Polymer Masked−Unmasked Protein Therapy. 1. Bioresponsive

Mar 19, 2008 - Centre for Polymer Therapeutics, Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff, CF10 3XF, U.K., and Pol...
2 downloads 4 Views 2MB Size
1146

Biomacromolecules 2008, 9, 1146–1154

Polymer Masked-Unmasked Protein Therapy. 1. Bioresponsive Dextrin-Trypsin and -Melanocyte Stimulating Hormone Conjugates Designed for r-Amylase Activation Ruth Duncan,*,† Helena R. P. Gilbert,† Rodrigo J. Carbajo,‡ and María J. Vicent†,§ Centre for Polymer Therapeutics, Welsh School of Pharmacy, Cardiff University, King Edward VII Avenue, Cardiff, CF10 3XF, U.K., and Polymer Therapeutics Laboratory and Laboratory of Structural Biology, Centro de Investigación Príncipe Felipe, Av. Autopista del Saler 16, E-46012 Valencia, Spain Received September 26, 2007; Revised Manuscript Received December 12, 2007

Polymer-protein conjugation, particularly PEGylation, is well-established as a means of increasing circulation time, reducing antigenicity, and improving the stability of protein therapeutics. However, PEG has limitations including lack of polymer biodegradability, and conjugation can diminish or modify protein activity. The aim of this study was to explore a novel approach for polymer-protein modification called polymer-masking-unmaskingprotein therapy (PUMPT), the hypothesis being that conjugation of a biodegradable polymer to a protein would protect it and mask activity in transit, while enabling controlled reinstatement of activity at the target site by triggered degradation of the polymeric component. To test this hypothesis, dextrin (R-1,4 polyglucose, a natural polymer degraded by R-amylase) was conjugated to trypsin as a model enzyme or to melanocyte stimulating hormone (MSH) as a model receptor-binding ligand. The effect of dextrin molecular weight (7700, and 47200 g/mol) and degree of succinoylation (9–32 mol %) on its ability to mask/unmask trypsin activity was assessed using N-benzoyl-L-arginine-p-nitroanilide (L-BAPNA). Dextrin conjugation reduced enzyme activity by 34–69% depending on the molecular weight and degree of succinoylation of dextrin. However, incubation with R-amylase led to reinstatement of activity to a maximum of 92–115%. The highest molecular dextrin (26 mol % succinoylation) gave optimum trypsin masking-unmasking. This intermediate was used to synthesize a dextrin-MSH conjugate (dextrin Mw ) 47200 g/mol; MSH content 37 wt %), and its biological activity ((R-amylase) was assessed by measuring melanin production by murine melanoma (B16F10) cells. Conjugation reduced melanin production to 11%, but addition of R-amylase was able to restore activity to 33% of the control value. These were the first studies to confirm the potential of PUMPT for further application to clinically important protein therapeutics. The choice of masking polymer, activation mechanism, and the rate of unmasking can be tailored to therapeutic application.

Introduction Although the number of antibody and protein therapeutics entering routine clinical use is increasing exponentially,1,2 there are still significant challenges for protein formulation and clinical use. These include poor solubility, susceptibility to chemical and proteolytic inactivation, and suboptimal pharmacokinetics at the whole body level (due to rapid reticuloendothelial system (RES) clearance and, for small proteins and peptides, rapid renal elimination) and the cellular level (inability to reach intracellular pharmacological targets). Additionally, some proteins can be antigenic causing anaphylactic reactions. Polymer-protein conjugation can be used to overcome these problems and poly(ethylene glycol) (PEG) conjugation (PEGylation; reviewed in refs 3 and 4) has been particularly successful. PEGylation (reviewed in ref 5) has generated novel treatments for cancer (e.g., Oncaspar, Neulasta), viral diseases (e.g., PEGASYS, PEGINTRON), acromegaly (Pegvisomant), and age-related macular degeneration (MACUGEN). Despite the ability of contemporary PEGylation chemistry to give well-defined con* Corresponding author: tel, 029 20876419; fax, 029 20874536; e-mail, [email protected]. † Centre for Polymer Therapeutics, Welsh School of Pharmacy. ‡ Laboratory of Structural Biology, Centro de Investigación Príncipe Felipe. § Polymer Therapeutics Laboratory, Centro de Investigación Príncipe Felipe.

jugates (reviewed in ref 6) and to introduce biodegradable PEG-protein linkers,7 the use of PEG does have disadvantages, not least its lack of biodegradability which brings the risk of polymer accumulation following chronic administration. The aim of this study was to develop a novel approach for polymer–protein modification, called polymer masking– unmasking protein therapy (PUMPT). It was hypothesized that coupling a biodegradable polymer to a protein would create a conjugate that would be inactivate in transit due to protein masking but, by triggered degradation of the polymer, would reinstate protein activity at a rate tailored to suit its mode of action. Additionally, careful choice of polymer and the activating mechanism may allow preferential localization of pharmacological activity to the desired target for action (summarized in Figure 1). To test the feasibility of PUMPT the polysaccharide dextrin, R-1,4-poly(glucose) was selected as the first model polymer (preliminary data were reported in refs 8 and 9). Dextrin is approved for clinical use as the peritoneal dialysis solution (Icodextrin),10 as a solution to prevent postoperative adhesions (Adept),11 and as a formulation solution for peritoneal administration of 5-fluorouracil.12 Thus its clinical safety is well documented.13 Degradation to maltose and isomaltose is mediated by R-amylase, and we have previously shown that dextrin modification by succinoylation14 can be used to introduce the reactive groups necessary for covalent drug conjugation (e.g., to doxorubicin15 or amphotericin B16). Moreover, the degree

