A miRNA screen identifies a transcriptional program controlling the fate of adult stem cell

  1. Jacques Robert
  2. Efstathios Vlachavas
  3. Gilles Lemaître
  4. Aristotelis Chatziioannou
  5. Michel Puceat
  6. Frederic Delom
  7. Delphine Fessart

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

The 3D cultures provide more insight into cell-to-cell and cell-to-matrix interactions, better mimicking the environment where stem cells reside compared to traditional 2D cultures. Although the precise molecular pathways involved in the regulation of stem and progenitor cell fate remain unknown, it is widely accepted that transcription factors play a crucial role as intrinsic regulators in these fate decisions.

In this study, we carried out a microRNA screen to track the behaviour of adult stem/progenitor cells derived from human mammary epithelial cells grown in 3D cultures. We identified miR-106a-3p, which enriches the adult stem cell-like lineage and promotes the expansion of 3D cultures. Transcriptomic analysis showed that this miRNA regulates transcription factors such as REST, CBFB, NF-YA, and GATA3, thereby enhancing the maintenance of adult stem/progenitor cells in human epithelial cells. These data reveal a clear transcriptional program that governs the maintenance of adult stem/progenitor cells and controls their fate.

Article activity feed

  1. Version published to 10.1101/314658v5 on bioRxiv
  6. Review Commons

    Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Reviewer #1 (Evidence, reproducibility and clarity):

    Major comments:

    • 1/ The model system used in this work is referred to as "organoids", with the premise that organoids, as representatives of the original tissue, can be used to study tissue development. However, the organoids presented in the study are spherical structures, and the paper does not provide any information about how and to what extent these organoids represent the original tissue. Furthermore, it would be difficult to expect these organoids to accurately represent human breast tissue, as they are derived, to the best of the reviewer's understanding (it is not explicitly noted, but rather the text refers the reader to several previous works), from primary human breast epithelial cells that had been cultured for 6 passages in 2D culture. The CD24/CD44 flow cytometry profile of these 2D cultures, as well as the organoids derived from them, shows a unimodal distribution of CD24 and CD44 expression, and does not show distinct cell populations, as typical in primary breast epithelial cells that have not been cultured in 2D

    • Response 1:

      We understand the concern raised by the reviewer. Before initiating the study we have carefully characterized our model (See Revision Section part 2.2, for the revisions that have already been carried out).

      In parallel, to further address this point, we plan to perform an extensive characterization of our mammary primary cells as well as our 3D-matrigel embedded culture. This will be achieved through flow cytometry using other multiple lineage-specific surface markers for luminal progenitor cells (such as CD49f−/low/EpCAM+), and basal/myoepithelial cells (such as CD49f+/EpCAM−/low).

      Thus, overall our experiments will confirm that our growth conditions (which we intend to describe in greater details in the methods section) for multipotent mammary stem cells can generate multi-lineage organoids.

    2/ There appears to be confusion between concepts: primarily confusing a basal phenotype with a stem cell phenotype, and progenitor activity with stem cell activity. Because the 3D organoid model does not display a replication of luminal differentiation, it cannot be used as a proxy for stem cell function. The results from the miRNA screen and the subsequent experiments support two conclusions: 1. miR-160b-3p leads to an enhanced mesenchymal phenotype, manifested by reduced CD24 and enhanced CD44 expression. 2. miR-160b-3p leads to increased organoid formation. The latter may be interpreted, at best, as higher basal progenitor activity, but is not a measure of stem cell activity, as that would require the cells to give rise to more than one lineage, which is not shown here.

    • Response 2:

    We agree with the reviewer on the confusions between several concepts that we did not discriminate enough. The planned characterization of our cultures will allow us to better define the nature of our primary cell population and avoid any confusion (See Response 1). Indeed, an organoid is a self-organized 3D tissue that typically originates from stem cells (pluripotent, fetal or adult). Therefore, we will replace the term “stem cell activity” through the article by “stem/progenitor activity”.

    3/ Overall, the information from figures 1-4 indicate that miR-160b-3p is driving and is associated with mesenchymal differentiation, possibly with EMT.

    Response 3:

    We agree with the reviewer that our data suggest that miR-160b-3p is driving and is associated with mesenchymal differentiation. As suggested by reviewer, to further evaluate the possible involvement of EMT, we will analysed using RT-qPCR, as we described in Fessart et al, (May 30, 2016 https://doi.org/10.7554/eLife.13887), the expression of the main EMT genes in miR-106a-3p cells. These results will be also confirmed at the protein level.

    4/ In contrast to the work in the breast 3D culture model, the experiments with hESCs are interesting and do support a role of this miR in stemness.

    Response 4:

    We would like to extend our gratitude to the reviewer for their valuable comments on this aspect of our work.

    There are several relatively minor comments that cumulatively somewhat undermine the strength of the work:

    5/ Abstract: the sentence "organoids can be directly generated from human epithelial cells by only one miRNA, miR-106a-sp" needs better clarification.

    • We agree with the reviewer, this sentence will be modified accordingly.

    6/ Page 5 line 103: HMECs is a term that generally refers to human mammary epithelial cells, not a specific derivation or subpopulation thereof.

    • We will replace HMEC by human primary mammary epithelial cells.

    7/ The graph in Fig. 1A is unnecessary, it shows only one bar.

    • We will remove this panel in the revised version of the manuscript.

    8/ It is unclear if all the HMECs were derived from the same donor, or several donors. There is no information about the donor and how the tissue and cells were derived. In general, it is not entirely clear how the cells were collected, processed, stored, and cultured from the time they were obtained from the donor until their use in the current study.

    • We will include in the material section the details information on our primary cells and our culture conditions

    9/ The key for Fig. 2D is unclear. The axes read "density" but the text refers to "intensity". Fluorescence intensity in flow cytometry is usually measured on a log scale. Differences on a linear scale are not usually considered meaningful. The authors should clarify why they chose a linear scale for this screen.

    • The screening was conducted using a high throughput microscope that measures the relative signal intensity, thus the scale is based on linear signal intensity.

    10/ In the miRNA screen, how long were the cells cultured after transfection, and was it enough time for them to shift phenotype?

