Loss of connexin40 is associated with decreased endothelium-dependent relaxations and eNOS levels in the mouse aorta.

Upon agonist stimulation, endothelial cells trigger smooth muscle relaxation through the release of relaxing factors such as nitric oxide (NO). Endothelial cells of mouse aorta are interconnected by gap junctions made of connexin40 (Cx40) and connexin37 (Cx37), allowing the exchange of signaling molecules to coordinate their activity. Wild-type (Cx40(+/+)) and hypertensive Cx40-deficient mice (Cx40(-/-)), which also exhibit a marked decrease of Cx37 in the endothelium, were used to investigate the link between the expression of endothelial connexins (Cx40 and Cx37) and endothelial nitric oxide synthase (eNOS) expression and function in the mouse aorta. With the use of isometric tension measurements in aortic rings precontracted with U-46619, a stable thromboxane A(2) mimetic, we first demonstrate that ACh- and ATP-induced endothelium-dependent relaxations solely depend on NO release in both Cx40(+/+) and Cx40(-/-) mice, but are markedly weaker in Cx40(-/-) mice. Consistently, both basal and ACh- or ATP-induced NO production were decreased in the aorta of Cx40(-/-) mice. Altered relaxations and NO release from aorta of Cx40(-/-) mice were associated with lower expression levels of eNOS in the aortic endothelium of Cx40(-/-) mice. Using immunoprecipitation and in situ ligation assay, we further demonstrate that eNOS, Cx40, and Cx37 tightly interact with each other at intercellular junctions in the aortic endothelium of Cx40(+/+) mice, suggesting that the absence of Cx40 in association with altered Cx37 levels in endothelial cells from Cx40(-/-) mice participate to the decreased levels of eNOS. Altogether, our data suggest that the endothelial connexins may participate in the control of eNOS expression levels and function.

Magnetic resonance-guided, real-time targeted delivery and imaging of magnetocapsules immunoprotecting pancreatic islet cells.

In type I diabetes mellitus, islet transplantation provides a moment-to-moment fine regulation of insulin. Success rates vary widely, however, necessitating suitable methods to monitor islet delivery, engraftment and survival. Here magnetic resonance-trackable magnetocapsules have been used simultaneously to immunoprotect pancreatic beta-cells and to monitor, non-invasively in real-time, hepatic delivery and engraftment by magnetic resonance imaging (MRI). Magnetocapsules were detected as single capsules with an altered magnetic resonance appearance on capsule rupture. Magnetocapsules were functional in vivo because mouse beta-cells restored normal glycemia in streptozotocin-induced diabetic mice and human islets induced sustained C-peptide levels in swine. In this large-animal model, magnetocapsules could be precisely targeted for infusion by using magnetic resonance fluoroscopy, whereas MRI facilitated monitoring of liver engraftment over time. These findings are directly applicable to ongoing improvements in islet cell transplantation for human diabetes, particularly because our magnetocapsules comprise clinically applicable materials.

Cell Adhesion Peptide RGDSP and Laminin Support Proliferation and Migration of Postnatal Retinal Stem Cells in a 3D Hydrogel Culture System

Purpose: Retinal stem cells (RSCs) can be isolated from radial glia population of the newborn mouse retina (Angénieux et al., 2006). These RSCs have great capacity to renew and generate neurons including cells differentiated towards the photoreceptor lineage (Mehri-Soussi et al., 2006). However, our published results showed poor integration and survival rate after cell grafting into the retina. The uncontrollable environment of retina seems to be the problem. To bypass this, we are trying to generate hemi-retinal tissue in vitro that can be used for transplantation. Methods: Expanded RSCs were seeded in a mixture of poly-ethylene-glycol (PEG)-polymer-based hydrogels crosslinked by peptides that also serve as substrates for matrix metalloproteinases. Different doses of crosslinker peptides were tested. Several growth factors were studied to stimulate cell proliferation and differentiation. Results: Cells were trapped in hydrogels and cultured in the presence of FGF2 and EGF. Spherical cell clusters indicating proliferation appeared within several days, but there was no cell migration within the gel. We then added cell adhesion molecules integrin ligand RGDSP, or laminin, or a combination of both, into the gel. Cells grown with laminin showed the best proliferation. Cells grown with RGDSP proliferated a few times and then started to spread out. Cells grown with the combination of RGDSP and laminin showed better proliferation than with RGDSP alone and larger spread-outs than with laminin alone. After stimulations with first FGF2 and EGF, and then only FGF2, some cells showed neuronal morphology after 2 weeks. The neuronal population was assessed by the presence of neuronal marker b-tubulin-III. Glial cells were also present. Further characterizations are undergoing. Conclusions: RSC can grow and migrate in 3D hydrogel with the addition of FGF2, EGF, RGDSP and laminin. Further developments are necessary to form a homogenous tissue containing retinal cells.