10.1021/bm701073n CCC: $40.75  2008 American Chemical Society Published on Web 03/19/2008

Dextrin-Protein Conjugates

Biomacromolecules, Vol. 9, No. 4, 2008

1147

of succinoylation can be used to tailor amylase degradation rate.17 Trypsin was used as a model enzyme, and melanocyte stimulating hormone (MSH) as a model receptor binding ligand. The effect of dextrin molecular weight (7700, and 47200 g/mol) and degree of succinoylation (9–32 mol %) on its ability to mask/unmask trypsin activity was first assessed. Conjugates were characterized by NMR, FTIR, gel permeation chromatography (GPC), SDS PAGE, and circular dichroism (CD) spectroscopy, and the biological activity of trypsin conjugates was assessed (R-amylase using N-benzoyl-L-arginine-p-nitroanilide (L-BAPNA) as a substrate. The dextrin-MSH conjugate was synthesized using dextrin of Mw ) 47200 g/mol (26 mol % succinoylation). In this case biological activity ((R-amylase) was assessed by measuring stimulation of melanin production in B16F10 murine melanoma cells.

Materials and Methods Materials. A 40% solution of acrylamide/bis(acrylamide) (ratio 37.5:1),ammonium persulfate, salivary R-amylase, L-BAPNA, biocinchoninic acid (BCA) solution, blue dextran, bromothymol blue, copper(II) sulfate, dextrin (type 1 from corn, Mw 7700 g/mol), anhydrous N,Ndimethylformamide (DMF), anhydrous succinic anhydride, tris(hydroxymethyl)aminomethane HCl (trizma HCl), trypan blue solution (0.4% w/v), MSH, and porcine pancreatic trypsin (type 1X-S) were from Sigma-Aldrich (U.K.). Pullulan gel filtration standards (Mw ) 5800–186000 g/mol) were from Polymer Laboratories (U.K.). Bromophenol blue, coomassie brilliant blue G-250, 2-mercaptoethanol, sodium dodecyl sulfate (SDS), prestained SDS-PAGE standards, and N,N,N,N′-tetramethylethylenediamine (TEMED) were from Bio-Rad (USA). 4-Dimethylaminopyridine (DMAP) was from Fisher Scientific (U.K.). Glycine (electrophoresis grade) was from ICN Biomedicals (USA), and N-hydroxysulfosuccinimide (sulfo-NHS) and N-ethyl-N′(3-dimethylaminopropyl)carbodiimide (EDC) were from Pierce Chemical Co. (USA). All solvents were of general reagent grade (unless stated) and were from Fisher Scientific (U.K.). Samples were dialysed using Spectra/Por Membranes (molecular weight cutoff 25000 or 2000 g/mol from BDH Merck (U.K.)). The B16F10 murine melanoma cell line was from the American type culture collection (ATCC) (USA). RPMI 1640 with phenol red (PR), glutamax, fetal bovine serum (FBS), and trypsin–ethylenediaminetetraacetic acid (EDTA) (0.05% w/w trypsin, 0.53 mM EDTA) were all from Invitrogen Life Technologies (U.K.). CO2 and N2 (medical grade, 95% v/v) and liquid N2 were supplied by BOC (U.K.). Synthesis and Characterization of Succinoylated Dextrin. Succinoylated dextrins (9–32 mol % carboxyl groups) were prepared as previously described by Hreczuk-Hirst et al.14 (Scheme 1a). Briefly,

Figure 1. Concept of PUMPT showing masking and then theoretical regeneration of protein activity in the presence of activating enzyme. Scheme 1. Synthesis of Dextrin-Trypsin Conjugates

Table 1. Characteristics of Succinoylated Dextrins and Conjugates dextrins

dextrin-trypsin conjugates

compound

Mwa (g/mol) (Mw/Mn)

succinoylation (mol %)

Mwa (g/mol) (Mw/Mn)

trypsin content (wt %)

trypsin dextrin succ-dextrin (7700 g/mol)

23400 7700 (1.4) 14600 (1.3)

15

72

succ-dextrin (7700)

13500 (1.4)

23

succ-dextrin (7700)

15500 (1.5)

32

25700 (1.7) 55000b 14000 (1.5) 39300b 19800 (1.5) 47700b

dextrin succ-dextrin (47200)

47200 (1.5) 61300 (1.4)

9

succ-dextrin (47200)

57800 (1.4)

18

succ-dextrin (47200)

39300 (1.4)

26

a

Mw estimated by GPC using pullulan standards.

b

84400 (1.6) 110600b 101800 (1.6) 122900b 49800 (1.3) 48200b

Mw estimated by GPC using protein standards.

67 81 51 78 76

1148

Biomacromolecules, Vol. 9, No. 4, 2008

Figure 2. Characterization of succinolylated dextrin and dextrin-trypsin conjugates. (a) FTIR spectra showing the increase in peak intensity at 1720 cm-1 with increasing incorporation of carboxyl groups. (b) FPLC traces of a dextrin-trypsin conjugate prepared with dextrin having different degrees of succinoylation. (c) SDS page electrophoresis of dextrin-trypsin conjugates showing absence of free trypsin.