    • The timeline of the screening process is detailed in the Material section. The cells were cultured for 6 days following transfection for the CD44/CD24 staining and 8 days following transfection for the 3D culture, which provides sufficient time for shifting the phenotype.

    11/ Page 6 line 154: The authors likely mean z-score, not z factor (two different things).

    • We thank the reviewer for pointing this, this will be corrected.

    12/ Page 7 line 161: "mir-106a-3p directly promotes the "transdifferentiation" of CD44low/CD24high cells phenotype into CD44high/CD24low cell phenotype" - is an unsupported statement, given that there could be several alternative explanations for the observed change in population ratios, including effects on survival or growth of cells of a certain population.

    • Transdifferentiation (also known as lineage reprogramming, or -conversion), is a process in which one mature, specialized cell type changes into another without entering a pluripotent state. This process involves the ectopic expression of transcription factors and/or other stimuli. We agree with the reviewer's observation that we did not fully demonstrate the transdifferentiation process in terms of lineage. Therefore, we will use flow cytometry to analyse multiple lineage-specific surface markers for luminal progenitor cells (such as CD49f−/low/EpCAM+), and basal/myoepithelial cells (such as CD49f+/EpCAM−/low) following miR-106a-3p expression.
    • Additionally, we acknowledge that there may be other alternative explanations; so we have already assessed the impact of mir-106a-3p on the population doubling time in culture to determine whether it affects the cells' survival or growth of the cells (See Section Part 2.2 for a description of experiments already carried out).

    13/ There is need for quantification of the phenotypes described in Fig. 3C

    • We thank the reviewer for this valuable suggestion and we will indeed characterize the 3D structure using flow cytometry.

    14/ Figures 3 D-F it is not clear if the graphs display percentage or mean number (there is discrepancy between figure text and figure legend text), and when percentage, not clear for figure 3F out of what.

    • We appreciate the reviewer's suggestion, and we have replaced the term 'axis' with 'Organoid Formation Capacity (OFC),' which is the commonly used nomenclature for this measure. OFC corresponds to the percentage of organoids per cell seeded, as explained in the legend.

    15/ Fig. 5B, what is the statistical significance of the enrichment?

    • In Figure 5B, the statistical significance of the enrichment is an FDR value of 0.024 and it is based on the multiple rotation gene-set testing (mroast) from the Limma R package (https://doi.org/10.1093/bioinformatics/btq401). This information will be included in the figure legend.

    Reviewer #2 (Evidence, reproducibility and clarity):

    In this work, Robert et al. utilize mammary organoid culture as an in vitro model of stem cell renewal and maintenance. Authors show that a putative mammary stem cell population, characterized as CD44high CD24low, is enriched in organoid culture relative to 2D monolayer culture. They conduct a microRNA (miR) screen and identify miR-106a-3p as a miRNA that enriches for stem cell (CD44high CD24low) and organoid formation capacity, confirming these findings using miR-106a-3p overexpressing cells. Authors also show that CBX7 overexpression achieves a similar enrichment in CD44high CD24low and organoid formation capacity in an miR-106a-3p dependent manner, though the rationale behind the connection between CBX7 and miR-106a-3p is not well defined. Finally, the manuscript conducts a transcriptomic analysis on miR-106a-3p-OE cells and aims to functionally validate its role in embryonic stem cells (ESCs) as a model that is versatile for renewal and differentiation studies. How this relates to mammary adult stem cells (ASCs) is not entirely clear. Overall, the manuscript contains significant technical and conceptual limitations, and the data presented do not support the major conclusions of the study. There are two overarching issues that question the validity of most of the findings in this study. Firstly, there is no clear demonstration that the structures reported as 3D organoids are indeed driven by multipotent stem cells (in all data and experimental approaches associated with this). Secondly, there is a lack of evidence to support the claim that CD44high CD24low cells are stem cells with an exclusive or enhanced capacity to generate multi-lineage organoids.

    In addition to these general points, there are several specific major issues summarized below:

    1/ Authors base their findings about mammary ASC renewal using a poorly defined organoid culture system that they claim is driven by ASC renewal and differentiation. These organoid growth conditions were originally developed to support growth of bronchial epithelial cells, and the authors have not presented data to objectively assess the validity of this extrapolation to expansion of mammary organoids. The authors did not present data to support the notion that these growth conditions generate multi-lineage organoids that expand from multipotent mammary stem cells.

    • Response 1:
    • This concern has been also raised by the 1st reviewer (See Response 12 from Reviewer 1).
    • Our experiments will confirm that our growth conditions (which will be more comprehensively described in the methods section) can generate multi-lineage organoids from multipotent mammary stem cells. We will include these characterizations in the new version of the manuscript.

    2/ Authors claim that miR-106a-3p downregulates stem cell differentiation and utilize 'organoid' culture to track the temporal expression of OCT4, SOX2 and NANOG during phases of organoid renewal and differentiation. However, mammary stem cell differentiation is associated with the emergence of luminal progenitor, mature luminal and myoepithelial lineages that are characterized by the expression of a well-defined set of markers. The authors did not investigate the expression of these markers in their 'differentiation' settings. Furthermore, modulators of major pathways reported to be critical for mammary ASC maintenance are lacking, including modulators of WNT, TGF/BMP, Notch and other pathways. As such, it is difficult to ascertain that organoids emerging under such culture conditions are the result of stem cell renewal and differentiation, as opposed to lineage-restricted proliferative non-ASCs, thus questioning the validity of many of the findings in this work.

    • Response 2:

    • As suggested by the reviewer, we will investigate the expression of emergence of luminal progenitor, mature luminal and myoepithelial lineages in our 'differentiation' settings using flow cytometry. This analysis will involve the examination of multiple lineage-specific surface markers, including CD49f−/low/EpCAM+ for luminal progenitor cells and CD49f+/EpCAM−/low for basal/myoepithelial cells in the 3D context.

    • Secondly, we agree with the reviewer that major pathways such as Wnt, TGFb/BMP and Notch have been reported to be critical for mammary ASC maintenance. As suggested by the reviewer, to further evaluate the possible involvement of these pathways, we have analyzed our transcriptomic data, using PROGENy pathway, to investigate which pathways are regulated. The major pathways are depicted in a new figure and will be included in the manuscript (See Revision Section part 2.2, for the revisions that have already been carried out).

