Thermal Targeting of an Acid-Sensitive Doxorubicin Conjugate of Elastin-like Polypeptide Enhances the Therapeutic Efficacy Compared to the Parent Compound in vivo

Abstract

Elastin-like polypeptides (ELP) aggregate in response to mild hyperthermia, but remain soluble under normal physiological conditions. ELP macromolecules can accumulate in solid tumors because of the enhanced permeability and retention effect. Tumor retention of ELPs can be further enhanced through hyperthermia-induced aggregation of ELPs by local heating of the tumor. We evaluated the therapeutic potential of ELPs in delivering doxorubicin (Dox) in the E0771 syngeneic mouse breast cancer model. The ELP-Dox conjugate consisted of a cell penetrating peptide at the N-terminus and the 6-maleimidocaproyl hydrazone derivative of doxorubicin at the C-terminus of ELP. The acid-sensitive hydrazone linker ensured release of Dox in the lysosomes/endosomes after cellular uptake of the drug conjugate. ELP-Doxo dosed at 5 mg Dox equivalent/kg, extended the plasma half-life of Dox to 5.5 hours. Additionally, tumor uptake of ELP-Dox increased 2-fold when hyperthermia was applied, and was also enhanced compared to free Dox. Although high levels of Dox were found in the heart of animals treated with free Dox, no detectable levels of Dox were found in ELPDox treated animals indicating a correlation between tumor targeting and reduction of potential cardiac toxicity of ELP-Dox. At an optimal dose of 12 mg Dox equivalent/kg, ELP-Dox in combination with hyperthermia induced a complete tumor growth inhibition, which was distinctly superior to free drug which only moderately inhibited tumor growth. In summary, our findings demonstrate that thermal targeting of ELP increases the potency of doxorubicin underlying the potential of exploiting ELPs to enhance the therapeutic efficacy of conventional anticancer drugs.

Introduction

Chemotherapy is the mainstay for treating different types of breast cancer, especially in the neoadjuvant and palliative setting. The effective use of the most common chemotherapeutics including doxorubicin, epirubicin, paclitaxel is however, often limited by their narrow therapeutic window resulting in unacceptable toxicity, lack of tumor selectivity, and/or multidrug resistance. Targeting strategies that can deliver these drugs specifically to the tumor site is thus highly desirable and could result in an improved therapeutic index and concomitant benefit for the patient.

In this study, we describe a thermally responsive drug delivery platform technology based on the genetically engineered biopolymer elastin-like polypeptide (ELP) that targets solid tumors by combining passive tumor targeting strategy with the application of mild hyperthermia (1–6). ELPs undergo inverse phase transition at a specific temperature known as inverse transition temperature (T_t), below which they stay in solution and above which, due to collapse of the hydrophobic structure form aggregates (7). This process is completely reversible. ELP is composed of repetitive units of Val-Prol-Gly-Xaa-Gly, where Xaa is a guest residue that can be any amino acid except proline. The number of repeats as well as the molecular composition of Xaa directly influence the T_t of ELP (8). Compared to low-molecular weight drugs ELPs remain in blood circulation with a half-life of ~8.7 h (9) sufficient for passive tumor targeting through the enhanced permeability and retention (EPR) effect characterized by hypervasculature and an impaired lymphatic drainage system of the tumor tissue (10), and thermally responsive ELPs can be actively targeted to the tumor tissue by application of local hyperthermia on the tumor (6). ELPs are conveniently expressed in E. coli and purified in high yields using inverse thermal cycling (11), and they have a defined molecular weight, advantages that set ELPs apart from synthetic polymers with similar properties.