with dextrin of Mw 47200 g/mol, 9 mol % succinoylation was achieved as follows. Dextrin (1 g, 2.12 × 10-5 mol) and DMAP (30 mg, 2.41 × 10-4 mol) were placed in a sealed flask and dissolved in anhydrous DMF (10 mL) before purging with nitrogen. Succinic anhydride (70 mg, 7 × 10-4 mol) dissolved in anhydrous DMF (2.5 mL) was then added. The reaction was left for 16-18 h (overnight) at 50 °C under nitrogen, and then DMF was partially evaporated under reduced pressure. The product was purified by precipitation into rapidly stirring diethyl ether (∼250 mL) and allowed to stir for 4 h before the solution was filtered under vacuum. The solid obtained was redissolved in double distilled water (10 mL) and further purified by dialysis (molecular

Duncan et al. weight cutoff 2000 g/mol) against double distilled water (6 × 5 L) over 48 h. The solution was finally freeze-dried to yield succinoylated dextrin product as a white solid (∼60% yield) that was characterized by FTIR (Avatar 360 ESP spectrometer with EZ OMNIC ESP 5.2 software; Thermo Nicolet, Loughborough, U.K.) and 1H NMR to confirm identity, titration to determine the carboxyl group content, and by GPC to measure the approximate molecular weight and polydispersity. Titrimetric analysis was conducted using NaOH (5 × 10-4 M) as base and bromothymol blue (1% in ethanol) as an indicator. Samples for GPC were prepared in PBS (3 mg/mL), and maltose (Mw) 342 g/mol) was used to mark the Vb and blue dextran (Mw ) 2000000 g/mol) to mark the Vo. Pullulan standards were used as a reference. Synthesis and Characterization of Dextrin–Trypsin Conjugates. Succinoylated dextrins (Mw ∼ 7700 and 47200 g/mol, 9–32 mol % succinoylation) were then conjugated to trypsin,8,9,14 (Scheme 1b). Briefly, EDC (77 mg, 4.00 × 10-4 mol) was added as solid to a water solution (8 mL) of succinoylated dextrin (100 mg, 1.657 × 10-1 mol of COOH groups) and the reaction mixture stirred for 10 min at room temperature. Sulfo-NHS (88 mg, 4.04 × 10-4 mol) was added also as a solid, and the reaction continued with stirring for a further 40 min at room temperature. Trypsin (198 mg, 8.29 × 10-6 mol) dissolved in double distilled water (5 mL) was added dropwise and the pH of the reaction adjusted to ∼8 with 1 M NaOH. The reaction was allowed to continue for 16 h at room temperature. The precipitated urea was filtered off, and the conjugate solution was purified by dialysis (molecular weight cutoff 25000 g/mol) against double distilled water (5 L) over 48 h. After freeze-drying dextrin-trypsin conjugate was obtained as a white solid (yield ∼50%). Dextrin-trypsin conjugates were characterized by GPC (TSK G4000PWXL and G3000 PWXL columns in series, mobile phase PBS (0.1 M, pH 7.4), flow rate of 1 mL/min) to establish approximate molecular weight and purity (against pullulan and protein standards), the BCA assay to determine protein content, and by SDS-PAGE to establish purity. The BCA assay18 used trypsin standards. SDS-PAGE electrophoresis was used to assess levels of free protein. SDS gels (12.5%) were prepared to 0.75 mm thickness. Conjugate samples (1 mg/mL in dd H2O) and reference samples (trypsin and a mixture of dextrin and trypsin) were prepared by diluting (1:1) with denaturing solution (3.8 mL of H2O, 5 mL; 0.5 M Tris HCl (pH 6.8), 8 mL 10% w/v SDS, 4 mL of glycerol, 2 mL of 2-mercaptoethanol, 0.4 mL of bromophenol blue 1% w/v) and heated for 5 min at 100 °C. These samples and protein molecular weight markers were loaded (10 µL), and the gel was run for 1 h at 200 V. It was then rinsed and stained (1 h) with coomassie blue stain before clearing by soaking in destaining buffer (35% methanol/5% acetic acid) for 1 h and then rehydrating in 5% methanol/7% acetic acid. Finally, the gel was rinsed with distilled water and dried for 2 days. Characterization of Dextrin-Trypsin Conjugates by NMR and Circular Dichroism Spectroscopy. NMR Spectroscopy. NMR spectra were recorded at 27 °C on a Bruker Avance Ultrashield Plus 600 spectrometer equipped with 5 mm single-axis gradient TCI cryoprobe. Data were processed using the program Topspin 1.3 (Bruker GmbH, Karlsruhe, Germany). The samples were prepared to a concentration of 0.22 mM in 90/10% H2O/D2O. One-Dimensional 1H Experiments. Proton NMR spectra were acquired with 16K complex points and a spectral width of 8400 Hz. The total number of scans was 32, with a repetition delay of 1.2 s. A WATERGATE scheme was used to suppress the water signal. Diffusion Experiments. Pulsed field gradient NMR spectroscopy was used to measure translational diffusion by fitting the integrals or intensities of the NMR signals to the following equation

I ) I0 exp[-Dγ2g2δ2(∆ - δ ⁄ 3)]

(1)

where I is the observed intensity, I0 the reference intensity (unattenuated signal intensity), D the diffusion coefficient, γ the gyromagnetic ratio of the observed nucleus, g the gradient strength, δ the length of the gradient, and ∆ the diffusion time. Two-dimensional diffusion-ordered

Dextrin-Protein Conjugates

Biomacromolecules, Vol. 9, No. 4, 2008

1149

Table 2. Activity of Dextrin-Trypsin Conjugates ((R-amylase)

dextrin Mwa(g/mol) dextrin-trypsin conjugates 7700 7700 7700 47200 47200 47200 trypsin a

degree of succinoylation (mol %)

protein content (wt %)

remaining activity (%)

amylase restored activity (%)

15 23 32 9 18 26

72 67 81 51 78 76

55 69 51 63 46 34

67 52 92 115 64 58

Kcat (s-1) masked

unmasked

4.09 ( 2.19 2.54 ( 1.78 2.59 ( 0.66 2.69 ( 0.25 2.23 ( 0.78 1.25 ( 0.38 2.52–2.93

1.27 ( 0.05 1.26 ( 0.57 2.30 ( 0.54 2.49 ( 0.31 2.40 ( 0.71 2.49 ( 0.33

Mw was estimated by GPC using pullulan standards.