Polarity of endogenous and exogenous glycosyl-phosphatidylinositol-anchored membrane proteins in Madin-Darby canine kidney cells.

Madin-Darby canine kidney cells (MDCK) were transfected with a cDNA encoding the glycosyl-phosphatidylinositol (GPI)-anchored protein mouse Thy-1 in order to study the steady-state surface distribution of exogenous and endogenous GPI-linked proteins. Immunofluorescence of transfected cells grown on collagen-coated coverslips showed that expression of Thy-1 was variable throughout the epithelium, with some cells expressing large amounts of Thy-1 adjacent to very faintly staining cells. Selective surface iodination of cells grown on collagen-coated or uncoated transwell filters followed by immunoprecipitation of Thy-1 demonstrated that all the Thy-1 was present exclusively in the apical plasma membrane. Although cells grown on uncoated filters had much smaller amounts of Thy-1, it was consistently localized on the apical surfaces. Immunofluorescent localization of Thy-1 on 1 micron frozen sections of filter-grown cells demonstrated that all the Thy-1 was on the apical surface and there was no detectable intracellular pool. Phosphatidylinositol-specific phospholipase C digestion of intact iodinated monolayers released Thy-1 only into the apical medium, indicating that Thy-1 was processed normally in transfected cells and was anchored by a GPI-tail. In agreement with previous findings, endogenous GPI-linked proteins were found only on the apical plasma membrane. These results suggest that there is a common mechanism for sorting and targeting of GPI-linked proteins in polarized epithelial cells.

Aging of myelinating glial cells predominantly affects lipid metabolism and immune response pathways.

Both the central and the peripheral nervous systems are prone to multiple age-dependent neurological deficits, often attributed to still unknown alterations in the function of myelinating glia. To uncover the biological processes affected in glial cells by aging, we analyzed gene expression of the Schwann cell-rich mouse sciatic nerve at 17 time points throughout life, from day of birth until senescence. By combining these data with the gene expression data of myelin mouse mutants carrying deletions of either Pmp22, SCAP, or Lpin1, we found that the majority of age-related transcripts were also affected in myelin mutants (54.4%) and were regulated during PNS development (59.5%), indicating a high level of overlap in implicated molecular pathways. The expression profiles in aging copied the direction of transcriptional changes observed in neuropathy models; however, they had the opposite direction when compared with PNS development. The most significantly altered biological processes in aging involved the inflammatory/immune response and lipid metabolism. Interestingly, both these pathways were comparably changed in the aging optic nerve, suggesting that similar biological processes are affected in aging of glia-rich parts of the central and peripheral nervous systems. Our comprehensive comparison of gene expression in three distinct biological conditions including development, aging, and myelin disease thus revealed a previously unanticipated relationship among themselves and identified lipid metabolism and inflammatory/immune response pathways as potential therapeutical targets to prevent or delay so far incurable age-related and inherited forms of neuropathies.

Dormant and self-renewing hematopoietic stem cells and their niches.

