Availability of somatic SCs is crucial for regenerative medicine, disease modeling and to gain insight into SC properties. Here we present experimental strategies to reprogram, in vitro, differentiated adult cells into their corresponding expandable tissue-specific stem/progenitor cells by the transient expression of the single transcriptional co-activator YAP.
Here we present protocols to isolate primary differentiated cells and turn them into stem/progenitor cells (SCs) of the same lineage by transient expression of the transcription factor YAP. With this method, luminal differentiated (LD) cells of the mouse mammary gland are converted into cells that exhibit molecular and functional properties of mammary SCs. YAP also turns fully differentiated pancreatic exocrine cells into pancreatic duct-like progenitors. Similarly, to endogenous, natural SCs, YAP-induced stem-like cells (“ySCs”) can be eventually expanded as organoid cultures long term in vitro, without further need of ectopic YAP/TAZ, as ySCs are endowed with a heritable self-renewing SC-like state.
The reprogramming procedure presented here offers the possibility to generate and expand in vitro progenitor cells of various tissue sources starting from differentiated cells. The straightforward expansion of somatic cells ex vivo has implications for regenerative medicine, for understanding mechanisms of tumor initiation and, more in general, for cell and developmental biology studies.
Tissue-specific somatic stem cells (SCs) are critical for tissue renewal and repair after injury. The possibility to easily isolate and unlimitedly expand ex vivo somatic SCs represents a critical issue for potential regenerative therapies, as well as for SC applications in basic research and disease modeling. Progress in this direction, however, has been limited by the difficulty of capturing the SC state of various epithelial organs in vitro. Indeed, in several adult tissues resident SCs may not exist, or not be readily available, or their number and regenerative potential may be eroded by aging or disease conditions. In 2016, we started to fill this gap by reporting that expression of a single transcriptional coactivator, YAP (YES-associated protein) or its closely related protein TAZ (transcriptional activator with a PDZ motif), into terminally differentiated cells efficiently creates functional, expandable, non-tumorigenic, autologous cell populations that are operationally and molecularly indistinguishable from their corresponding tissue-specific SCs1. A pulse of sustained YAP or TAZ activity for few days is sufficient to induce the appearance of self-renewing somatic SCs. This is a stable condition that is no longer dependent on continuous transgene expression, as it can be transmitted through cell generations without further expression of ectopic YAP/TAZ1. The protocol presented here details the procedure used to generate de novo epithelial stem/progenitor cells of the mammary gland and pancreas, starting from differentiated cells of these tissues. This procedure fills a black box in the current reprogramming/transdifferentiation arena. Main efforts in these directions have indeed so far centered on cell transition to an induced pluripotent stem cell (iPSC) state, followed by conversion of these embryonic and pluripotent SCs into more differentiated cells. However, iPSCs are tumorigenic once introduced in adult tissues, raising the need of developing protocols for their complete and efficient differentiation2. However, this differentiation step, even when possible, comes at the price of long-term expandability, self-organization and organ repopulation potentials. These are essential attributes for organ regeneration that are in fact typical only of endogenous tissue-specific SCs and of the presently described YAP-induced SCs (ySCs). Similarly, direct transdifferentiation of one cell type into another by using cocktails of various transcription factors also generate differentiated cells that lack essential proliferative and stemness potential3.
The procedure described here also takes advantage of the recently introduced organoid technology, by which endogenous SCs can be expanded and differentiated ex vivo4. YAP-induced SCs may generate organoid-forming SCs even in experimental, biological or disease conditions in which endogenous SCs are not present. We would like to note that, at the difference with other reprogramming procedures, the type of cell plasticity imparted by YAP may correspond to the only form of the reversion to a SC-like status that occurs in living tissues. Reacquisition of SC-like traits has been associated with tissue repair or oncogenic activation5. Although dispensable for the homeostasis of several adult tissues, YAP and/or TAZ are absolutely essential for regeneration, tumor growth and expansion of somatic SCs in vitro1,6,7,8,9,10,11,12
All animal procedures were performed adhering to our institutional guidelines and approved by OPBA and the Ministry of Health
1. Generation of YAP-induced Mammary Stem-like Cells (yMaSCs)
NOTE: All media and solution compositions for section 1 are specified in Table 1.
