Regulation of c-Myc Expression by Ahnak Promotes Induced Pluripotent Stem Cell Generation*

We have previously reported that Ahnak-mediated TGFβ signaling leads to down-regulation of c-Myc expression. Here, we show that inhibition of Ahnak can promote generation of induced pluripotent stem cells (iPSC) via up-regulation of endogenous c-Myc. Consistent with the c-Myc inhibitory role of Ahnak, mouse embryonic fibroblasts from Ahnak-deficient mouse (Ahnak−/− MEF) show an increased level of c-Myc expression compared with wild type MEF. Generation of iPSC with just three of the four Yamanaka factors, Oct4, Sox2, and Klf4 (hereafter 3F), was significantly enhanced in Ahnak−/− MEF. Similar results were obtained when Ahnak-specific shRNA was applied to wild type MEF. Of note, expressionof Ahnak was significantly induced during the formation of embryoid bodies from embryonic stem cells, suggesting that Ahnak-mediated c-Myc inhibition is involved in embryoid body formation and the initial differentiation of pluripotent stem cells. The iPSC from 3F-infected Ahnak−/− MEF cells (Ahnak−/−-iPSC-3F) showed expression of all stem cell markers examined and the capability to form three primary germ layers. Moreover, injection of Ahnak−/−-iPSC-3F into athymic nude mice led to development of teratoma containing tissues from all three primary germ layers, indicating that iPSC from Ahnak−/− MEF are bona fide pluripotent stem cells. Taken together, these data provide evidence for a new role for Ahnak in cell fate determination during development and suggest that manipulation of Ahnak and the associated signaling pathway may provide a means to regulate iPSC generation.

It has been well established that the so-called Yamanaka factors, Oct4 (octamer-binding transcription factor 4), Sox2 (SRY (sex determining region Y)-box 2), Klf4 (Kruppel-like factor 4), and c-Myc, together stimulate the generation of induced pluripotent stem cells (iPSC) 3 from diverse types of somatic cells including embryonic fibroblast cells (1)(2)(3)(4)(5). Oct4 belongs to the family of POU transcriptional factors and binds to an octamer sequence motif, AGTCAAAT, leading to expression of target genes in developing germ cells (6,7). Sox2 is a member of Sox transcriptional factor family containing the high mobility group domain and is known to be required for maintenance of diverse types of stem cells throughout life (8). KLF4 is a member of Kruppel-like factor transcriptional factor family that binds to CACCC sequence regulating gene expression during embryonic development (9,10). c-Myc is a well known basic helixloop-helix transcriptional factor that plays important roles in cell cycle regulation, cellular transformation, and maintenance of pluripotency (11,12).
Ahnak is an exceptionally large protein (ϳ700 kDa) and has an extended central region composed of 36 repeat units that serve as a scaffolding motif for interaction of phospholipase C-␥ with PKC in regulation of cell proliferation and migration (13)(14)(15)(16). Moreover, the central repeated unit is known to bind to R-Smad protein in response to TGF-␤ and mediate down-regulation of c-Myc expression and cell growth retardation. Specifically, nuclear localization of Ahnak in complex with phospho-Smad3 and binding to promoter of c-Myc were strongly enhanced in response to TGF-␤, indicating that Ahnak downregulates the expression of c-Myc as Smad3 target genes. Consistently, Ahnak-deficient MEF cells show a significant up-regulation of c-Myc expression (17).
There have been efforts to exclude c-Myc from the mix of ectopically expressed transcription factors for iPSC generation mainly because of its potential transforming activity. Exclusion of c-Myc, however, led to a dramatic reduction in efficiency (18 -20). Therefore, developing a method to efficiently utilize endogenous c-Myc activity for iPSC generation could represent an important progress. Here, we present that Ahnak knock-out MEF or wild type MEF expressing shRNA specific to Ahnak can be converted to iPSC at a high efficiency without infection of c-Myc retrovirus. Further dissection of Ahnak pathway may provide novel strategies for regulation of iPSC generation and reducing the potential danger from cellular transformation.

Materials and Methods
Isolation of Mouse Embryonic Fibroblasts-MEF were isolated from embryos at embryonic day 13.5. The uterus was dissected out from each mouse and rinsed with PBS. After isolation of each fetus, the head and internal organs were removed to isolate only the trunk, which was subsequently finely minced by ejection through a 10-ml syringe. The mixture of cells and small tissue masses was incubated with 5 ml of trypsin-EDTA at 37°C with shaking for 30 min. Digestion was terminated with addition of FBS, and cells were resuspended in fresh medium (DMEM, 10% FBS) and transferred to 150-mm culture dishes. The MEF were grown to 70 -80% confluence prior to passage. MEF from two-or four-passage cultures were used in all experiments.
