B₆D₂-F₁ (C57BL/6J ♀ × DBA/2J ♂) mice were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and Charles River Laboratories (Wilmington, MA, USA) and housed at the Research Animal Resource Center of Weill Medical College of Cornell University, in a temperature- and light-controlled room on a 14 h light:10 h dark photoperiod with food and water ad libitum.

Females (7–9 weeks) were superovulated with 10 IU of pregnant mare's serum gonadotrophin (Sigma, St. Louis, MO) then 10 IU of hCG (Sigma) 48 h apart. They were next placed with males of the same strain and mating was confirmed by presence of a vaginal plug the following morning. Zygotes collected approximately 20 hours after hCG were cultured in KSOM^AA medium (Specialty Media, Phillipsburg, NJ) at 37°C with 6% CO₂ in humidified air. Our ESC medium consisted of Dubelcco's Modified Eagle's Medium (DMEM, Invitrogen, Carlsbad, CA, USA), 10% fetal bovine serum (FBS, HyClone, Logan, UT, USA), and 10% newborn calf serum (NCS, HyClone) supplemented with mouse leukemia inhibitory factor (LIF, 2000IU/ml, Chemicon International, Temecula, CA), 0.1 mM nonessential amino acids (NEAA, Specialty Media), 0.1 mM β-mercaptoethanol (β-ME, Sigma), and 50 U/ml penicillin/50 μg/ml streptomycin (Sigma). This investigation also utilized Knockout™ DMEM (Ko-M, Invitrogen) with 15% Knockout ™ serum replacement (Ko-S, Invitrogen) containing LIF, NEAA, β-ME, antibiotics, and 4 mM L-glutamine (Sigma) as the supplementary medium.

Feeder layers of mitomycin C-treated MEFs were prepared on either 0.1% gelatin treated 4-well culture dishes (Nalge Nunc International, Rochester, NY) or in 30 μl microdrops under oil in tissue culture dishes. Intact expanded blastocysts (controls) and also zona-free blastocysts and ICMs isolated after immunosurgery or microdissection (see below), were plated onto MEF monolayers and cultured in ESC medium in 6% CO₂ at 37°C. Blastocysts were observed every 24 h for hatching and attachment of the trophoblast to a single MEF layer, while monitoring ICM size in two dimensions. On day 4 or 5 after plating (D4 or D5), ICM of at least 100 μm were dissociated using a glass pipette, followed by trypsinization in PBS containing 250 U/ml trypsin and 1 mM EDTA (PBS-try-EDTA) to promote cell dispersion. After re-plating in fresh wells coated with feeder cells, ESC colonies developed in 2 or 3 days after which they were trypsinized and propagated by passaging every 2–3 days thereafter. All established cell lines were tested for mycoplasma using MycoAlert ™ Mycoplasma Detection Kit (Cambrex, Rockland, ME, USA).

Intact blastocysts were plated onto feeder layers on post fertilization day 4–5. Zona-free blastocysts were obtained from incubation of day 4 blastocysts for 4–5 min in 24 U/ml pronase (Sigma). These were washed x2 and cultured for at least 1 h in medium with serum supplementation. To obtain ICMs, zona-free blastocysts were incubated with heat-inactivated rabbit anti-mouse serum (Sigma) diluted with DMEM (Invitrogen) in 1:25 for 15 min then exposed to guinea pig complement (Cedarlane laboratories, Hornby, Ontario, Canada) at 1:10 dilution for 30 min in 6% CO₂ in air at 37°C (Fig. 1). The disrupted trophoblast was then removed by aspiration through a hand-pulled 50–60 μm (inner diameter) pipette 3. Laser-assisted ICM isolation was performed as follows: Blastocysts were secured by two holding pipette with the ICM being positioned at 9 o'clock. Once adequate tension was established, approximately 10 infrared laser pulses (300 mW × 1 ms, ZILOS-tk™, Hamilton Thorne Research, Beverly, MA USA) were fired to split the blastocyst into two unequal portions – the smaller consisting of ICM, the larger consisting exclusively of trophoblast (Fig. 2).

The laser dissected ICM components were randomly plated either on MEF feeder layer or directly on gelatin-coated dishes, and cultured in Ko-M with 15% of Ko-S (serum free) containing LIF, NEAA, β-ME, antibiotics, and L-glutamine as described above. Intact blastocysts on MEF were used as controls. ICM size was measured daily, ICM dissociation and ESC propagation was carried out as described earlier. Pluripotency was evaluated according to cell size, nucleus/cytoplasm ratio, and nucleolar patterns. ESC colonies were graded according to number, density, and quality. Colony character was judged under an inverted microscope equipped with phase-contrast optics and stratified as follows: good (GG), average (GA), or poor (GP) according to colony composition of >70%, 40–70%, and <40% pluripotent cells, respectively (Fig. 3). Typically, pluripotent cells are large and have a high nucleus/cytoplasm ratio with one or more distinct nucleoli. DAPI banding was performed on metaphase chromosomes for ploidy assessment.