In the mouse, over the last 20 years, a set of cell-surface markers and activities have been identified, enabling the isolation of bone marrow (BM) populations highly enriched in hematopoietic stem cells (HSCs). These HSCs have the ability to generate multiple lineages and are capable of long-term self-renewal activity such that they are able to reconstitute and maintain a functional hematopoietic system after transplantation into lethally irradiated recipients. Using single-cell reconstitution assays, various marker combinations can be used to achieve a functional HSC purity of almost 50%. Here we have used the differential expression of six of these markers (Sca1, c-Kit, CD135, CD48, CD150, and CD34) on lineage-depleted BM to refine cell hierarchies within the HSC population. At the top of the hierarchy, we propose a dormant HSC population (Lin(-)Sca1(+)c-Kit(+) CD48(-)CD150(+)CD34(-)) that gives rise to an active self-renewing CD34(+) HSC population. HSC dormancy, as well as the balance between self-renewal and differentiation activity, is at least, in part, controlled by the stem cell niches individual HSCs are attached to. Here we review the current knowledge about HSC niches and propose that dormant HSCs are located in niches at the endosteum, whereas activated HSCs are in close contact to sinusoids of the BM microvasculature.

Selectin ligands on leukemic cells

RESUME L'infiltration tissulaire par les cellules leucémiques, responsable de leucostase, est une complication grave de la leucémie aiguë hyperleucocytaire. Elle peut entraîner une détresse respiratoire et des troubles neurologiques de mauvais pronostic. Pendant longtemps, la prolifération intravasculaire des cellules leucémiques et l'augmentation de la viscosité étaient considérées comme en étant responsables, et le traitement reposait sur une cytoréduction rapide par leucaphérèse. Actuellement, l'interaction entre les cellules leucémiques et l'endothélium vasculaire est plutôt considérée comme la cause de ce phénomène. En effet, les cellules leucémiques peuvent induire l'expression des sélectives endothéliales. Les sélectives initient le roulement des leucocytes avant leur adhésion ferme et leur migration dans les tissus. Elles reconnaissent des ligands spécifiques exprimés à la surface des leucocytes, comme PSGL-1 qui est un ligand commun des sélectives. Cependant, plusieurs études suggèrent que d'autres ligands de la E-sélective soient exprimés par les leucocytes. L'interaction des cellules leucémiques avec la E- et la P- sélective est corrélée avec l'expression de la molécule CLA, reconnue par l'anticorps HECA-452. L'immunopurification des ligands de la E-sélective avec cet anticorps a permis d'isoler, des cellules THP1 et U937, une protéine de 170 kDa, ainsi qu'une autre protéine de 250 kDa des cellules U937, en plus de PSGL-1. Ces protéines ont également été purifiées avec la protéine de fusion Esélective/IgM. CD43 et CD44 semblent être des ligands de la E-sélective sur certaines lignées, mais leur interaction avec la E-sélective n'est pas toujours retrouvée. De plus, cette étude a permis de montrer que ces ligands de la E-sélectiné sont exprimés dans les rafts lipidiques, comme PSGL-1 et la L-sélective des neutrophiles. Ces deux nouveaux ligands sont en cours d'identification. Ils pourraient représenter une nouvelle cible dans le traitement de la leucostase, mais aussi lors d'inflammation chronique ou de métastases. ABSTRACT Leukostasis is alife-threatening complication of acute leukemia, that results from tissue infiltration of leukemic blasts that migrate out of blood flow and interfere with normal tissue functions. The process leading to these complications has been attributed to the overcrowding of leukemic cells in the microcirculation. However, leukostasis more likely results from the adhesive interactions between leukemic blasts and the endothelium. Activated endothelium express adhesion molecules like P- and E-selectin, and leukemic cells themselves can induce the expression of E-selectin on endothelial cells. Selectins are essential in initiating the rolling of intravascular cells on endothelium before firm adhesion and transmigration outside of blood vessels. They interact with specific ligands on leukocyte cell surface. P-selectin glycoprotein ligand-1 (PSGL-1) is common ligand for E-, P- and L-selectin. Recently, CD44, ESL-1 and CD44 were shown to cooperate. ìn supporting mouse neutrophil adhesion to E-selectin. Other E-selectin ligands remain to be identified in humans. Leukemic cells were screened in order to characterize human E-selectin ligands. The interactions of E- and P-selectin correlate with the expression of CLA epitope. Therefore, HECA-452 mAb that recognizes CLA was used for immunopurification. Aglycoprotein of 170 kDa was purified from THP1 and U937 cells, and a protein of 250 kDa from U937 cells. These proteins were also purified by affinity binding to E-selectin/IgM chimera. PSGL-1 bound to E-selectin as expected, but CD43 and CD44 were not always adsorbed on E-selectin chimera, depending on cell types. E-selectin ligands were also shown to be in lipid rafts in leukemic cells, like PSGL-1 and L-selectin in human neutrophils. The 170 kDa protein has been sequenced, and three interesting ligands were among the candidates: ESL-1, CD44 and podocalyxin. These ligands are under investigation, and may represent a new therapeutic target in leukostasis, inflammation or cancer metastasis.