Production of Retrovirus Encoding OCT4, SOX2, KLF4, or c-Myc-The pMXs-IRES-Puro-based retroviral expression vectors for mouse OCT4, SOX2, KLF4, and c-Myc were obtained from Addgene (Cambridge, MA). For knockdown of Ahnak in mouse and human cells, an shRNA targeting Ahnak in a microRNA scaffold (miR-shRNA) was utilized (21). A long duplex oligonucleotide containing an Ahnak targeting sequence, 5Ј-ATCTCCATGCCTGATGTGG-3Ј, within the mir-30 context was synthesized in pGEM-T Easy cloning vector (Bioneer Inc., Daejeon, Korea). To target human Ahnak for the control, a scrambled sequence was utilized. The construct was ligated 5Ј to IRES-GFP in LZRS retroviral vector as previously described.
HFF derived iPSC were generated by three factors (OSK) in combination with retrovirus encoding scrambled control shRNA or Ahnak-specific shRNA. HFF cells were cultured in DMEM containing 10% FBS. At day 2, about 5 ϫ 10 5 cells were infected with concentrated viral preparations with 8 g/ml Polybrene (Sigma-Aldrich). The next day, the medium was replaced with fresh medium and cultured for another 3 days. At the end of day 5, cells were transferred to mitotically inactivated MEF. Cells were subsequently cultured in DMEM containing 10% FBS for another 2 days, and the medium was replaced with human ES cell medium (DMEM-F12 with 20% KOSR (Knock-Out Serum Replacement, Gibco). Colonies were AP-stained and counted 14 days after the viral transduction.
Alkaline Phosphatase (AP) Staining-To detect AP activity, the cells were washed with PBS, fixed with 4% paraformaldehyde in PBS for 10 min at room temperature and washed with distilled water. They were stained with 5-bromo-4-chloro-3indolyl-phosphate/nitro blue tetrazolium color development substrate kit (Promega, Fitchburg, WI) according to the manufacturer's protocol. The stained cells were photographed, and the colonies were counted.
Embryoid Body (EB) Formation and in Vitro Differentiation-J1 mouse ES or iPSC colonies were harvested by trypsinization and suspended in mouse ES medium without leukemia inhibitory factor. The single cells were cultured in suspension in untreated Petri dishes to generate embryoid bodies. After 7 days, aggregated cells were plated onto gelatincoated coverslips in 12-well plates and incubated for another 14 days. The cells were stained with anti-␤III-tubulin (Millipore), anti-smooth muscle actin (clone 1A4; Dako, Glostrup, Denmark), and anti-␣-fetoprotein (Dako).
Immunofluorescence Staining-J1 mouse ES cells and iPSC were plated in 12-well plates on coverslips. Cells were fixed with 4% paraformaldehyde in PBS for 10 min at room temperature, washed with PBS, and then permeabilized in 0.5% Triton X-100 for 10 min. After blocking with PBS containing 1% bovine serum albumin for 1 h at room temperature, the cells were incubated with primary antibodies against anti-Oct4 (H-134; Santa Cruz Biotechnology, Santa Cruz, CA), anti-Sox2 (Millipore), and anti-stage-specific embryonic antigen 1 (SSEA1; Santa Cruz Biotechnology) in PBS overnight at 4°C. After washing three times with PBS and incubating with the secondary antibodies, rhodamine-conjugated goat anti-mouse IgG (KPL, Gaithersburg, MD), or Alexa Fluor 594-conjugated goat anti-rabbit IgG (Life Technologies, Inc.) for 1 h at room temperature in the dark, cells were counterstained with DAPI cells and examined with a LSM510 meta confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany).
Teratoma Formation and Histological Analysis-iPSC (Ahnak Ϫ/Ϫ -iPSC-3F-8) were suspended at 1 ϫ 10 7 cells/ml in DMEM containing 10% FBS. Nude mice were anesthetized with diethyl ether. Injection of 100 l of the cell suspension (1 ϫ 10 6 cells) was applied subcutaneously into the dorsal flank. Four weeks after the injection, tumors were surgically dissected from the mice. The samples were weighed, fixed in PBS containing 4% formaldehyde and embedded in paraffin. The sections were stained with hematoxylin and eosin for histological examination.