    • There was significant enrichment indicating down-regulation of TGFβ, MAPK, WNT, PI3K genes along with gene sets representing other oncogenic pathways, are up-regulated such as the hypoxia response, JAK/STAT pathway, and p53 pathway activity (Figure A and B). Stem cells possess self-renewal activities and multipotency, characteristics that tend to be maintained under hypoxic microenvironments [1], thus this is not surprising to observe an up-regulation of Hypoxia pathway and activation of HIF1α transcription factor. In mammary stem cells, it has also been shown that p53 is critical to control the maintenance of a constant number of stem cells pool [2, 3]. Remarkably, it has been shown that cell sorting of the cells with a putative cancer stem cell phenotype (CD44+/CD24 low) express a constitutive activation of Jak-STAT pathway [4], which we observed as up-regulated along with STAT1 and STAT2 transcription factors. In parallel, we observe a down-regulation of Wnt signaling as well as PI3K pathways. Wnt is known to play important role in the maintenance of stem cells; its inhibition has been shown to lead to the inactivation of PI3 kinase signaling pathways to ensures a balance control of stem cell renewal [5]. Moreover, we observe a down-regulation of SMAD4 and SMAD3 transcription factors correlated with a down-regulation of TGFβ pathway. Signals mediated by TGF-β family members have been implicated in the maintenance and differentiation of various types of somatic stem cells [6].

    • References

    1. Semenza GL. Dynamic regulation of stem cell specification and maintenance by hypoxia-inducible factors. Molecular aspects of medicine2016 Feb-Mar;47-48:15-23.
    2. Solozobova V, Blattner C. p53 in stem cells. World journal of biological chemistry2011 Sep 26;2(9):202-14.
    3. Cicalese A, Bonizzi G, Pasi CE, Faretta M, Ronzoni S, Giulini B et al. The tumor suppressor p53 regulates polarity of self-renewing divisions in mammary stem cells. Cell2009 Sep 18;138(6):1083-95.
    4. Hernandez-Vargas H, Ouzounova M, Le Calvez-Kelm F, Lambert MP, McKay-Chopin S, Tavtigian SV et al. Methylome analysis reveals Jak-STAT pathway deregulation in putative breast cancer stem cells. Epigenetics2011 Apr;6(4):428-39.
    5. He XC, Zhang J, Tong WG, Tawfik O, Ross J, Scoville DH et al. BMP signaling inhibits intestinal stem cell self-renewal through suppression of Wnt-beta-catenin signaling. Nature genetics2004 Oct;36(10):1117-21.
    6. Watabe T, Miyazono K. Roles of TGF-beta family signaling in stem cell renewal and differentiation. Cell research2009 Jan;19(1):103-15.

    3/ The authors base their work on the claim that CD44high CD24low cells represent bona fide mammary ASCs. This claim is not support by functional work to show that organoid generation is exclusive to or enhanced in CD44high CD24low cells relative to other cells. The claim of differentiation of this stem cell population in vitro is not supported by data to show emergence of luminal progenitor, mature luminal and myoepithelial lineages that are characterized by the expression of a well-defined set of reported markers as abovementioned. Although miR-106a-3p is claimed to downregulate stem cell differentiation based on a gene set enrichment analysis, authors did not investigate the expression of mammary differentiation markers or the association of miR-106a-3p-transfected HMECs with gene set pathways involved in mammary gland differentiation.

    • Response 3:
    • This concern has been also raised by the 1st reviewer (See Response 1 from Reviewer 1). As explained in response 1, to address this point, we will perform an extensive characterization of our mammary primary cells as well as our 3D-matrigel embedded cultures using flow cytometry This analysis will involve multiple lineage-specific surface markers, including luminal alveolar progenitor (such as CD49f−/low/EpCAM+) and basal/myoepithelial cells (such as CD49f+/EpCAM−/low).
    • Secondly, as suggested by the reviewer, we will investigate the expression of mammary differentiation markers or the association of miR-106a-3p-transfected human mammary epithelial cells with gene set pathways involved in mammary gland differentiation by bioinformatics using our transcriptomic data.

    4/ Line 129: "Together, these results indicate that cells grown as organoids acquired a CD44high / CD24low expression pattern similar to that of stem/progenitor cells, which suggests that 3D organoids can be used to enrich breast stem cell markers for further screening". This conclusion is not supported by the data presented. It is not clear if the expression of these markers was acquired upon 3D culture. It could be that 3D culture better maintained and expanded already existing CD44high CD24low cells in 2D culture. It is also unclear if these cells are indeed organoid-forming. Authors have not isolated and tested the organoid-forming capacity of CD44high CD24low cells relative to other cells. Along the same line, the conclusion that " mir-106a-3p directly promotes the "transdifferentiation" of CD44low/CD24high cell phenotype into CD44high/CD24low cell phenotype" isn't justified. Experiments required to conclude that 'transdifferentiation' is involved are lacking. miR-106a-3p overexpression could be creating conditions that are permissive for expansion of already existing CD44high CD24low cells as opposed to 'transdifferentiation' of other cell types into this phenotype. The authors could FACS-isolate CD44low CD24low cells and treat these with control or miR-106a-3p to conclusively establish 'transdifferentiation'.

    • Response 4:
    • This concern has been also raised by the 1st reviewer (See Response 12 from Reviewer 1). Transdifferentiation (lineage reprogramming, or -conversion), is a process in which one mature, specialized cell type changes into another without entering a pluripotent state. This process involves the ectopic expression of transcription factors and/or other stimuli. We agree with the reviewer that we did not demonstrate the transdifferentiation process in terms of lineage. Therefore, we will use flow cytometry to analyze multiple lineage-specific surface markers for luminal progenitor (CD49f−/low/EpCAM+), and basal/myoepithelial cells (CD49f+/EpCAM−/low) following miR-106a-3p expression.
    • Additionally, we cannot exclude the possibility that the 3D culture better maintained and expanded the pre-existing CD44high CD24low cells, as suggested by the reviewer. The reviewer recommends FACS-isolating CD44low CD24low cells. We apologise for any lack of clarity in our description. Technically, our primary cells have a low percentage of colony-forming efficiency (CFE) in 3D and not enough cells for FACS isolation of CD44low and CD24low cells. Therefore, we leveraged our knowledge that the expression of CBX7 potentiates the growth of 3D structures. We decided to FACS-isolate the different subpopulations of CD44 and CD24 cells to further elucidate the expression of miR-106a-3p in the different CD44/CD24 cell subpopulations. CBX7-transfected human mammary epithelial cells showed enrichment in CD44high/CD24low cells as compared to empty vector-transfected human mammary epithelial cells (Figure 4A-B). Subsequently, we separated the CD44high/CD24low (green) population from the CD44high/CD24high cell populations (blue) using flow cytometry to analyze the role of the endogenous expression of miR-106a-3p. The CD44high/CD24low population was the only one to exhibit endogenous expression of miR-106a-3p (Figure 4C). Blocking the endogenous expression of miR-106a-3p with LNA-anti-miR-106a-3p or LNA-control (Figure 4D) impacted organoid generation (Figure 4E-F). We will modify the text accordingly to include this point and this figure will be moved in the supplemental figure.