T cell receptor specificity is critical for the development of epidermal gammadelta T cells.

A particular feature of gammadelta T cell biology is that cells expressing T cell receptor (TCR) using specific Vgamma/Vdelta segments are localized in distinct epithelial sites, e.g., in mouse epidermis nearly all gammadelta T cells express Vgamma3/Vdelta1. These cells, referred to as dendritic epidermal T cells (DETC) originate from fetal Vgamma3+ thymocytes. The role of gammadelta TCR specificity in DETC's migration/localization to the skin has remained controversial. To address this issue we have generated transgenic (Tg) mice expressing a TCR delta chain (Vdelta6.3-Ddelta1-Ddelta2-Jdelta1-Cdelta), which can pair with Vgamma3 in fetal thymocytes but is not normally expressed by DETC. In wild-type (wt) Vdelta6.3Tg mice DETC were present and virtually all of them express Vdelta6.3. However, DETC were absent in TCR-delta(-/-) Vdelta6.3Tg mice, despite the fact that Vdelta6.3Tg gammadelta T cells were present in normal numbers in other lymphoid and nonlymphoid tissues. In wt Vdelta6.3Tg mice, a high proportion of in-frame Vdelta1 transcripts were found in DETC, suggesting that the expression of an endogenous TCR-delta (most probably Vdelta1) was required for the development of Vdelta6.3+ epidermal gammadelta T cells. Collectively our data demonstrate that TCR specificity is essential for the development of gammadelta T cells in the epidermis. Moreover, they show that the TCR-delta locus is not allelically excluded.

Immune response to mouse mammary tumor virus in mice lacking the alpha/beta interferon or the gamma interferon receptor.

Mouse mammary tumor virus (MMTV) is a retrovirus which induces a strong immune response and a dramatic increase in the number of infected cells through the expression of a superantigen (SAg). Many cytokines are likely to be involved in the interaction between MMTV and the immune system. In particular, alpha/beta interferon (IFN-alpha/beta) and gamma interferon (IFN-gamma) exert many antiviral and immunomodulatory activities and play a critical role in other viral infections. In this study, we have investigated the importance of interferons during MMTV infection by using mice with a disrupted IFN-alpha/beta or IFN-gamma receptor gene. We found that the SAg response to MMTV was not modified in IFN-alpha/betaR(0/0) and IFN-gammaR(0/0) mice. This was true both for the early expansion of B and T cells induced by the SAg and for the deletion of SAg-reactive cells at later stages of the infection. In addition, no increase in the amount of proviral DNA was detected in tissues of IFN-alpha/betaR(0/0) and IFN-gammaR(0/0) mice, suggesting that interferons are not essential antiviral defense mechanisms during MMTV infection. In contrast, IFN-gammaR(0/0) mice had increased amounts of IL-4 mRNA and an altered usage of immunoglobulin isotypes with a reduced frequency of IgG2a- and IgG3-producing cells. This was associated with lower titers of virus-specific antibodies in serum early after infection, although efficient titers were reached later.

Brain-derived neurotrophic factor enhances the expression of the monocarboxylate transporter 2 through translational activation in mouse cultured cortical neurons.

