Scientists have demonstrated that it is possible to reprogram the genome from a fully differentiated adult human cell to a pluripotent state by transfetring its nucleus into an unfertilized human oocyte, as long as the oocyte’s own haploid genome is left in as well. A team at New York Stem Cell Foundation Laboratory (SCFL), the University of California at San Diego, and Columbia University, found that carrying out the modified somactic cell nuclear transfer (SCNT) process allowed cell development to progress to the blastocyst stage. Human stem cells derived from these blastocysts contained both a haploid genome derived from the oocyte and a diploid somatic cell genome that analysis demonstrated was reprogrammed to a pluripotent state.
The researchers claim their achievement provides new insights into why previous approaches to SCNT have failed and pointers toward negotiating the hurdle: Effectively, it seems that taking out the oocyte nucleus removes factors that are necessary for the transferred adult genome to be reprogrammed. When the somatic cell nucleus is used to replace the oocyte nucleus, development stops at the 6–10 cell stage and transcription is inhibited.
SCFL’s Scott Noggle, Ph.D., Dieter Egli, Ph.D., and colleagues, describe the technique and results in Nature in a paper titled “Human oocytes reprogram somatic cells to a pluripotent state.”
Born in 1996, Dolly the sheep was the ultimate proof that the nucleus from a fully differentiated adult cell could drive the process of generating a complete animal. The SCNT process used by Roslin Institute scientists involved placing a sheep mammary gland cell nucleus into an enucleated oocyte (from a different sheep) and implantating that oocyte into a surrogate mother.
Dolly’s birth combined with separate research demonstrating that pluripotent stem cells could be derived from human blastocyts, hinted at the promise of generating supplies of patient-specific stem cells for human regenerative medicine. However, Dr. Noggle et al. report, even though studies have continued along this path in humans, despite legal and ethical considerations associated with the use of human oocytes and embryos, “none have achieved the derivation of a stem cell line.”
The ability to induce pluripotent stem cell formation from differentiated cells by forcing the expression of transcription factors has provided an alternative approach to generating patient-derived stem cells, the authors admit. Nevertheless, research suggests that induced pluripotent stem cells (iPSCs) may differ from human embryonic stem cells (hESCs) in terms of gene expression, DNA methylation, and differentiation potential. iPSCs also have a tendency to develop de novo mutations and copy number variations, and studies indicate that they retain epigenetic memory of their origins.
To try and figure out why SNCT in humans has not yet been successful, the team needed a source of unfertilized oocytes. Admitting that very few women will agree to donate their oocytes for research without payment, the investigators developed protocols (that were approved by relevant boards of Columbia University), which allowed women participating in the reproductive egg donation program to choose between donation for either reproductive purposes or research.
They were provided with equal levels of payment, whichever they chose. “Consequently, the decision to donate was before and independent of their decision to donate for research,” the investigators state. As a result, 270 mature oocytes were collected for research.
Somatic cell reprogramming in a number of species has been achieved by replacing the oocyte genome at metaphase II (MII) of meiosis, with a somatic cell nucleus. Thirty five MII oocytes were thus enucleated and replaced with fluorescently tagged somatic cell genomes obtained from the skin cells of a male type 1 diabetes (T1D) patient or a healthy adult male.
All oocytes condensed the somatic chromosomes, and upon activation, 71% of 31 oocytes continued normal cleavage development. However, the team notes, “as we had previously observed following nuclear transfer into human zygotes, development arrested at a stage of 6–10 cells.”
They then tried leaving the oocyte genome in for a defined period of time after the GFP-tagged somatic cell nucleus was transferred. Six to eight hours after artificial activation two interphase nuclei had formed within each single cell, and at this point either the somatic cell nucleus or oocyte nucleus was removed. Interestingly, activated oocytes in which the somatic genome was left formed cells, but all arrested without reactivation of the GFP transgene. Conversely, activated oocytes left with just the oocyte genome cleaved: Four out of seven of these developed to the blastocyst stage and allowed the generation of pluripotent parthenogenetic stem cells.
Development of activated cells to the blastocyst stage could also be achieved when the oocyte genome was replaced by a blastomere. Parthenogenic development to the blastocyst stage resulted when the oocyte’s own genome was removed and then retransferred back into the same oocyte.
The 4–8 cells stage of development is the point at which extensive transcriptional activity by the zygote genome normally kicks in and is also the stage at which development arrests following somatic genome transfer, the researchers note. Transcriptome analysis interestingly showed that transcription profiles and transcript abundance at the 6–12 cell stage after genome exchange most closely resembled a state of inhibition of transcriptional activity. Further analysis suggested that after genome transfer, neither the transcriptional program of a somatic cell nor that of a blastomere was being expressed.
It was this finding that led the team to test whether developmental arrest after classical SNCT was caused by a basic inability of the somatic cell genome to be expressed and replicated or whether molecules specific to the oocyte genome were required to allow that expression and subsequent development. They therefore transferred the fibroblast nucleus to an oocyte, but this time left the oocyte genome in situ indefinitely. When this approach was taken, development to the blastocyst occurred in about 21% of the manipulated oocytes, “indicating that the somatic cell genome did not interfere with development to the blastocyst stage,” the authors remark.
They then derived two cell lines from the inner cell mass of resulting triploid blastocysts. SoPS1 contained the genome of a male type 1 diabetes patient, and soPS2 contained the genome of a healthy male adult. Both also contained the oocyte genome. A proportion of soPS1 developed some genetic changes over 23 passages, including translocations and additional copies of chromosomes 12 and 17. soPS2, on the other hand, was karyotypically stable over more than 20 passages. Both cell lines completed over 30 passages (over 100 population doublings) over six months, without replicative crisis.
Importantly, the soPS cell lines expressed molecular markers characteristic of pluripotent stem cells, and after differentiation in vitro or injection into immunocompromised mice, generated cell types representative of all three germ layers. “The global gene-expression profile of both soPS cell lines clustered closely with that of other pluripotent cell types including NYSCF1, a stem cell line derived from an IVF blastocyst,” the authors note.
In fact, over 1,327 genes were identified that were differentially expressed between soPS2 cells and their donor fibroblasts. “Among the genes with the most significant upregulation in soPS1 and soPS2 were genes typically expressed in pluripotent stem cells but not in fibroblasts, such as LIN28A, POU5F1, SOX2, NANOG, and LEFTY1,” they continue.
“Genes that were most downregulated included those typically expressed in fibroblasts, such as fibroblast activating protein (FAP), pappalysin (PAPPA), metallopeptidase (MMP3), a collagen triple helix-containing protein (CTHRC1), and a mesoderm-specific transcription factor (SNAI2).”
Using a genome-wide allelotyping approach and gene-expression profiling to compare gene expression from the haploid oocyte genome with gene expression from the effectively reprogrammed diploid somatic genome, the researchers could find no evidence that the reprogrammed somatic cell nucleus retained any epigenetic memory.
The authors claim their studies demonstrate “the feasibility of reprogramming human cells using oocytes and identifies removal of the oocyte genome as the primary cause of developmental failure after genome exchange.” They also suggest that given a reliable source of human oocytes, “it should be possible to overcome the requirement of the oocyte genome for somatic cell reprogramming, allowing the generation of diploid pluripotent stem cells.”