NMR spectroscopy (DOSY) was performed with a stimulated echo sequence using bipolar gradient pulses19 and with a WATERGATE scheme to suppress the water signal. The lengths of pulses and delays were held constant, and 20 spectra of 32 scans each were acquired with the strength of the diffusion gradient varying between 5% and 100%. The lengths of the diffusion gradient and the stimulated echo were optimized for each sample. Typical values were δ ) 5–7 ms, ∆ ) 50–150 ms. Processing and analysis of the data were performed with the DOSY protocol included in the Topspin 1.3 software package. Circular Dichroism (CD) Measurements. CD spectra of dextrin (Mw 47200 g/mol), succinoylated dextrin (Mw 47200 g/mol; 26 mol %), trypsin, succinoylated dextrin (Mw 47200 g/mol; 26 mol %) plus trypsin, and the dextrin-trypsin conjugate were acquired on a Jasco-810 spectropolarimeter equipped with a Peltier temperature control in quartz cells of 0.1 cm path length at two different temperatures (4 and 37 °C) and in two different buffers (Tris at pH 8.2 (40 mM Tris, 16 mM CaCl2) and H2O at pH 5.5) trying to mimic biological conditions. Compounds were analyzed at 2 mM trypsin-equiv in all cases; when control polymers were studied, the same amount of polymeric carrier as in dextrin-trypsin conjugate was used. CD spectra were the average of 40 scans made at 0.2 nm intervals, and always the same buffer without compounds, used as baseline, was subtracted. Results are expressed as mean molar residue ellipticities [Q]MR (deg cm2/dmol). Degradation of Dextrin and Succinoylated Dextrins by Amylase.17 Dextrin (Mw ) 7700 and 47200 g/mol) or succinoylated dextrins (47200 g/mol and ∼15 and ∼30 mol % succinoylation) were dissolved in PBS (4 mL, pH 7.4) at a concentration of 3.75 mg/mL. Amylase (400 µL; 2.5 units/mL) was added and the mixture incubated at 37 °C for 5 h. At various times, samples (100 µL) were taken, immediately snap frozen in liquid nitrogen to stop the reaction, and stored at –20 °C until analysis. Before GPC analysis, samples were placed for 5 min in a boiling water bath to denature enzyme activity and precipitate the protein. The supernatant was then analyzed by GPC to determine the change in molecular weight of the sample over time. Analysis of Trypsin Activity.20 L-BAPNA (7 mg/mL stock in DMSO) and the trypsin and dextrin-trypsin conjugate samples (0.1 mg/mL stock in Tris buffer; pH 8.2, 40 mM Tris, 16 mM CaCl2) were prepared. L-BAPNA to give a final concentration in the range 1.61 × 10-4 to 9.66 × 10-4 M was added to a 1 mL cuvette followed by trypsin (200 ng or 2 µg) or dextrin-trypsin conjugates (200 ng or 2 µg of trypsin-equiv) to give a total volume of 1 mL. The release of p-nitroanilide (NAp) (410 nm ) 8800 L/mol per cm) was then measured at 37 °C for 20 min at 400 nm and NAp released expressed as a percentage of that seen for native trypsin. The data obtained were also transformed using a Hanes-Woolfe plot and Kcat calculated.21 To study the effect of R-amylase on the activity of the dextrin-trypsin conjugates, they (300 µL of 1 mg/mL stock) were first incubated with R-amylase (120 µL of 1 mg/mL stock) for 16 h at 37 °C in Tris buffer (2580 µL, pH 8.2). Then samples (2 µg trypsin-equiv) were added to a 1 mL cuvette containing L-BAPNA (concentrations from 1.61 × 10-4 to 9.66 × 10-4 M) and NAp release measured at 400 nm as described above. Synthesis of Dextrin-MSH Conjugates. Dextrin (47200 g/mol; 26 mol % succinoylation) was conjugated to MSH as described above