    5/ Line 242: "These data demonstrate that miR-106a-3p is involved in the early cell differentiation process into the three germ layers ". The authors have not conducted functional or mechanistic work to show that miR-106a-3p is involved in morphological or transcriptomic differentiation changes in ESCs. This and other data would be necessary to substantiate this conclusion.

    • Response 5:
    • Indeed, we agree that we cannot conclude that the miR-106a-3p is involved in the early cell differentiation process into the three germ layers without demonstrating the differentiation changes in ESCs through transcriptomic analysis. To clarify the take-home message, we did not include the transcriptomic characterization of the differentiation changes in ESCs. Instead, we focused on assessing the impact on Oct4, Sox2, and Nanog expression. However, to further understand the impact of miR-106a-3p depletion on hESCs differentiation, we have also monitored the expression of specific genes upon induction of the three embryonic germ layers (See Section Part 2.2 for a description of the experiments already carried out).

    Selected technical points:

    1. Figure 1A: The Y-axis label is misleading and suggests multiple organoids seeded per cell? Organoid Formation Efficiency (OFE) or Organoid Formation Capacity (OFC) are the commonly used nomenclature for this.
    • Response 6:
    • Since reviewer 1 found that the graph in Fig. 1A to be unnecessary, we have decided to remove this panel in the revised version of the manuscript. Furthermore, as suggested, we will replace the axis “Number of organoids per cell seeded” with “Organoid Formation Capacity” (OFC), which is the commonly used nomenclature for this measure in the article.
    1. Figure 2G: X-axis unclear. Is this meant to investigate the percentage of cells expressing CD44/CD24 as double high, double low, high/low and low/high?
    • Response 7:
    • We thank the reviewer for this careful examination of the manuscript. We apologize for this error, and will make the corresponding adjustment in Figure 2G.
    1. Figure 4C: In HMEC-CBX7 cells, it is unclear whether the high miR-106a-3p levels in CD44high CD24low cells are due to CBX7 expression. An important control, HMEC-Control Vector, is missing.
    • Response 8:
    • This control has been done (see Section 3, for details on experiments that have been carried out).

    Reviewer #3 (Evidence, reproducibility and clarity):

    In this study the authors identify miR-106a-3p as a potent inducer of organoid formation from HMECs. Overexpression of miR-106a-3p induced formation of more organoids, increased the number of stem/progenitor cells and overall positively affected the stemness properties of the organoids, likely by affecting SOx2, Oct4 and Nanog expression.

    Major comments:

    1/ the flow of the paper is confusing, it appears that the authors are trying to combine several non-completed studies into one paper. It is not immediately evident how is the rationale or the conclusion supported by published data. For example, is there published evidence that Sox2/Oct4/Nanog are expressed in healthy mammary gland stem cells in vivo, or whether they have a role in establishment of stem cell population?

    • Response 1:
    • There is still a controversy about the existence of unipotent, bipotent or multipotent stem cells in mammary gland tissue [7-9]. Notably, OCT4, SOX2, and NANOG collectively form the core transcriptional network responsible for maintaining pluripotency in embryonic stem cells [10]. Evidence has accumulated over the past few years, accumulating evidence has supported the presence of stem cells in both mouse and human mammary [11]. Various strategies have been used to identify and isolate human breast stem/progenitor cells, including FACS sorting based on cell surface antigen expression. In addition, an in vitro cell culture system has been described allowing the propagation of human mammary epithelial cells in an undifferentiated state through their ability to proliferate in suspension as non-adherent mammospheres [12].. Simoe et al have demonstrated that stem cells isolated from both normal human breast and breast tumor cells display an increased expression of the embryonic stem cell genes NANOG, OCT4 and SOX2 [13]. Moreover, they have shown that the ectopic expression of any one of these factors, but in particular NANOG and SOX2, in breast cancer cells increases the pool of stem cells and enhances the cells' ability to form mammospheres. They observed higher expression of NANOG, OCT4, and SOX2 in the stem cell populations CD44+CD24−/low and EMA+CALLA+ compared to the rest of the sample population. Cells overexpressing these factors displayed an increase in the stem cell populations, thereby confirming the role of Nanog, Oct4, and Sox2 in the maintenance of human mammary stem cells. We will include this rationale in the manuscript.
    • References
    1. Van Keymeulen A, Rocha AS, Ousset M, Beck B, Bouvencourt G, Rock J et al. Distinct stem cells contribute to mammary gland development and maintenance. Nature2011 Oct 9;479(7372):189-93.
    2. Deome KB, Faulkin LJ, Jr., Bern HA, Blair PB. Development of mammary tumors from hyperplastic alveolar nodules transplanted into gland-free mammary fat pads of female C3H mice. Cancer research1959 Jun;19(5):515-20.
    3. Shackleton M, Vaillant F, Simpson KJ, Stingl J, Smyth GK, Asselin-Labat ML et al. Generation of a functional mammary gland from a single stem cell. Nature2006 Jan 5;439(7072):84-8.
    4. Boyer LA, Lee TI, Cole MF, Johnstone SE, Levine SS, Zucker JP et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell2005 Sep 23;122(6):947-56.
    5. LaMarca HL, Rosen JM. Minireview: hormones and mammary cell fate--what will I become when I grow up? Endocrinology2008 Sep;149(9):4317-21.
    6. Dontu G, Abdallah WM, Foley JM, Jackson KW, Clarke MF, Kawamura MJ et al. In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes & development2003 May 15;17(10):1253-70.
    7. Simoes BM, Piva M, Iriondo O, Comaills V, Lopez-Ruiz JA, Zabalza I et al. Effects of estrogen on the proportion of stem cells in the breast. Breast cancer research and treatment2011 Aug;129(1):23-35.

