Serum Protein Adsorption Modulates the Toxicity of Highly Positively Charged Hydrogel Surfaces

Abstract

Hydrogels formed from peptide self-assembly are a class of materials that are being explored for their utility in tissue engineering, drug and cell delivery, two- and three-dimensional cell culture, and as adjuvants in surgical procedures. Most self-assembled peptide gels can be syringe-injected in vivo to facilitate the local delivery of payloads, including cells, directly to the targeted tissue. Herein, we report that highly positively charged peptide gels are inherently toxic to cells, which would seem to limit their utility. However, adding media containing fetal bovine serum, a common culture supplement, directly transforms these toxic gels into cytocompatible materials capable of sustaining cell viability even in the absence of added nutrients. Multistage mass spectrometry showed that at least 40 serum proteins can absorb to a gel’s surface through electrostatic attraction ameliorating its toxicity. Further, cell-based studies employing model gels having only bovine serum albumin, fetuin-A, or vitronectin absorbed to the gel surface showed that single protein additives can also be effective depending on the identity of the cell line. Separate studies employing these model gels showed that the mechanism(s) responsible for mitigating apoptosis involve both the pacification of gel surface charge and adsorbed protein-mediated cell signaling events that activate both the PI3/Akt and MAPK/ERK pathways which are known to facilitate resistance to stress-induced apoptosis and overall cell survival.

Graphical Abstract

INTRODUCTION

The cytocompatibility of peptide-based hydrogels formed by self-assembly has been widely studied, and many have been found to be non-toxic in 2D and 3D cell culture systems.^1–27 However, there is an interesting consideration when using peptide gels for cell-based applications that involves the electrostatic potential of the hydrogel material. Some peptides used for constructing self-assembled gels can carry significant electrostatic charge. There are several reasons for this, but an important one for our lab is that charged peptides, especially those that are positively charged, have favorable solubility traits that allow for facile purification via chromatography after their preparation by solid-phase synthesis. In general, the electrostatic potential of a monomeric peptide is greatly magnified in a hydrogel composed of an enormous copy number of self-assembled peptides. Thus, gels prepared from peptide self-assembly can be highly charged. One might expect such surfaces to be toxic to cells. Given the prevalence of peptide gels being developed for cell-based applications, we set out to determine if positively charged gel surfaces are inherently toxic and if so are they pacified by adsorbing proteins from calf serum, a common additive to cell growth media. Although there are reports in the literature concerning protein adsorption to nanoparticles and the effect on cell toxicity,²⁸ to our knowledge, there are no reports concerning peptide-based hydrogels.

We use three model gels of varying charge prepared from different self-assembling peptides developed previously by our group.^3,4,29 The peptides belong to a class of amphiphilic cationic sequences that undergo triggered self-assembly forming a network of fibers that constitute the formation of hydrogel. Fibrils are composed of a bilayer of folded β-hairpin peptides that hydrogen-bond along the long axis of a given fibril.³⁰ The bilayer is formed by the association of the hairpins’ hydrophobic faces, shielding them from water. Import to the study presented here, the hydrophilic faces of the assembled hairpins display charged amino acid side chains on both sides of the fibril along its entire length. Thus, the charge of the bulk gel is defined by the charge state of the monomeric peptide used for assembly. Peptide HLT2 (VLTKVKTKVpPTKVEVKVLV-NH₂) carries a net charge of +5 per monomer at neutral pH, MAX8 (VKVKVKVKVpPTKVEVKVKV-NH₂, +7), and MAX1 (VKVKVKVKVpPTKVKVKVKV-NH₂, +9). The electropositive surface charge of their corresponding gels increases in the order HLT2 < MAX8 < MAX1. Interestingly, we have used both MAX8^2,3,31,32 and HLT2⁴ gels for 2D and 3D cell culture experiments but had never specifically investigated the role of calf serum, which was used as a common additive to the cell media. As will be shown, protein adsorption plays a dramatic role in cell viability. This is in contrast to a recently developed family of cytocompatible negatively charged hairpin peptide gels, some of which can facilitate 2D and 3D cell culture in chemically defined media in the absence of fetal bovine serum (FBS).^9,11 All the three positively charged peptides used in the current study were synthesized by Fmoc-based solid-phase peptide synthesis, purified by reverse-phase high-performance liquid chromatography (RP-HPLC) and characterized by liquid chromatography-mass spectrometry (LCMS), Figure S1.

