We have analyzed the growth of USSCs in three serum-free media formulations; HEScGRO, PSM and USSC growth medium^ACF. Figures 1A clearly demonstrate that cells cultured in HEScGRO and PSM do not grow, whereas cells grown in USSC growth medium^ACF proliferate at the same rate as cells grown in stem cell proliferation medium containing 30% fetal calf serum.

However, after one to three passages in USSC growth medium^ACF cells start to form small clusters, which become tight spheres after a further 24 hours (Figure 1B). These spheres are initially heterogeneous in shape, but with continued culturing, rounded spheres with defined borders are observed. Seven days after their first appearance, the size of the spheres varies. Figure 1C shows that the majority of spheres (64%) are small in diameter (<120 μm), 31% are medium (120-240 μm) and the remaining spheres (5%) are large (>240 μm). The smallest sphere observed was 30 μm, and the largest was 475 μm.

Figure 1D shows that the spheres can be mechanically dissociated and can reform spheres upon culturing in USSC growth medium^ACF. Evidence of proliferation was observed as spheres could be passaged and new spheres were observed to form.

As the culturing of stem cells as spheres is important for a number of applications, we set out to confirm whether USSCs could form spheres using one of the standard sphere formation methods 3132. Using the serum-free hanging drop method, we cultured cells either in USSC growth medium^ACF, or stem cell proliferation media without fetal calf serum. Figure 2A shows that cells grown in stem cell proliferation media without fetal calf serum in the presence or absence of bFGF do not form spheres, but cells grown in USSC growth medium^ACF always form spheres. There also appears to be a differential effect on the morphology of the sphere depending on whether bFGF is used, in that spheres derived using bFGF are smaller, but tighter, and have less of a halo.

To determine their structure, spheres were serially sectioned and stained with haematoxylin and eosin. Figure 2B shows representative images of a serial section of a sphere. The cells in the spheres appear homogenous in size and morphology and 48% of the spheres also contained cavities.

Altogether, these data suggest that USSC growth medium^ACF enables USSC proliferation, that this growth occurs simultaneously with the loss of the cells' ability to adhere to plastic, and that these non-adherent cells can continue to grow in suspension as spheres.

The ability of mechanically dissociated spheres to revert to monolayer growth and continue to proliferate was studied after culture in USSC growth medium^ACF on fibronectin or gelatin coated surfaces, or in the presence of stem cell proliferation media containing fetal calf serum. Figure 1D shows that the cells from the spheres re-adhere and morphologically resemble the parental USSCs in the presence of fetal calf serum, or when the surface is coated with extracellular matrix proteins. These cells were cultured for up to two weeks, during which they were passaged three times and expanded up to six-fold. This confirms that cells within the spheres, which have been cultured in USSC growth medium^ACF, can revert to monolayer growth and continue to proliferate.

To determine if day seven spheres express pluripotency-associated markers, the expression of Oct4 and Sox2 was determined by RT-PCR. Figure 3 shows that USSCs grown in adherent culture do not express Oct4, but do express Sox2, and that USSCs grown as spheres maintain this expression profile. In view of recent reports questioning accurate detection of Oct4 mRNA expression 33, we confirmed that the primers used in this study do not bind to Oct4B (isoform 2) and are specific to the cDNA of the nuclear-localized Oct4A (isoform 1). We also confirmed this result using antibody staining. The antibody staining clearly detects nuclear-localized Oct4 in hES cells, but expression could not be detected in USSCs nor seven day spheres (Figure 4).

To determine if the cells in the spheres initiate a differentiation program, we analysed markers associated with commitment of stem cells. Figures 3 and 5 show there is no change in the expression profile of markers associated with ectodermal (Pax6 and Nestin), mesodermal (Brachyury and Goosecoid) and endodermal (FoxA2, Cytokeratin 8 and Gata6) differentiation in day seven spheres when compared to USSCs. This expression profile is also maintained after five passages as determined by RT-PCR for Sox2, Pax6, Gata6, and Brachyury (Figure 6).

To confirm that the differentiation capacity of the cells in the spheres had been maintained; seven day spheres were dissociated and incubated in lineage-specific differentiation medium. Figure 7 shows that the cells within the spheres maintain the ability to differentiate and express markers representative of the three germline layers. Cells expressing neuronal-specific β-tubulin III (a marker of ectodermal cells), that stain positive for mineralization as detected by Alizarin Red S staining (a marker of mesodermal cells) and express epithelial-specific surfactant protein-C (a marker of endodermal cells) were all observed. The differentiation capacity of the cells is also maintained after five passages in USSC growth medium^ACF (Figures 8 and 9).