    2/ The organoids are all spherical, while mammary gland is characterized by branching. Therefore, the organoids are not recapitulating the gland morphology and the validation should include wider range of molecular markers.

    • Response 2:
    • This concern has been also raised by both reviewers 1 and 2. As explained in response 1 to reviewer 1, we will perform a comprehensive characterization of our mammary primary cells as well as our 3D-matrigel embedded culture culture using flow cytometry to examine multiple lineage-specific surface markers, including luminal alveolar progenitor (such as CD49f−/low/EpCAM+), and basal/myoepithelial cells (such as CD49f+/EpCAM−/low). As a result, our experiments will collectively confirm that our growth conditions (which will be more thoroughly described in the methods section) for multipotent mammary stem cells can indeed generate multi-lineage organoids.

    3/ The choice of control is not clear. For example, why was miR-106a-5p chosen as a control? And why choose miR-106a-5p when a better candidate would be miR-106b-3p, that produces very high number of organoids as well? And how did the other miRNAs affected the CD44/24 profile?

    • Response 3:
    • We apologize for any lack of clarity in our description. It's important to note that miRNAs consist of two strands, -5p and -3p, and both strands can coexist and play distinct roles. For example, paired species of members in the let-7 and mir-126 families coexist and have different regulatory functions in reprogramming and differentiation of embryonic stem cells [1]. Several deep sequencing studies have demonstrated the coexistence of 5p/3p pairs in approximately half of the miRNA populations analyzed [2, 3]. Therefore, to determine whether the biological effect specifically resulted from the -3p strand, we also assessed the -5p strand.
    • References
    1. Koh W, Sheng CT, Tan B, Lee QY, Kuznetsov V, Kiang LS et al. Analysis of deep sequencing microRNA expression profile from human embryonic stem cells derived mesenchymal stem cells reveals possible role of let-7 microRNA family in downstream targeting of hepatic nuclear factor 4 alpha. BMC genomics2010 Feb 10;11 Suppl 1(Suppl 1):S6.
    2. Jagadeeswaran G, Zheng Y, Sumathipala N, Jiang H, Arrese EL, Soulages JL et al. Deep sequencing of small RNA libraries reveals dynamic regulation of conserved and novel microRNAs and microRNA-stars during silkworm development. BMC genomics2010 Jan 20;11:52.
    3. Kuchenbauer F, Mah SM, Heuser M, McPherson A, Ruschmann J, Rouhi A et al. Comprehensive analysis of mammalian miRNA* species and their role in myeloid cells. Blood2011 Sep 22;118(12):3350-8.

    4/ What is the CD44/CD24 profile in the organoid cultures from Figure 7?

    • Response 4:
    • The CD44/CD24 profile from the organoid cultures from Figure 7 will be assessed in the revised manuscript.

    Part 2.2 Following the experiment that have been already carried out that will be included in the manuscript after revisions.

    Reviewer #1 (Evidence, reproducibility and clarity):

    Major comments:

    The model system used in this work is referred to as "organoids", with the premise that organoids, as representatives of the original tissue, can be used to study tissue development. However, the organoids presented in the study are spherical structures, and the paper does not provide any information about how and to what extent these organoids represent the original tissue. Furthermore, it would be difficult to expect these organoids to accurately represent human breast tissue, as they are derived, to the best of the reviewer's understanding (it is not explicitly noted, but rather the text refers the reader to several previous works), from primary human breast epithelial cells that had been cultured for 6 passages in 2D culture. The CD24/CD44 flow cytometry profile of these 2D cultures, as well as the organoids derived from them, shows a unimodal distribution of CD24 and CD44 expression, and does not show distinct cell populations, as typical in primary breast epithelial cells that have not been cultured in 2D.

    • Response:
    • First, we assessed the aldehyde dehydrogenase activity (ALDH) [14] in our primary cells to demonstrate that these culture conditions preserve stem/progenitor properties (See Figure A, below). Additionally, we conducted immuno-staining of primary cells in 2D culture for lineage-specific luminal markers (CK18, MUC1) and basal markers (CK14, CK5), which revealed the heterogeneity of our population (See Figure B, below).

    FIGURE FOR REVIEWERS

    • Reference
    1. Ginestier C, Hur MH, Charafe-Jauffret E, Monville F, Dutcher J, Brown M et al. ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell stem cell2007 Nov;1(5):555-67.
    • For the characterization of the 3D culture, we have presented early time points of 3D cell growth, which are mainly spherical (until Day 8). However, differentiation occurs in a stepwise manner, where single stem cells proliferate to form small spheroids before undergoing multilineage differentiation into organoids that initiate branching (Day 10). Subsequently, ducts undergo budding and lobule formation (Day 20). We have included a time-course representation of this 3D growth to illustrate the differentiation process (See Figure 1C).
    • Under these culture conditions, the 3D structure retains its self-renewal activity and the ability to reseed and regenerate secondary tissues. We have also assessed the self-renewal capacity of our 3D structure (See Figure 1D).

    **FIGURE FOR REVIEWERS
    **

    Page 7 line 161: "mir-106a-3p directly promotes the "transdifferentiation" of CD44low/CD24high cells phenotype into CD44high/CD24low cell phenotype" - is an unsupported statement, given that there could be several alternative explanations for the observed change in population ratios, including effects on survival or growth of cells of a certain population.

    • Indeed, we cannot exclude the possibility of other alternative explanations. Therefore, we have already assessed the impact of miR-106a-3p on population doubling in culture to determine whether it has an effect on the survival or growth of the cells. We observed that miR-V cells stopped growing after approximately 15 population doublings, indicating their limited proliferative potential. In contrast, we found that miR-106a extended the lifespan of the cells, suggesting an effect on cell survival (Figure below). We will include this information in the manuscript.

