The primary structures of [V₇]₂₀ and are shown in Figure 1A. In [RGD-V₆]₂₀, 20 RGD motifs were evenly distributed throughout the whole molecular structure. Typically, from a 40 l fermentation batch of E. coli, approximately 600 g of wet cell cake was obtained. After 3 rounds of inverse transition cycling, 1,036 and 1,080 mg of [V₇]₂₀ and [RGD-V₆]₂₀, respectively, were recovered from a 40 liter batch with a purity greater than 95%, as judged by densitometry on an SDS-PAGE gel (Figure 1B). When a UV-visible spectrum of [RGD-V₆]₂₀ in phosphate-buffered saline (PBS) was scanned from 800 to 200 nm (Figure 1C), its maximum absorbance occurred at 280 nm, and 2 shoulders were detected at 273 (ε = 5,505 M^-1 cm^-1) and 288 nm (ε = 4,666 M^-1 cm^-1). No absorbance was detected in the visible region.

To investigate the effect of RGD modification on cell adhesion, a series of comparative assays was performed on the surfaces of TCPS, [V₇]₂₀, [RGD-V₆]₂₀, and fibronectin. As an initial test, protein-coated surfaces were prepared by isothermal adsorption of 5 μM protein solutions at 4°C.

The relative percentages of cell attachment on TCPS, [V₇]₂₀, and [RGD-V₆]₂₀ were 12% ± 2.8%, 43% ± 7.4% and 88% ± 9.5% of fibronectin, respectively, suggesting efficient cell adhesion to the [RGD-V₆]₂₀ coated surface. To determine the cell-binding constants, the plots of cell adhesion versus protein-coating concentration were calculated and are displayed in Figure 2. As the coating concentration increased from 0.1 to 10 μM, no noticeable cell-binding pattern was observed on the plot for [V₇]₂₀; the data did not fit to either a linear first order (r² = 0.582) or hyperbolic (r² = 0.816) regression. In contrast, N2a adhesion to either [RGD-V₆]₂₀ (r² = 0.979) or fibronectin (r² = 0.980) resulted in a hyperbolic graph with a saturation around 2 μM protein concentration. The apparent maximum cell binding to [RGD-V₆]₂₀ was approximately 96% of fibronectin, with half-maximal adhesion on [RGD-V₆]₂₀ and fibronectin occurring at a coating concentration of 2.4 × 10^-7 and 1.4 × 10^-7 M, respectively.

The saturated adsorption density of the RGD-VPGIG polypeptide (31 kDa) and fibronectin (225 kDa) were adopted from the published literature as being 30 ng/mm² (1 pmol/mm²) 11 and 8 ng/mm² (36 fmol/mm², as monomeric form) 13 , respectively. Based on the report that ELPs with different amino acid content showed a similar levels of adsorption 11 , we numerically estimated the saturated surface density of [RGD-V_{6 20} as 500 fmol/mm², which in turn led us to interpret that both the B _max and K _d values of [RGD-V_{6 20} for N2a cell binding would be 14-fold lower than those respective values for fibronectin.

The effect of protein-coated surfaces on cell spreading morphology was inspected by phase contrast microscopy. Morphological characteristics of N2a cells were divided into two shapes: round shape and extended shape. Round cells were dominant on both TCPS and (Figure 3A and B), whereas almost equal populations of two classes of cell shapes were observed on the microscopic images of the cells adhered on [RGD-V₆]₂₀-adsorbed substrates (Figure 3C). For fibronectin-coated substrate, extended cells were dominant and relatively small population of less spread cells was observed (Figure 3D). Consequently, the percentage of cell spreading increased in the following order of proteins: TCPS (less than 10%) < [V₇]₂₀ (15.5% ± 3.2% < [RGD-V₆]₂₀ (42.9% ± 6.5%) < fibronectin (84.3% ± 6.9%).

