We derived a fibroblast line from a skin biopsy from a healthy adult male (HUF1) () and used immunocytochemistry to characterize the expression of cell surface markers commonly expressed on pluripotent stem cells (). Unexpectedly, we observed that, even prior to reprogramming, the HUF1 line possessed cells that demonstrated heterogeneous expression of stage specific embryonic antigen 3 (SSEA3; ). SSEA3 is a cell surface glycosphingolipid considered an embryonic/pluripotency marker 
. Overlaying phase contrast and SSEA3 immunofluorescence images revealed that the SSEA3 expression was detected across the entire cell surface () and using confocal microscopy we observed that the SSEA3 expression was primarily localized to the cellular membrane (). Additional small and localized regions of SSEA3 fluorescence were also detected around the peri-nuclear region, possibly reflecting the intracellular processing and packaging of SSEA3 on peri-nuclear endoplasmic reticulum and/or Golgi bodies (). Notably, in positive controls, strong cell surface expression of SSEA3 was observed in H9 human embryonic stem cells (hESCs)() and no expression was observed in the negative controls ().
Expression of SSEA3 from primary human dermal fibroblasts.
We next examined whether the expression of SSEA3 in a subset of fibroblasts was specific to HUF1 or a more general observation. Eight additional primary adult human fibroblast lines were derived from skin biopsies and immunoassayed. We observed that all eight lines contained a subpopulation of cells that were positive for SSEA3 (). Fluorescence activated cell sorting (FACS) analysis of HUF1 cells stained with the SSEA3/488 antibody complex, revealed a larger subpopulation of cells with little or no SSEA3 expression and a smaller subpopulation with detectable SSEA3 expression (). Subsequently, we isolated (through FACS) and cultured the top 10% and bottom 10% of the SSEA3/488 fluorescing cells as our SSEA3-positive and negative populations respectively (). Immunofluorescence analysis of the two populations, following overnight adherence to exclude analysis of non-viable cells, revealed that >97% of the SSEA3-positive population expressed detectable SSEA3/488 fluorescence and 0% of the SSEA3-negative population expressed detectable SSEA3/488 fluorescence (), demonstrating that the fluorescence activated cell sorting process can purify viable subpopulations of cells from a heterogeneous somatic population. These subpopulations were then used for reprogramming to iPSCs.
FACS analysis and isolation of SSEA3-positive and SSEA3-negative primary adult human fibroblasts.
Previous reprogramming work demonstrated that we could reprogram the entire, unsorted population of HUF1 somatic cells using retroviral vectors that express OCT4, SOX2, KLF4 and cMYC to generate iPSCs that express the same pluripotency markers as control H9 ESCs (). Reprogrammed cells possessed a normal karyotype (), differentiated into beating cardiomyocytes in vitro
) and differentiated into representatives of all three germ layers in vivo
(). We transduced our SSEA3-positive and SSEA3-negative populations with the same retroviral vectors, under identical experimental conditions, and seeded the transduced cells onto inactivated mouse embryonic fibroblasts (MEFs). After three weeks of culture under standard hESC conditions, plates were examined in a double-blind analysis by three independent hESC biologists for iPSC colony formation. Colonies with iPSC morphology were picked and expanded. We observed that all three biological replicates with the transduced SSEA3-negative cells formed many large background colonies (10–27 per replicate, ) but no iPSC colonies emerged; in contrast, all three biological replicates with the transduced SSEA3-positive cells resulted in the formation of iPSC colonies (4–5 per replicate, ) but very few large background colonies (0–1 per replicate, ). When we further characterized the cell lines derived from the iPSC-like colonies, we observed that they possessed hESC-like morphology, growing as flat colonies with large nucleo-cytoplasmic ratios, defined borders and prominent nucleoli (). When five lines were further expanded and characterized, all demonstrated expression of key pluripotency markers expressed by hESCs, which included: alkaline phosphatase, Nanog, SSEA3, SSEA4, TRA160 and TRA181 (). The SSEA3-selected iPSCs also demonstrated a normal male karyotype (46, XY)(), the ability to differentiate into functional beating cardiomyocytes in vitro
) and differentiate into representatives of all three germ layers in vivo
(). Most importantly, since we observed no iPSC colony formation or line derivation from the transduced SSEA3-negative cells, this suggests that these cells possess significantly lower or even no reprogramming potential relative to the SSEA3-expressing cells (). Additionally, a 10-fold enrichment of primary fibroblasts that strongly express SSEA3 resulted in a significantly greater efficiency (8-fold increase) of iPSC line derivation compared to the control derivation rate (p<0.05, ).
Characterization of unsorted HUF1 derived induced pluripotent stem cells (HiPS-1 control).
Morphology of colonies and lines following retroviral transduction of HUF1 cells.
Derivation of human iPSCs from SSEA3 sorted primary dermal fibroblasts.
Characterization of SSEA3-selected HiPSCs.
We next examined the expression of genes that might potentially confer the enhanced reprogramming to the SSEA3-positive population, including Nanog 
, Sall4 
and hTert 
as well as the control housingkeeping gene Gapdh. In addition to the SSEA3-positive and negative populations of cells, which represented the top 10% and bottom 10% of SSEA3 expressing cells respectively, we also included the intermediary SSEA3-expressing cells, which represented the remaining 80% of the total HUF1 cell population. Three biological replicates for each of the three subpopulations were analyzed. While no significant differences in gene expression were observed for Sall4, hTert or Gapdh (), the analysis revealed that expression of Nanog was significantly increased (p<0.05) in the SSEA3-positive cell population compared to either the SSEA3-intermediate or SSEA3-negative population ().
Transcriptional analysis of primary dermal fibroblast subpopulations with differential SSEA3 expression.