cross Cancer Institute. Stannous octoate was purchased from MP Biomedicals Inc. (Germany). Spectra/por dialysis tubing (MWCO, 3.5kDa) was purchased from Spectrum Laboratories (Rancho Dominguez, CA). Cy3 azide, Cy5.5 NHS dye, and Cu(II) (Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine) (TBTA) complex were purchased from Lumiprobe (Hallandale Beach, Florida). All other chemicals and reagents used were of analytical grade.

Cell Culture

Cell culture media RPMI 1640, fetal bovine serum (FBS), and penicillin −streptomycin-L-glutamine were purchased from GIBCO, Life Technologies INC. (Burlington, ON, Canada). KG-15 cell line (CD20+) and SUP-M2 cell line (CD20–) were received from the laboratory of Dr. Raymond Lai, Department of Pathology and Medicine, University of Alberta. Epstein-Barr virus positive KG-15 cells were generated from patient samples as part of an approved protocol from the Human Research Ethics Board, University of Alberta (#Pro00058140) to Dr Raymond Lai’s lab. Anaplastic large cell lymphoma SUP-M2 cells were purchased from ATCC.

Synthesis of PEO-PCL and PEO-PBCL Diblock Copolymers

Methoxy poly(ethylene oxide)-b-poly (ε-caprolactone) (PEO-PCL) with different DPs of the PCL block was synthesized by ring opening polymerization of ε-caprolactone using methoxy PEO (5000 Da) as the initiator and stannous octoate as catalyst.[40] Briefly, melted PEO (0.5 g) and dried ε-caprolactone (0.285 mg or 0.170 mg) were mixed in a vial with stannous octoate. The vials were then sealed under vacuum. The reaction was preceded by placing the vial in the oven at 140 °C for 6 h and vortexing every 30 min for the first 2 h. Cooling the vials at room temperature terminated the reaction. For the PEO-PBCL block copolymer, methoxy PEO (5000 Da) (0.5 g), α-benzylcarboxylate ε-caprolactone (BCL) (0.545 g), and stannous octoate were used as initiator, monomer, and catalyst, respectively, under identical reaction conditions as described above. The product of this reaction was purified by dissolving it in dichloromethane and then precipitation in n-hexane. The product was then washed with n-hexane twice and dried in a vacuum oven overnight at room temperature. The samples were dissolved in deuterated chloroform (CDCl3) and then subjected to 1H NMR 600 MHz (Bruker; Billerica MA) for characterization. The DP of PCL and PBCL blocks was determined from the 1H NMR based on the peak intensity ratio of the protons from the ethylene moiety of PEO (δ = 3.65 ppm) to peak intensity of the proton from the methylene of the caprolactone (δ = 4.06 ppm) considering a molecular weight of 5000 Da for the PEO block. The DP for each block is represented as a subscript in the abbreviated name of polymers in the manuscript.

Preparation of α-Propargyl-ε-caprolactone End-Capped PEO-PCL and PEO-PBCL Block Copolymers

The α-propargyl-carboxylate-ε-caprolactone (PrCL) (60 mg) was polymerized at the end of PEO114-PCL22, PEO114-PCL15 or PEO114-PBCL22 (200 mg) through reflux in dry toluene (10 mL) for 24 h using stannous octoate as catalyst. The prepared PEO114-PCL22-PPrCL4, PEO114-PCL15-PPrCL4, and PEO114-PBCL22-PPrCL4 were purified by precipitating the solution in n-hexane. The precipitate was collected and dried in a vacuum oven overnight. The samples were dissolved in CDCl3 and then subjected to 1H NMR 600 MHz (Bruker 300 AM; Billerica MA) for characterization. The degree of polymerization of PPrCL was calculated based on the peak intensity ratio of the protons from the ethylene moiety of the PEO (δ = 3.65 ppm) to peak intensity of the proton from the methylene of the propargyl (δ = 2.50 ppm), considering a molecular weight of 5000 Da for the PEO block.

Conjugation of Cy3 Azide to PEO-PCL-PrPCL and PEO-PBCL-PrPCL

Cy3 azide was conjugated to the propargyl end of the PEO-PCL-PrPCL and PEO-PBCL-PPrCL block copolymers by click chemistry (Figure ). The reaction is copper-catalyzed azide-alkyne cycloaddition that allowed the synthesis of 1, 4 substituted trizole. PEO114-PCL22-PPrCL4, PEO114-PCL15-PPrCL4 or PEO114-PBCL22-PPrCL4 polymer (0.01 mM), Cy3 azide (0.001 mM), Cu(II) TBTA (0.0001 mM), and ascorbic acid (0.0001 mM) were added to degassed DMSO. Cu(I) is the active form of the catalyst; however, Cu(II) was used because Cu(I) is unstable and cannot be stored; so it is generated in situ by reducing Cu(II) to Cu(I) with ascorbic acid. The TBTA complex was used to stabilize Cu(I) toward disproportionation. The reaction was kept stirring in dark for 24 h.[41] The solution was purified by dialysis against double distilled water for 24 h changing water every hour for the first 4 h. The solution was then centrifuged for 5 min to remove any remaining free dye. The Cy3 conjugated polymers were then freeze-dried and stored until further use.