Recently, Mackay et al. reported that conjugates of hydrophilic ELP (T_t >> 42 °C) and doxorubicin have enhanced anti-tumor activity over doxorubicin against C26 murine colon carcinoma tumors in BALB/c mice (12). However, since the ELP variant used in this study was not responsive to mild hyperthermia, the thermal targeting properties of ELP could not be exploited. Because ELP is a thermally responsive molecule, drug delivery by ELP should be enhanced in response to the temperature difference between the systemic environment and the tumor region which is exposed to hyperthermia. Therefore, we used a thermally responsive ELP1 variant (59 kDa) that has T_t close to 40 °C to evaluate the enhancement of doxorubicin delivery by the application of focused mild hyperthermia on the tumor. Additionally, we modified the N-terminus of ELP1 with the cell penetrating peptide, SynB1 for enhanced intratumoral and intracellular uptake (13–15). The cell penetrating thermally responsive SynB1-ELP1 polypeptide was conjugated to a thiol-reactive doxorubicin prodrug. This prodrug of doxorubicin, (6-maleimidocaproyl) hydrazone of doxorubicin (DOXO-EMCH, 1) binds rapidly and selectively to cysteine-34 of circulating albumin in situ and has marked anti-tumor activity in pre-clinical studies (16–19) and in a phase I clinical trial (20). DOXO-EMCH, meanwhile renamed INNO-206 is scheduled to enter a sarcoma phase 2 study (see www.cytrx.com).

DOXO-EMCH is a valid candidate to evaluate thermal targeting by ELP because the maleimide moiety on DOXO-EMCH allows for selective conjugation to cysteine residues at the C-terminus of ELP. Additionally, the acid-sensitive hydrazone linker acts as a predetermined breaking point for ensuring effective cleavage of the drug either in the slightly acidic tumor microenvironment or after intracellular uptake in the acidic compartments of the endosomes and/or lysosomes (20). Since Doxo-EMCH delivered by SynB1-ELP ultimately results in the delivery of doxorubicin (Dox), the ELP-doxorubicin delivery system is simply referred to as SynB1-ELP-Dox. In this report, we show that thermal targeting of Dox delivered by SynB1-ELP enhances its tumor uptake, plasma kinetics, and maximum tolerated dose compared to free Dox, resulting in a potent anti-cancer agent capable of stabilizing disease progression.

Materials and Methods

Synthesis and conjugation of DOXO-EMCH to SynB1-ELP

The thermally responsive SynB1-ELP1 and the thermally insensitive SynB1-ELP2 constructs were cloned as described (21) with three Gly-Gly-Cys repeats at the C-terminus for conjugation with DOXO-EMCH (Figure 1A). ELP2 (61 kDa, T_t ~ 70 °C) is a size-matched control for ELP1 and is used to parse the non-specific effects of hyperthermia (21). The polypeptides were expressed in E. coli using the hyperexpression protocol and purified by inverse thermal cycling (11).

Figure 1. Characteristics of SynB1-ELP1-Dox.

(A) The macromolecular carrier of Dox is composed of the SynB1 cell penetrating peptide (underlined), the thermally responsive biopolymer elastin-like polypeptide, ELP1 (within parentheses, where X_aa indicates a V:G:A ratio of 5:3:2 and n indicates 150 repeats), and three terminal cysteine residues where up to three molecules of (6-maleimidocaproyl)hydrazone derivatives of doxorubicin, 1 (B) can bind. For the thermally insensitive ELP variant, ELP2 the V:G:A ratio is 1:7:8 and n is 160. (C) The effect of Dox labeling on T_t of ELP shows a logarithmic relationship with the ELP concentration.

Doxorubicin was derivatized at its C-13 keto position with a maleimidocaproyl hydrazide linker yielding (6-maleimidocaproyl)hydrazone derivative of Dox (DOXOEMCH, 1, Figure 1B) which was covalently linked to the three cysteine residues on ELP as described (17, 22). Briefly, the SynB1-ELP1 or SynB1-ELP2 polypeptide was conjugated to DOXO-EMCH in a Michael addition by incubating 125 μM protein with 10-fold molar excess of tris-(2-carboxyethyl)phosphine (Invitrogen) at +4 °C for 20 min and then with 4-fold molar excess of Doxo-EMCH at 27 °C for 1 h in the dark in a final volume of 10 mL of 0.1 M Na₂CO₃ buffer pH 9. The majority of unreacted label was removed by inverse thermal cycling, and the drug-protein conjugate was resupsended in phosphate buffered saline (PBS) and passed through a desalting spin column (Thermo Scientific). Dox concentration was determined using Dox absorbance at 495 nm and extinction coefficient, 9250 M⁻¹cm⁻¹. Dox absorbance at 280 nm was factored out and the protein concentration was calculated using absorbance at 280 nm and the protein's molar extinction coefficient, 7190 M⁻¹cm⁻¹ based on the following equation:

$$\left[\text{Protein}\right]=\frac{({\mathrm{Abs}}_{280nm}-(0.713\times {\mathrm{Abs}}_{495nm}))}{7190{\mathrm{M}}^{-1}{\mathrm{cm}}^{-1}}$$

