Stem cell defects in parthenogenetic peri-implantation embryos

In vivo and in vitro differentiation of uniparental embryonic stem cells into hematopoietic and neural cell types

The biological role of genomic imprinting in adult tissue is central to the consideration of transplanting uniparental embryonic stem (ES) cell-derived tissues. We have recently shown that both maternal (parthenogenetic/gynogenetic) and paternal (androgenetic) uniparental ES cells can differentiate, both in vivo in chimeras and in vitro, into adult-repopulating hematopoietic stem and progenitor cells. This suggests that, at least in some tissues, the presence of two maternal or two paternal genomes does not interfere with stem cell function and tissue homeostasis in the adult. Here, we consider implications of the contribution of uniparental cells to hematopoiesis and to development of other organ systems, notably neural tissue for which consequences of genomic imprinting are associated with a known bias in development and behavioral disorders. Our findings so far indicate that there is little or no limit to the differentiation potential of uniparental ES cells outside the normal developmental paradigm. As a potentially donor MHC-matching source of tissue, uniparental transplants may provide not only a clinical resource but also a unique tool to investigate aspects of genomic imprinting in adults.

Ectopic expression of Fgf3 leads to aberrant lineage segregation in the mouse parthenote preimplantation embryos

BACKGROUND

Parthenogenetic mammalian embryos were reported to die in utero no later than the 25-somite stage due to abnormal development of both embryonic and extraembryonic lineages. Interestingly, it has been shown that parthenogenetic ICM cells tend to differentiate more into primitive endoderm cells and less into epiblast and ES cells. Hence we are interested in studying the molecular mechanisms underlying lineage defects of parthenotes.

RESULTS

We found that parthenote inner cell masses (ICMs) contained decreased numbers of Sox2(+) /Nanog(+) epiblast cells but increased numbers of Gata4(+) primitive endoderm cells, indicating an unusual lineage segregation. We demonstrate for the first time that the increased Gata4 level in parthenotes may be explained by the strong up-regulation of Fgf3 and Fgfr2 phosphorylation. Inhibition of Fgfr2 activation by SU5402 in parthenotes restored normal Nanog and Gata4 levels without affecting Fgf3, indicating that Fgf3 is upstream of Fgfr2 activation. In parthenote trophectoderm, we detected normal Cdx2 but ectopic Gata4 expression and reduced Elf5 and Tbr2(Eomes) levels.

CONCLUSIONS

Taken together, our work provides for the first time the insight into the molecular mechanisms of the developmental defects of parthenogenetic embryos in both the trophectoderm and ICM.

Germline competency of parthenogenetic embryonic stem cells from immature oocytes of adult mouse ovary

Parthenogenetic embryonic stem cells (pESCs) have been generated in several mammalian species from parthenogenetic embryos that would otherwise die around mid-gestation. However, previous reports suggest that pESCs derived from in vivo ovulated (IVO) mature oocytes show limited pluripotency, as evidenced by low chimera production, high tissue preference and especially deficiency in germline competence, a critical test for genetic integrity and pluripotency of ESCs. Here, we report efficient generation of germline-competent pESC lines (named as IVM pESCs) from parthenogenetic embryos developed from immature oocytes of adult mouse ovaries following in vitro maturation (IVM) and artificial activation. In contrast, pESCs derived from IVO oocytes show defective germline competence, consistent with previous reports. Further, IVM pESCs resemble more ESCs from fertilized embryos (fESCs) than do IVO pESCs on genome-wide DNA methylation and global protein profiles. In addition, IVM pESCs express higher levels of Blimp1, Lin28 and Stella, relative to fESCs, and in their embryoid bodies following differentiation. This may indicate differences in differentiation potentially to the germline. The mechanisms for acquisition of pluripotency and germline competency of IVM pESCs from immature oocytes remain to be determined.

Human fetal liver stromal cells that overexpress bFGF support growth and maintenance of human embryonic stem cells

In guiding hES cell technology toward the clinic, one key issue to be addressed is to culture and maintain hES cells much more safely and economically in large scale. In order to avoid using mouse embryonic fibroblasts (MEFs) we isolated human fetal liver stromal cells (hFLSCs) from 14 weeks human fetal liver as new human feeder cells. hFLSCs feeders could maintain hES cells for 15 passages (about 100 days). Basic fibroblast growth factor (bFGF) is known to play an important role in promoting self-renewal of human embryonic stem (hES) cells. So, we established transgenic hFLSCs that stably express bFGF by lentiviral vectors. These transgenic human feeder cells — bFGF-hFLSCs maintained the properties of H9 hES cells without supplementing with any exogenous growth factors. H9 hES cells culturing under these conditions maintained all hES cell features after prolonged culture, including the developmental potential to differentiate into representative tissues of all three embryonic germ layers, unlimited and undifferentiated proliferative ability, and maintenance of normal karyotype. Our results demonstrated that bFGF-hFLSCs feeder cells were central to establishing the signaling network among bFGF, insulin-like growth factor 2 (IGF-2), and transforming growth factor β (TGF-β), thereby providing the framework in which hES cells were instructed to self-renew or to differentiate. We also found that the conditioned medium of bFGF-hFLSCs could maintain the H9 hES cells under feeder-free conditions without supplementing with bFGF. Taken together, bFGF-hFLSCs had great potential as feeders for maintaining pluripotent hES cell lines more safely and economically.

