Quantitative Assessment of Neuronal Differentiation
in Three-dimensional Collagen Gels Using Enhanced Green
Fluorescence Protein Expressing PC12 Pheochromocytoma Cells

evaluating the morphological differentiation of neuronal cells in a three-dimensional (3-D) system to assist in quality control of neural tissue engineering constructs for use in reparative medicine. Neuronal cells tend to aggregate in the 3-D scaffolds, hindering the application of two-dimensional (2-D) morphological methods to quantitate neuronal differ- entiation. To address this problem, we developed a stable transfectant green fluorescence protein (GFP)-PC12 neuronal cell model, in which the differentiation process in 3-D can be monitored with high sensitivity by fluorescence microscopy. Under 2-D conditions, the green cells showed collagen adherence, round morphology, proliferation properties, ex- pression of the nerve growth factor (NGF) receptors TrkA and p75NTR, stimulation of extracellular signal-regulated kinase phosphorylation by NGF and were able to differentiate in a dose-dependent manner upon NGF treatment, like wild-type (wt)-PC12 cells. When grown within 3-D collagen gels, upon NGF treatment, the GFP-PC12 cells differentiated, expressing long neurite outgrowths. We describe here a new validated method to measure NGF-induced differentiation in 3-D. Having properties similar to those of wt-PC12 and an ability to grow and differentiate in 3-D structures, these highly visualized GFP-expressing PC12 cells may serve as

The PC12 cell line has been extensively used in neuroscience research as a neuronal model for studying neuronal signaling (Vaudry et al.2002) and has become the model of choice for the study of neuronal differentiation (Fujita et al.1989; Ravni et al.2006). When treated in conventional two dimensional (2-D) tissue culture conditions with nanomolar concentrations of nerve growth factor (NGF), PC12 cells stop dividing, elaborate neurite processes (outgrowths), and become electrically excitable (Fujita et al.1989). NGF biological effects are mediated by two types of receptors: the NGF high-affinity tyrosine kinase (TrkA) receptor (Kaplan and Miller2000) and the low-affinity neurotrophin receptor p75 (p75NTR), a member of the tumor necrosis factor receptor superfamily (Chao et al.1998). Activation of these receptors by NGF initiates differentiation of the cells with a time course of several weeks, characterized by processes such as proliferation arrest and initiation of the neurite outgrowths in the first 2 days of treatment; continuous, progressive elongation of the neurites for up to 3 weeks of treatment, including extensive sprouting (branching) and synapse connections between the neurites; and generation of neurite networks and ganglia-like cell organization due to aggregation of neuronal cell bodies (Fujita et al.1989).

H. Arien-Zakay: S. Lecht: P. Lazarovici (*)
Department of Pharmacology and Experimental Therapeutics,
School of Pharmacy, Faculty of Medicine,
The Hebrew University of Jerusalem,
P.O. Box 12065, Jerusalem 91120, Israel
e-mail: philipl@ekmd.huji.ac.il

A. Perets:B. Roszell :P. I. Lelkes :P. Lazarovici
School of Biomedical Engineering, Science and Health Systems,
Drexel University,
Philadelphia, PA 19122, USA

such as growth, branching, differentiation, and synapse formation were found to be influenced by different matrix- resident biophysical cues such as micro- and nanotopog- raphy (Foley et al.2005; Moxon et al.2007), including substrate stiffness (Leach et al.2007). Furthermore, PC12 cells were used to explore the microenvironmental adequacy for neuronal cells of new bioengineering materials such as agarose scaffolds (Yu et al.1999), peptide scaffolds (Holmes et al.2000), neuron\u2013microelectrode interfaces (Bieberich and Anthony2004; Moxon et al.2007), microchannel-containing substrates (Mahoney et al.2005), and microfabricated silicon-based nanostructures (Lopez et al.2006; Moxon et al.2007).