Cord blood was collected with informed consent from mothers undergoing elective Caesarean section. The protocol was approved by the University of Melbourne and the Royal Women's Hospital Human Ethics Review Committees. After the delivery of the baby, the cord was clamped and the blood was collected from the umbilical cord vein in bags containing 21 ml citrate phosphate dextrose (Macopharma, Mouvaux, France).

The USSC lines were successfully generated as described by Kögler et al. 1. Briefly, the mononuclear cell fraction was separated on a Ficoll gradient (density 1.077 g/cm³, GE Healthcare, Uppsala, Sweden) at 400 g for 25 min followed by lysis of erythrocytes in ammonium chloride buffer (0.14 M ammonium chloride, 0.25 mM ethylene-diamine-tetra-acetic acid (EDTA), 10 mM tris(hydroxymethyl)aminomethane pH7.4) and two washing steps in calcium- and magnesium-free phosphate-buffered saline (PBS), pH 7.4. To initiate the growth of the adherent USSC colonies, the mononuclear cell suspension was plated at 2 × 10⁶ cells/cm² in initiation medium containing low-glucose Dulbecco's Modified Eagle's Medium (Lonza, Walkersville, MD, USA) supplemented with 30% fetal calf serum (Hyclone, Logan, UT, USA), 10^-7 M dexamethasone (Sigma, St Louis, MO, USA), 100 U/ml penicillin (Invitrogen, Chadsworth, CA, USA), 0.1 mg/ml streptomycin (Invitrogen) and 2 mM Ultraglutamine (Lonza).

Cells were incubated at 37°C in 5% CO₂ in a humidified incubator. Every five days, the non-adherent fraction was removed and fresh medium was added. Initiation medium was used until the first colonies appeared or for a maximum of three weeks. Expansion of the cells was carried out using initiation medium without dexamethasone (stem cell proliferation medium). After reaching 80% confluency, the cells were subcultured using 0.25% trypsin/EDTA (Invitrogen), and replated at a 1:2 dilution in stem cell proliferation medium. Low passage cells were frozen down at 1 × 10⁶ cells/ml in 10% dimethyl sulfoxide (Sigma) and 50% fetal calf serum (Hyclone). For each experiment, cells were thawed and used before they reached passage ten.

The phenotype of the lines was as previously published 18. All the lines were negative for CD34, CD24, CD31, CD45 and MHCII, and positive for CD44, CD71 and MHCI. The multilineage capacity of the lines was confirmed by differentiation into cells representative of the three germline layers.

Cells were cultured in three different media; PSM, HEScGRO™ or USSC growth medium^ACF. PSM; is a basal medium containing low-glucose Dulbecco's Modified Eagle Medium (Cambrex BioScience), 100 U/ml penicillin/0.1 mg/ml streptomycin (Invitrogen), and 2 mM Ultraglutamine (Cambrex BioScience) with or without 10 μM of S1P (Biomol, Exeter, UK) and 20 ng/ml PDGF (Peprotech, NJ, USA) as described in Pebay et al., 2007. HEScGRO™ (Millipore) was prepared according to the manufactures instructions. USSC growth medium^ACF (Stem Cell Sciences, UK) was supplemented with 10 ng/ml recombinant human basic fibroblast growth factor (bFGF) (Millipore) before use.

Cells were seeded at 5.2 × 10⁴ cells/cm² and split 1:2 when they reached 80% confluency. However, after one to three passages, the cells cultured in USSC growth medium^ACF grew as non-adherent spheres without requiring any intervention to initiate aggregation. To maintain sphere proliferation, spheres were dissociated mechanically or using 0.25% trypsin/EDTA every four days, and allowed to reform in USSC growth medium^ACF. Cells were cultured for five passages and then characterized further. If spheres are not dissociated within a week they become cystic and resistant to disaggregation by trypsinization.

Spheres were also generated using the hanging drop formation 31. Briefly, a total of 1 × 10⁴ cells were placed in 30 μl of USSC growth medium^ACF (plus or minus bFGF), or in stem cell proliferation medium without fetal calf serum (plus or minus bFGF), and dropped onto the inside of a Petri dish lid. Drops were incubated upside down in a Petri dish that contained PBS in the base to prevent evaporation.

To confirm the cells could revert to monolayer growth, dissociated and non-dissociated spheres were grown in stem cell proliferation medium that contains fetal calf serum. To determine if the cells could adhere while cultured in USSC growth medium^ACF, dissociated and non-dissociated spheres were plated in Petri dishes coated with 1% gelatin (Sigma) in PBS, or 10 μg/ml fibronectin (Becton Dickinson, Franklin Lakes, NJ, USA).