Thermal Characterization of ELP-Dox

SynB1-ELP1 labeled with DOXO-EMCH was diluted to various concentrations of ELP in 100% serum. The solution was heated at a rate of 1 °C/min from 20 to 80 °C, and the resulting turbidity was measured as a function of temperature (Cary, Varian Instruments). The percentage of maximum absorbance at 600nm was calculated for each dataset, and T_t was defined as the temperature at which 50% of the maximum was observed. T_t versus concentration curves were fitted to a logarithmic fit to estimate the thermal targeting range of the SynB1-ELP-Dox conjugates.

Cell line and tumor model

The E0771 cells originally isolated from a spontaneous cancer in C57BL/6 were obtained from Dr. F.M. Sirotnak, Memorial Sloan-Kettering Cancer Center, New York (23). They are characterized as estrogen receptor positive medullary breast adenocarcinoma and closely resemble the human disease (24). No further authentication was done for this cell line. Cells were cultured in RPMI 1640 medium (Mediatech) supplemented with 10% fetal bovine serum, 100 units/mL penicillin, 100 μg/mL streptomycin, and 25 μg/mL amphotericin B (Invitrogen), and 2 g/L of NaHCO₃ at 37 °C in a humidified atmosphere containing 5% CO₂.

Female C57BL/6 mice were purchased from the National Cancer Institute. All animal experiments were conducted in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals as approved by the University of Mississippi Medical Center Institutional Animal Care and Use Committee.

Intracellular distribution

E0771 cells at ~50% confluency were exposed to 40 μM free Dox or SynB1-ELP1-Dox for 1 min or 20 μM free Dox or SynB1-ELP1-Dox for 30 or 240 min at 37 °C. To determine the effect of temperature on Dox distribution, cells were exposed to 20 μM SynB1-ELP1-Dox or free Dox for 1 h at 37 or 42 °C, after which the drugs were replaced with fresh medium and cells were incubated at 37 °C for 24 h. All treated cells were rinsed with PBS and fixed with 4% paraformaldehyde for 10 min at room temperature. After rinsing with PBS for three times, cells were visualized for Dox fluorescence using a laser scanning confocal microscope (TS2 Leica). All treatment concentrations were based on Dox equivalent.

Tumor implantation

E0771 cells grown to ~80% confluence were harvested by trypsin digestion. Cells were washed in PBS, counted using a Coulter counter, and diluted in PBS. With the mice under isofluorane anesthesia, 1×10⁶ cells/200 μL/mouse were injected subcutaneously near the fat pad of the fourth mammary gland in the lower right abdomen.

Pharmacokinetics study

Mice bearing ~250 mm³ tumors were anesthetized by isoflurane, and a Micro-Renathane catheter (Braintree Scientific) was placed in the femoral artery. Animals were intravenously (IV) injected via the femoral vein with 5 mg Dox equivalent/kg of either SynB1-ELP1-Dox or free Dox. Blood samples were collected over a 4 hour period via the arterial catheter in heparinized microhematocrit glass tubes. A 6 h sample was taken at the time of euthanasia by cardiac puncture. Tumors of the animals in the hyperthermia group were heated immediately following drug administration with infrared light generated by a laser emitting diode device (Mettler Electronics) that raises the core tumor temperature to 42 °C (6). The heating protocol consisted of a thermal cycling procedure with 20 min of heat application followed by 10 min heat withdrawal repeated 4 times up to 2 h to ensure maximum accumulation of ELP in the tumor (6, 25). Plasma from the cells was separated by centrifuging the samples at 15.7 × g for 5 min. Protein-bound doxorubicin was cleaved and extracted from plasma as described (12): briefly, 10 μL plasma sample was incubated with 490 μL of acidified 90% isopropanol (75 mM HCl) overnight at 4 °C in the dark. The isoproponal extract was collected by centrifugation at 15.7 × g for 10 min at 4 °C and loaded in black 96-well plates in duplicate at 200 μL/well. Dox fluorescence was measured at 485 nm excitation and 590 nm emission in a plate reader (BioTek). The raw data was fitted to a free Dox standard curve to estimate Dox plasma concentration with respect to time. Plasma clearance data were fitted to a two-compartmental model using Microcal Origin as described (6).