2. Generation of yDucts
NOTE: All media and solution compositions for section 2 are specified in Table 2.
Generation of yMaSCs
An overview of the experimental strategy to reprogram primary mammary LD cells by transient expression of YAP is presented in Figure 1A. Primary mammary LD epithelial cells are purified by fluorescent-activated cell sorting13. Figure 1B represents a typical sorting procedure to obtain three distinct subpopulations: Basal cells (EpCAMlowCD49fhighCD61–), Luminal Progenitor (LP) cells (EpCAMhighCD49flowCD61+) and LD Cells (EpCAMhighCD49flowCD61–). Careful gating of the three subpopulations is essential to isolate a pure preparation of LD cells, that are fully differentiated and completely growth arrested when seeded in mammary gland colony forming conditions (see Figure 1C, left panel). Conversely, when induced to express exogenous YAP, LD cells start proliferating to form easily recognizable dense epithelial colonies in 5% basement membrane matrix suspension cultures (Figure 1C). The efficiency of reprogramming, attested around 3% for a typical experiment, can be scored by counting the number of colonies over the number of single cells originally seeded in basement membrane matrix suspension cultures (Figure 1D). Reprogrammed luminal cells (yMaSCs) can then be passage into 100% basement membrane matrix organoid culture conditions (see scheme in Figure 1A), self-organizing into complex organoid-like structures that develop around multiple lumens and display remarkable self-renewal ability even in absence of doxycycline (i.e. in absence of transgenic YAP expression) (Figure 1E). Histologically, yMaSC-derived organoids display a basal layer (K14 positive), facing the basement membrane matrix reconstituted ECM and a luminal layer (K8 positive), facing the lumen-like cavities within the organoid (Figure 1F). This architecture is indistinguishable from that of organoids formed by native MaSCs (Figure 1F).
Generation of yDucts
An overview of the experimental strategy to reprogram primary pancreatic acini by transient expression of YAP is presented in Figure 2A. Entire acinar clusters are isolated from the bulk of the pancreatic tissue by a combination of mild dissociation and size exclusion through filtration. A typical preparation is presented in Figure 2B. After isolation, the acinar cell clusters should appear as a suspension of exocrine acinar units of homogeneous size, with no contamination by endocrine Langerhans islets or fragments of the pancreatic ductal tree and minimal dissociation to single cells. Contamination by endocrine islets or ductal fragments is an indication of deficient selective filtration (step 2.1.10), possibly due to harsh handling; unwanted dissociation of acinar clusters to single cells might be due to excessive collagenase treatment or unbuffered activity of proteolytic enzymes released by the tissue, which can be curbed by additional SBTI treatment.
A typical acinar reprogramming experiment is presented in Figure 2C; within 5-7 days of culture in 3D Collagen-I based hydrogel in the presence of doxycycline, pancreatic acini derived from R26-rtTAM2/tetO-YAPS127A mice readily turn into duct-like clusters (that we named yDucts), composed by a thin monolayer of epithelial cells that proliferate around an expanding central cavity. The reprogramming efficiency, which is around 70% for a typical experiment, can be easily measured by scoring the number of duct-like clusters over the total number of seeded acini (Figure 2D). The negative control cells, that is R26-rtTAM2/+ cells or R26-rtTAM2/tetO-YAPS127A cells left without doxycycline, invariably remain as post-mitotic acinar clusters in these culture conditions, as previously reported1,14,15. Reprogrammed yDucts can then be passaged at the single cell level into Matrigel-based organoid culture conditions16 (see scheme in Figure 2A), displaying remarkable self-renewal ability even in absence of doxycycline (i.e. in absence of transgenic YAP expression) (Figure 2E).