MATERIALS AND METHODS

Peptide Synthesis and Purification.

Peptides were synthesized on PL-rink resin (Agilent Technologies, Santa Clara, CA, USA) using an automated ABI 433A peptide synthesizer (Applied Biosystems, Foster City, CA, USA). Syntheses were performed via solid-phase Fmoc-based chemistry with 1H-benzotriazolium-1-[bis-(dimethylamino)methylene]-5-chloro-hexafluorophosphate-(1-),3-oxide (HCTU, Peptides International, Louisville, KY, USA) activation. Fmoc-protected amino acids were purchased from Novabiochem (Merck, Darmstadt, Germany). Dried resin-bound peptides were cleaved from the resin and side-chain-deprotected with a trifluoroacetic acid (TFA)/thioanisole/ethanedithiol/anisole (90:5:3:2) cocktail for 2 h under an argon atmosphere. Crude peptides were precipitated with cold diethyl ether and lyophilized. HLT2, MAX8, and MAX1 peptides were purified by RP-HPLC using a Vydac C18 Column with solvents consisting of solvent A (0.1% TFA in water) and solvent B (0.1% TFA in 90% acetonitrile) and lyophilized. AcVES3-RGDV peptide was purified as described previously.¹¹ Analytical HPLC and electron spray ionization (positive mode) mass spectrometry confirmed the purity and composition of peptides, Figure S1.

Fetuin-A Purification.

Fetuin-A (Fet-A) was purified from commercially available bovine fetuin (FETB-16-N-1, Alpha Diagnostic International, San Antonio, TX, USA). First, size exclusion chromatography was performed with ÄKTApurifier (GE Healthcare, Chicago, IL, USA) using a HiPrep 26/60 Sephacryl S-200 HR column and phosphate buffer saline. Then, Fet-A was further purified by ionexchange chromatography with a 5 mL HiTrap Q HP. Buffer A (20 mM N-methylpiperazine, pH 4.8) and buffer B (20 mM N-methylpiperazine, 1 M NaCl, pH 4.8) were used as running buffers. Proteins bound to the column were eluted by a linear gradient of 0–30% buffer B. Fractions from the size exclusion chromatography and the ionexchange chromatography were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Coomassie Brilliant Blue (CBB) staining.

Cell Culture.

Human neonatal dermal fibroblasts (HDFs, Cell Applications San Diego, CA, USA), HEK293 human embryonic kidney cells (ATCC, Manassas, VA, USA), HS-5 human bone marrow stromal cells (ATCC), and A549 human lung carcinoma cells (ATCC) were maintained in Dulbecco’s modified Eagle’s medium (DMEM, low glucose, GlutaMAX Supplement, Thermo Fisher Scientific, Waltham, MA, USA) containing 10% FBS (Life Science Seradigm Premium Grade, VWR, Radnor, PA, USA), 100 U/mL penicillin, and 100 μg/mL streptomycin.

Preparation of FBS- and Protein-Adsorbed Peptide Gels and Cell Culture Experiments on Gels.

Peptides were initially dissolved in water at 1 wt % on ice and mixed with an equal volume of ice-cold 2× buffer [50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 300 mM NaCl at pH 7.4 for HLT2 and MAX8 and 250 mM borate, 20 mM NaCl at pH 9 for MAX1]. The 0.5 wt % peptide solution was transferred to non-tissue culture-treated 96-well plates (50 μL/well) and incubated at 37 °C for 1 h to form gels. Then, the gels were incubated with 200 μL of 10% FBS-supplemented DMEM or DMEM containing 2.5 mg/mL of bovine serum albumin (BSA, Sigma, St. Louis, MO, USA), 2 mg/mL of Fet-A, or 0.03 mg/mL of recombinant human vitronectin (VN, PEPROTECH, Rocky Hill, NJ, USA) overnight at 37 °C to allow protein adsorption. The amounts of BSA, Fet-A, and VN used for gel surface adsorption are comparable to those found in 10% FBS/DMEM.^33–36 Quantitative analysis using fluorescently labeled proteins (see below) indicate that 72% of the added BSA, 85% of the added Fet-A, and 87% of the added VN had adsorbed to the gels, Figure S2. Thus, the relative concentrations of each protein adsorbed with respect to total gel volume was 7.2 mg/mL (BSA), 6.8 mg/mL (Fet-A), and 0.1 mg/mL (VN). Control gels (gel alone) were prepared by incubating with serum-free DMEM. After incubation, the gels were washed with HEPES buffer (25 mM HEPES, 150 mM NaCl, pH 7.4, 200 μL/well) twice. Then, cells (2 × 10⁴ cells/200 μL/well) were cultured on the peptide gels for 4 days. For the data in Figure 1B investigating gel surface toxicity, the HLT2 gel was prepared in a portion of one well, affording gel and tissue culture polystyrene (TCPS) surface next to each other in the well (24-well TCPS plate, 10 μL/well). After gelation for 30 min, HDFs (1.2 × 10⁵ cells/1 mL/well) in serum-free DMEM were seeded to the well and cultured for 4 days.