Conjugation of Rituximab or Cy5.5-labeled Rituximab to the NHS-PEG-DSPE Micelles

Rituximab (Mwt 145 KDa) was labeled with Cy5.5 NHS ester at an applied 1:20 molar ratio (mAb: Cy5.5). RTX contains 49 lysines in total. The NHS ester dye was reacted with the primary amine from the antibody in PBS at slightly alkaline media of pH = 8.3 for 4 h at room temperature in the dark to yield a stable amide bond. The conjugate was then purified by size exclusion chromatography on the Sepharose CL-6B column using PBS (pH 7.4) as the mobile phase. About 40% of the lysine’s primary amines in RTX can react with Cy5.5. Therefore, 60% of remaining primary amines from lysine in the RTX is free for the reaction of NHS-PEG-DSPE.

Micelles were prepared from the NHS-PEG-DSPE phospholipid by cosolvent evaporation method. Briefly, NHS-PEG-DSPE (0.001 mM) was dissolved in acetone, followed by drop-wise addition of this solution to 2 mL of phosphate buffer saline (PBS) while stirring. The solution was left overnight to allow evaporation of the acetone. The RTX or Cy5.5-labeled RTX was added to NHS-PEG-DSPE micelles at a 1:100 molar ratio. The NHS group from the micelles was reacted with the primary amine from the antibody at pH 8 for 4 h in PBS. The conjugate was then purified by Sepharose CL-6B column using PBS (pH = 7.4) as the mobile phase.

Formation of Mixed Micelles by Postinsertion Method

Mixed micelles were formed by incubating PEO114-PCL15-PPrCL4-Cy3, PEO114-PCL22-PPrCL4-Cy3, or PEO114-PBCL22-PPrCL4-Cy3 micelles with the RTX-PEG-DSPE (plain or labeled with Cy5.5 NHS dye) at a 1:1 molar ratio in PBS for 24 h (Figure ).

Quantification of the Conjugated mAb on the Micelles

Mixed micelles composed of RTX-PEG-DSPE/PEO114-PCL15-PPrCL4 (RTX-MM-PCL15), RTX-PEG-DSPE/PEO114-PCL22-PPrCL4 (RTX-MM-PCL22), or RTX-PEG-DSPE/PEO114-PBCL22-PPrCL4 (RTX-MM-PBCL22) with no dye were prepared using the method described under the method described above. The mixed micelles were then purified from free unconjugated RTX by passing through Sepharose column using PBS as the mobile phase. The eluent was collected in 32 fractions of 1 mL each, from the column. Each fraction was subjected to dynamic light scattering (DLS) and protein quantification by Micro BCA assay to distinguish and separate the parts containing free RTX, RTX modified micelles, and those containing plain micelles. Micro BCA assay using absorbance at 560 nm (by Synergy Hybrid plate reader, Winooski) was also used to quantify the concentration of mAb in fractions containing RTX conjugated micelles and those containing free RTX.

Micelle Size

The hydrodynamic diameter and polydispersity of PEO114-PCL22-PPrCL4, PEO114-PCL15-PPrCL4, PEO114-PBCL22-PPrCL4, and NHS-PEG-DSPE micelles and that of mixed micelles without any conjugated dye (1 mg/mL) were measured using DLS (ZETA-Sizer Nano. Malvern Instruments Ltd., U.K.). The analysis was performed at a scattering angle of 173° at 25 C°.

Micelle Stability

Critical micelle concentration (CMC) of PEO114-PCL22-PPrCL4, PEO114-PCL15-PPrCL4 PEO114-PBCL22-PPrCL4, and their mixed micelle counterparts with NHS-PEG-DSPE (without dye or antibody) as well as NHS-PEG-DSPE micelles alone was determined by DLS.[42] Single attenuator index was used throughout all measurements. The scattered light was detected at an angle of 173°. Measurements were carried out in polystyrene cells at 25 °C. Using the scattered light intensity as a function of polymer concentration, a series of micelle solutions (0.060–1000 μg/mL) in deionized water were prepared for different micellar formulations. The count rate (Kcps) as a function of the intensity of the scattered light was plotted against log concentration of a polymer. The intersection of the two linear graphs in the sigmoidal curve was used to determine the CMC on the x axis.