An N2a migration assay was performed on TCPS- and protein-coated surfaces prepared at 20 μg/ml coating concentration. The typical images of cell migration across the scratched zone are shown in Figure 4A, and the number of cells that migrated during the 30 h culture period has been compared in Figure 4B. When cells were grown on TCPS and [V₆]₂₀, the number of cells in the scratched area steadily increased over the 30-h period, giving CMI values of (2.1 ± 0.4)/h and (2.7 ± 0.7)/h, respectively. When cell motility on [RGD-V₆]₂₀ and fibronectin was measured after 24 h of culture, the N2a cells migrated with the same speed of (3.8 ± 0.6)/h, which was 1.8-fold faster than the mobility of the cells on TCPS. However, when the number of migrated cells were counted at 30 h of culture, the CMI was significantly (P < 0.01) enhanced to (7.5 ± 0.7)/h. Interestingly, during the 24–30 h culture period, the CMI values measured for [RGD-V₆]₂₀ and fibronectin were (22.8 ± 2.7)/h and (22.2 ± 1.8)/h, respectively, which were approximately 8-fold faster than the migration speed of the cells on the [V₇]₂₀-coated surface during the same period.

We evaluated whether [RGD-V_{6 20} could affect N2a differentiation by measuring the expression levels of several neuronal proteins such as NSE and Tuj1, NF68, MAP2, and GFAP, which have been widely used as biomarkers for neuronal cells, axonal development, dendritic cells, and astrocytes, respectively 14 . Figure 5 compares the mRNA levels of these markers in N2a cells grown on TCPS or [RGD-V_{6 20} in the presence and absence of retinoic acid (RA). The expression of TuJ1 (1.5-fold), MAP2 (1.5-fold), and GFAP (1.4-fold) was enhanced significantly (P < 0.05) in the cells cultured on [RGD-V_{6 20}, whereas the expression of NSE and NF on [RGD-V_{6 20} did not differ from that of the cells grown on TCPS. The mRNA levels of marker proteins were further elevated following RA treatment, the highest expression being observed in the cells cultured on [RGD-V_{6 20} treated with RA.

Immunofluorescence staining of Tuj1 expression was performed to further validate the stimulatory effect of [RGD-V₆]₂₀ on neuronal differentiation. Figure 6 shows the immunofluorescence images of N2a cells on TCPS, [V₇]₂, [RGD-V₆]₂₀, and fibronectin after 5 days of culture. Initially, culture surfaces were prepared using 0.1 μM of protein solutions. In response to 20 μM RA treatment, most N2a cells extended neurites on the TCPS, [V₇]₂₀, and [RGD-V₆]₂₀ surfaces (Figure 6a, b, c). In the absence of RA, less than 4% of cells showed neurite formation on TCPS and [V₇]₂₀ (Figure 6c, d), whereas on [RGD-V₆]₂₀, 10.6% ± 2.2% (P < 0.05) of cells exhibited tadpole-like morphological features with sprouting neurites (Figure 6f). We subsequently analyzed the neuritogenesis at 5 μM coating concentration, and the percentage of neuritogenesis was found to be 12.5% ± 3.1%, which was significantly different from the values obtained for TCPS and [V₇]₂₀ (P < 0.05) but was not statistically significant (P > 0.05) from the value obtained at the 0.1 μM [RGD-V₆]₂₀ coating concentration. Interestingly, N2a cells bearing 2–4 nuclei within a single cytosol body with multiple neurites were seen (Figure 6g). In parallel experiments, the effect of fibronectin on N2a differentiation was analyzed at the 5 μM coating concentration, and 15.2% ± 2.4% of neurite formation was noted (Figure 6h). Figure 6i compares the neuritogenesis of N2a cells on TCPS and protein-coated surfaces and clearly shows that [RGD-V₆]₂₀ promotes neuronal differentiation.