with minor modification. Briefly, 30 mg (5.15 × 10-7 mol) succinoylated dextrin was reacted with 5 mg (2.6 × 10-6 mol) of MSH for 16 h at room temperature. The resulting dextrin-MSH conjugates were purified by FPLC fractionation (Superdex HR 10/30; mobile phase PBS 0.1 M (pH 7.4) and a flow rate of 0.5 mL/min), desalted using vivaspin centrifugal filters and lyophilized. Total MSH content was determined using the BCA assay18 (with an MSH standard curve), and purity was assessed using SDS-PAGE electrophoresis and analytical FPLC. In addition, degradation of dextrin-MSH conjugate, dextrin, and succinoylated dextrin (all 1 mg/mL) was monitored during incubation in tissue culture media (RPMI + 10% FBS or RPMI + 10% heat inactivated FBS diluted 1:20 in PBS; pH 7.4). Aliquots (200 µL) of incubation medium were withdrawn at t ) 0 and 16 h and snap frozen in liquid nitrogen before GPC analysis (TSK-gel columns G4000 PWXL and G3000 PWXL in series). The change in molecular weight was expressed as a percentage of the starting molecular weight of the sample. Melanin Production in B16F10 Cells Incubated with MSH and Dextrin-MSH. B16F10 cells were routinely maintained in RPMI 1640 (+ Glutamax) media + 10% FBS and split twice weekly at a ratio of 1:9. Because of the complexity of this assay, it was first necessary to conduct a number of preliminary experiments to optimize the assay conditions. As was shown by O’Hare et al.,22 it was found here that MSH-induced melanin production decreased with increasing time after B16F10 subculture (results not shown), and the optimum time for MSH addition was 24 h after subculture. Melanin production decreased as incubation time increased, reaching a plateau after 72 h. This might have been due to receptor down-regulation and/or medium depletion of the substrate tyrosine.23 An optimized assay used to determine the ability of dextrin-MSH conjugates ((R-amylase; 0.2 mg/mL; 10 ui/mL) to induce melanin production was as follows. Twenty four hours after cell seeding (2.5 × 104 cells/mL) into a 96 well plate, the incubation media was removed and MSH or dextrin-MSH (10-4M MSH-equiv) dissolved in RPMI + 10% heat-inactivated FBS and sterile filtered (0.2 µm) was added (100 µL). The cells were then incubated for 72 h before absorbance at 405 nm was measured using a Tecan plate reader to quantify melanin production. Values were expressed as a percentage of the free MSH control.

Results Dextrin Conjugates, Synthesis and Characterization. The characteristics of the library of succinoylated dextrins synthesized are summarized in Table 1. 1H NMR, 13C NMR and FTIR confirmed succinoylation, and the FTIR peak seen at 1720 cm-1 increased with increasing modification (Figure 2a) (also see ref 14). Titration gave dextrin succinoylation values of between 9 and 32 mol % (Table 1), and typically the reaction conversion efficiency was 70–100%. GPC suggested an increase in dextrin molecular weight following succinoylation (Mw of ∼12000–16000 g/mol and ∼60000 g/mol) without any change in polydispersity (1.3–1.5) (Table 1), but this likely reflects only a difference in polymer coil conformation.

1150

Biomacromolecules, Vol. 9, No. 4, 2008

Figure 3. 1H NMR and DOSY spectra of trypsin, dextrin-trypsin, a mixture of succinoylated dextrin and trypsin, and succinoylated dextrin. (Prepared using succinoylated dextrin (47200 g/mol; 26 mol %)). Expansion of the NH/aromatic region from the 1D 1H spectra are shown for (a) trypsin, (b) succinoylated dextrin, (c) a mixture of succinoylated dextrin and trypsin, and (d) dextrin-trypsin conjugate. (e) Diffusion DOSY spectrum for the dextrin-trypsin conjugate sample. Typical experimental parameters: diffusion time 50 ms, gradient time 5 ms, gradient strength 5–100%. The horizontal axis projection shows the 1H spectrum of the corresponding sample.

Using the succinoylated dextrin intermediates, dextrin-trypsin conjugates were prepared and their characteristics are summarized in Tables 1 and Table 2. The trypsin content was 50–80 wt % (neither dextrin nor succinoylated dextrin interfered with the BCA protein assay, results not shown). Both GPC (not shown) and FPLC analysis (Figure 2b) indicated no free trypsin in conjugates produced from the higher molecular weight intermediate with up to 26 mol % succinoylation, and this was confirmed for both families of conjugate by SDS-PAGE electrophoresis (Figure 2c). However, there was some free trypsin in the lower molecular weight dextrin conjugates. GPC estimation of molecular weight characteristics was made using both pullulan and protein standards knowing that neither would give an absolute molecular weight but only some indication of

Duncan et al.

hydrodynamic volume. Apparent molecular weights were ∼14000–102000 g/mol depending on the starting molecular weight of the dextrin used and polydispersity by GPC was 1.3–1.6 (Table 1). SDS page electrophoresis, however, showed that a broad range of conjugates were present (Figure 2c). Dextrin-trypsin conjugate formation was confirmed by comparing the 1H NMR spectra acquired for trypsin, succinoylated dextrin, the dextrin-trypsin conjugate and an equimolecular mixture of succinoylated dextrin and trypsin (Figure 3). A close look at the aromatic/amide proton region of the 1H NMR spectra (6–11 ppm) shows a significant difference between the three samples, with the profile of the NH region of trypsin (Figure 3a) being dramatically different from that of the putative conjugate (Figure 3d) indicating that the protein is surrounded by a completely different chemical environment. Noncovalent polymer encapsulation/interaction with trypsin can be discounted by comparing the spectra for trypsin, conjugate, and the simple mixture of succinoylated-dextrin plus trypsin. In the latter case (Figure 3c), the 1H signals from the amide/aromatic region show small changes when compared with the trypsin spectrum, the chemical shift values being maintained although affected by a general broadening, probably due to unspecific interactions of the protein with the polymer. DOSY identifies species depending on their molecular size/ hydrodynamic radius.24 The spectrum of the succinoylated dextrin-trypsin conjugate gives valuable qualitative information about the number of species present in solution and shows at least three diffusion coefficients (Figure 3e, dashed lines). From the 1H projection on the F2 axis it can be concluded that two of the species contain trypsin as they show cross peaks in the NH/aromatic region for proteins (above 6 ppm). As the SDSPAGE analysis of the same sample shows no free trypsin present, these two diffusion coefficients must correspond to two polymer-protein conjugates. The remaining species only correlates with 1H signals located in the region of the polymer moiety (2.5–5.5 ppm) and therefore could be ascribed to the remaining free polymer. CD spectroscopy (Figure 4) also showed clear differences between the dextrin-trypsin conjugate and the simple mixture of the two components. In good agreement with the NMR data, the presence of free succinoylated dextrin in the sample seemed to result in nonspecific interactions with the protein chain leading to a greater structural mobility and loss of ∼10% of the trypsin R-helix conformational parameter. This was not seen in the conjugate sample, as it displayed a spectrum very similar to that of trypsinsanother clear proof of protein conjugation to the polymeric carrier. As expected, dextrin and succinoylated dextrin showed a complete unstructured conformation. GPC evaluation of dextrin and succinoylated dextrin degradation by R-amylase revealed a decrease in molecular weight with time (Figure 5). The extent of degradation of native dextrins of starting molecular weight of 47200 and 7700 g/mol at 5 h was 95% and 30%, respectively. The succinoylated dextrins were degraded more slowly than parent polymer, the higher the succinoylation, the slower the degradation rate. Using these incubation conditions a plateau in molecular weight was reached in all cases. Dextrin-Trypsin Activity ((r-Amylase). When compared at the same protein concentration, the activity of the dextrin– trypsin conjugates was reduced compared to free trypsin in all cases (Figure 6; Table 2). For those conjugates prepared from 47200 g/mol dextrin, enzyme activity fell significantly from 63 to 34% in relation to an increase (from 9 to 26 mol %) in the degree of intermediate succinoylation. In contrast, for those