Kinetic stability was measured with DLS by mixing the micelles formed from individual polymers and those of mixed micelles (without dye or antibody) at a total concentration of 1 mg/mL with sodium dodecyl sulfate (SDS) (20 mg/mL) in a 2:1 v/v ratio. The scattered light intensity and the PDI were measured at different incubation time intervals.[43]

Measuring the Cellular Association of Micelles by Flow Cytometry

To evaluate the cellular association of micelles, in vitro, two different lymphoma cell lines KG-15 (CD20+) and SUP M2 (CD20–) were used. KG-15 and SUP M2 cells (1×105 per well) were suspended in the media in a 24-well plate and incubated at 37 °C until the confluence was around 70%. The cells were treated with the Cy5.5-labeled RTX-MM PCL15-Cy3, RTX-MM PCL22-Cy3, or RTX MM PBCL22-Cy3 as well as their plain (no RTX) counterparts for 4 h. A competition study was also carried out by pretreating the cells with free RTX at a concentration equivalent to 10 times the concentration of the conjugated antibody on micelles, 2 h prior to the micellar treatment. After the incubation, cells were centrifuged and washed twice with cold PBS before analysis using a BD FACS Canto II flow cytometer. The cells were gated using forward versus side scatter to exclude debris and dead cells before the analysis of 10 000 cells. The data were analyzed with FlowJo cell analysis software version 10 (FlowJo, LLC, Ashland, Ore.) and plotted as a Quadrant plot.

Encapsulation of PTX in Polymeric Micelles and Their Characterization

Paclitaxel (PTX)-loaded polymeric micelles were prepared by the dialysis method.[44] The block copolymers (PEO114-PCL15-PPrCL4, PEO114-PCL22-PPrCL4, or PEO114-PBCL22-PPrCL4) and PTX were dissolved in N,N-dimethyl formamide (0.5 mL). This solution was added to doubly distilled water (4 mL) in a drop-wise manner under moderate stirring for 1 day, followed by organic solvent removal by dialysis against double distilled water for another day (Spectrapor, MW cut-off 3500 Da). The micellar solution was then centrifuged at 11 600g for 5 min to remove any free unimers and unencapsulated PTX. The PTX-loaded PMs were than incubated with NHS-PEG-DSPE for another day to allow the formation of mixed micelles. An aliquot from the micellar solution was diluted with acetonitrile allowing the micelles to break. The solution was analyzed for PTX concentration using LC-10AD SHIMADZU HPLC system at a flow rate of 1.0 mL/min at room temperature. The detection was performed at 227 nm using a SPD-10Avp SHIMADZU UV–VIS detector. Reversed phase chromatography was carried out with a Microsorb-MV 5μ C18-100 Å column (4.6 mm × 250 mm) with 40 μL of the sample injected in a gradient elution using 0.1% trifluoroacetic acid aqueous solution and acetonitrile. The percent of acetonitrile was 40% at time zero and increased to 100% within 15 min.[45] The amount of loaded PTX was measured and used to calculate PTX encapsulation efficiency and loading as described in the following equations

Release of PTX from Polymeric Micelles

Release of PTX from MM-PCL15, MM-PCL22, or MM-PBCL22 and micelles prepared from individual PEO-PCLs or PEO-PBCL or PEG-DSPE polymer was determined in 0.01 M phosphate buffer (pH 7.4) containing 10% fetal bovine serum (FBS). The micellar solution (3 mL) was transferred into a dialysis bag (Spectrapor, MW cut-off 3500 Da) and placed into 500 mL of 0.01 M phosphate buffer (pH 7.4) with 10% FBS. The release study was performed at 37 °C in a Julabo SW 22 shaking water bath (Germany). At selected time intervals, the whole release media has been replaced with a fresh one and aliquots of 200 μL were withdrawn from the inside of the dialysis bag for HPLC analysis and replaced with double distilled water. The amount of PTX released was calculated by subtracting the amount of PTX remaining in the dialysis bag from the initial amount of PTX. The release profiles were compared using the similarity factor, f2, and the profiles were considered significantly different if f2 < 50.where n is the sampling number and R and T are the percent released of the reference and test formulations at each time point “j”.

In Vitro Cytotoxicity Study

The cytotoxicity of free PTX and PTX encapsulated in RTX-MM PCL15, RTX-MM PCL22, or RTX-MM PBCL22 and their plain counterparts (without RTX modification) against 2 different cell lines (KG-15 and SUP-M2) was tested using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The cells were grown in RPMI media supplemented with 10% FBS and 1% P/S (Penicillin and Streptomycin) antibiotics and maintained at 37 °C with 5% CO2 in a tissue culture incubator. Growth media containing 5000 cells was placed in each well of a 96-well plate and incubated until the confluence was around 50%. PTX-loaded micellar solutions, free PTX, and empty micelles at different concentrations were incubated with the cells for 24 or 72 h. MTT solution (5 mg/mL) was added to each well, and the plates were then incubated for another 4 h. The plates were centrifuged at 1000g for 10 min, and the media was aspirated and replaced by DMSO to dissolve the formazan crystals. The cell viability was detected by measuring the optical absorbance at 570 nm using a Synergy Hybrid plate reader (Winooski). The mean absorbance of each treatment was determined and converted to the percentage of viable cells relative to the control.

Statistical Analysis

Statistical analysis was performed using either unpaired Student’s t-test or one-way ANOVA with Tukey post-test analysis, as indicated in the text. The significance level (α) was set at 0.05. For nonlinear regression analysis, Graphpad prism was used (version 7.00, Graphpad Software Inc., La Jolla, CA). All experiments were conducted in triplicate unless mentioned otherwise. In tables or graphs, data points are represented as mean ± standard deviation (SD).