Figure 1.

Effect of serum on the viability of cells cultured on positively charged peptide gels. (A) HDFs were cultured on 0.5 wt % HLT2, MAX8, and MAX1 peptide gels with serum-free DMEM or 10% FBS-supplemented DMEM for 4 days. Live cells and dead cells were visualized using calcein AM (green) and ethidium homodimer-1 (red), respectively. Scale bar = 100 μm. All panels are at the same scale. (B) Viability of HDFs cultured on adjacent HLT2 gel and TCPS surfaces in serum-free DMEM at day 4. Scale bar = 500 μm. (C) Percent of live (green bar) and dead (red bar) HDFs cultured on the HLT2 gel under serum-free conditions as a function of time. Data are represented as mean ± SD of three independent replicates.

LIVE/DEAD Cell Viability Assay on Peptide Gels.

Cell viability was assessed using a LIVE/DEAD Cell Viability Kit (Thermo Fisher Scientific). After cell culture on peptide gels, the medium was replaced with serum-free DMEM containing 1 μM calcein-AM and 2 μM ethidium homodimer-1. After incubation at 37 °C for 30 min, live and dead cells were visualized using an EVOS FL Auto Cell Imaging System (Thermo Fisher Scientific). The percentage of live and dead cells were obtained by counting the cells in a central field of three wells and presented as an average.

Lactate Dehydrogenase Cytotoxicity Assay on Peptide Gels.

Cell viability on the HLT2 gel was also measured using a Cytotoxicity LDH Assay Kit-WST (Dojindo Molecular Technologies, Kumamoto, Japan). HDFs (2 × 10⁴ cells/200 μL/well) were cultured on peptide gels for 4 days. Lactate dehydrogenase (LDH) release was measured following the instructions at day 1 and day 4 by measuring absorbance at 490 nm with an Epoch Microplate Spectrophotometer (BioTek Instruments, Winooski, VT, USA). The results are presented as an average of three wells.

Detection of Activated Caspase-3/7.

Activation of caspase-3/7 was detected using a CellEvent Caspase-3/7 Green Detection Reagent (Thermo Fisher Scientific). HDFs (2 × 10⁴ cells/200 μL/well) were cultured on the peptide gels with serum-free DMEM containing 10 μM of the CellEvent Caspase-3/7 Green Detection Reagent for 2 days. Then, the cells were visualized using an EVOS FL Auto Cell Imaging System. HDFs treated with 1 mM staurosporine and 10% Triton X-100 on tissue culture 96-well plates were used as a positive and negative control, respectively (data not shown).

Separation of Adsorbed Proteins by SDS-PAGE.

HLT2 gels (0.5 wt %, 100 μL) were prepared in glass vials (12 × 35 mm, Thermo Fisher Scientific) and incubated with 10% FBS in HEPES buffer (1 mL) for 24 h. Then, the FBS-adsorbed gels were washed with HEPES buffer (1 mL) at 37 °C for 10 min twice and dissolved in 0.1% TFA in water (900 μL). The dissolved FBS-HLT2 was applied to SDS-PAGE and analyzed by LC-MS/MS as described below. SDS-PAGE was carried out in an XCell SureLock Mini-Cell Electrophoresis System (Thermo Fisher Scientific). Bolt 8% Bis-Tris Plus Gels (10 wells) and MOPS buffer were used. Gels were run at 200 V for 35 min. SeeBlue Plus Pre-Stained Protein Standard was used as the molecular mass markers. Gels were stained with a solution of 0.25 wt % CBB G-250, 40% methanol, and 7% acetic acid in water. The major two bands in the gel were separated using a blade, digested by trypsin (trypsin from porcine pancreas, Sigma), and analyzed by LC-MS/MS.