Figure 1: Isolation of primary mammary LD cells and induction of mammary stem cells. (A) Schematic representation of the experimental procedure adopted to reprogram primary mammary LD cells. (B) Representative FACS-plots illustrating a typical sorting procedure to purify LD cells. i) dissociated cells are gated according to forward and side scatter for live cells (P1; blue); ii) population P1 is then further gated according to its Lin profile: the subpopulation of Lineage-negative cells (P2; grey) is selected, excluding Lineage-positive hematopoietic cells; iii) population P2 is then separated into an EpCAMhigh (P3; yellow + green) and an EpCAMlow (P6; red) subpopulations; iv) P3 and P6 are then further gated according to their CD61/CD49f profile into three subpopulations: EpCAMlowCD49fhighCD61– Basal cells (P7; red), EpCAMhighCD49flowCD61+ LP cells (P8; yellow) and EpCAMhighCD49flowCD61– LD cells (P9; green). (C) Images are illustrative of the ability of LD cells, infected with the indicated constructs, to form mammary colonies 15 days after seeding in mammary colony medium. Only YAP-expressing cells turn into colony-forming cells, whereas negative control cells (EGFP-infected) remain as growth-arrested single cells. Scale bar = 50 μm. (D) Quantification of the colony forming ability of the indicated cells, as in (C). Data are presented as mean + s.d. and are representative of five independent experiments, each with six technical replicates. (E) Representative image of YAP-reprogrammed mammary stem cell-like cells (yMaSCs) after 12 days into organoid culture conditions in fresh three-dimensional 100% basement membrane matrix hydrogel in the absence of Doxycycline. Scale bar = 100 μm. (F) Representative immunofluorescence images for the basal marker K14 (green) and the luminal marker K8 (red) of organoids derived from the indicated cells, after 12 days into organoid culture conditions. Scale bar = 10 μm. This figure is reproduced from Panciera et al., 20161. Please click here to view a larger version of this figure.
Figure 2: Isolation of primary pancreatic acinar cells and induction of pancreatic progenitors. (A) Schematic representation of the experimental procedure adopted to reprogram primary pancreatic exocrine acinar cells. (B) Representative image of primary pancreatic acini just after the isolation procedure (step 2.1.14). The acinar preparation should appear as a homogeneous suspension of acinar clusters, with minimal presence of single cells. Scale bar = 400 μm. (C) Representative images of primary pancreatic acini derived from R26-rtTAM2 (upper panels)or R26-rtTAM2; tetO-YAPS127A (lower panels) mice and cultured in 3-D Collagen I -based hydrogel for 5 days with or without Doxycycline (doxy), as indicated. Only YAP-expressing primary acini convert to cells growing as cyst-like organoid after Doxycycline addition. Scale bars = 70 μm. (D) Quantification of the ability of pancreatic acini to form ductal organoids upon transgenic YAP overexpression as in (C). Data are presented as mean + s.d. and are representative of five independent experiments, performed with four technical replicates. (E) Representative image of YAP-reprogramed ductal-like cells (yDucts) after three passages in fresh three-dimensional 100% basement membrane matrix hydrogel in the absence of Doxycycline. Scale bar = 130 μm. Please click here to view a larger version of this figure.
Isolation of primary mammary cells | |
Ca2+ chelating solution | Store at 4 °C |
EDTA | 0.02% w/V |
PBS | |
Collagen I coating solution | |
Acetic Acid 0.02N, pH 3,23 | |
Rat Tail Collagen (coating) | 1:50 |
Dispase solution | Store at 4 °C |
Dispase | 5 mg/ml |
PBS | |
Dissociation Medium | |