    FIGURE FOR REVIEWERS

    • Why was GATA3 not included in the last analysis depicted in Fig. 7?
    • We apologize for any lack of clarity in our description. It's important to note that GATA3 was not included in Figure 7. As explained in the text, GATA3 plays a role in regulating Nanog expression. However, it's worth noting that the cells did not form organoids when GATA3 was depleted, as illustrated in the results below. We will include these results in the revised version of the manuscript.

    FIGURE FOR REVIEWERS

    Reviewer #2:

    2/ Authors claim that miR-106a-3p downregulates stem cell differentiation and utilize 'organoid' culture to track the temporal expression of OCT4, SOX2 and NANOG during phases of organoid renewal and differentiation. However, mammary stem cell differentiation is associated with the emergence of luminal progenitor, mature luminal and myoepithelial lineages that are characterized by the expression of a well-defined set of markers. The authors did not investigate the expression of these markers in their 'differentiation' settings. Furthermore, modulators of major pathways reported to be critical for mammary ASC maintenance are lacking, including modulators of WNT, TGF/BMP, Notch and other pathways. As such, it is difficult to ascertain that organoids emerging under such culture conditions are the result of stem cell renewal and differentiation, as opposed to lineage-restricted proliferative non-ASCs, thus questioning the validity of many of the findings in this work.

    • We agree with the reviewer that major pathways such as Wnt, TGFβ/BMP and Notch have been reported to be critical for mammary ASC maintenance. As suggested by the reviewer, to further evaluate the possible involvement of these pathways, we have analyzed our transcriptomic data, using the PROGENy R package (version 1.16.0) (https://doi.org/10.1038/s41467-017-02391-6) and DoRothEA (version 1.6.0)-decoupleR (version 2.1.6) computational pipeline (https://doi.org/10.1101/gr.240663.118, https://doi.org/10.1093/bioadv/vbac016), to investigate the differential activation of major signaling pathways and transcriptional factors. The major pathways are depicted in the figure below and will be included in the manuscript.

    FIGURE FOR REVIEWERS

    • There was significant enrichment indicating down-regulation of TGFβ, MAPK, WNT, PI3K genes along with gene sets representing other oncogenic pathways, are up-regulated such as the hypoxia response, JAK/STAT pathway, and p53 pathway activity (Figure A and B). Stem cells possess self-renewal activities and multipotency, characteristics that tend to be maintained under hypoxic microenvironments [15], thus this is not surprising to observe an up-regulation of Hypoxia pathway and activation of HIF1α transcription factor. In mammary stem cells, it has also been shown that p53 is critical to control the maintenance of a constant number of stem cells pool [16,17]. Remarkably, it has been shown that cell sorting of the cells with a putative cancer stem cell phenotype (CD44+/CD24 low) express a constitutive activation of Jak-STAT pathway [18], which we observed as up-regulated along with STAT1 and STAT2 transcription factors. In parallel, we observe a down-regulation of Wnt signaling as well as PI3K pathways. Wnt is known to play important role in the maintenance of stem cells; its inhibition has been shown to lead to the inactivation of PI3 kinase signaling pathways to ensure a balance control of stem cell renewal [19]. Moreover, we observe a down-regulation of SMAD4 and SMAD3 transcription factors correlated with a down-regulation of TGFβ pathway. Signals mediated by TGF-β family members have been implicated in the maintenance and differentiation of various types of somatic stem cells [20].
    • References

    1. Semenza GL. Dynamic regulation of stem cell specification and maintenance by hypoxia-inducible factors. Molecular aspects of medicine2016 Feb-Mar;47-48:15-23.

    2. Solozobova V, Blattner C. p53 in stem cells. World journal of biological chemistry2011 Sep 26;2(9):202-14.

    3. Cicalese A, Bonizzi G, Pasi CE, Faretta M, Ronzoni S, Giulini B et al. The tumor suppressor p53 regulates polarity of self-renewing divisions in mammary stem cells. Cell2009 Sep 18;138(6):1083-95.

    4. Hernandez-Vargas H, Ouzounova M, Le Calvez-Kelm F, Lambert MP, McKay-Chopin S, Tavtigian SV et al. Methylome analysis reveals Jak-STAT pathway deregulation in putative breast cancer stem cells. Epigenetics2011 Apr;6(4):428-39.

    5. He XC, Zhang J, Tong WG, Tawfik O, Ross J, Scoville DH et al. BMP signaling inhibits intestinal stem cell self-renewal through suppression of Wnt-beta-catenin signaling. Nature genetics2004 Oct;36(10):1117-21.

    6. Watabe T, Miyazono K. Roles of TGF-beta family signaling in stem cell renewal and differentiation. Cell research2009 Jan;19(1):103-15.

    1. Line 242: "These data demonstrate that miR-106a-3p is involved in the early cell differentiation process into the three germ layers ". The authors have not conducted functional or mechanistic work to show that miR-106a-3p is involved in morphological or transcriptomic differentiation changes in ESCs. This and other data would be necessary to substantiate this conclusion.
    • Response:
    • Indeed, we cannot conclude that miR-106a-3p is directly involved in the early cell differentiation process into the three germ layers without demonstrating the differentiation changes in ESCs through transcriptomic analysis. To clarify the take-home message, it's important to note that we did not include the transcriptomic characterization of differentiation changes in ESCs. Instead, we focused on assessing the impact of miR-106a-3p depletion on the expression of Oct4, Sox2, and Nanog. However, to gain a deeper understanding of the effects of miR-106a-3p depletion on hESCs differentiation, we also monitored the expression of specific genes upon the induction of the three embryonic germ layers (see Figure A, B, and C below). We observed that the expression of endodermal genes was not or only weakly affected by the level of miR-106a-3p expression (Figure A), whereas the expression of mesoderm- and ectoderm-specific genes increased upon miR-106a-3p down-regulation (Figure B and C). We will include these findings in the manuscript.

    FIGURE FOR REVIEWERS

    Reviewer 2 (Minor points)

    1. Figure 4C: In HMEC-CBX7 cells, it is unclear whether the high miR-106a-3p levels in CD44high CD24low cells are due to CBX7 expression. An important control, HMEC-Control Vector, is missing.
    • Response
    • The control HMEC-Vector is shown in panel Figure 4A for the FACS analysis. In Figure 4C, we did not include the control since there is no expression of miR-106a-3p, but it's important to note that this control was included in the experiment. As illustrated, there is no miR-106a-3p expression in the HMEC control cells. We intend to include this panel in Figure 4 of the manuscript.