Biodistribution study

Dox distribution was measured using tumors and hearts harvested 6 h after drug administration described above. Tumors and hearts were excised, snap-frozen in liquid nitrogen and stored at −80 °C. Dox concentration was determined as described (12): briefly, tissues were weighed, suspended in 1.5 mL of acidified isopropanol, homogenized (PowerGen 125 tissue homogenizer, Fisher Scientific) and incubated overnight at +4 °C with continuous mixing. Dox was extracted in the supernatant fraction by centrifugation at 15.7 × g for 10 min at +4 °C. The homogenates were resuspended in 1.0 mL of 90% isopropanol, centrifuged, and the supernatant was collected. This process was repeated twice to ensure quantitative Dox extraction. The supernatant fractions were evaporated to dryness in a SpeedVac concentrator. The precipitate was pooled in 600 μL of 90% isopropanol. The resulting sample was centrifuged at 15.7 × g for 10 min at 4 °C to remove tissue debris and then loaded in duplicate in black 96-well plates at 200 μL/well. % injected dose per gram (%ID/g) was calculated by fitting Dox fluorescence from each sample to a free Dox standard curve after subtracting autofluorescence from saline control samples.

Tumor reduction at equimolar dose

Mice bearing ~150 mm³ tumor were injected with 5 mg/kg Dox equivalent of free Dox, thermally responsive SynB1-ELP1-Dox, or thermally insensitive SynB1-ELP2-Dox on d0, d2, and d4 intravenously via the femoral vein. Additionally, 280 mg/kg of SynB1-ELP1, the equivalent dose of protein present in the SynB1-ELP1-Dox treatment, was administered to the animals to rule out the toxicity of the unlabeled carrier. Animals in the hyperthermia group were heated using the heating protocol as described above. Tumor measurement (width² × length/2) and body weight were recorded every day after treatment. Animals were sacrificed on d14, when the control tumor size reached 2000 mm³. Tumors were harvested and weighed after euthanasia.

Maximum tolerated dose study

In a dose finding study, non-tumor bearing mice were injected with free drug at 8, 10, 20, or 40 mg/kg, or with SynB1-ELP1-Dox at 10, 15, 20, or 40 mg/kg Dox equivalent every other day three times via the femoral vein. Additionally, in order to determine the effect of hyperthermia on treatment tolerance, animals bearing tumors were treated with 12 or 15 mg/kg of SynB1-ELP1-Dox (Dox equivalent), followed by heating of the tumors. Animals were weighed every day to monitor the change in body weight following treatment. Mice that lost excess of 20% of their pre-treatment body weight were euthanized. Animals that showed < 20% weight loss were monitored up to a 30 day period. The highest dose at which no animal mortality and no weight loss > 20% was observed was defined as the maximum tolerated dose (MTD).

Anti-tumor efficacy at the maximum tolerated dose

Mice bearing tumors at ~150 mm³ were injected with the MTD of free drug (8 mg/kg), or the MTD of SynB1-ELP1-Dox (12 mg/kg) on d0, d2, and d4 intravenously via the femoral vein. Additionally, 650 mg/kg of SynB1-ELP1, the equivalent dose of protein present in the SynB1-ELP1-Dox treatment, was administered to the animals. Animals in the hyperthermia group were heated using the heating protocol described above. Tumor volume and body weight were measured daily up to day 14 at which point animals were sacrificed, and tumors were harvested and weighed.