Nano-LC–MS/MS Analysis.

The total FBS-adsorbed HLT2 gel sample and the two protein samples obtained by SDS-PAGE were digested by trypsin according to manufacturer instructions. Each tryptic digest was diluted to an estimated concentration of 0.25 pmol/μL with LC/MS-grade water based on the measured amount of pre-digested sample. Tryptic peptide mixtures were analyzed by nano-LC–MS/MS on a system (Thermo Fisher Scientific) consisting of an Easy nLC 1000 liquid chromatograph interfaced to an LTQ-Orbitrap-XL mass spectrometer equipped with a nanoSpray Flex electrospray ionization (ESI) source. Peptide mixtures were trapped, desalted, and separated on a two-column system consisting of a trap column connected to an analytical column through a low-dead-volume tee that could be either diverted to waste for sample loading and desalting or to the analytical column for component separation. The two columns were identical Thermo Easy 75 μm ID × 10 cm fused silica capillary columns packed with 3 μm C18-A2. A multi-slope, binary gradient consisting of combinations of mobile phase A, composed of 2% acetonitrile/water with 0.1% formic acid, and mobile phase B, composed of 90% acetonitrile/water with 0.1% formic acid, was used to trap and elute the peptide mixtures at a flow rate of 300 nL/min in the following manner. After injection of a 2 μL aliquot of tryptic digest, the trap column was loaded and washed with 100% A for 20 min with the flow diverted to waste. The flow was then directed to the analytical column and a shallow 2-h gradient from 0 to 50% B was used to elute and separate the trapped peptides through the two-column system. This was then followed by a 20 min gradient from 50 to 100% B, which was then maintained for an additional 25 min to wash the column. The total analysis time was 185 min.

ESI mass spectra (MS1) were obtained over the range m/z 200–2000 at an Orbitrap resolution of 30,000 using an automatic gain control target of 5 × 10⁵. Data-dependent acquisition for collision-induced dissociation (CID) MS/MS analysis was carried out under data system control using Xcalibur software (version 2.1.0, SP1). Precursor ion selection required a threshold of 5 × 10⁴ and a charge state of 2 or 3 within the range m/z 350–1600. A top-5 precursor ion selection was employed with an isolation width of 2.0 Da and a repeat count of 3 with dynamic exclusion enabled using an exclusion window of 90 s and a mass tolerance of±1.5 Da. CID was carried out at a relative collision energy of 30 and product ion spectra (MS/MS) were obtained at a resolution of 15,000 to keep maximum cycle time <3 s. Each sample was analyzed in duplicate.

Mass spectra raw files were searched against the Swiss-Prot bovine and human (for possible sample contamination) protein databases using MaxQuant. The bovine database contains 6253 protein entries. The minimum required peptide length was seven amino acids, and two missed cleavages were permitted for fully tryptic peptides. Mass tolerances for first search peptide tolerance and main search peptide tolerance were 20 and 5 ppm, respectively. The mass tolerance for fragment ion masses was 10 ppm. Variable modifications of methionine oxidation were also allowed. The maximum false discovery rate of peptide-spectrum matches and protein were set to 1% for identification.

Quantitative Measurement of Protein Adsorption.

UV-based difference measurements were used to determine the amount of BSA, Fet-A, and VN bound to the hydrogel surface. HLT2 gels (50 μL) were prepared in glass vials (12 × 35 mm) and incubated with 2.5 mg/mL of BSA fluorescein isothiocyanate conjugate (FITC-BSA, sigma A9771), 2 mg/mL of fluorescein-Fet-A, or 0.03 mg/mL of fluorescein-VN in HEPES buffer (200 μL) at 37 °C for 24 h. After the incubation, UV absorbance at 494 nm of each supernatant was measured to determine % protein adsorbed by difference. Fluorescein-labeled Fet-A and fluorescein-labeled VN were prepared using NHS-fluorescein (Thermo Fisher Scientific).

Transmission Electron Microscopy.