    FIGURE FOR REVIEWERS

  7. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    In this study the authors identify miR-106a-3p as a potent inducer of organoid formation from HMECs. Overexpression of miR-106a-3p induced formation of more organoids, increased the number of stem/progenitor cells and overall positively affected the stemness properties of the organoids, likely by affecting SOx2, Oct4 and Nanog expression.

    Major comments: the flow of the paper is confusing, it appears that the authors are trying to combine several non-completed studies into one paper. It is not immediately evident how is the rationale or the conclusion supported by published data. For example, is there published evidence that Sox2/Oct4/Nanog are expressed in healthy mammary gland stem cells in vivo, or whether they have a role in establishment of stem cell population?
    The organoids are all spherical, while mammary gland is characterized by branching. Therefore, the organoids are not recapitulating the gland morphology and the validation should include wider range of molecular markers.
    The choice of control is not clear. For example, why was miR-106a-5p chosen as a control? And why choose miR-106a-5p when a better candidate would be miR-106b-3p, that produces very high number of organoids as well? And how did the other miRNAs affected the CD44/24 profile?
    What is the CD44/CD24 profile in the organoid cultures from Figure 7?

    Significance

    The study provides a novel methodology to enrich for the mammary stem cells. However, many of the experiments need further clarification.

  8. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    In this work, Robert et al. utilize mammary organoid culture as an in vitro model of stem cell renewal and maintenance. Authors show that a putative mammary stem cell population, characterized as CD44high CD24low, is enriched in organoid culture relative to 2D monolayer culture. They conduct a microRNA (miR) screen and identify miR-106a-3p as a miRNA that enriches for stem cell (CD44high CD24low) and organoid formation capacity, confirming these findings using miR-106a-3p overexpressing cells. Authors also show that CBX7 overexpression achieves a similar enrichment in CD44high CD24low and organoid formation capacity in an miR-106a-3p dependent manner, though the rationale behind the connection between CBX7 and miR-106a-3p is not well defined. Finally, the manuscript conducts a transcriptomic analysis on miR-106a-3p-OE cells and aims to functionally validate its role in embryonic stem cells (ESCs) as a model that is versatile for renewal and differentiation studies. How this relates to mammary adult stem cells (ASCs) is not entirely clear.

    Overall, the manuscript contains significant technical and conceptual limitations, and the data presented do not support the major conclusions of the study. There are two overarching issues that question the validity of most of the findings in this study. Firstly, there is no clear demonstration that the structures reported as 3D organoids are indeed driven by multipotent stem cells (in all data and experimental approaches associated with this). Secondly, there is a lack of evidence to support the claim that CD44high CD24low cells are stem cells with an exclusive or enhanced capacity to generate multi-lineage organoids.

    In addition to these general points, there are several specific major issues summarized below:

    1. Authors base their findings about mammary ASC renewal using a poorly defined organoid culture system that they claim is driven by ASC renewal and differentiation. These organoid growth conditions were originally developed to support growth of bronchial epithelial cells, and the authors have not presented data to objectively assess the validity of this extrapolation to expansion of mammary organoids. The authors did not present data to support the notion that these growth conditions generate multi-lineage organoids that expand from multipotent mammary stem cells.
    2. Authors claim that miR-106a-3p downregulates stem cell differentiation and utilize 'organoid' culture to track the temporal expression of OCT4, SOX2 and NANOG during phases of organoid renewal and differentiation. However, mammary stem cell differentiation is associated with the emergence of luminal progenitor, mature luminal and myoepithelial lineages that are characterized by the expression of a well-defined set of markers. The authors did not investigate the expression of these markers in their 'differentiation' settings. Furthermore, modulators of major pathways reported to be critical for mammary ASC maintenance are lacking, including modulators of WNT, TGF/BMP, Notch and other pathways. As such, it is difficult to ascertain that organoids emerging under such culture conditions are the result of stem cell renewal and differentiation, as opposed to lineage-restricted proliferative non-ASCs, thus questioning the validity of many of the findings in this work.
    3. The authors base their work on the claim that CD44high CD24low cells represent bona fide mammary ASCs. This claim is not support by functional work to show that organoid generation is exclusive to or enhanced in CD44high CD24low cells relative to other cells. The claim of differentiation of this stem cell population in vitro is not supported by data to show emergence of luminal progenitor, mature luminal and myoepithelial lineages that are characterized by the expression of a well-defined set of reported markers as abovementioned. Although miR-106a-3p is claimed to downregulate stem cell differentiation based on a gene set enrichment analysis, authors did not investigate the expression of mammary differentiation markers or the association of miR-106a-3p-transfected HMECs with gene set pathways involved in mammary gland differentiation.
    4. Line 129: "Together, these results indicate that cells grown as organoids acquired a CD44high / CD24low expression pattern similar to that of stem/progenitor cells, which suggests that 3D organoids can be used to enrich breast stem cell markers for further screening".

    This conclusion is not supported by the data presented. It is not clear if the expression of these markers was acquired upon 3D culture. It could be that 3D culture better maintained and expanded already existing CD44high CD24low cells in 2D culture. It is also unclear if these cells are indeed organoid-forming. Authors have not isolated and tested the organoid-forming capacity of CD44high CD24low cells relative to other cells. Along the same line, the conclusion that " mir-106a-3p directly promotes the "transdifferentiation" of CD44low/CD24high cell phenotype into CD44high/CD24low cell phenotype" isn't justified. Experiments required to conclude that 'transdifferentiation' is involved are lacking. miR-106a-3p overexpression could be creating conditions that are permissive for expansion of already existing CD44high CD24low cells as opposed to 'transdifferentiation' of other cell types into this phenotype. The authors could FACS-isolate CD44low CD24low cells and treat these with control or miR-106a-3p to conclusively establish 'transdifferentiation'.

    1. Line 242: "These data demonstrate that miR-106a-3p is involved in the early cell differentiation process into the three germ layers "

    The authors have not conducted functional or mechanistic work to show that miR-106a-3p is involved in morphological or transcriptomic differentiation changes in ESCs. This and other data would be necessary to substantiate this conclusion.