DMEM:F12 | |
Hyaluronidase Stock Solution | 400 U/mL |
Pen/Strep | 1x |
Stock solution collagenase I | 600 U/mL |
Haemolytic Solution | Store at 4 °C |
NH4Cl Solution | 1 parts |
TrisBase 20.6 g/L | 9 parts |
adjust pH to 7.2 | |
HBSS/PS | Store at 4 °C |
HBSS | |
Pen/Strep | 2x |
Hyaluronidase Stock Solution | Filter 0.2 µm, store at 4 °C |
Hyaluronidase from bovine testes (powder) | 2,000 U/mL |
Sodium phosphate Buffer 1M pH7.3 | |
NH4Cl Solution | Store at T. amb. |
H2O | |
NH4Cl | 7.1 g/L |
adjust pH to 7.65 | |
Sorting Solution | Filter 0.2 µm, store at 4 °C |
BSA | 0.1% |
EDTA | 1 mM |
HEPES pH 7 | 25 mM |
PBS | |
Wash Medium #1 | |
DMEM/F12 | |
Pen/Strep | 1x |
Wash Medium #2 | |
DMEM/F12 | |
FBS | 5% |
Pen/Strep | 1x |
Mammary 2D culture medium | |
DMEM/F12 | |
FBS | 2% |
heparin | 4 mg/mL |
L-Glutamine | 1x |
murine bFGF | 10 ng/mL |
murine EGF | 10 ng/mL |
Pen/Strep | 1x |
Induction and Passaging of yMaSCs | |
Mammary Colony Medium | |
DMEM:F12 | |
FBS | 5% |
heparin | 4 µg/mL |
L-Glutamine | 1x |
Matrigel (add immediately before seeding) | 5% |
murine bFGF | 20 ng/mL |
murine EGF | 10 ng/mL |
Pen/Strep | 1x |
Mammary Organoid Medium | |
Advanced DMEM:F12 | |
B27 | 1x |
GlutaMax | 1x |
heparin | 4 µg/mL |
Hepes | 1x |
human Noggin | 100 ng/mL |
murine bEGF | 20 ng/mL |
murine EGF | 50 ng/mL |
R-Spondin 1 | 1 µg/mL |
Table 1: Generation of yMaSCs. Composition of all different culture media and solutions required for isolation of primary mammary LD cells and induction of yMaSCs (section 1.)
Isolation of primary pancreatic acini | |
Acinar Culture Medium | |
BPE | 50 µg/mL |
BSA | 0.1% |
Dexamethasone | 1 µg/mL |
FBS | 0.1% |
ITS-X | 1x |
Pen/Strep | 1x |
SBTI | 0.2 mg/mL |
Waymouth’s Medium | |
Acinar Wash Medium | |
BSA | 0.1% |
Pen/Strep | 1x |
RPMI Medium | |
SBTI | 0.2 mg/mL |
Acinar Recovery Medium | |
Acinar Culture Medium | |
FBS | 30% |
Collagenase I solution A | |
Acinar Wash Medium | |
Stock solution collagenase I | 360 U/mL |
PBS/PS | Store at 4 °C |
PBS (Phosphate-buffered saline) | |
Pen/Strep | 1x |
Stock solution collagenase I | Store at -20 °C |
Collagenase, type I (powder) | 6,000 U/mL |
PBS | |
Passaging of pancreatic organoids | |
Collagenase I solution B | |
PBS 1x | |
Stock solution collagenase I | 240 U/mL |
Pancreatic Organoid Medium | |
Advanced DMEM/F12 | |
B27 | 1x |
gastrin | 10 nM |
human FGF10 | 100 ng/mL |
human Noggin | 100 ng/mL |
murine EGF | 50 ng/ml |
N-Acetylcysteine | 1.25 mM |
Nicotinamide | 10 mM |
Pen/Strep | 1x |
R-Spondin 1 | 1 mg/mL |
SBTI | 0.2 mg/mL |
Table 2: Generation of yDucts. Composition of all different culture media and solutions required for isolation of primary acinar cells and induction and passaging of yDucts (section 2.)
Here we present protocols to reprogram ex vivo terminally differentiated epithelial cells of different tissues into their corresponding tissue-specific progenitor cells (or ySCs) by transient expression of YAP, as reported previously1. We have detailed two procedures: one allowing reprogramming of FACS-purified cells through lentiviral vectors and a second one that avoids viral infection and takes advantage of transgenic YAP expression. Each protocol presents an efficient strategy to isolate and culture primary differentiated cells and a strategy to force exogenous YAP gene expression in differentiated cells, generating de novo somatic tissue-specific expandable stem cells (see schemes in Figures 1A and 2A).
We demonstrated that the isolation strategies here presented effectively isolate a pure population of differentiated cells, as demonstrated by the fact that we never detected any outgrowth from negative control samples (Figures 1C and 2C).