A major issue in neuroengineering is the investigation of the biocompatibility of scaffolds with neuronal cells, aimed at generating a three-dimensional (3-D) neuronal constructs for implantation in patients in order to repair defective neural pathways. To achieve this goal, it is important to use in vitro models in which it is possible to visualize cell growth in the 3-D construct. In general, the methods used to date include fixation of the neuronal- containing scaffolds and histological evaluation of the morphology of the fixed cells, allowing temporal\u201csnap- shots\u201d at defined time points, but not a continual real time monitoring of neuronal development. As an alternative, nondestructive fluorescent labeling of the living cells populating the 3-D construct may be advantageous. Such labeling methods must also be compatible with the histochemical methods commonly used for routine mor- phological analysis (Tohill et al.2004) and facilitate noninvasive visualization of neuronal development.

The goal of the present study was to generate a fluorescent PC12 cell model to be used in 3-D neuronal engineering in order to address fundamental issues of optimization of neuronal differentiation. In this study, we generated a stable PC12 cell line expressing the green fluorescent protein (GFP) from the jellyfishAequorea

species and can be easily detected by UV light and fluorescence microscopy. It also has the advantage of being a\u201cvital dye\u201d whose presence can be monitored in living organisms and unfixed tissue (Okabe et al.1997). It has the potential to elucidate nervous system development (Niell and Smith2004) and to clarify the complex biological process of neurite outgrowth regulation (Laketa et al.2007). In this study, using the GFP-PC12 cells, we defined a new morphological analysis for measurement of neuronal cell differentiation in a 3-D environment and validated this method according to a commonly used bioassay for NGF-induced differentiation under 2-D conditions.

PC12 cells (NIH clone, originally provided by Dr. G. Guroff, NICHD, NIH) were grown in 25 cm2flasks in Dulbecco\u2019s modified Eagle\u2019s medium (DMEM) supple- mented with 7% fetal bovine serum (FBS), 7% horse serum, 20 U/ml penicillin and 20\u03bcg/ml streptomycin, all purchased from Biological Industries (Beit Haemek, Israel). Cells were incubated at 37\u00b0C in 6% CO2(Katzir et al.

Transduced 293T (human kidney epithelial, ATCC, Manassas, VA, USA) cells were cultured in DMEM containing 400\u03bcg/ml of a neomycin analogue, G418 (Gibco BRL, Rockville, MD, USA), 10% FBS, 20 U/ml penicillin, and 20\u03bcg/ml streptomycin (Kosaka et al.2004). All experiments were carried out in a clean room, according to ISO7 requirements (10,000 particles/m3) using a P2+ facility and lentiviral vector generation and transduction according to NIH biosafety guidelines.

HIV-1-derived lentiviral vectors were generated by transient transfection of a human kidney 293T cell line by a three- plasmid expression system. The packaging plasmid encodes HIV-1 gag, pol, tat, and rev proteins; the envelope plasmid encodes the VSV/G glycoprotein gene; and the transfer vector (TK113) encodes a CMV promoter driving the expression of GFP (Kosaka et al.2004). YTK 113 (vector), VSV-G (packaging), and deltaNRF (envelope) plasmids were transiently transfected into 293T cells using a calcium phosphate transfection method (Kosaka et al.2004). All transfections were carried out using 60-mm dishes coated with poly-L-lysine where 1.5\u20132.5\u00d7106 293T cells were plated in 5 ml of complete medium 24 h before transfection. One hour before transfection, the medium was replaced with serum-free DMEM containing antibiotics. The DNA/ CaCl2solution was prepared and left at room temperature for 30 min before being added to the cells. The cells were then returned to the 37\u00b0C incubator (5% CO2) for 24 h when the medium was replaced with 7 ml of fresh complete medium. After 48 h, the medium was removed and filtered through a 0.45-\u03bcm filter to collect the viral particles secreted into the medium. The viral titer was adjusted to 4.0\u00d7107IU/ml by dilution.

cells/well in six-well tissue culture plates coated with 10% poly-L-lysine (Sigma-Aldrich, St. Louis, MO, USA). The cells were transduced with 1 ml of 4.0\u00d7107IU/ml TK113 viral particles (HIV-based vector from which the GFP reporter gene is constitutively expressed under the control of the cytomegalovirus promoter) and supplemented with 2 ml of PC12 medium. The cells were incubated for 3 days post transduction in a 37\u00b0C incubator, 6% CO2, at high humidity.