Multipotency of the USSCs and the spheres was confirmed by differentiation into bone-, neuronal- and epithelial-like cells. Spheres were dissociated with 0.25% trypsin/EDTA (Invitrogen) and plated on 0.1% gelatin (Sigma) in PBS-coated plates overnight before induction.

To test for mesodermal differentiation capacity of the cells, bone differentiation assays were performed. Cells were plated at 7 × 10³ cells/cm² cultured for ten days in stem cell proliferation medium containing 10^-7 M dexamethasone (Sigma), 10 mM glycerol-6-phosphate (Sigma) and 50 μg/ml ascorbic acid (Sigma) as described by Jaiswal et al. 53. Medium was replaced every five days. To confirm successful mineralisation, cells were fixed for ten minutes with 70% ethanol at 4°C, and stained for ten minutes with 1% Alizarin Red S (Sigma) in distilled water, pH 4.2. The cells were then washed five times in distilled water, layered in PBS (minus calcium and magnesium) and phase images were captured as described below.

To demonstrate the neuronal differentiation capacity of the cells, the differentiation protocol described by Dottori et al. was adapted for USSCs 54. Briefly, cells were plated on 10 μg/ml poly-D-lysine (Sigma), 10 μg/ml fibronectin (Becton Dickinson) coated glass coverslips and cultured in neural basal medium consisting of Neurobasal A with 2% B-27, 1% insulin-transferrin-selenium-A, 1% N2, 2 mM L-glutamine, 100 U/ml penicillin and 0.1 mg/ml streptomycin (all sourced from Invitrogen). Neural basal medium was supplemented with 20 ng/ml basic fibroblast growth factor (R&D Systems, Minneapolis, MN, USA) and 20 ng/ml epidermal growth factor (R&D Systems) prior to adding to the cells. Media was changed every two to three days, for a total culture period of 12 days at 37°C with 5% CO₂ in a humidified incubator. Differentiated cells were fixed in 4% paraformaldehyde (BDH, Dorset, United Kingdom) for ten minutes at 4°C and immunostained overnight with 1 μg/ml of the mouse IgG₁ anti-human β-tubulin III (Invitrogen); or 1 μg/ml mouse IgG₁negative isotype control (Dako, Glostrup, Denmark). Cells were stained with 2 μg/ml of the goat anti-mouse Alexa568 or 488 IgG secondary antibody (Invitrogen) and counterstained with 0.5 μg/ml of 4, 6-diamidino-2-phenylindole (DAPI, Sigma) mounted in Vectashield (Vector Laboratories, Burlingame, CA, USA). Fluorescent images were captured as described below.

To test for epithelial differentiation capacity of the cells, cells were grown in Small Airway Growth Medium (SAGM, Lonza) as described by Berger et al. 55. Cells were seeded at 5 × 10⁴/cm² in stem cell proliferation medium the day before the addition of SAGM and then were differentiated for eight days and medium was changed after four days. Differentiation was assessed by nested RT-PCR for Surfactant Protein C (SPC). RNA extraction and PCR protocols are detailed below. The PCR conditions for SPC were as described by Berger et al. 55, and the human bronchial epithelial cell line, 16HBE14o-, was used as a positive control for SPC expression 56.

Total RNA was isolated from 5-10 × 10⁵ trypsinized USSCs and from mechanically dislodged pools of spheres using RNeasy Mini Kit (Qiagen, Germantown, MD, USA) and RNase-Free DNase I Set (Qiagen) according to the manufacturer's instruction, and quantified using a NanoDrop 1000 Spectrophotometer. Reverse transcripts were prepared from 500 ng of total RNA using the SuperScript III First-Strand Synthesis System (Invitrogen), and cDNA was amplified using the Taq PCR Core Kit (Qiagen). Briefly, 0.4 μM of of forward and reverse primers, 0.25 U of Taq, 1× Q-solution, 1× PCR buffer containing 1.5 μM of MgCl₂, 200 μM of dNTPs and 2 μl of cDNA were used. Table 1 contains the primer sequences, amplification conditions and expected RT-PCR product sizes of all genes investigated. The PCR conditions were as follows: 94°C for 2 minutes; 28-38 cycles of 94°C for 30 seconds, 30 seconds annealing at 51-61°C, 72°C for 1 minute; and a final extension step at 72°C for 2 minutes.

Microscopy was performed on a fluorescent inverted microscope (Carl Zeiss AG, Oberkochen, Germany). Images were captured using Axiom photo system (Carl Zeiss).