*This website was produced as an assignment for an undergratuate course at Davidson College.*

Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibrolast Cultures by Defined FactorsKazutoshi Takahashi and Shinya Yamanaka
Cell 126, 663-676, August 25, 2006

Premise of Study, Basic Approach Taken, and Summary of Findings:

Previous to the publication of this paper, it was known that somatic cells could be reprogrammed to exhibit stem cell properties, such as pluripotency.  This reprogramming was achieved by transferring the nuclear components of these somatic cells with oocytes, or by fusing them with embryonic stem (ES) cells.  Thus, the authors hypothesized that oocytes and ES cells “contain factors that can confer totipotency or pluripotency to somatic cell” (Takahashi et al., 2006).

The authors based their research on this hypothesis, and set out to find the factors that were responsible for the reprogramming.  The authors began with 24 genes that served as candidate factors, and tested the effects of the introduction of varying combinations of these genes into mouse embryonic fibroblasts (MEFs).  In order to determine whether these MEF cells had become induced pluripotent (iPS) cells under varying circumstances and with varying combinations of the 24 factors, the authors:

Analysis of the data listed above, as well as analysis of induced tail-tip fiber (TTF) cells, allowed the authors to determine which cells had been induced into pluripotent states, both morphologically and through expression of ES marker genes.  Four factors were found to be necessary for induced pluripotency: Oct3/4, Sox2, c-Myc and Klf4.  The authors were surprised that Nanog was not required for induction, as it has been shown to be important for maintainence of pluripotency in ES cells.

The findings of this research could have a large impact.  It is believed that stem cells could help provide cures for diseases such as Parkinson’s disease and diabetes.  However, study and use of stem cells has been limited, because, at the time of this study, human stem cells needed to be extracted from human embryos.  The ability to essentially create stem cells from adult, somatic cells would circumvent the use of embroys and ensure that stem cell research could be pursued.

Analysis and Critique of Data:

Figure 1: Generation of iPS cells from MEF cultures and Detection of Pluripotent State


Figure 2: Narrowing Down the Candidate Factors:
In order to determine which factors were necessary for the induction of pluripotency, cells were transduced with combinations of the 24 factors and their resistance to G418 (indicating activation of Fbx15/pluripotency) was assessed.

Figure 3: Gene-Expression Profiles of iPS cells

Though in many cases, iPS cells expressed ES marker genes, expression levels were very low for some genes, even in some iPS cells containing all 10 factors. Examples of low expression are found in iPS-MEF10 clone 3 for Gdf3, Soc2, Cripto. Other examples include iPS-MEF4 clone 7 for ERas and Sox2.

For the Oct3/4 promoter, methylation levels were high in MEF but significantly lower in ES, iPS-MEF4-7 and iPS-MEF10-6 cells. This indicates that the Oct3/4 promoter is much more readily available for transcription in ES and iPS cells than in MEF cells. Furthermore, acetylation levels of the Oct3/4 promoter were significantly higher in ES and iPS cells than in MEF cells.

Similar patterns were seen with the Nanog promoter, further indicating that iPS cells had been induced to express ES-related genes, whereas MEF cells were preventing these ES genes from being transcribed.

Figure 4: Global Gene-Expression Analysis

This figure shows global gene-expression via DNA microarrays of these cells types:

The figure shows a Pearson correlation analysis, and displays relative expression levels of genes, with red indicating increased expression compared to median levels, and green indicating decreased expression.  Overall, expression levels iPS-MEF4-7 and iPS-MEF10-6 cells showed the most similarity to ES cells.  iPS-MEF3-3 cells also showed some similarities, but not to the same degree.  Genes common up-regulated in ES cells and all iPS cells included Myb, Kit, Gdf3, and Zic3 (group I in panel 4b).  Genes up-regulated more efficiently in ES cells, iPS-MEF4, and iPS-MEF10 than iPS-MEF3 cells included Dppa3, Dppa4, Dppa5, Nanog, Sox2, Esrrb, and Rex1 (Group II in panel 4b).  This ones again points to the inability of iPS-MEF3 cells to completely take on ES cell qualities, as has been exhibited with iPS-MEF3 cells’ morphology.

Though iPS-MEF4 and iPS-MEF10 cells did express many genes at similar levels as ES cells, many genes were up-regulated more prominently in ES cells than iPS-MEF4 and iPS-MEF10 cells.  These genes include Dnmt3a, Dnmt3b, Dnmt3l, Uft1, Tcl1, and the LIF receptor gene (Group III in panel 4b).

Though there are some clear correlation patterns seen in figure 4, it would be helpful if the authors had pointed out what regions of DNA correspond to what gene, so that readers could draw conclusions themselves on the matter of relative expression of specific, ES-related genes.

Figure 5: Determining Pluripotency of iPS cells by Examining Differentiation

Here, the authors examine tumors derived from injected mice, and determine whether they have been able to differentiate into all three germ layers.

Tumors were obtained with 5 iPS-MEF10 clones, 3 iPS-ME4 clones, 1 iPS-MEF4wt clone, and 6 iPS-MEF3 clones.  Clones with tumors that differentiated into all three germ layers included 2 iPS-MEF10 clones (3 and 6), 2 iPS-MEF4 clones (2 and 7), and three iPS-MEF4wt clones.

Data not shown indicates that iPS-MEF10-6 “could give rise to all three germ layers even after 30 passages,” as confirmed by immunostaining and RT-PCR. Some iPS-MEF10 and iPS-MEF4 clones, however, were not able to differentiate into all three germ layers – some were not able to differentiate at all. Additionally, no iPS-MEF10-3 clones showed signs of differentiation.