Statistical Analysis

A one-way ANOVA with Bonferroni tests for pair-wise comparison of treatment groups was performed to analyze the statistical differences between the treatment groups and the untreated control.

Results and Discussion

Thermal characterization of SynB1-ELP-Dox

To attach the cargo drug DOXO-EMCH to ELP, we designed three cysteine residues at the carboxy terminus of ELP that were separated by diglycine spacers (Figure 1A). The cysteines ideally allowed for conjugation of up to three molecules of DOXO-EMCH, while the diglycine spacers ensured that Doxo-EMCH attachment was carried out with minimum steric hindrance. A Michael addition of DOXO-EMCH to ELP resulted in an average molar label ratio of 2.10. A turbidity assay done to determine the effect of drug labeling on ELP phase transition showed a logarithmic relationship between T_t and concentration of ELP. Although the attachment of Doxo-EMCH to SynB1-ELP1 decreased the T_t slightly, SynB1-ELP-Dox T_t remained above 37 °C in the lower micromolar range suitable for thermal targeting in animals (Figure 1C). At the same time, the T_t of SynB1-ELP2-Dox was detected at T>>37 °C, confirming that it is a suitable control to assess the nonspecific effect of hyperthermia.

SynB1-ELP1-Dox intracellular distribution

Previous studies with conjugates of DOXO-EMCH and albumin have shown that Dox does not enter the nucleus but localizes in the cytoplasm, primarily in mitochondria and golgi apparatus (26). Therefore, we evaluated the subcellular distribution of Doxo-EMCH delivered by SynB1-ELP1 by confocal microscopy. Dox delivered by SynB1-ELP1 exhibited a time-dependent nuclear distribution compared to free Dox that displayed nuclear localization immediately after treatment (Figures 2A–B). In this time course experiment, 1 min after exposure, Dox delivered by SynB1-ELP1 was mainly detected in the cytoplasm or on the outer surface of the cell, whereas free Dox localized primarily in the nucleus. Although some Dox from SynB1-ELP1 was still detected in the cytoplasm after 30 min exposure, by 4 h almost all of Dox from SynB1-ELP1 was found in the nucleus. Interestingly, the distribution of free Dox shifted from an exclusively nuclear localization after 30 min to both nuclear and cytoplasmic distribution by 4 h, which is similar to other findings (26). Furthermore, addition of hyperthermia did not alter the nuclear distribution of Dox delivered by SynB1-ELP1 since Dox fluorescence was detected in the nucleus even after 24 h post-treatment (Figure 2C). The SynB1 peptide is derived from the protegrin family of antimicrobial peptides and is known to facilitate cellular uptake of its cargo by adsorptive-mediated endocytosis (27), which likely places the ELP-Dox conjugates inside the acidic environment of endosomes/lysosomes where the cleavage of the acid-labile hydrazone linker releases Dox inside the cell. The cytoplasmic distribution of SynB1-ELP1 observed in a previous study (15) and the results of this data indicate that Dox is being cleaved from the ELP carrier intracellularly and being diffused to the nucleus where it acts in analogy to free Dox.

Figure 2. Delivery of Dox to the nucleus by SynB1-ELP1.

Laser scanning confocal images of E0771 cells exposed to (A) SynB1-ELP1-Dox or (B) free Dox at 40 µM Dox equivalent for 1 min or 20 µM Dox equivalent for 30 min and 4 h show a time-dependent nuclear distribution of Dox delivered by SynB1-ELP1. Additionally, images of E0771 cells exposed to (C) SynB1-ELP1-Dox or (D) free Dox at 20 µM Dox equivalent for 1 h at 37 °C or 42 °C indicate that hyperthermia does not affect the distribution of Dox delivered by SynB1-ELP1. Scale bar = 20 µm.