Dextrin-Protein Conjugates

Biomacromolecules, Vol. 9, No. 4, 2008

1151

Figure 4. Circular dichroism spectra of dextrin, succinoylated dextrin, free and conjugated trypsin, and succinoylated dextrin plus trypsin at 4 °C in Tris buffer (pH 8.2). Samples were prepared using succinoylated dextrin (47200 g/mol; 26 mol %).

Figure 6. Trypsin activity before and after dextrin conjugation: Effect of the degree of succinoylation.

Figure 5. Degradation of dextrin and succinoylated dextrins by R-amylase.

conjugates synthesized using the lower molecular weight dextrin, the degree of succinoylation did not correlate with retained enzyme activity. This was always in the range 51–69%. Calculation of the rate constants Km, Vmax, and Kcat showed that conjugation of trypsin to dextrin typically did not alter Km, but the Vmax was reduced ∼10-fold. The Kcat values calculated typically were 2–3 s-1(Table 2). The enzyme activity of conjugates exposed to R-amylase was usually increased (Figure 7, Table 2). For example, for the lower molecular weight dextrin modified to 32 mol %, trypsin activity was restored to 92% of the control following R-amylase exposure. In the case of the higher molecular weight dextrin having 26 mol % succinoylation, R-amylase exposure increased trypsin activity from 34% to 58% of the free trypsin control. In all cases, Kcat was typically in the range of 1.2–2.5 s-1. Synthesis and Characterization of Dextrin-MSH. Several batches of dextrin-MSH conjugate were synthesized, and after

FPLC fractionation (Figure 8), GPC analysis confirmed conjugate formation and absence of free MSH (also confirmed by SDS-PAGE, data not shown). The total MSH content was ∼14 wt % in pilot scale batches, but after scale-up of conjugate production for the biological assays the conjugate contained 37 wt % MSH loading. Stimulation of Melanin Production by Dextrin-MSH ((r-Amylase). The melanin production assay was also optimized according to the B16F10 cell seeding density, the MSH concentration used (Figure 9a), and by characterizing the MSH stability and conjugate stability in tissue culture media (Figure 9b). As FBS contains amylase, and indeed the dextrin-MSH conjugates were found to degrade in the presence of 10% FBS (Figure 9b), all melanin induction experiments were conducted in heat inactivated 10% FBS. These conditions ((amylase) did not adversely affect B16F10 viability (results not shown). Melanin production was related to MSH concentration (Figure 9a). It was shown that the dextrin-MSH conjugate (37 wt % MSH) stimulated melanin production to a value of 11% the free MSH control when analyzed using the optimized melanin assay in RPMI + 10% heat-inactivated FBS (Figure 9c). Following incubation of the dextrin-MSH conjugate with R-amylase (0.2 mg/mL), melanin production was increased to 33%. Significant masking and unmasking of activity was seen (p < 0.05, ANOVA and Bonferroni post hoc test).

Discussion The two models explored here demonstrated the feasibility of PUMPT using dextrin for protein/peptide masking, with unmasking possible by addition of R-amylase. As R-amylase

1152

Biomacromolecules, Vol. 9, No. 4, 2008

Duncan et al.

Figure 8. FPLC purification of dextrin-MSH.

Figure 7. Reinstatement of trypsin activity following incubation of dextrin-trypsin conjugates in the presence of R-amylase.

is present in serum and extracellular fluids, the PUMPT concept has physiological relevance. Succinoylation was chosen for dextrin functionalization as this chemistry should yield a nontoxic intermediate, the resultant linker is relatively stable,16,25 and dextrin succinoylation had been previously optimized.16 Not surprisingly, it was easier to regenerate enzyme activity by addition of R-amylase than MSH-receptor binding owing to the easier accessibility of the low molecular weight substrate to the enzyme’s active site. As shown previously for both chemically modified dextrans26 and for dextrin,17 the degree of modification governed the rate of degradation. The extent of regeneration of protein/peptide activity in these in vitro models is complex to interpret. It will be related to (i) the rate of polymer degradation, (ii) liberation of active protein with minimal or no glucose oligomers attached, and (iii) maintenance of protein or peptide integrity during the conjugation and regeneration process. A particular challenge for design of the in vitro experiments designed to evaluate PUMPT is the difficulty of mimicking the optimal physiological conditions for polymer degradation while