The authors claim, “These data demonstrate that the majority of, but not all, iPS-MEF10 and iPS-MEF4 clones exhibit pluripotency”(Takahashi et al., 2006).  Although some examples of differentiation are shown, the authors do not show pluripotency in all the clones they claim can achieve it – they only show iPS-MEF4-7 and iPS-MEF10-6.  The data is compelling, but does not allow the reader to decide if they agree that the majority of these clones exhibited pluripotency.


Figure 6: Characterization of iPS Cells Derived from Adult Mouse Tail-Tip Fibroblasts

Next, the authors introduced the four factors into mouse tail-tip fibroplasts (TTFs).  One set of cells (iPS-TTF4) were obtained from a 7-week old male mouse transgenic for the Fbx15βgeo/βgeo construct.  Another set of cells (iPS-TTFgfp4) were obtained from a 12-week-old female (also transgenic for the Fbx15βgeo/βgeo construct) which constitutively expressed GFP from the CAG promoter.  In addition, one clone was obtained in which the four factors were flanked with two loxP sites in the transgene.  The authors hoped to use this loxP site to determine if iSP induction could occur in the absence of transgenic gene products, but were unable to, because there were too many loxP sites on multiple chromosomes, which would’ve led to inter and intrachromosomal rearrangements.


Figure 7: Further Chracterization of Expression in iPS Cells

Real-time PCR data not shown shows that endogenous expression levels of Oct3/4 and Sox2 (two ES marker genes that showed limited expression in previous RT-PCR expression gels) was lower in iPS cells than in ES cells.  The authors point out that total amounts of Oct3/4 and Sox2 expressed (including both endogenous and transgenic products) were higher in iPS cells than in ES cells.

Levels of endogenous Oct3/4 RNA are detectable in all undifferentiated iPD cells, but less than in ES cells. No endogenous Oct3/4 RNA is detected in differentiated iPS cells, but transgenic RNA is present. Total levels of Oct3/4 RNA were higher in iPS cells than ES cells. Oct3/4 protein was detectable in all iPS and ES cells, but not in MEF or TTF cells.

Sox2 showed the same patterns as Oct3/4, except in the ratio of endogenous/transgenic RNA. Some clones did not show any endogenous RNA for Sox2.

Nanog RNA was detected for all cells except for differentiated iPS-MEF4-7 cells. Nanog protein was present in small but detectable levels in all iPS cells.



Given the data, the authors conclude that Oct3/4, Sox2, c-Myc and Klf4 are essential for the generation of iPS cells.  These factors may work together, suppressing certain cell functions and enhancing others such that all necessary ES genes can operate and induce pluripotency.  The authors suggest mechanisms by which these factors may work together to induce pluripotency:




In their discussion, the authors note concerns about the origins of the iPS cells studied.  This concern stems from the fact that, of all their transfected cells, only a small portion became iPS cells.  They worry these could’ve just been multipoint stem cells (which have been isolated before).  But, because they received even fewer iPS cells from bone marrow, which should technically contain more of these multipoint stem cells, they do not believe the iPS cells the obtained were naturally occurring multipoint stem cells.  In addition, because they saw so much variation across their different constructs, it seems as though the iPS properties were due to their constructs and not from randomly extracting already existing iPS cells.

Possible reasons for low frequency of iPS cell derivation include the fact that levels of factors needed for pluripotency induction may have narrow range (especially in light of the theory that these factors work together to induce pluripotency), as well as the fact that minor chromosomal alterations may occur during transfection.

Proposed Further Research:

The data presents a compelling case for the ability to create iPS cells capable of differentiating into a variety of tissues.  Among the most compelling data presented is that of figure 6D, in which varying cell types are shown to contain iPS cells.  While the authors are able to show that this is possible in mice cells, the rate at which pluripotency occurred was very low when you consider how many cells they transfected with the four factors proposed to be sufficient for induced pluripotency.

Some of their data also proved to be troubling, especially their analysis of methylation.  While gene products were apparent in RT-PCR gels, methylation levels of necessary genes still appeared to be fairly high – a circumstance which does not occur in ES cells.  Further studies should investigate these methylation levels, and determine how pluripotency is achieved if these genes are still fairly protected from transcription by the cell.

As the authors mentioned, the construct in which the four factors were flanked by loxP sites could be refined in such a way that will shed light on how much these iPS cells rely on transgenic proteins, as opposed to endogenous ones.  This could be done by regulating the activity of the cre enzyme, which cleaves at loxP sites.  A reporter complex would be necessary in order to determine which factors had been removed with the cre/loxP interaction.  Given the heavy methylation of some necessary genes, including Oct3/4 (one of the four factors introduced), it would be important to know whether or not iPS cells could adequately produce Oct3/4 once the transgenic Oct3/4 product was removed.

Additionally, the authors were not able to outline any mechanism by which these factors interact to induce pluripotency.  In order to gain more information on possible mechanisms, the four factor proteins could be tagged and followed within iPS cells.  Information gained from these experiments could lead to more refined studies, more capable of identifying specific mechanisms.

Because stem cells may be able to cure human diseases, it is also important that future studies be conducted to test whether it is possible to create human iPS from adult, somatic cells.  Presumably, such a study could use this paper as a road map for transfecting cells with the four factors.

Though these future studies could help push forward knowledge of stem cells, the work presented by Takahashi and Yamanaka provides a fantastic foundation, and presents an exciting possibility to the scientific community.


Takahashi, K., Yamanaka, S. Induction of Pluripotent Stem Cells from Mouse Embryonic and Adult Fibroblast Cultures by Defined Factors. Cell 2006; 126: 663-676.


Back to Alexis' Homepage

Molecular Biology Homepage

Davidson College Homepage

Please direct questions or comments to alvalauriorton@davidson.edu