MCT2 is the predominant neuronal monocarboxylate transporter allowing lactate use as an alternative energy substrate. It is suggested that MCT2 is upregulated to meet enhanced energy demands after modifications in synaptic transmission. Brain-derived neurotrophic factor (BDNF), a promoter of synaptic plasticity, significantly increased MCT2 protein expression in cultured cortical neurons (as shown by immunocytochemistry and western blot) through a translational regulation at the synaptic level. Brain-derived neurotrophic factor can cause translational activation through different signaling pathways. Western blot analyses showed that p44/p42 mitogen-activated protein kinase (MAPK), Akt, and S6 were strongly phosphorylated on BDNF treatment. To determine by which signal transduction pathway(s) BDNF mediates its upregulation of MCT2 protein expression, the effect of specific inhibitors for p38 MAPK, phosphoinositide 3-kinase (PI3K), mammalian target of rapamycin (mTOR), mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) kinase (MEK), p44/p42 MAPK (ERK), and Janus kinase 2 (JAK2) was evaluated. It could be observed that the BDNF-induced increase in MCT2 protein expression was almost completely blocked by all inhibitors, except for JAK2. These data indicate that BDNF induces an increase in neuronal MCT2 protein expression by a mechanism involving a concomitant stimulation of PI3K/Akt/mTOR/S6, p38 MAPK, and p44/p42 MAPK. Moreover, our observations suggest that changes in MCT2 expression could participate in the process of synaptic plasticity induced by BDNF.

Mouse interleukin-2 receptor alpha gene expression. Delimitation of cis-acting regulatory elements in transgenic mice and by mapping of DNase-I hypersensitive sites.

The alpha chain of the interleukin-2 receptor (IL-2R alpha) is a key regulator of lymphocyte proliferation. To analyze the mechanisms controlling its expression in normal cells, we used the 5'-flanking region (base pairs -2539/+93) of the mouse gene to drive chloramphenicol acetyltransferase expression in four transgenic mouse lines. Constitutive transgene activity was restricted to lymphoid organs. In mature T lymphocytes, transgene and endogenous IL-2R alpha gene expression was stimulated by concanavalin A and up-regulated by IL-2 with very similar kinetics. In thymic T cell precursors, IL-1 and IL-2 cooperatively induced transgene and IL-2R alpha gene expression. These results show that regulation of the endogenous IL-2R alpha gene occurs mainly at the transcriptional level. They demonstrate that cis-acting elements in the 5'-flanking region present in the transgene confer correct tissue specificity and inducible expression in mature T cells and their precursors in response to antigen, IL-1, and IL-2. In a complementary approach, we screened the 5' end of the endogenous IL-2R alpha gene for DNase-I hypersensitive sites. We found three lymphocyte specific DNase-I hypersensitive sites. Two, at -0.05 and -5.3 kilobase pairs, are present in resting T cells. A third site appears at -1.35 kilobase pairs in activated T cells. It co-localizes with IL-2-responsive elements identified by transient transfection experiments.

Intracortical connectivity of layer VI pyramidal cells in the barrel cortex of mice