SynB1-ELP1-Dox pharmacokinetics and tumor distribution

Since unmodified ELP1 has a terminal plasma half-life of nearly 8 h (9), we evaluated the pharmacokinetics of SynB1-ELP1-Dox to determine whether it could also extend the plasma half-life of Dox. As anticipated, SynB1-ELP1-Dox displayed a biphasic Dox clearance and a longer systemic circulation presence (Figures 3A). In contrast, free drug displayed a rapid plasma clearance with relatively lower plasma concentrations (Figure 3B). Since we were only able to observe the distribution phase of free Dox using our assay method, the terminal half-life of SynB1-ELP1-Dox was compared to that of free Dox values from the literature. While the fit of SynB1-ELP1-Dox with no hyperthermia yielded a terminal half-life of 305.5±40.5 min, the fit of SynB1-ELP1-Dox in combination with hyperthermia yielded a terminal half-life of 336.5±70.2 min (Supplementary Table S1). The reported terminal half-life of Dox in C57BL/6 mice is 31.8 min (28), nearly 10-fold faster than ELP-conjugated Dox suggesting that ELP conjugation with Dox extends its systemic circulation time compared to free Dox.

Figure 3. Pharmacokinetics and tissue distribution of SynB1-ELP1-Dox.

C57BL/6 black mice were treated with either 5 mg Dox equivalent/kg of either SynB1-ELP1-Dox in the absence or presence of hyperthermia, or free Dox. (A) Dox plasma concentration after SynB1-ELP1-Dox treatment was fitted to a two-compartment model. (B) Due to relatively faster plasma clearance and lower detectable plasma concentrations, the plasma clearance of free Dox could only be confirmed up to 10 min; therefore, the pharmacokinetics data of free Dox (28) was used for comparison with SynB1-ELP1-Dox. Dox concentration in (C) tumors and (D) hearts at 6 h post-injection. * indicates p<0.01 (ANOVA, Bonferroni contrast, mean ± s.d; n = 5). ND indicates that Dox fluorescence signal was not detected over the control background.

To evaluate whether Dox tumor levels were enhanced by thermal targeting of ELP over free Dox, tumor samples were analyzed for Dox fluorescence. Additionally, since Dox accumulation in the heart is a major concern, we also assayed heart samples from the treated animals. With respect to Dox levels in the tumor, more than 2-fold increase in Dox uptake was observed after SynB1-ELP1-Dox treatment in combination with hyperthermia compared to treatment with only free drug (Figure 3C). Thermal targeting with SynB1-ELP1 also significantly enhanced Dox uptake in the tumor by a factor of 2 compared to when no hyperthermia was applied (Figure 3C) indicating that hyperthermia is necessary for increased tumor accumulation of SynB1-ELP1-Dox. This is because application of hyperthermia at the tumor site causes ELP1 to phase transition into aggregates, which essentially pools it in the tumor microenvironment. The thermal cycling hyperthermia application protocol, where the heat source is applied and withdrawn periodically, ensures that the aggregates trapped in the tumor vasculature resolubilize and diffuse into the tumor extravascular space (25). Subsequently, the SynB1 moiety on ELP1 facilitates its endocytosis in the tumor cells (13). Therefore, the function of hyperthermia is to maximize the local concentration of ELP that is delivered inside the tumor cells (6, 14, 15).

With respect to Dox levels in the heart, the fluorescence intensity of Dox from animals treated with SynB1-ELP1-Dox could not be detected over the control background signal. However, at the same dose of free Dox, we detected high levels of Dox in the heart (Figure 3D). Given that Dox related cardiotoxcitiy is a dose-limiting side-effect, this observation can have meaningful implications in the therapeutic use of ELP-Dox. We expect that through the means of thermal targeting, ELP-Dox can potentially reduce cardiotoxicity of Dox by steering it away from the heart and targeting it to the tumor. This is further supported by the findings that chronic treatment with DOXO-EMCH resulted in reduced cardiomyopathy compared to free Dox in rats (19).