simultaneously presenting the most favorable environment for activity of liberated enzyme or ligand–receptor binding. Dextrin-Trypsin Model. The experiments designed to evaluate R-amylase activation of the dextrin-trypsin conjugates used a 16 h R-amylase/conjugate incubation period prior to measurement of trypsin activity and assay conditions described previously for the comparison of ovine and porcine trypsin using 27 L-BAPNA. Although some immediate regeneration of trypsin activity has been seen on addition of R-amylase to specific dextrin-trypsin conjugates (results not shown), an exposure time of 16 h was used here to allow better comparison of the library of conjugates, some of which had such a high degree of succinoylation that the polymer degradation was expected to be relatively slow. Not surprisingly, the highest degree of trypsin masking was achieved using the higher molecular weight dextrin due to the greater steric hindrance caused by the larger bulkier polymer chain. Increasing dextrin succinoylation led to greater masking. Although masking was achieved using the lower molecular weight dextrin, it was less efficient and there was no obvious relationship in masking between degree of succinoylation and conjugate activity. This was not surprising as the average weight percent of trypsin in conjugates (thus number of trypsin molecules per dextrin chain) did not vary greatly between samples. Interestingly the Kcat values estimated for all conjugates were comparable to that seen for free trypsin both here and for porcine trypsin determined under almost identical conditions using L-BAPNA (Kcat ) 2.89 s-1).27 Importantly, as far as we are aware, this is the first time that DOSY NMR spectroscopy was used as an analytical technique to define polymer-protein conjugate composition. Using NMR it was possible to differentiate between covalent binding and simple polymer-protein complexation. Measuring diffusion using DOSY NMR also opens up the possibility of analyzing complex polymer-protein mixtures and reaction kinetics. Dextrin-MSH Conjugates. Dextrin-MSH conjugates prepared using FPLC fractionation did not contain free MSH. The establishment of controlled conditions to enable assay of activity in vitro was, however, complicated. Both succinoylated dextrin and the dextrin-MSH conjugates were degraded over time when incubated in RPMI + 10% FBS, and although use of heatinactivated FBS reduced degradation, it did still occur (Figure 9b). This was not surprising as MSH degradation by serum enzymes is well-known (reviewed in ref 28). As the in vitro assay measured MSH induction of melanin production by B16F10 cells after 72 h, it was important to establish that addition of R-amylase would not itself affect cell viability or melanin production. When incubated in RPMI containing heat-

Dextrin-Protein Conjugates

Biomacromolecules, Vol. 9, No. 4, 2008

1153

activity compared to the dextrin-trypsin conjugates might be due to MSH degradation during the 72 h incubation and/or irreversible inactivation of the MSH by dextrin conjugation. One potential site for conjugation is the MSH lysine separated from the active site by only one amino acid. Any remaining glucose oligomers could potentially cause significant steric hindrance of receptor binding. It should be noted that the extent of regeneration of activity of both conjugate models is only an indication of the potential extent and rate of unmasking in an in vivo setting. In these experiments only one exposure time was used in each model (16 or 72 h) to enable comparisons across a library of compounds to be made. PUMPT, Opportunities and Challenges. PUMPT provides a new strategy with potential to improve delivery of proteins having such high inherent toxicity that safe parenteral administration is precluded, to enable the protection of proteins normally inactivated while traveling to their therapeutic target and/or to allow controlled release with time of active proteins at the target site. PUMPT should be widely applicable to protein and peptide therapeutics having various kinds of pharmacological targets, e.g., enzymes, cytokines, growth factors, and antibodies. Our recent studies have successfully applied the dextrin-R-amylase strategy for the delivery of epidermal growth factor for use in the context of wound repair29 and for phospholipase A2 as an anticancer agent.30 Moreover, experiments with hyaluronic acid and hyaluronidase31 have verified the suitability of other polymers and activation strategies for PUMPT. For further PUMPT development it will be essential to optimize the conjugation chemistry on a case-by-case basis depending on the protein therapeutic to be modified, to minimize polymer polydispersity thus minimizing conjugate heterogeneity, and improve the purification techniques to ensure removal of free polymer. The trypsin and MSH conjugates described here were pure in respect of free protein, but DOSY NMR suggested presence of free dextrin in the dextrin-trypsin conjugates. In addition, the identity and biocompatibility of the primary metabolites of PUMPT must be characterized to ensure safe elimination. Acknowledgment. H.G. thanks BBSRC for supporting her PhD studentship and M.J.V. is grateful for a Marie Curie Fellowship, Contract No. HPMF-CT-2002-01555, for support. R.J.C. and M.J.V. are currently researchers from the Ramón y Cajal program (MEC, Spain), and they would also like to thank Dr. A. Pineda-Lucena for helpful discussions. Figure 9. Experiments designed to investigate the biological activity of MSH (control) and dextrin-MSH conjugates in B16F10 cells. (a) Control studies designed to define the effect of MSH concentration on melanin production (72 h) in B16F10 cells (seeded to a density of 2.5 × 104 cells/mL), and data represent the mean (n ) 6) ( standard deviation. (b) Effect of media composition on the degradation (by GPC) of dextrin (47200 g/mol), succinoylated dextrin (26 mol %), and dextrin-MSH when incubated for 16 h at 37 °C. (c) Melanin production (72 h) in B16F10 cells incubated with MSH or dextrin-MSH ((amylase 0.2 mg/mL) (10-5 M MSH equiv). Cells were incubated in RPMI + 10% heat-inactivated FBS at a seeding density of 2.5 × 104 cells/mL. Data shown represents the mean (n ) 12–18) ( SEM. Statistical significance (/) indicates p < 0.05 measured using ANOVA and Bonferroni post hoc test and comparing each treatment and the control.