ABSTRACT : The whisker-to-barrel pathway of rodents is formed by a series of somatotopic projections from the mystacial whisker follicles to the layer IV of the primary somatosensory cortex such that each follicle corresponds to a cluster of cortical neurons called barrel. Barrels are present in layer IV but form part of functional columns that comprise the entire depth of the somatosensory cortex. Interestingly, the cortex of the barrelless mouse strain (BRL) is organized such a manner that thalamocortical afferents do not remodel their projections in layer IV and barrels fail to appear. Nevertheless, functionally, a columnar organization persists, indicating that functional columns are not only provided by thalamocortical projections and layer IV cells. Since in the visual cortex of cats, layer VI cells contribute to the response properties of layer IV neurons, we wonder whether layer VI pyramidal cells could contribute to the columnar organization of the primary somatosensory cortex of mice. To address -this question, we morphologically analyzed the distribution of intracortical axon collaterals of layer VI neurons after in-vivo juxtacellular injections of biocytin in the C2 barrel column. Injected hemispheres were tangentially serial cut and intracortical collaterals of individual layer VI neurons were reconstructed at the light microscopic level. The position of axonal boutons was recorded to evaluate the distribution of presumed synaptic contacts. In normal (NOR) mice, cluster analysis shows that layer VI pyramidal cells can be classified in four statistically different clusters of neurons. Moreover, we assume that two classes are formed by cortico-cortical neurons and two classes are formed by cortico-thalamic neurons. Looking at the direction of the main axon in the white matter, we noticed that its orientation correlates perfectly with the type of neuron: cortico-cortical neurons send main axon medially whereas cortico-thalamic neurons send main axon laterally. Performing the same study in the BRL strain, we showed that the BRL mutation affects layer VI pyramidal cells tangentially and radially: the effects of the mutation are illustrated by a significant decrease of the index of colurnnarization and a significant decrease of percentage of boutons in granular and supragranular layers comparing to NOR neurons. In spite of these differences, the same four classes of layer VI neurons have been found in BRL mice. Using a tangential analysis of the boutons distribution, we showed that putative synapses are distributed mainly in the C2 barrel column. This was observed for each layer, type of neuron, cluster or strain, indicating that layer VI pyramidal cells could participate to the functional columnar organization of the barrel cortex. To determine post-synaptic partners of layer VI neurons in layer IV, we conducted an ultrastructural analysis of layer VI-to-IV contacts. We showed that synapses principally occur on spines and spiny dendritic shafts, supposed to belong to excitatory neurons. We furthermore showed that pre-synaptic elements are significantly different between en passant and terminaux contacts, which support hypothesis that terminaux boutons should show longer duration of facilitation than en passant boutons. RÉSUMÉ : Le «whisker-to-barrel pathway» des rongeurs est caractérisé par une série de projections somatotopiques depuis les follicules des moustaches ('whiskers') jusqu'à la couche IV de l'aire somatosensorielle primaire, de telle façon que chaque follicule corresponde à un groupe de neurones corticaux appelés tonneaux (`barrels'). Les tonneaux sont seulement présents en couche IV mais font partie de colonnes fonctionnelles qui s'étendent sur toute la profondeur du cortex somatosensoriel. Chez les souris mutantes barrelless (BRL), le cortex somatosensoriel est organisé de façon telle que lés afférences thalamocorticales ne remodellent pas leurs projections en couche IV et que les tonneaux n'apparaissent pas. Fonctionnellement, pourtant, une organisation en colonnes persiste, ce qui indique que les colonnes fonctionnelles ne sont pas uniquement produites par les projections thalamocorticales et par les cellules de la couche IV. Puisque les cellules de la couche VI contribuent à influencer les réponses des cellules de la couche IV dans le cortex visuel du chat, nous nous sommes demandé si ces cellules ne pourraient pas aussi contribuer à l'organisation en colonnes du cortex somatosensoriel primaire de la souris. Pour répondre à cette question, nous avons analysé de façon morphologique la distribution intracorticale des collatéraux axonaux de neurones de la couche VI. Suite à des injections juxtacellulaires de biocytine in-vivo dans la colonne C2, les hémisphères cérébraux ont été tangentiellement coupés en série et les collatéraux intracorticaux des neurones de la couche VI ont été reconstruits en microscopie optique. La position des boutons axonaux a aussi été enregistrée pour évaluer la distribution des contacts synpptiques potentiels. Chez les souris NOR, une analyse multivariée montre que les cellules pyramidales de la couche VI sont distribuées en quatre classes. Deux de ces classes sont probablement formées de neurons cortico-corticaux, alors que les deux autres sont probablement formées de neurones corticothalamiques. En observant la direction de l'axone principal dans la matière blanche, nous avons noté que son orientation est parfaitement corrélée avec le type supposé de neurone : les neurones corticocorticaux envoient leurs axones principaux médiallement, alors que les neurons cortico-thalamiques envoient leurs axones principaux latéralement. En menant la même étude chez les souris BRL, nous avons montré que la mutation affecte les cellules pyramidales de la couche VI de façon tangentielle, mais aussi radiaire : les effets de 1a mutation se traduisent par une diminution significative de l'index de « columnarization » et de la connectivité en couches granulaire et supragranulaire. Malgré ces différences, les quatre mêmes classes de neurones ont été retrouvées. En utilisant une analyse tangentielle de la distribution des boutons, nous avons montré que les synapses potentielles sont distribuées principalement dans la colonne C2. Cette observation a été faite dans chaque couche, chaque type de neurones, chaque classe de neurones et chaque souche de souris, indicant que les cellules de la couche VI participent certainement à l'organisation en colonne du cortex somatosensoriel. Pour déterminer les partenaires post-synaptiques des cellules de la couche VI en couche IV, nous avons conduit une analyse ultrastructurelle de ces contacts. Nous avons montré que les synapses interviennent principalement sur les épines et sur les dendrites supposés appartenir à des cellules excitatrices. Nous avons aussi montré que les éléments pré-synaptiques de ces synapses sont significativement differents selon le type de bouton, en passant ou terminal, ce qui supporte l'hypothèse que les boutons terminaux seraient capables d'une plus longue facilitation.