Hyperthermia enhances tumor inhibition by SynB1-ELP1-Dox

The antitumor potential of SynB1-ELP1-Dox in combination with hyperthermia was evaluated and compared with that of standard doxorubicin at an equimolar dose of 5 mg/kg given on day 0, 2 and 4. Fourteen days after treatment, mice treated with SynB1-ELP1-Dox in conjunction with hyperthermia had a mean tumor volume that was nearly 4-fold smaller than the saline treated control tumors and nearly 2.5-fold smaller than the SynB1-ELP1-Dox without hyperthermia treated group (Figure 4A). In the presence of hyperthermia both SyB1-ELP1-Dox and free Dox exerted similar anti-tumor effect. A 2-fold enhancement of anti-tumor activity of free Dox was seen in the presence of hyperthermia compared to when no hyperthermia was used. This could be explained by the synergistic effect of hyperthermia on doxorubicin toxicity through increased perfusion and vascular fenestration (29–31). However, at the same dose, the thermally insensitive variant SynB1-ELP2-Dox had no effect on tumor inhibition either in the presence or absence of hyperthermia (Figure 4C) indicating that hyperthermia-mediated aggregation of ELP1 is necessary for the enhanced anti-tumor activity of SynB1-ELP1-Dox. This also underlies possibly difference in response of free Dox versus Dox-conjugated to SynB1-ELP to hyperthermia. Furthermore, the protein carrier, SynB1-ELP1 had no effect on tumor inhibition compared to the untreated control (Figure 4C) indicating that the carrier is non-toxic. Most importantly, under our heating protocol, hyperthermia alone had no effect on tumor growth. Tumor weights from these animals correlated with the tumor growth curve data (Supplementary Figure 1S). Body weight change in response to treatment was observed well above the 80% pre-treatment weight limit indicating that the animals are able to tolerate the therapy (Figures 4B and 4D).

Figure 4. Hyperthermia enhances the anti-tumor activity of SynB1-ELP1-Dox.

Tumor bearing mice were intravenously injected with (A, B) saline, 5 mg/kg of Dox equivalent of SynB1-ELP1-Dox or free Dox, or (C, D) SynB1-ELP2-Dox or SynB1-ELP1 on d0, d2 and d4 (↓); mean ± S.D, n = 6 – 10. (A, C) Tumor volume and (B, D) body weight in response to treatment up to day 14 are shown. * indicates significant difference between treatment and saline control at 37 °C, p<0.0001.

Maximum tolerated dose of SynB1-ELP1-Dox

Even though SynB1-ELP1-delivered Dox and free Dox had similar potencies for tumor reduction in equimolar comparison (Figure 4), we expected that the ELP-fused form of the drug would have higher a maximum tolerated dose (MTD). Based on our definition of MTD, both free Dox and SynB1-ELP1-Dox were found to be highly toxic at a dose higher than 20 mg/kg Dox equivalent (Table 1). Treatment with 10 mg/kg free Dox did not result in complete survival. However, at 8 mg/kg all the free Dox treated animals survived with weight loss < 20%. Thus, the MTD of free Dox was set at 8 mg/kg. Animals were able to tolerate both 10 mg/kg and 15 mg/kg of SynB1-ELP1-Dox treatments (Table 1). To evaluate the effect of hyperthermia on the MTD of SynB1-ELP1-Dox, tumor bearing animals were used so that focused heat could be applied to the tumor. In conjunction with hyperthermia, a dose of 15 mg/kg of SynB1-ELP1-Dox was poorly tolerated by the animals. At a slightly lower dose of 12 mg/kg however, animal tolerance of the treatment was acceptable resulting in a complete survival (Table 1). Therefore, the MTD of SynB1-ELP1-Dox was set at 12 mg/kg. However, this dose was given three times so that the cumulative dose of SynB1-ELP1-Dox was 36 mg/kg (Table 1). Although the ELP-Dox designed by Mackay et al. had an MTD of 20 mg/kg (12), this dose was only given once indicating that SynB1-ELP1-Dox is better tolerated, which may be due to the enhanced ELP tumor penetration mediated by SynB1.

Table 1.