ELPs were expressed from pET-25b(+)-1 containing the [V_{7 20} or [RGD-V_{6 20} gene in 40 l culture of E. coli BLR(DE3) (Novagen). Protein expression was induced at an OD₆₀₀ of about 0.6 with 1 mM β-isopropyl thiogalactoside (Sigma) and allowed to express for 4 h post induction. The ELPs were by phase transition cycling 19 . The purified ELPs were dissolved in phosphate-buffered saline (PBS, pH 7.2), and their purity was analyzed by visualization on a 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel using Coomassie Blue staining. SDS-PAGE gel images were captured on an IQuant Capture 350 system and analyzed by ImageQuant TL 9 (GE Health Care). The protein concentration was determined by UV spectrophotometry using the molar extinction coefficient 5,690 M^-1 cm^-1 of a single Trp residue at 280 nm. The molecular weights (MWs) of [V_{7 20} or [RGD-V_{6 20} were computed by the Compute pI/MW software available from the ExPASy Proteomics Server and were 59,545 and 58,044 Da, respectively. Throughout this study, all experiments including biological assays were carried using the non-reduced, as-purified form of ELPs 12 .

The N2a cell line (ATCC CCL-131) was used as a test model for in vitro neuronal cell culture. N2a cells were maintained as a monolayer in EMEM medium (Gibco) supplemented with 10% (w/v) FBS, 2 mM glutamine, 1 mM sodium pyruvate, 1.5 g/l sodium bicarbonate, 100 unit/ml penicillin, and 100 μg/ml streptomycin at 37°C in a humidified atmosphere of 95% air and 5% CO₂. For biological assays, N2a cells at 60–80% confluence were plated out in specified culture plates at a density of 10⁴–10⁶ cells per well.

Cell adhesion was measured by a hexosaminidase activity assay as described previously 20 . Wells in a 96-well polystyrene plate (SPL Life Science) were treated with 100 μL of [V_{7 20}, [RGD-V_{6 20}, and fibronectin (R&D Systems) solutions (at concentrations of 0.1, 0.2, 0.5, 1, 2, 5, or 10 μM) at 4°C. After overnight protein adsorption, the wells were rinsed 3 times with 100 μl PBS (pH 7.2, Gibco) and blocked with 100 μl 0.5% heat-inactivated (60°C for 1 h) BSA for 1 h at 37°C. The cells were treated with trypsin and suspended in the culture media at a density of 3 × 10⁵ cells/ml; 100 μL of the cell suspension was then added to each well and incubated with 100 μl EMEM containing 2% FBS for 30 min. After incubation for 1 h at 37°C, the medium was removed, and the wells were rinsed twice with PBS to remove unattached cells. After rinsing, attached cells were incubated with 50 μl citrate buffer (50 mM, pH 5.0) containing 3.75 mM p-nitrophenyl-N-acetyl-β-D-glucosaminide (Sigma) and 0.25% Triton X-100 for 30 at 37°C. The reaction was stopped by adding 50 μl glycine buffer (50 mM, pH 10.4) containing 5 mM EDTA, and the absorbance was measured at 405 nm in a Titertek microplate reader. The adsorption data were fit to a Langmuir model for single species adsorption to a single site on the substratum, A = A_maxC/(K _d + C), where A is the absorbance at 405 nm, C is the coating concentration, A_max is the maximum absorbance, and K _d is the coating concentration required to achieve half-maximal absorbance 13 .

A 96-well plate was coated with 100 μl of the protein solution (20 μg/ml) at 4°C overnight and blocked with 100 μl of heat-inactivated 0.5% BSA for 1 h at 37°C. N2a cells were seeded (2 × 10⁴ cells/well) and incubated in 100 μl EMEM containing 10% FBS at 37°C in a 5% CO₂ incubator for 24 h. After removing the supernatant, the wells were washed 3 times with 200 μl PBS, and the adherent cells were fixed and stained in 1 step with 500 μl of 0.03% crystal violet (w/v) in 20% methanol for 10 min. Using a Nikon Ts-100 light microscope, phase contrast images (at least 20 cells in an image) were taken from 3 different cell locations for each experimental sample. The number of spreading cell population was determined by manual counting of the cells showing extended cell body, i.e., ratio of long to short diameter is greater than 2. And cells that exhibited lamellipodia were also scored as spread cells. Percentage of spreading cells was defined as (spreading cells/total number of adherent) X 100.