Breast stem cells and hormonal mechanisms in carcinogenesis

A woman's risk of breast cancer is strongly affected by her reproductive history. The hormonal milieu is also a key determinant of the course of the disease. Combining mouse genetics with tissue recombination techniques, we have established that the female reproductive hormones, estrogens, progesterone, and prolactin, act sequentially on the mammary epithelium to trigger distinct developmental steps. The hormones impinge directly on a subset of luminal mammary epithelial cells that express the respective hormone receptors and act as sensor cells translating and amplifying systemic signals into local stimuli. Local signaling is stage and age specific. During puberty, estrogens promote proliferation using the EGF family member, amphiregulin, as essential paracrine mediator. In adulthood, progesterone, rather than estrogen, is the major inducer of stem cell activation and cell proliferation of the mammary epithelium. Hormonal signaling modulates crucial developmental pathways that impinge on mammary stem cell populations, while Notch signaling, by inhibiting p63, is central to mammary cell fate determination. Cell proliferation occurs in two waves. The first results from direct stimulation of the small fraction of hormone receptor positive cells. It is followed by a second wave of progesterone-induced proliferation involving mostly hormone receptor negative cells, in which RANKL is a key mediator. A model in which repeated activation of paracrine signaling by progesterone with resulting stem cell activation promotes breast carcinogenesis is proposed.

Quantification, self-renewal, and genetic tracing of FL1+ tumor-initiating cells in a large cohort of human gliomas.

Evidence has emerged that the initiation and growth of gliomas is sustained by a subpopulation of cancer-initiating cells (CICs). Because of the difficulty of using markers to tag CICs in gliomas, we have previously exploited more robust phenotypic characteristics, including a specific morphology and intrincic autofluorescence, to identify and isolate a subpopulation of glioma CICs, called FL1(+). The objective of this study was to further validate our method in a large cohort of human glioma and a mouse model of glioma. Seventy-four human gliomas of all grades and the GFAP-V(12)HA-ras B8 mouse model were analyzed for in vitro self-renewal capacity and their content of FL1(+). Nonneoplastic brain tissue and embryonic mouse brain were used as control. Genetic traceability along passages was assessed with microsatellite analysis. We found that FL1(+) cells from low-grade gliomas and from control nonneoplasic brain tissue show a lower level of autofluorescence and undergo a restricted number of cell divisions before dying in culture. In contrast, we found that FL1(+) cells derived from many but not all high-grade gliomas acquire high levels of autofluorescence and can be propagated in long-term cultures. Moreover, FL1(+) cells show a remarkable traceability over time in vitro and in vivo. Our results show that FL1(+) cells can be found in all specimens of a large cohort of human gliomas of different grades and in a model of genetically induced mouse glioma as well as nonneoplastic brain. However, their self-renewal capacity is variable and seems to be dependent on the tumor grade.