For this investigation, ESC pluripotency was determined by two positive markers: alkaline phosphatase (AP) activity 814 and Oct-4 15161718. TROMA-1 monoclonal antibody (Ab) directed against cytokeratin-like filaments present in trophectoderm and endodermal cells was used as a negative marker 8192021. All markers were tested on control blastocysts. Specimens were fixed with 2% paraformaldehyde (Sigma) and permeabilized with 0.5% Triton X-100 (Sigma). AP activity in fixed cells was detected using an azo-dye technique with a Texas-Red filter under fluorescence. The stain solution contained naphthol AS-MX phosphate (Sigma) and fast red TR salt (Sigma) 22. To detect expression of Oct-4 and TROMA-1, after incubation at 20°C for 60 min in 0.1% bovine serum albumin (BSA, Sigma) and goat serum, ESCs were exposed to Oct-4 polyclonal Ab (Santa Cruz Biotechnology, Santa Cruz, CA) at 1:100 dilution and to monoclonal TROMA-1 Ab (Developmental Studies hybridoma Bank, Iowa City, IA) at 1:6 dilution. After rinsing unbound Abs with PBS/BSA, specimens were next incubated with Alexa Fluor^® 488 (Invitrogen) or FITC conjugated secondary Abs (Chemicon). The nuclei were counterstained with 0.125 μg/ml of DAPI (Molecular Probes, Eugene, OR) in antifade solution or Topro-3 (Molecular Probes) at 1:500 dilution. Specimens were examined via phase-contrast, fluorescent, or laser confocal microscopy.

Pluripotency confirmation of putative ESCs was by morphological criteria, specific molecular markers, and expression of typical marker genes. Nanog (a divergent homeodomain protein that directs propagation of undifferentiated cells) is down-regulated during early de-differentiation and becomes silent in completely differentiated cells 232425 while transthyretin (Ttr, a protein in visceral yolk sac endoderm in vivo) 262728 is expressed in differentiated endoderm cells 2930. RNA was isolated from ESCs by Absolutely RNA^® Nanoprep Kit (Stratagene, La Jolla, CA) with MEFs serving as a negative control. RNA was stored at -80°C for qualification analysis. Primers were custom-designed by OligoPerfect Designer software (Invitrogen) for the target sequences of Nanog and Ttr genes, while Act-β and Gapdh were used as normalizers. Quantitative real-time PCR (qRT-PCR) was performed using SuperScript™ III Platinum^® Two-Step qRT-PCR Kit with SYBR^® Green (Invitrogen). Analysis was performed using an ABI Prism 7900 HT (Applied Biosystems, Foster City, CA). The qPCR results were plotted by the Sequence Detection System Analysis Software (Version 2.0, Applied Biosystems). Gene expression was reported as ratio data, calculated from the cycle threshold against Act-β considered at 100% expression 31.

ESCs (n ~10⁷) were subsequently treated with trypsin at 37°C in 5% CO₂ for 5–7 min, resulting in detachment/release of intact ESC colonies from underlying cells. Cell aggregates of approximately 50–60 ESCs were placed in 20 μl hanging droplets (DMEM, 20% FBS, 2 mM L-glutamine, 0.1 mM NEAA, antibiotics, and 1 mM β-ME) on a 100 mm² non-tissue culture lid flipped over a dish containing 10 ml PBS. ESCs were cultured for two days to allow aggregation into spheroid embryoid bodies (EBs). On the third day EBs were transferred to a 60 mm bacteriological petri dish in 5 ml of DMEM + 20% FCS, then after two days placed in gelatin coated dishes for five days 32. Differentiation into myocardiocytes was determined by presence of pulsatile contractility at day 8–9 of culture, at which point the EBs were processed for histological sections and transmission electron microscopy.

Induction of testicular teratoma formation was used to assess differentiation capacity of experimental cell lines. Approximately 1 to 4 × 10⁶ undifferentiated cells were injected into the testes of six- to eight- week-old severe combined immunodeficiency (SCID) mice (C.B-Igh-1^b/IcrTac-Prkdc^scid, Taconic, Germantown, NY, USA). Three to five weeks later tumors were fixed in 4% paraformaldehyde, embedded in paraffin, and examined histologically after hematoxylin and eosin staining.