Statistics-The data are expressed as the means Ϯ S.D. or Ϯ S.E. of values from three to five independent experiments. Statistical analyses were performed with a two-tailed Student's t test. p values of Ͻ0.05 were considered statistically significant. All Western blots in the figures were representative of three independent experiments.

Mutually Exclusive Expression of Ahnak and c-Myc-We
have previously demonstrated that Ahnak plays an inhibitory role in c-Myc expression through TGF␤-Smad3 signaling cascade in epithelial cells (17). To evaluate whether Ahnak regulates c-Myc expression in embryonic fibroblasts, we examined Ahnak knock-out (Ahnak Ϫ/Ϫ ) MEF. It was readily seen that c-Myc expression was significantly enhanced in Ahnak Ϫ/Ϫ MEF compared with that in wild type MEF (Fig. 1A). Because c-Myc is expressed in ES cells and is ectopically expressed for efficient iPSC generation, we sought to examine the possible utility of Ahnak in induction of iPSC. We first questioned whether Ahnak expression is dynamically regulated during EB formation from ES cells. Indeed, Ahnak expression was sharply increased during EB formation from ES cells (Fig. 1B), whereas expression of Oct4, Nanog, and c-Myc showed changes in the opposite direction. These results are consistent with the idea that the inhibition of Ahnak may provide a favorable condition for maintenance of ES cells in undifferentiated states and possibly induce iPSC in the absence of expressing c-Myc, a proto-oncogene.
Efficient Generation of Ahnak Ϫ/Ϫ iPSC without Ectopic c-Myc-We sought to check whether ectopic expression of three of the Yamanaka factors (Oct4, Sox2, and Klf4) is suffi-cient for efficient iPSC formation in Ahnak Ϫ/Ϫ MEF without ectopic expression of c-Myc. Indeed, infection of Ahnak Ϫ/Ϫ MEF with the retroviruses encoding the three factors resulted in significantly increased iPSC generation compared with the level seen with wild type MEF (Fig. 2, A and B). This may have resulted from the fact that c-Myc expression was increased in Ahnak deficient cells. To evaluate the effect of c-Myc on iPSC generation, c-Myc gene was transfected in wild type MEF cells in various doses along with the other three transcription factors. Efficiency in iPSC generation clearly increased in a c-Myc dose-dependent manner (Fig. 2, C and D).
We next questioned whether knockdown of Ahnak has similar effects on iPSC generation. A miR-shRNA specific to Ahnak was constructed for retroviral delivery via infection of wild type MEF cells. We found that knockdown of Ahnak expression resulted in a sharp up-regulation of c-Myc expression as we saw in Ahnak Ϫ/Ϫ cells (Fig. 3A). Consistently, infection of wild type MEF cells with retrovirus expressing Ahnak-specific miR-shRNA led to a significantly increased level of iPSC generation compared with infection of control miR-shRNA retrovirus (Fig. 3,  B and C). The difference was seen either with all four Yamanaka factors or with three of them (i.e. without c-Myc) consistent with the c-Myc dose-dependent efficiency of iPSC generation. We next examined whether knockdown of Ahnak has similar effects in human fibroblasts. As was the case in MEF, we saw clear increase in c-Myc protein level and a concomitant increase in AP-positive colonies (Fig. 3, D and E), indicating that the Ahnak regulation can be applied in human iPSC generation as well.
To further examine the pluripotency of iPSC from wild type and Ahnak Ϫ/Ϫ MEF cells generated either with four or three of Yamanaka factors, we analyzed the expression of representative markers for endoderm, mesoderm, and ectoderm after differentiation. The iPSC were induced to form EBs and then examined by immunocytochemistry with antibodies to ␤III tubulin (for ectoderm), smooth muscle actin (for mesoderm), and ␣-fetoprotein (for endoderm). We found that all three markers were expressed in WT-iPSC-4F-#19 and Ahnak Ϫ/Ϫ -iPSC-4F-#31, as well as in WT-iPSC-3F-#6 and Ahnak Ϫ/Ϫ -iPSC-3F-#8, indicating pluripotency of the four different iPSC lines we generated (Fig. 5A). To evaluate transcriptional expression of three germ layer marker genes, expression levels of ectoderm marker genes (FGF5 and Nestin), mesoderm marker genes (Eomes, Brachyury, and Mixl1), and endoderm marker genes (Gata4 and Sox17) during EB formation (up to day 10) were investigated. Expression levels of representative marker genes for the three germ layers during EB differentiation of four established cell lines were shown to be similar to those seen with mES (Fig. 5B).