HLT2 gel (0.5%) in HEPES buffer was prepared as described above in a non-tissue culture-treated 48-well plates (100 μL/well). Following 1 h of incubation at 37 °C, 400 μL of HEPES buffer supplemented with 2.5 mg/mL of BSA was added on top of the gels for 24 h of incubation at 37 °C to allow protein adsorption. Control HLT2 gel was prepared similarly, incubated with only HEPES buffer. To allow visualization of distinct fibers, gel samples were diluted ×40 into water and mixed thoroughly. A 4 μL drop of the diluted peptide solution was placed on a 200-mesh copper grid covered by carbon film (Electron Microscopy Science, Hatfield, PA, USA) for 1 min and blotted by filter paper. Subsequently, grid was washed by adding 4 μL of water for several seconds. Water was blotted by a filter paper, and 4 μL of 0.75% uranyl formate was immediately added to the grid for 1 min, then blotted with a filter paper, and left to air dry. Images were taken with a Technai T12 at an 80 kv accelerating voltage. Average fibril width was measured via ImageJ software by taking 200 (HLT2 or BSA-HLT2) independent measurements from distinct fibrils in the field of view of fibrils observed at 3 (HLT2) or 6 (BSA-HLT2) separate micrographs, representing different locations of the fibrils.

Western Blot.

BSA-, Fet-A-, and VN-adsorbed HLT2 gels (500 μL) were prepared in non-tissue culture-treated 35 mm dishes as described above. HDFs (1 × 10⁶ cells/2 mL/dish) in serum-free DMEM were seeded and cultured on the gels. After 24 h, the cells were lysed with cell lysis buffer (RIPA buffer, Thermo Fisher Scientific, 500 μL). Cell lysates were separated by SDS-PAGE and transferred to nitrocellulose membranes. The membranes were blocked by the StartingBlock Blocking Buffer (Thermo Fisher Scientific), incubated with either anti-phospho-Akt (Ser473, D9E), anti-Akt (pan, C67E7), anti-phospho-ERK1/2 (Thr202/Tyr204, D13.14.4E), or anti-ERK1/2 (137F5) monoclonal antibody (1:1000, Cell Signaling, Danvers, MA) overnight at 4 °C, and detected by horseradish peroxidase-conjugated secondary antibody (1:2000, Cell Signaling) using an ECL substrate (Thermo Fisher Scientific). The relative phosphorylation of Akt and ERK was assessed using ImageJ software.

RESULTS AND DISCUSSION

Two-dimensional cell culture experiments show the stark effect of calf serum on cell viability. Figure 1A shows a live/dead assay performed 4 days after human neonatal dermal fibroblasts (HDFs) are introduced to all the three hydrogel surfaces. HDFs are important in producing connective tissues and central to wound healing, making them a good model to study material–cell interactions. Cells are attached avidly to the gels irrespective of the presence or absence of serum. However, in serum-free DMEM, almost all of the cells undergo cell death irrespective of the hydrogel identity. In contrast, when cells are cultured in DMEM containing 10% FBS, nearly all are viable, indicating that the serum is extremely effective at rescuing the cells. This data is in agreement with that previously published by our group, showing that these gels are cytocompatible under FBS-supplemented conditions.^{1–4,31,32,37} Figure 1A also shows that there is little dependence of the degree of hydrogel-positive charge on cytotoxicity under serum-free conditions, suggesting that in general, highly charged surfaces are toxic to cells.

We next briefly investigated possible mechanisms of cell death. Since the cells seem to behave similarly on all three gels, we focused our attention on only one for the remaining studies, namely, HLT2. This gel carries the least charge and best represents the more moderately charged peptide-based gels being developed in the field. Figure 1B shows that cell death is indeed mediated by a surface–contact mechanism. Here, hydrogel is added to a spatially defined portion of a tissue culture polystyrene (TCPS) well creating two distinct surfaces (HLT2 gel and TCPS) directly next two each other. Cells were introduced to the well and viability assessed after 4 days of culture in serum-free DMEM. Clearly, cells that attached to the hydrogel have died, whereas cells on the TCPS surface remain viable with a clear demarcation between the two surfaces. This experiment also shows that cell death is not due to any free peptide that might dissociate from the gel or from serum starvation. Figure 1C shows that the rate at which cells die after being introduced to the HLT2 gel is somewhat slow. We were surprised to observe that most of the cells were viable after 24 h, speculating that they would have undergone rapid necrosis perhaps through a mechanism involving cell depolarization and/or membrane disruption after coming into contact with the gel’s surface. This observed slower rate of cell death is more consistent with an apoptotic mechanism, which was confirmed by the presence of activated caspases 3 and 7 in cells cultured on the gel at day 2, Figure 2A. Cells undergoing apoptosis in culture in the absence of phagocytes complete their death with secondary necrosis, which is characterized by cell membrane permeabilization.³⁸ Figure 2B shows the percent of LDH released from the cytoplasm over time. The amounts of LDH released at days 1 and 4 are consistent with the data in Figure 1C showing the relative amounts of cell death on these days. Last, the ethidium homodimer dye used in the live/dead assays (Figure 1) can only access cells with compromised membranes to yield its red fluorescence. Thus, taken together, the data suggest that cells coming into contact with the HLT2 gel surface under serum-free conditions undergo apoptosis accompanied by secondary necrosis. The exact molecular mechanism leading cells to apoptosis is not yet known.