Maximum tolerated dose

Treatment (d0, d2, d4)	Dosea (mg/kg)	Day of >20% of pre-treatment body weight lossb	Number of survivors/Sample size
SynB1-ELP1-Dox no hyperthermia	10	-	3/3
	15	-	3/3
	20	5, 5, 5	0/3
	40	3, 3, 3	0/3
SynB1-ELP1-Dox with hyperthermia	12	-	5/5
	15	3, 4, 5	0/3
Free Dox	8	-	3/3
	10	0, 2	4/6
	20	0, 2, 4	0/3
	40	2, 3, 3	0/3
	Vehicle c	-	3/3

^aDoxorubicin equivalent.

^bIndividual animal.

^c8% DMSO-phosphate buffer.

Complete tumor inhibition by SynB1-ELP1-Dox

The therapeutic efficacy of SynB1-ELP1-Dox against the E0771 murine breast tumors was evaluated and compared to free doxorubicin at the respective MTD. Fourteen days after the start of the treatment, the mean tumor volume of the animals treated with SynB1-ELP1-Dox without hyperthermia was at 1002 mm³, and 6.5-fold smaller at 154 mm³ when combined with hyperthermia. This was also significantly smaller than the saline treated control tumors by nearly 13-fold (Figure 5A). Under hyperthermic condition, SynB1-ELP1-Dox outperformed free Dox by approximately 5-fold, resulting in a complete inhibition of tumor progression (Figure 5A). In contrast, free Dox resulted in partial tumor inhibition. Overt toxicity was not evident in the treated animals (Figure 5B). Tumor weight data from the treated animals were consistent with the tumor growth curve (Figure 5C). Furthermore, the drug free carrier SynB1-ELP1 did not have any effect on tumor inhibition (data not shown) either with or without hyperthermia indicating that the ELP vector by itself is non-toxic. The cumulative effect of improved pharmacokinetics, enhanced tumor uptake and concomitant increase of MTD of SynB1-ELP1-Dox over free Dox resulted in a significant enhancement in the efficacy of SynB1-ELP1-Dox compared to the free drug.

Figure 5. At the MTD SynB1-ELP1 in combination with hyperthermia completely inhibits tumor growth.

Tumor bearing animals were treated on d0, d2, and d4 (↓) at the MTD with saline with or without hyperthermia, free Dox (8 mg/kg) with or without hyperthermia, and SynB1-ELP1-Dox (12 mg/kg) with or without hyperthermia; mean ± S.D, n = 6 – 10. (A) Tumor size and (B) body weight was measured daily up to 14 days after treatment. (B) Animals were sacrificed 14 days after treatment, and the tumors were harvested and weighed. * indicates significant difference between treatment and saline control at 37 °C, p<0.001.

In conclusion, thermal targeting of SynB1-ELP1-Dox resulted in complete tumor growth inhibition and substantially higher therapeutic benefit of the drug in an animal model that is otherwise partially responsive to standard doxorubicin treatment. To the best of our knowledge, this is the first pre-clinical demonstration of thermal targeting of a chemotherapeutic using the ELP carrier system. The advantage of combining hyperthermia with ELP is that hyperthermia increases permeability of tumor vasculature compared to normal vasculature resulting in enhanced extravasation of macromolecules (32–34)), and enhanced cellular uptake (35, 36). Furthermore, due to the built-in acid-labile linker on Doxo-EMCH, delivery by SynB1-ELP ensures that the drug is released inside the cells. In contrast, liposomal formulation of doxorubicin such as poly(ethylene glycol) encapsulated liposomal doxorubicin (PEG-Dox; Doxil®), although has superior therapeutic efficacy compared to standard doxorubicin (37), shows mostly perivascular accumulation and limited tumor uptake (38). Additionally, cellular delivery of the drug depends on the release of the drug from the liposomes, which in turn depends on the concentration gradient between the tumor vasculature and the extravascular space. Our current findings encourage us to evaluate the ELP drug delivery system in other tumor-bearing models to optimize the hyperthermia regimen in order to obtain tumor regressions. Tumor-specific SynB1-ELP-Dox could have an important impact in tumors that respond poorly to standard doxorubicin, and the potential to be combined in multimodal cancer therapy. Furthermore, since the use of hyperthermia is already established in clinical practice, precise heating of deep-seated tissues can extend thermal targeting of ELPs to internal organs as well (reviewed in (39–41)).