A χ² test was utilized for comparisons involving blastocyst culture, attachment, ESC derivation, and gene expression analysis. A two-tailed test was used to assess significance, considered at 5% probability. Statistical comparisons were reported in text and tables only when significance was reached. Data tables show numbers within rows with different superscripts as significantly different. All statistical computations were performed with StatView 512+ (BrainPower Inc., Calabasas, CA, USA).

In a series of experiments, 34 mated mice produced a total of 613 zygotes, 566 (92.3%) of which developed to blastocyst stage after four days of culture. Five preliminary studies were performed in which 126 intact blastocysts were plated on MEFs and cultured in DMEM supplemented with serum. Two days later, 82 (65.1%) ICMs had hatched and attached to the feeder cell layer. On plating day 5 (D5), 71 ICMs (56.3%) were successfully dissociated and replated, However, no ESC lines were established. Optimal medium for murine ESCs was determined from nine experiments where 126 intact blastocysts were allocated to DMEM + 20% serum (Fig 4a–e) vs. Ko-M + 15% Ko-S (Table 1) (Fig.4f–j). In both media, over 60% of the plated blastocysts attached to the MEF layer and it was possible to isolate between 20 to 30% of the ICMs. While only 1 ESC line (1.6%) was harvested from the standard DMEM system, 11 (17.5%) were obtained from the Ko-M medium (p < 0.01). The culture carried out in microdrops yielded comparable hatching and attachment rates, although it facilitated easier monitoring of ICM attachment and assessment of development than the 4-well plate system. Although gelatin treatment was technically difficult for microdrop stabilization and positioning, gelatinization of the area before preparation of the microdrop areas made this easier.

ICM attachment ability and consequent ESC colony harvest according to the different embryo manipulation methods vs. control (n = 84) are depicted in Table 2. Zona removal enhanced ICM attachment and became significant (p < 0.05) when operative isolation of the ICM was carried out. Embryo manipulation including operative isolation of the ICM did not affect harvesting rate nor quality and grading of ESC colonies. For these studies, morphology was a consistent and reliable criterion for evaluating ESC characteristics, the morphological characteristics of stemness correlated closely with cell markers and gene expression assays. A proportion of GG colonies were not influenced by culture systems or ICM isolation method. Colonies graded as GG had four normal male and one female karyotype. All established ESC lines remained euploid after several passages. Colony pluripotency was confirmed by molecular markers. Outcomes using AP and Oct-4 were in agreement (χ² = 0.105) for determining pluripotency of derived ESC lines. TROMA-1 served as a negative control for endodermal and trophoblastic cell contamination (Fig. 5). Pluripotency marker assessment was concordant in all ESC colonies (Table 3). Multimarker assay confirmed the ability of the morphological criteria and grading system to identify pluripotent cells. We processed 10,000 to 30,000 cells to generate 0.3 to 0.9 μg of total RNA that displayed two ribosomal subunits. Reverse transcription and qRT-PCR were successful in all cases. As expected, Nanog was expressed in 74.6% of ESCs, but not in any MEFs. The latter manifested a higher expression (69.6%, p = 0.0001) of Ttr (denoting differentiation) compared to ESC colonies (40.7%).

Using morphological determinants and molecular markers, at least 80% of ESC colony cells were undifferentiated. Culture in the absence of feeder cells and LIF permitted formation of tight, round aggregates. After day 2 of culture, a distinctive outer layer of endodermal cell differentiation became apparent, each droplet containing a spherical embryoid body of ~320 μm diameter (Fig. 6). Reaching ~450 μm diameter by day 9 of culture, about 20% EBs displayed foci of pulsatile contraction typical of cardiomyocyte differentiation (Fig. 6). Histology and ultrastructural examination confirmed cell components of primordial embryonic germ layers (Fig. 7). EBs cultured for 16 days displayed widespread epithelial cell differentiation as well as cells lining the lumen of cavity found between the EBs. Both simple and stratified epithelial cells were present. Testes injected with ESCs uniformly developed tumors containing derivatives of all three embryonic germ layers including mucus-producing columnar epithelium, ciliated columnar epithelium, glandular epithelium arranged in acini, endothelium lined vessels within intraluminal neutrophils, striated muscle, bone, cartilage, neural tissue (often forming rosettes), and keratinizing epithelium (Fig. 8).

Intact blastocysts (n = 131) were laser dissected and 65 of these were plated on MEF with the remainder (n = 66) on gelatin-coated dishes (Table 4). Laser dissection was successful in all cases. In the absence of MEF, attachment of ICM occurred at a similar rate than control (81.8% vs. 92.3%), and observed growth patterns were also comparable. A total of 6 (9.1%) ESC lines was established on gelatin, while 15 (23.1%) on MEF. ESC harvest efficiency was significantly higher on feeder cells (p < 0.05). Cell lines obtained in feeder cell-free conditions maintained pluripotency to the same degree as that observed in controls