Figure 2.

(A) HDFs at day 2 cultured on the HLT2 gel under serum-free conditions contain activated caspases-3/7 (green). (B) LDH released from HDFs cultured on HLT2 in serum-free DMEM at days 1 and 4.

The ability of serum to rescue cells is likely due to the adsorption of proteins onto the gel surface as opposed to the action of soluble protein. We tested this assertion by examining the cytocompatibility of FBS-adsorbed HLT2 gels across several cell lines. FBS-adsorbed gels were prepared by incubating preformed gels with 10% FBS for 1 day, followed by extensive washing with serum-free buffer. Cells were then introduced to the gel surface and cultured in serum-free DMEM. Figure 3 shows the percent of live versus dead cells after 4 days of culture for HDFs, HEK293 human embryonic kidney cells, HS-5 human bone marrow stromal cells, and A549 human lung carcinoma cells on serum-free and serum-adsorbed gels. The data is striking for all the cell lines showing a clear effect of protein adsorption on cell viability. For example, more than 80% of the HDFs cultured on the serum-adsorbed gel remain viable after 4 days with no additional nutrients added to the culture. All of the cells died on the serum-free gel surface. This effect was universal across all the cell lines studied. Further, Figure 4A shows that when HDFs are cultured on the FBS-adsorbed gel, relatively few cells fluoresce positive for activated caspases-3/7 at day 4. Importantly, panel B quantitates and compares the number of cells with activated caspase-3/7 for serum-adsorbed versus non-adsorbed surfaces at day 2, a critical time point for cells on the bare HLT2 surface, Figure 1C. Clearly, the FBS-adsorbed HLT2 gel significantly limits apoptosis.

Figure 4.

FBS-adsorbed HLT2 gel limits apoptosis. (A) HDFs were cultured on the FBS-adsorbed HLT2 gel with serum-free DMEM for 4 days and activation of caspase-3/7 was visualized (green). (B) Percent of caspase-3/7 positive cells on the HLT2 gel alone versus the FBS-adsorbed HLT2 gel. Data are represented as mean ± SD of three independent replicates.

Figure 3.

Effect of adsorption of FBS components on cell viability. FBS-adsorbed HLT2 gels were prepared as described in the Materials and Methods section. The HLT2 gel alone was used as a control. HDFs (A), HEK293 cells (B), HS-5 cells (C), and A549 cells (D) were cultured on FBS-HLT2 gels with serum-free DMEM for 4 days. Live cells (green) and dead cells (red) were quantified using calcein AM and ethidium homodimer-1, respectively. Data are represented as mean ± SD of three independent replicates.

Multistage mass spectrometry (MSn) was used to identify the proteins bound to the material’s surface. In this experiment, the FBS-adsorbed HLT2 gel was entirely dissolved with 0.1% TFA and proteolyzed with trypsin. Mass analysis identified 44 proteins (Figure 5A), 37 of which were acidic having isoelectric points (pIs) < 7. Thus, electrostatics is the likely modus operandi responsible for protein adsorption. This was confirmed by a control experiment employing a negatively charged hydrogel prepared from the peptide AcVES3-RGDV (Ac-VEVSVSVEVpPTEVSVEVEVGGGGRGDV-NH₂), which carries a charge of −6 per monomer at neutral pH.¹¹ DMEM containing 10% FBS was added to this gel and allowed to incubate for 24 h. After washing, SDS-PAGE showed that only a small amount of protein had bound to any appreciable level, Figure S3. To narrow the pool of proteins for future studies on the HLT2 gel, SDS-PAGE couple with mass analysis was performed to identify those that had adsorbed at high concentration. Figure 5B shows the presence of large concentrations of BSA (66 kDa, measured pI 4.7–5.6³⁹) and Fet-A (43–54 kDa, measured pI 3.3–4.3⁴⁰), both major carrier proteins in serum. Further, both proteins are the main components of FBS, which contains 20–36 mg/mL of BSA and 10–21 mg/mL of Fet-A.^33–35 Thus, we choose BSA and Fet-A for further study. In addition, we also included VN, a critical cell-adhesive protein in FBS.^36,41,42

Figure 5.

