Letter pubs.acs.org/NanoLett
Spontaneous Coassembly of Biologically Active Nanoparticles via Affinity Binding of Heparin-Binding Proteins to Alginate-Sulfate Emil Ruvinov,† Inbar Freeman,† Roei Fredo,† and Smadar Cohen*,†,‡,§ †
Avram and Stella Goldstein-Goren Department of Biotechnology Engineering, ‡Regenerative Medicine and Stem Cell (RMSC) Research Center, and §The Ilse Katz Institute for Nanoscale Science and Technology, Ben-Gurion University of the Negev, Beer-Sheva 8410501, Israel S Supporting Information *
ABSTRACT: Controlled delivery of heparin-binding (HB) proteins represents a challenge in regenerative medicine strategies. Here, we describe the features of novel nanoparticles (NPs), spontaneously coassembled due to affinity interactions between HB proteins and the semisynthetic anionic polysaccharide, alginate-sulfate. The NPs efficiently encapsulated and protected the proteins from proteolysis. Injection of a combination of NPs encapsulating multiple therapeutic growth factors promoted effective and long-term tissue repair in animal models of severe ischemia (murine model of hindlimb ischemia and acute myocardial infarction in rats). This simple yet efficient NP fabrication method is amenable for clinical use. KEYWORDS: Affinity binding, alginate-sulfate, coassembly, heparin-binding proteins, nanoparticles, repair
H
ranging from 17 to 80 kDa and with binding constants (KA) in the range of 106−107 M−1, led to the spontaneous formation of NPs. By atomic force microscopy (AFM), the coassembly of alginate-sulfate with basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF), or hepatocyte growth factor (HGF) revealed a topography of nearly spherical particles spread on the mica surface (Figure 1b−d) compared to the fiber-like structure of soluble alginate-sulfate (Figure 1a). Such nanostructure of pristine alginate, possibly due to side-byside association of individual polymer chains, was previously reported.9 By contrast, coincubation of pristine alginate with HB proteins did not yield NPs (Supporting Information (SI) Figure S1), pointing to the involvement of sulfate groups in the affinity interactions between HB proteins and alginate-sulfate. Nearly spherical NPs were also evident in the dry and cryogenic transmission electron microscopy (TEM) images following incubation of gold-labeled alginate-sulfate with the HB proteins (Figure 1e−g,i), compared to the fiber-like structure of soluble alginate-sulfate (Figure 1h). The diameter
eparin-binding (HB) proteins are critical effectors in tissue regeneration and repair. Although proven to be therapeutically effective in preclinical studies, the clinical translation of HB proteins as therapeutics has been limited due to the use of supraphysiological dosages, thus increasing the risk for systemic toxicity.1,2 Delivery of HB proteins in nanoparticles (NPs) could potentially reduce the dosage required to achieve therapeutic efficacy and improve their clinical outcome.1 In conceiving the appropriate nanoparticulate delivery system for the task, we attempted to mimic the affinity interactions existing between HB proteins and sulfated sequences in heparan-sulfate glycosaminoglycans (GAGs),3 while avoiding the disadvantages associated with their prolonged in vivo use, such as fast degradation rate (∼30 min to 2.5 h)4 and possible side effects (e.g., bleeding, toxicity). For this purpose, alginatesulfate5 was synthesized from the algae-derived polysaccharide alginate,6,7 and the affinity binding of several HB proteins to alginate-sulfate was validated by surface plasmon resonance (SPR) analysis.5,8 We now show that in aqueous solution, the affinity interactions between alginate-sulfate and several HB proteins, representing heparin-binding proteins with molecular weights © XXXX American Chemical Society
Received: September 7, 2015 Revised: December 20, 2015
A
DOI: 10.1021/acs.nanolett.5b03598 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
size range of the particles observed in the dry TEM images was similar to that determined by AFM analysis, in the order of several tens of nm (SI Table S1). In comparison, cryo-TEM analysis (Figure 1i) of hydrated NPs, demonstrated that the coassembled NPs of HGF with gold-labeled alginate-sulfate were ∼6-fold larger (224.7 ± 43.1 nm, n = 20). The larger size is attributed to NP swelling in aqueous solutions, as previously reported for other polysaccharide-based hydrogel NPs.10,11 Of note, as typical cryogenic TEM specimen width is limited to ∼250 nm,12 larger sized NPs cannot not be identified and analyzed by this technique. To accurately measure the hydrated NP size and distribution, we used nanoparticle tracking analysis (NTA).13,14 Data analysis of particle concentration as a function of particle diameter revealed that compared to soluble alginate-sulfate, NP formation due to the coassembly of alginate-sulfate with bFGF or HGF was associated with a reduction in particle concentration (6 × 106/ml for alginate-sulfate vs 2.4 × 106/ ml and 1 × 106/ml, for HGF NPs and bFGF NPs, respectively) without significant shifts in diameter size (∼100 nm) (Figure 2a). However, the size distribution profile of both NP samples featured smaller peaks of larger diameters, which were absent in alginate-sulfate (Figure 2b). We then reanalyzed the data for volumetric changes by showing volume concentration (particle volume based on the estimated sphere diameter/ml), instead of particle concentration, as a function of particle diameter.15 Compared to soluble alginate-sulfate, which exhibited one major peak (∼200 nm, 5 × 1012 nm3/ml volume concentration), NPs coassembled from bFGF or HGF and alginatesulfate exhibited a smaller alginate-sulfate peak (due to the reduced total