A scratch assay was used to measure cell population motility 21 . The wells in a 6-well plate were treated with 2 ml of the protein solutions (20 μg/ml) at 4°C overnight and were blocked with 0.5% heat-inactivated BSA at 37°C for 1 h. N2a cells (5 × 10⁵ cells/well) were seeded and incubated with 1.2 ml EMEM supplemented with 10% FBS for 24 h at 37°C. After removing the FBS media and washing the adherent cells with 2.0 ml PBS, a cell-free gap was created by a scratch to the confluent cell layer by using a sterile Gilson 1 ml pipette tip. Three non-overlapping regions of the scratched area were taken at 0 h by using a Nikon EcliTCPSe TE 300 microscope. The cells were incubated at 37°C in EMEM containing 0.5% FBS with mitomycin C (10 μg/ml, Sigma) to block proliferation. Phase contrast microscopic images of each scratched zone were taken at a designated time during the 30 h culture period. Cell migration index (CMI) was defined as [(N_t2 - N₀) - (N_t1 - N₀)]/(t₂ - t₁), where N_t2 and N_t1 are the number of cells detected in the scratched zone at 2 different culture times, and N₀ is the number of cells at 0 h. The number of cells in the scratched area was determined by manual counting of the spots in the microscopic images.

A 12-well plate was coated with 500 μl [RGD-V₆]₂₀ (10 μM or 600 μg/ml) at 4°C overnight. After removing the protein solution, wells were washed twice with PBS. N2a cells were seeded (5 × 10⁵ cells/well) and incubated at 37°C in a 5% CO₂ incubator for 48 h in 3 ml EMEM containing 10% FBS. Total RNA was isolated from the conditioned N2a cells using Trizol Reagent (Invitrogen). cDNA was synthesized from 4 μg of total RNA by using a High-Capacity cDNA reverse transcription kit according to the manufacturer’s instructions. qRT-PCR was performed using a SYBR Green PCR master mix kit (Applied Biosystems) in an ABI 7500 Real Time PCR System. The cycling conditions were as follows: 50°C for 2 min; 95°C for 10 min; 40 cycles of 95°C for 15 s and 1 min at 60°C. The primer sets were designed using the Primer Express 3.0 software based on the sequences from GenBank, and are shown in Table 1. The housekeeping gene GAPDH was used as an internal standard. Reaction specificity was confirmed by melting curve analysis.

Lab-Tek chamber slides were treated with 100 μl [V₇]₂₀ and [RGD-V₆]₂₀ (0.1 or 5 μM) and 5 μM fibronectin solutions at 4°C overnight. The culture plate was rinsed 3 times with 200 μl PBS twice and was blocked with 100 μl of 0.5% heat-inactivated BSA for 1 h at 37°C. N2a cells were plated at a density of 1 × 10⁴ cells per well in EMEM containing 10% FBS and was then cultured for 5 days. The cells were fixed with 4% (v/v) paraformaldehyde for 15 min at room temperature, washed 3 times in PBS buffer, permeabilized in 0.3% Triton X-100 (v/v) for 10 min, and then washed with PBS buffer. Non-specific binding was blocked with 5% normal goat serum for 30 min. The cells were then incubated with anti-tubulin beta III isoform at a 1:100 (v/v) dilution for 2 h at room temperature. After 3 washes in PBS, cells were incubated in Alexa Fluor 488- or 568-conjugated anti-mouse IgG secondary antibody (Invitrogen) for 2 h at room temperature in the dark. Nuclear sections were counterstained with DAPI (Sigma). The preparations were then mounted in Prolong Gold antifade reagent (Invitrogen). Using a Leica DMI 3000 fluorescence microscope, fluorescence images (at least 100 cells in an image) were taken from 3 different cell locations for each experimental sample. The percentage of neurite formation was determined by manual counting of the cells bearing neurites longer than the cell’s diameter.

Data are expressed as mean ± S.D obtained from at least three independent experiments. The statistical significance of the differences between groups was determined by one-way analysis of variance (ANOVA) and the Tukey post-hoc test. All statistical analyses were performed using GraphPad InStat (Ver. 3.05). Statistical significance was set at P < 0.05 or P < 0.01.