Identification of FBS components adsorbed to the HLT2 gel. (A) Table of FBS components adsorbed to the HLT2 gel identified by LC-MS/MS analysis. The right column represents the theoretical isoelectric points (pIs) of the proteins. pI values lower than 7 are shown in red. (B) SDS-PAGE of FBS alone (left lane) and adsorbed proteins isolated from the HLT2 gel (right lane). The two most abundant proteins adsorbed to the gel are BSA and FET-A.

Next, the protective abilities of BSA, Fet-A, and VN were examined by individually adsorbing each protein to the surface of the HLT2 gel. Each protein was introduced by using DMEM as a vehicle at concentrations mimicking those found in DMEM cell culture media containing 10% FBS (see Materials and Methods section). Quantitative analysis showed that the relative concentrations of each protein adsorbed with respect to total gel volume was 7.2 mg/mL (BSA), 6.8 mg/mL (Fet-A), and 0.1 mg/mL (VN), Figure S2. HDFs, HEK293, HS-5, and A549 cells were cultured on each of the gels for 4 days, Figure 6A–D. In general, adsorption of each individual protein improved the cytocompatibility of the HLT2 gel toward all the cell lines studied. The data also show that for a given cell line, each protein can have a differential effect on viability. For example, adsorbed Fet-A and VN significantly increased HDF viability in comparison to the moderate influence of BSA. A caspase assay quantitating apoptotic cells shows a similar trend, with Fet-A and VN substantially mitigating HDF cell death, Figure 6E. With respect to the identity of the adsorbed protein, each protein can impact individual cell lines differently. For example, although BSA increases HEK293 viability substantially, its effect on the other cell lines is moderate. Last, the data shows that VN can be quite potent at mitigating the toxicity of the HLT2 surface, especially toward HDF and HEK293 cells. Given that the total amount of VN adsorbed to the surface is an order of magnitude less than that of BSA and Fet-A, the ability of this protein to protect the cells is striking. The data in Figure 6 stimulates questions concerning the mechanism(s) by which these adsorptive coatings are exerting their effect. One obvious mechanism is that the negatively charged proteins bind to the positively charged gel surface, effectively reducing the charge experienced by the cells. In fact, FITC-labeled BSA was used as a model to visualize protein adsorption at the surface of the gel. Figure 7A shows that the protein effectively coats the gel’s surface penetrating about 1 mm into the gel network. TEM was then used to gain an understanding of the distribution of unlabeled BSA adsorbed to the gel’s fibrils. Figure 7B,C shows micrographs of HLT2 fibrils isolated from gel surfaces in the absence and presence of BSA. Fibril widths were measured for each (panel D), clearly showing that BSA adsorbs to the fibrils increasing their average diameter by ~2 nm. We also measured fibril widths as a function of axial distance every 8 nm along the long axis of individual fibers to gain an understanding of how protein molecules might be arranged along the fiber. Figure 7E shows a constant fibril width (~5–6 nm) for the entire 100 nm fibril length measured, indicating that BSA coats the entire fiber. Taken together, the data in Figure 7 suggests that the gels’ surface charge should be effectively passivated by the adsorbed protein. However, this is not likely the only mechanism at play. For example, large amounts of BSA (7.2 mg/mL) and Fet-A (6.8 mg/mL) coat their respective gels, presumably neutralizing the surface to similar extents. Yet HDFs react differently to the two surfaces and the same is true for A549 cells. Thus, in addition to charge pacification, adsorbed protein-mediated cell signaling events may also be playing a role.

Figure 6.

Effect of BSA-, Fet-A-, and VN-adsorbed HLT2 gels on cell viability. HDFs (A), HEK293 cells (B), HS-5 cells (C), and A549 cells (D) were cultured on the BSA-, Fet-A-, and VN-adsorbed HLT2 gels with serum-free DMEM for 4 days. The HLT2 gel alone was used as a control. Live cells and dead cells were quantified using calcein AM and ethidium homodimer-1, respectively, and (E) suppression of caspase-3/7 activity by adsorption of BSA, Fet-A, and VN to the HLT2 gel. HDFs were cultured on the gels with serum-free DMEM for 2 days and % of caspase-3/7 positive cells measured. Data are represented as mean ± SD of three independent replicates.