Finally, we tested whether Ahnak Ϫ/Ϫ -iPSC-3F-#8 cells can form teratoma. To this end, the cells were injected into immunodeficient SCID/beige mice, and induced nodules were histologically examined for the presence of all three embryonic germ layers (ectoderm, mesoderm, and endoderm). Indeed we found that neural-like tissues (Fig. 6, arrowheads in panels a, b, e, and f) and epidermis (Fig. 6, arrowheads in panels c, d, g, and h) representing ectoderm, cartilage (Fig. 6, arrowheads in panels i, j, m, and n) and muscle tissues (Fig. 6, arrowheads in panels k, l,  o, and p) representing mesoderm, and gut-like epithelium (Fig.  6, arrowheads in panels q-x) representing endoderm in nodules from seven of eight mice. Together with results from examination of EB, these data indicate that Ahnak Ϫ/Ϫ -iPSC-3F-#8 cells are bona fide pluripotent stem cells.

Discussion
Because the four transcription factors Oct4, Sox2, Klf4, and c-Myc have been shown to activate reprogramming of somatic cells to pluripotent stem cells, efforts have been continually made to improve the efficiency of the induction and safety of the iPSC application. The chief safety concern for iPSC is the potential for tumorigenesis. In particular, the inclusion of c-Myc, a powerful proto-oncogene, in the induction process has represented the major hurdle for medical application of these cells (2). Specifically, it has been reported that tumors from iPSC often show reactivation of exogenous c-Myc (2).
Multiple studies have reported successful iPSC generation without exogenous c-Myc. For example, generation of human iPSC has been successfully carried out with the combination of Oct4, Sox2, Nanog, and Lin28 (22). Also, N-Myc or L-Myc can replace c-Myc for the generation of iPSC (18,19). Inclusion of   UTF1 and siRNA for p53, as well as expression of poly(ADPribose) polymerase 1, also could replace c-Myc and lead to efficient iPSC generation (23,24). Still, studies involving direct comparisons demonstrated that efficiency of iPSC generation decreases in the absence of c-Myc (18 -20, 25).
Ahnak, a scaffold protein, binds to R-Smad, potentiates TGF␤-mediated cell signaling, and thereby induces down-regulation of c-Myc expression (13). Given that Ahnak expression is increased during the differentiation of ES cells to EB, whereas Oct4 and Nanog, representative markers of iPSC, are downregulated, it was possible to hypothesize that Ahnak is a gene with an anti-stemness property. Our success in improving the efficiency in the absence of ectopic expression of c-Myc via down-regulating Ahnak suggests another possible way of generating safer iPSC.
We have in fact found multiple reports indicating that inhibition of TGF-␤ signaling promotes induction of pluripotency (26 -29). This is consistent with our results that at least during the initial phase of induction, TGF-␤ signaling should be inhibited rather than enhanced for an efficient induction of pluripotency. It is likely that some TGF-␤ signaling is in taking place in ES or iPSC given that some p-Smad3 is detected (Fig. 4C). However, it is clear that the signaling is not strong enough to turn off c-Myc expression. It could be in this context that the p-Smad3 level we observed can be interpreted (Fig. 4C). In fact, given that Ahnak Ϫ/Ϫ mice are viable and that p-Smad3 is detected in Ahnak Ϫ/Ϫ cells, we could safely conclude that Ahnak-independent mechanism can mediate a certain level of TGF-␤ signaling, and this is not mutually exclusive with c-Myc expression. Alternatively, it is also possible that once iPSC is established, c-Myc expression attains a measure of independence from TGF-␤ signaling, which allows coexistence of TGF-␤ signaling and c-Myc expression in iPSC.
Importantly, the iPSC (Ahnak Ϫ/Ϫ -iPSC-3F-#8) generated from Ahnak Ϫ/Ϫ MEF without c-Myc could differentiate into all three germ layers in vitro and could generate teratoma efficiently in vivo. Although Ahnak plays an important role in many biological processes, deficiency of Ahnak apparently did not affect manifestation of pluripotency. That transient knockdown expression of Ahnak also improved iPSC generation without c-Myc implies the potential for a broad application of Ahnak regulation in generation and differentiation of iPSC. With the detailed molecular mechanism by which Ahnak regulates c-Myc in various context elucidated, Ahnak and molecular components of pathways Ahnak regulates could develop into important targets in regulating generation of safe iPSC with efficiency.