inactivated FBS (10%), the activity of the masked conjugate was reduced to 11%. Addition of R-amylase elevated melanin to 33% the control value. The relatively poor reinstatement of

References and Notes (1) Baker, M. Nat. Biotechnol. 2005, 23, 1065–1072. (2) Nagle, T.; Berg, C.; Nassr, R.; Pang, K. Nat. ReV. Drug DiscoVery 2003, 2, 75–79. (3) Harris, J. M.; Chess, R. B. Nat. ReV. Drug DiscoVery 2003, 2, 214– 221. (4) Pasut, G.; Veronese, F. M. Prog. Polym. Sci.,in press. (5) Veronese, F. M.; Harris, J. M. Peptide and Protein PEGylation III: Advances in Chemistry and Clinical Applications. AdV. Drug DeliVery ReV. 2008, 60 (1), 1–88. (6) Roberts, M. J.; Bentley, M. D.; Harris, J. M. AdV. Drug DeliVery ReV. 2002, 54, 459–476. (7) Filpula, D.; Zhao, H. AdV. Drug DeliVery ReV. 2008, 60 (1), 29–49. (8) Gilbert, H. R. P.; Vicent, M. J.; Duncan, R. J. Pharm. Pharmacol. 2004, S-17, 46. (9) Gilbert, H. R. P.; Vicent, M. J.; Duncan, R. Proc. Annu. Meet. Controlled Release Soc., 32nd 2005, 32, 455–456. (10) Cada, D. J.; Levien, T.; Baker, D. E. Hosp. Pharm. 2003, 38, 669– 677.

1154

Biomacromolecules, Vol. 9, No. 4, 2008

(11) Verco, S. J. S.; Rodgers, K. E.; Roda, N.; Peers, E. M.; Brown, C. B.; diZerega, G. S. Hum. Reprod. 1999, 14, 269–269. (12) Kerr, D. J.; Young, A. M.; Neoptolemas, J. B.; Sherma, M.; VanGeene, P.; Stanley, A.; Ferry, D.; Dobbei, J. M.; Vinche, B.; Gilker, J.; Eime, E.; Dombros, N.; Fontzilas, G. Br. J. Cancer 1996, 74, 2032– 3235. (13) Guo, A.; Wolfson, M.; Holt, R. Kidney Int. 2002, 62 (Suppl. 81), S72– S79. (14) Hreczuk-Hirst, D.; German, L.; Duncan, R. J. Bioact. Compat. Polym. 2001, 16, 353–365. (15) Hreczuk-Hirst, D.; German, L. A.; Duncan, R. Proc. Int. Symp. Controlled Release Bioact. Mater., 25th 1999, 25, 1086–1087. (16) German, L. A.; Tupper, J.; Hreczuk-Hirst, D.; Dagini, B.; Humber, D. P.; Shaunak, S.; Duncan, R. J. Pharm. Pharmacol. 2000, 52, 37. (17) Hreczuk-Hirst, D.; Chicco, D.; German, L.; Duncan, R. Int. J. Pharm. 2001, 230, 57–66. (18) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal. Biochem. 1985, 150, 76–85. (19) Brand, T.; Cabrita, E. J.; Berger, S. Prog. NMR Spectrosc. 2005, 46, 159–196. (20) John, R. A. In Enzyme Assays. A Practical Approach; Eisenthal, R., Danson, M. J., Ed.; Oxford University Press: Oxford, U.K., 1998; pp59–92.

Duncan et al. (21) Cornish-Bowden, A. Fundamentals of Enzyme Kinetics; Portland Press: London, U.K., 2004. (22) O’Hare, K. B.; Duncan, R.; Strohalm, J.; Ulbrich, K.; Kopeckova, P. J. Drug Targeting 1993, 1, 217–229. (23) Eberle, A. N. The Melanotropins. Chemistry, Physiology and Mechanisms of Action; Karger: London, U.K., 1988. (24) Barjat, H.; Morris, G. A.; Smart, S.; Swanson, A. G.; Williams, S. C. R. J. Magn. Reson., Ser. B 1995, 108, 170–172. (25) Bruneel, D.; Schacht, E. Polymer 1994, 35, 2656–2658. (26) Vercauteren, R.; Bruneel, D.; Schacht, E.; Duncan, R. J. Bioact. Compat. Polym. 1990, 5, 4–15. (27) Johnson, K. D.; Clark, A.; Marshall, S. Comp. Biochem. Physiol., Part B: Biochem. Mol. Biol. 2002, 131, 423–431. (28) Sawyer, T. K.; Sanfilippo, P. J.; Hruby, V. J.; Engel, M. H.; Heward, C. B.; Burnett, J. B.; Hadley, M. E. Biochemistry 1980, 77, 5754– 5758. (29) Hardwicke, J.; Ferguson, E. L.; Moseley, R.; Stephens, P.; Thomas, D. W.; Duncan, R. 7th International Symposium on Polymer Therapeutics: Lab to Clinic, 2008, Valencia, in press. (30) Ferguson, E. L.; Schmaljohann, D.; Duncan, R. Proc. Annu. Meet. Controlled Release Soc., 33rd 2006, 33, 660. (31) Gilbert, H. R. P.; Duncan, R. Proc. Annu. Meet. Controlled Release Soc., 33rd 2006, 33, 601.

BM701073N