Flow cytometric analysis of the PC12-GFP/lentivirus- transduced PC12 cell population was performed using a FACS with fluorescence excitation at 488 nm wavelength (Simons et al.2006). In brief, the cell culture was dissociated from the dish by agitation and resuspended in PBS without magnesium and calcium, supplemented with 3% FBS and filtered through a 20-\u03bcm filter to generate individually dispersed cells. Cell density was adjusted to 20\u00d7106cells/ml. For analysis, 50,000 events were acquired on a FACSCanto (Becton-Dickinson, San Jose, CA, USA) flow cytometer using FACS Diva software (Becton- Dickinson). Evaluation of flow data was performed using FlowJo (Tree Star, Ashland, OR, USA). GFP-positive cells were collected under aseptic conditions in 5 ml FACS polystyrene tubes (BD Biosciences, Bedford, MA, USA) containing 400\u03bcl of FBS. Wild-type PC12 (wt-PC12) cells, at the same density, served as the negative control.

GFP-positive cells (3.5\u00d7105), aseptically sorted by the FACS, were returned to culture. After 7 days of culture, cloning of individual GFP-PC12 expressing cells was performed by limiting dilution. The cells were then diluted to a concentration of 1 cell/100\u03bcl and aliquoted into 480 wells of five 96-well plates. One day later, a well containing a single fluorescent cell was identified by fluorescence microscopy to ensure that each colony would be derived from a single cell. To promote proliferation, the GFP-PC12 cells were grown in PC12 growth medium supplemented with 10% conditioning medium collected from a confluent wt-PC12 cell culture. The clones were monitored every other day and split when the colony reached 50% confluence. After 5 weeks of expansion, the chosen clones were collected and frozen in PC12 medium supplemented with 5% dimethylsulfoxide (Sigma-Aldrich). For biological characterization, different parameters were evaluated over a period of several weeks in samples of each

clone and wild-type cells. GFP expression was estimated by fluorescence microscopy. Clone adherence was assessed by monolayer generation; nonadherent cells grew in suspen- sion or partially attached to the dish. The wt-PC12 cultures contained six different morphotypes (round, monopolar, bipolar, tripolar, multipolar, and fibroblast like). The selected GFP-PC12 clones showed a similar heterogeneous morphology.

wt-PC12 or GFP-PC12 cells were seeded on 24-well plates coated with 200\u03bcg/ml type I rat tail collagen (BD Biosciences) and cultured as described above.

Sandwiched 3-D type I collagen gels at a final concentra- tion of 1.5 mg/ml solution were constructed as previously described (Dietrich and Lelkes2006; Mondrinos et al.

calculated amounts of 10\u00d7 Dulbecco\u2019s PBS (Mediatech, Herndon, VA, USA), 1 N NaOH, cell culture grade H2O, and type I rat collagen (BD Biosciences) were mixed, and the pH was adjusted under sterile conditions to 7.4\u00b10.05. To generate the bottom layer of the gel, 400\u03bcl of the collagen solution was added to each well of a 24-well plate and allowed to polymerize for 15 min at 37\u00b0C. For the second layer, wt-PC12 or GFP-PC12 cells were transferred into one volume of serum-free medium, which was then mixed with the collagen solution. A 500-\u03bcl volume of this suspension was added on top of the first layer. The cells were evenly distributed by brief agitation of the culture plates. These were then incubated for 15 min in a hybridization oven (37\u00b0C). After complete polymerization of the various gel assemblies, 1 ml of growth medium, supplemented as described for the 2-D cell culture conditions, was added to the sandwich-like gel assemblies. The cells were then cultured for various periods of time in a conventional CO2incubator (6% CO2, 37\u00b0C), and the medium was replaced every 2 days.

The proliferation of wt-PC12 and GFP-PC12 cells was estimated daily over a period of up to 14 days using the Alamar blue (BD Biosciences) staining method essentially as previously described (Nikolaychik et al.1996). Briefly, in both 2-D and 3-D cell culture conditions, the wells were incubated for 4 h with the Alamar blue reagent (10%v/v in complete medium). At the end of the incubation period,