The lentiviral vectors used in this study for the reprogramming of primary mammary LD cells are doxycycline inducible, offering the possibility of a tight control of the transgene expression; this allows to turn on and off exogenous YAP expression at will. Particular attention should be placed in avoiding the use of an excessive viral titer, as this can be detrimental in terms of reprogramming efficiency. In the case of primary acinar cells, we switched to a fully transgenic approach to obtain a YAP-dependent reprogramming with minimal manipulations. This latter strategy is also particularly appropriate for primary pancreatic acini, as isolated acinar clusters are scarcely amenable to lentiviral infection and very fragile. The transgenic strategy employed offers the same advantage of doxycycline-dependent lentiviral vectors for the tight control of gene expression. Moreover, the transgenic strategy exploited with primary pancreatic acini bears the additional advantage of a much higher reprogramming efficiency compared to the viral-induced reprogramming of mammary LD cells. Beyond the different intrinsic plasticity associated to cells derived from different tissues, the higher rate of pancreatic reprogramming might be derived from the higher efficiency of expression associated to uniform and autonomous YAP expression in all explanted cells. Notably, we have demonstrated that exogenous YAP is no longer required after generation of ySCs (yMaSC colonies and yDucts), without affecting their self-renewal capacity. This is because ySCs reactivate endogenous YAP/TAZ and use them for self-renewal when exogenous YAP is turned off1.
We validated the notion that ySCs indeed emerge from differentiated cells by controlling the cell of origin of our reprogramming experiments through genetic lineage-tracing validations1.
Extensive characterization of ySCs shows that YAP-induced reprogramming generates normal somatic SCs1 as i) at the transcriptomic level, ySCs display massive overlaps with native SCs; ii) ySCs display differentiation potential and can generate a multilineage progeny always restricted to the identity of their tissue of origin; iii) ySCs are non-transformed and non-tumorigenic when transplanted in vivo.
Here we also describe procedures to maintain and expand in culture both yMaSCs and yDucts as organoids embedded in 100% basement membrane matrix hydrogels. These conditions allow for the self-organization of ySCs into three-dimensional organoids that ensure the maintenance of stemness properties long-term in culture, enabling to expand these stem populations at will for downstream analyses and applications. For unknown reasons, we failed to obtain yMaSC organoids by placing infected LD cells directly in organoid culture conditions 7 days after doxycycline treatment in plastic tissue culture plate; in other words, the intermediate growth step in mammary colony conditions is essential. In our hands, even native MaSCs require mammary colony conditions before passaging in organoid culture. Furthermore, the most efficient organoid outgrowth is obtained when we avoid dissociating the primary colonies into single cells, but rather transfer the intact colonies into organoid culture conditions.
Organoid culture conditions also bear the advantage of giving the possibility to cryopreserve ySCs, provided that the organoids are recovered from their matrices, avoiding cell dissociation prior to cryopreservation in nitrogen bath.
The YAP reprogramming procedure presented can convert distinct differentiated cell types derived from different adult tissues into their corresponding tissue-specific stem cells (we have tested it using mammary, pancreatic and neuronal cells)1. At difference from iPSCs or other reprogramming efforts, YAP/induced SCs can maintain the memories of their tissue of origin. Of note, de-differentiation of somatic cells into cells endowed with stem-like properties is the only form of cell fate plasticity and reprogramming observed in vivo, for example after tissue damage and to support wound healing5,17,18,19,20. It is noteworthy that YAP and TAZ are largely dispensable for normal homeostasis but crucial for tissue repair in multiple tissues11,21. Consistently with a physiological function of the reprogramming steps here described, YAP/TAZ have been recently shown to be required in intestinal regeneration in mouse models of ulcerative colitis patients by causing a conversion of adult intestinal cells into a repairing epithelium that displays features of the fetal gut19. YAP reprogramming thus expands the current induced cell plasticity strategies by providing a means to generate somatic stem cells, a state that has been so far challenging to capture in vitro. This approach, if extended also to human-derived cells, might have broad relevance from regenerative medicine applications to the study of the somatic stemness state and for expansion of somatic stem cells in vitro.
The authors have nothing to disclose.
We thank F. Camargo for the gift of tetO- YAPS127A mice; R26-rtTAM2 mice (stock #006965) were purchased from The Jackson Laboratory. We thank Chiara Frasson and Giuseppe Basso for help with FACS procedures. This work is supported by AIRC Special Program Molecular Clinical Oncology ''5 per mille'' and an AIRC PI-Grant to S.P, and by Epigenetics Flagship project CNR-Miur grants to S.P. This project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (grant agreement DENOVOSTEM No. 670126).