    Selected technical points:

    1. Figure 1A: The Y-axis label is misleading and suggests multiple organoids seeded per cell? Organoid Formation Efficiency (OFE) or Organoid Formation Capacity (OFC) are the commonly used nomenclature for this.
    2. Figure 2G: X-axis unclear. Is this meant to investigate the percentage of cells expressing CD44/CD24 as double high, double low, high/low and low/high?
    3. Figure 4C: In HMEC-CBX7 cells, it is unclear whether the high miR-106a-3p levels in CD44high CD24low cells are due to CBX7 expression. An important control, HMEC-Control Vector, is missing.

    Significance

    The findings in this report do not represent a sufficient conceptual advance in mammary stem cell biology. There are some interesting data on the role of miR-106a-3p in differentiation and regulation of pluripotency-inducing factors OCT4, SOX2 and NANOG. This is, however, more relevant to ESC biology. The data presented do not sufficiently explain the proposed transcriptional and molecular influences of miR-106a-3p on ESC maintenance and differentiation. Furthermore, there are multiple instances in the manuscript of overstating conclusions in a manner that is not supported by the data presented, as well as several instances of a conceptually premature extrapolation of various aspects of ESC biology to mammary ASC biology. It is not clear what literature or experimental findings serve as the basis for the author's connection between these two developmentally and temporally distinct aspects of tissue biology.

  9. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    Major comments:

    • The model system used in this work is referred to as "organoids", with the premise that organoids, as representatives of the original tissue, can be used to study tissue development. However, the organoids presented in the study are spherical structures, and the paper does not provide any information about how and to what extent these organoids represent the original tissue. Furthermore, it would be difficult to expect these organoids to accurately represent human breast tissue, as they are derived, to the best of the reviewer's understanding (it is not explicitly noted, but rather the text refers the reader to several previous works), from primary human breast epithelial cells that had been cultured for 6 passages in 2D culture. The CD24/CD44 flow cytometry profile of these 2D cultures, as well as the organoids derived from them, shows a unimodal distribution of CD24 and CD44 expression, and does not show distinct cell populations, as typical in primary breast epithelial cells that have not been cultured in 2D.
    • There appears to be confusion between concepts: primarily confusing a basal phenotype with a stem cell phenotype, and progenitor activity with stem cell activity. Because the 3D organoid model does not display a replication of luminal differentiation, it cannot be used as a proxy for stem cell function. The results from the miRNA screen and the subsequent experiments support two conclusions: 1. miR-160b-3p leads to an enhanced mesenchymal phenotype, manifested by reduced CD24 and enhanced CD44 expression. 2. miR-160b-3p leads to increased organoid formation. The latter may be interpreted, at best, as higher basal progenitor activity, but is not a measure of stem cell activity, as that would require the cells to give rise to more than one lineage, which is not shown here.
    • Overall, the information from figures 1-4 indicate that miR-160b-3p is driving and is associated with mesenchymal differentiation, possibly with EMT.
    • In contrast to the work in the breast 3D culture model, the experiments with hESCs are interesting and do support a role of this miR in stemness.
      There are several relatively minor comments, that cumulatively somewhat undermine the strength of the work:
    • Abstract: the sentence "organoids can be directly generated from human epithelial cells by only one miRNA, miR-106a-sp" needs better clarification.
    • Page 5 line 103: HMECs is a term that generally refers to human mammary epithelial cells, not a specific derivation or subpopulation thereof.
    • The graph in Fig. 1A is unnecessary, it shows only one bar.
    • It is unclear if all the HMECs were derived from the same donor, or several donors. There is no information about the donor and how the tissue and cells were derived. In general, it is not entirely clear how the cells were collected, processed, stored, and cultured from the time they were obtained from the donor until their use in the current study.
    • The key for Fig. 2D is unclear. The axes read "density" but the text refers to "intensity". Fluorescence intensity in flow cytometry is usually measured on a log scale. Differences on a linear scale are not usually considered meaningful. The authors should clarify why they chose a linear scale for this screen.
    • In the miRNA screen, how long were the cells cultured after transfection, and was it enough time for them to shift phenotype?
    • Page 6 line 154: The authors likely mean z-score, not z factor (two different things).
    • Page 7 line 161: "mir-106a-3p directly promotes the "transdifferentiation" of CD44low/CD24high cells phenotype into CD44high/CD24low cell phenotype" - is an unsupported statement, given that there could be several alternative explanations for the observed change in population ratios, including effects on survival or growth of cells of a certain population.
    • There is need for quantification of the phenotypes described in Fig. 3C
    • Figures 3 D-F it is not clear if the graphs display percentage or mean number (there is discrepancy between figure text and figure legend text), and when percentage, not clear for figure 3F out of what.
    • Fig. 5B, what is the statistical significance of the enrichment?
    • Why was GATA3 not included in the last analysis depicted in Fig. 7?

    Significance

    This manuscript describes experiments that aim to explore the role of miR-160b-3p in stem cells. It uses primarily a model of breast epithelial cells in 3D Matrigel culture, termed organoids.

    The first part of the paper describes a screen that identified miR-160b-3p as changing the expression profile of the two surface markers CD24 and CD44 in breast epithelial cells, which the authors refer to as a stem cell population, and enhancing organoid formation capability, which the authors interpret as stem cell capacity. In the second part of the paper, the experiments use another cell model - human embryonic stem cells (ESCs). In the second part, the authors link the expression of miR-160b-3p to the expression of Nanog, Sox2, and Oct4, which are key transcription factors that play essential roles in maintaining pluripotency and self-renewal of ESCs.
    The premise of the work is strong, namely that genetic screens that use an organoid model have the potential to uncover genes and pathways that direct tissue development and regulate stem cell fate and function. Several issues make it difficult to draw conclusions about stem cell function from the first part of the paper, namely the experiments with breast epithelial organoids. The second part of the paper is stronger and more convincing of the main claim, which is that miR-160b-3p has a role in stem cell maintenance.

  10. Version published to 10.1101/314658v4 on bioRxiv
  11. Version published to 10.1101/314658v3 on bioRxiv
  12. Version published to 10.1101/314658v2 on bioRxiv
  13. Version published to 10.1101/314658v1 on bioRxiv