volume of free polymer) and peaks at significantly larger diameters (bFGF NPs, at 290 nm, 0.7 × 1012 nm3/ml and at 470 nm, 1.5 × 1012 nm3/ml; HGF NPs, at 440 nm, 1.6 × 1012 nm3/ml) (Figure 2c). These results clearly point to a relative increase in the volume concentration of larger particles in solution due to NP formation. The appearance of two peaks in the bFGF NP sample could be attributed to the existence of two NP species with different stoichiometries of components. Of note, the assumption of completely spherical particles used by NTA and the resulting particle diameter should be treated with caution in the case of the soluble linear alginate-sulfate. Surface charge analysis by zeta (ζ) potential measurement (Figure 2d) revealed a lesser negative charge for the coassembled NPs (−25.5 ± 1.5 and −32.5 ± 0.5 mV for HGF NPs and bFGF NPs, respectively, p < 0.05) compared to that of free alginate-sulfate in aqueous solution (−44.7 ± 1.3 mV). Taking into account the particulate nature of the alginatesulfate/protein complexes and net positive charge of HB proteins at pH ∼ 7 (pI of 9.9 and 7.7 for bFGF and HGF, respectively; calculated at http://isoelectric.ohv.org based on amino-acid sequences of mature proteins), these results confirm the TEM data by showing that the alginate-sulfate chains are an integral part of the coassembled NPs, bestowing a net negative charge on the particle surface. Negative surface charge is generally preferable for NP formulations, as it reduces nonspecific adsorption of serum proteins, improves colloidal stability, and prevents aggregation.16 To substantiate NP coassembly due to affinity binding of HB proteins to alginate-sulfate, we employed the fluorescence resonance energy transfer (FRET) technique using connective tissue growth factor (CTGF), a known member of the matrixassociated and heparin-binding protein family.17 Interactions between CTGF (used as FRET acceptor) and alginate-sulfate
Figure 1. Images of spontaneously coassembled NPs formed due to affinity interactions between alginate-sulfate and HB proteins, obtained using high-resolution microscopy. (a−d) Tapping-mode AFM images of (a) soluble alginate-sulfate and (b−d) bFGF NPs, VEGF NPs, and HGF NPs, respectively. (e−g). Dry TEM images of bFGF NPs, VEGF NPs, and HGF NPs with nanogold-labeled alginate-sulfate. Several NPs are encircled for convenience. (h,i) Cryogenic TEM images of alginate-sulfate and HGF NPs with nanogold-labeled alginate-sulfate. AlgS, alginate-sulfate. B
DOI: 10.1021/acs.nanolett.5b03598 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. NP size and surface charge (a−d) and interaction analysis proving NP formation (e). (a−c) NTA analysis of the size distribution of NPs formed due to coassembly of HB proteins and alginate-sulfate. (a) Mean diameter changes after NP formation as a function of particle concentration. (b) Size distribution range of 200−800 nm, showing populations with larger diameters in NP samples, indicated by arrows. (c) Mean diameter changes after NP formation as a function of particle-volume concentration. (d) Surface charge (ζ-potential) measurements of various NPs and alginate-sulfate. *, p < 0.05 (Tukey’s posthoc test). (e) Intermolecular interaction measurement proving NP formation by FRET. Mean normalized FRET values are presented as a function of donor concentration (alginate-sulfate-488 nm or unmodified alginate-488 nm, 1.3, 13.3, and 133.3 nM) at constant acceptor concentration (CTGF-555 nm, 80 nM). Asterisks designate statistically significant difference (t test, p < 0.05). Negative values in CTGF-pristine alginate sample are obtained due to overcorrection of donor and acceptor cross-talk.18,19AlgS, alginate-sulfate.
(used as FRET donor) resulted in a significant FRET signal, compared to the lack of signal in samples of pristine alginate, further confirming the close association of alginate-sulfate and the protein, due to specific affinity binding (Figure 2e and SI). Next, we tested whether the coassembled NPs protect the entrapped HB protein from enzymatic proteolysis by exposing the samples to trypsin and evaluating the degree of HB protein degradation by MALDI-TOF mass spectroscopy. The exposure of suspended bFGF-NPs to the enzyme in external medium yielded less digestion fragments compared to the protein’s soluble state (SI Figure S2). The peak of bFGF in the sample of bFGF NPs could not be identified. This could be readily explained by the interaction of AlgS with bFGF and NP formation, which significantly reduces the amount of free protein susceptible to degradation. Apart from trypsin peaks (identified in previous studies),20 in several spectral ranges (i.e., 950−1000, 1100−1500, 6000−20000 m/z) there are no digestion peaks at all. This suggests that bFGF in the NP form is less susceptible to enzymatic digestion, as previously shown for other encapsulated HB proteins.20,21
The culmination of the current work was to explore the therapeutic potential of the coassembled NPs in two models of severe ischemia: hindlimb ischemia and myocardial infarction (MI). For the treatment of hindlimb ischemia in mice, three coassembled NPs, each containing one angiogenic factor, VEGF, PDGF-BB, or TGF-β1, were fabricated. This triple factor combination has been shown to produce a mature blood vessel network.8,22,23 The NPs were combined, dispersed in calcium cross-linked alginate solution,20,24,25 and then injected into the ischemic muscle, 24 h after femoral artery occlusion. At this time point before treatment, the Laser Doppler image (Figure 3a, left image) demonstrates no blood flow due to the artery occlusion. Quantitative analysis further confirmed that blood perfusion after occlusion of the femoral artery is low,