Figure 7.

(A) FITC-labeled BSA adsorbed to the surface of an HLT2 gel. Light absorption and fluorescence emission images shown at left and right, respectively. (B,C) TEM micrographs showing fibrils isolated from 0.5 wt % HLT2 gel (B) in the absence of protein or (C) following 24 h of incubation with BSA. Scale bar = 100 nm. (D) Widths of individual fibrils of each sample were determined using ImageJ software by measuring the width of fibrils from three separate micrographs from the gel alone (black) or six separate micrographs from the BSA gel (red), representing different locations of the fibrils on the grid, n = 200 for each gel sample. (E) The widths of discrete fibrils were measured at increments of 8 nm along an axial distance of 100 nm for fibrils from HLT2 gel alone (black) or following incubation with BSA (red). n = 8 fibrils for each group. Additional representative micrographs for BSA-free and adsorbed HLT2 fibrils can be found in Figures S4 and S5.

Protein kinase B (Akt) and extracellular signal-regulated kinase (ERK) are involved in multiple pathways of resistance to stress-induced apoptosis and overall cell survival.^43,44 The BSA-, Fet-A-, and VN-adsorbed HLT2 gels were examined for their ability to elicit Akt and ERK phosphorylation in HDF cells, Figure 8. Akt phosphorylation was observed for both the Fet-A and VN gels, but not the BSA gel. In contrast, ERK phosphorylation was observed for all the three protein-adsorbed gels, with the BSA gel showing the strongest effect. These results are consistent with previous reports showing that all three proteins can activate these kinases. For example, albumin is capable of activating ERK-dependent cell growth and proliferation,^45,46 and Fet-A can elicit both Akt- and ERK-dependent adhesion and proliferation across multiple cell types.^47,48 The extracellular matrix (ECM) can influence apoptotic events through integrin signaling.⁴³ VN, found in both serum and ECM, binds to αvβ3 and αvβ5, membrane-bound integrins, via an -RGD- motif found in the protein’s primary sequence. In addition to its role in adhesion, VN has been found to inhibit apoptosis through integrin binding and activation of both the PI3/Akt and MAPK/ERK pathways.⁴⁹ Collectively, the data suggests that all the three adsorbed proteins are capable of inducing cell signaling events resisting apoptosis and/or promoting cell proliferation. Similar cell signaling pathways could be playing a role in the serum-dependent reduction of nanoparticle toxicity previously reported,²⁸ a possibility not yet explored.

Figure 8.

Phosphorylation of Akt and ERK1/2 in HDFs cultured on the protein-absorbed HLT2 gels. (A) HDFs were cultured on BSA-, Fet-A-, and VN-HLT2 gels with serum-free DMEM for 24 h, at which time Akt/pAkt and ERK1/2/pERK were measured. (B,C) Relative optical density (OD) of pAkt and pERK as a function of gel identity. Quantification performed using ImageJ software. The OD of pAkt and pERK was normalized to the OD of Akt and ERK for each sample. Data are represented as mean ± SD of three independent replicates. *p < 0.01.

CONCLUSIONS

The adsorption of serum proteins effectively mitigates the cytotoxicity of positively charged peptide-based hydrogels. Multistage mass spectrometry showed that at least 40 serum proteins adsorb to the gel surfaces, most of which are negatively charged. Cell-based assays employing model HLT2 gels coated individually with BSA, Fet-A, or VN suggest that the mechanism(s) of cell rescue likely involve both surface charge pacification and distinct cell signaling events resulting in resistance to apoptosis that are mediated by the absorbed protein. In this study, we only examined the effects of whole serum and three of its protein components likely to influence cell survival. We are cognizant that FBS contains a plethora of other proteins, including growth factors, as well as vitamins, lipids, and hormones that can contribute to cell survival. With respect to utilizing these positively charged gels for 2D and 3D cell culture, serum or a cocktail of essential adsorptive proteins capable of mitigating apoptosis are necessary additives. The choice of protein(s) to use will be determined by the intended application of the gel. This is especially important for gels being developed for immunological applications where adsorbed protein could impact immune responses.