10 mL sterile syringes | Rays | 10LC | |
100 mm cell strainer | Corning | 352360 | |
15 mL sterile conical tubes | Corning | 430052 | |
24-well ultra low attachment plates | Costar | 3473 | |
40 mm cell strainers | Corning | 352340 | |
48-well multiwell plates | Corning | 353078 | |
50 mL sterile conical tubes | Corning | 430290 | |
6-well multiwell plates | Corning | 353046 | |
Advanced DMEM/F12 | Gibco | 12634028 | |
B27 supplement (50x) | Gibco | 17504001 | |
BPE | Gibco | 13028014 | |
BSA | Sigma | A9418 | |
Collagenase, type I | Sigma | 17018029 | |
dexamethasone | Sigma | D4902 | |
Dispase | Gibco | 1705-041 | |
Disposable scalpels | Swann-Morton | 0503 | |
DMEM/F12 | Gibco | 11320033 | |
DMSO | Sigma | D2650 | |
DnaseI | Roche | 11284932001 | |
doxycycline hyclate | Sigma | D9891 | |
EDTA | Sigma | E5134 | |
Ethanol 100% | Sigma | 51976 | |
FACS tubes (with strainer caps) | Falcon | 352235 | |
FBS | Gibco | 10270106 | |
FITC anti-mouse CD326 (Ep-CAM) | BioLegend | 118208 | |
FUdeltaGW-rtTA | Addgene | #19780 | |
FUW-tetO-EGFP | Addgene | #84041 | used as negative control |
FUW-tetO-MCS | Addgene | #84008 | used as negative control |
FUW-tetO-wtYAP | Addgene | #84009 | |
FUW-tetO-YAPS94A | Addgene | #84010 | used as negative control (transcriptionally dead YAP mutant) |
GlutaMax | Gibco | 35050061 | |
HBSS | Gibco | 24020117 | |
HCl | Sigma | 30721 | |
heparin sodium salt | Sigma | H3149 | |
HEPES buffer solution (1M) | Gibco | 15630-056 | |
human R-Spondin1 (His Tag) | Sino Biological | 11083-H08H-5 | |
Hyaluronidase from bovine testes | Sigma | H3506 | |
ITS-X | Gibco | 51500056 | |
K-14 antibody | Life Technologies | Ab7800 | |
K-8 antibody | Life Technologies | Ab14053 | |
L-Glutamine | Gibco | 25030081 | |
Lin (allophycocyanin [APC] mouse lineage antibody cocktail) | BD Biosciences | 51-9003632 | |
Matrigel® Growth Factor Reduced Basement Membrane Matrix, Phenol Red-Free | Corning | 356231 | |
N-Acetylcysteine | Sigma | A9165 | |
NaOH | J.T.Baker | 0402 | |
NH4Cl | Sigma | A9434 | |
Nicotinamide | Sigma | 72340 | |
non-cell adhesive 10 cm dishes (sterile polystirol petri dish ø 94) | ROLL | 18248 | |
PBS 10x | Euroclone | ECM4004XL | |
PE Hamster Anti-Mouse CD61 | BD Biosciences | 553347 | |
PE-Cy5 Rat Anti-Human CD49f | BD Biosciences | 551129 | |
PE/Cy7 anti-mouse/rat CD29 Antibody | BioLegend | 102222 | |
Pen/Strep (10,000 U/mL) | Gibco | 15140122 | |
Rat Tail Collagen I (coating) | Sigma | 122-20 | |
Rat Tail Collagen I for 3D culture | Cultrex | 3447-020-01 | |
recombinant human FGF10 | Peprotech | 100-26 | |
recombinant human Noggin | Peprotech | 120-10C | |
recombinant murine EGF | Peprotech | 315-09 | |
recombinant murine FGF basic (bFGF) | Peprotech | 450-33 | |
RPMI 1640 medium | Gibco | 31870025 | |
SBTI (Trypsin inhibitor from Glycine max) | Sigma | T6522 | |
Tris BASE | Roche | 11814273001 | |
Trypsin-EDTA 0,05% | Gibco | 25300054 | |
Waymouth medium | Gibco | 31220023 |