Ribonucleoprotein localization in mouse oocytes

RNA molecules rarely function alone in cells. For most RNAs, their function requires formation of various ribonucleoprotein (RNP) complexes. For example, mRNP composition can determine mRNA localization, translational repression, level of translation or mRNA stability. RNPs are usually studied by biochemical methods. However, biochemical approaches are unsuitable for some model systems, such as mammalian oocytes and early embryos, due to the small amounts that can be obtained for experimental analysis. In such cases, microscopic techniques are often used to learn about RNPs. Here, we present a review of immunostaining, fluorescence in situ hybridization with subcellular resolution and a combination of both, with emphasis on the mouse oocyte and early embryos models. Application of these techniques to whole-mount fixed oocytes and early embryos can provide information about RNP composition and localization with three-dimensional resolution.

Correlation of expression and methylation of imprinted genes with pluripotency of parthenogenetic embryonic stem cells

Mammalian parthenogenetic embryos (pE) are not viable due to placental deficiency, presumably resulting from lack of paternally expressed imprinted genes. Pluripotent parthenogenetic embryonic stem (pES) cells derived from pE could advance regenerative medicine by avoiding immuno-rejection and ethical roadblocks. We attempted to explore the epigenetic status of imprinted genes in the generation of pES cells from parthenogenetic blastocysts, and its relationship to pluripotency of pES cells. Pluripotency was evaluated for developmental and differentiation potential in vivo, based on contributions of pES cells to chimeras and development to day 9.5 of pES fetuses complemented by tetraploid embryos (TEC). Consistently, pE and fetuses failed to express paternally expressed imprinted genes, but pES cells expressed those genes in a pattern resembling that of fertilized embryos (fE) and fertilized embryonic stem (fES) cells derived from fE. Like fE and fES cells, but unlike pE or fetuses, pES cells and pES cell-fetuses complemented by TEC exhibited balanced methylation of Snrpn, Peg1 and U2af1-rs1. Coincidently, global methylation increased in pE but decreased in pES cells, further suggesting dramatic epigenetic reprogramming occurred during isolation and culture of pES cells. Moreover, we identified decreased methylation of Igf2r, Snrpn, and especially U2af1-rs1, in association with increased contributions of pES cells to chimeras. Our data show that in vitro culture changes epigenetic status of imprinted genes during isolation of pES cells from their progenitor embryos and that increased expression of U2af1-rs1 and Snrpn and decreased expression of Igf2r correlate with pluripotency of pES cells.

Biological differences between in vitro produced bovine embryos and parthenotes

Parthenotes may represent an alternate ethical source of stem cells, once biological differences between parthenotes and embryos can be understood. In this study, we analyzed development, trophectoderm (TE) differentiation, apoptosis/necrosis, and ploidy in parthenotes and in vitro produced bovine embryos. Subsequently, using real-time PCR, we analyzed the expression of genes expected to underlie the observed differences at the blastocyst stage. In vitro matured oocytes were either fertilized or activated with ionomycin +6-DMAP and cultured in simple medium. Parthenotes showed enhanced blastocyst development and diploidy and reduced TE cell counts. Apoptotic and necrotic indexes did not vary, but parthenotes evidenced a higher relative proportion of apoptotic cells between inner cell mass and TE. The pluripotence-related POU5F1 and the methylation DNMT3A genes were downregulated in parthenotes. Among pregnancy recognition genes, TP-1 was upregulated in parthenotes, while PGRMC1 and PLAC8 did not change. Expression of p66(shc) and BAX/BCL2 ratio were higher, and p53 lower, in parthenotes. Among metabolism genes, SLC2A1 was downregulated, while AKR1B1, PTGS2, H6PD, and TXN were upregulated in parthenotes, and SLC2A5 did not differ. Among genes involved in compaction/blastulation, GJA1 was downregulated in parthenotes, but no differences were detected within ATP1A1 and CDH1. Within parthenotes, the expression levels of SLC2A1, TP-1, and H6PD, and possibly AKR1B1, resemble patterns described in female embryos. The pro-apoptotic profile is more pronounced in parthenotes than in embryos, which may differ in their way to channel apoptotic stimuli, through p66(shc) and p53 respectively, and in their mechanisms to control pluripotency and de novo methylation.

Parthenogenesis as an approach to pluripotency: advantages and limitations involved

Embryonic stem cells (ESCs) are invaluable cells derived from the inner cell mass of the mammalian blastocyst. They have nearly indefinite self-renewal, retain their developmental potential after prolonged periods in culture and display great plasticity that allow them to differentiate into all cell types of the body. They provide exciting opportunities to develop unique models for developmental research and hold great potential for cell and tissue replacement therapy. However, these unique cells cannot be obtained without destroying an embryo and, despite the potential therapeutic usefulness, their derivation in the human raises substantial ethical as well as legal and political concerns because it unavoidably involves the destruction of viable embryos. In the recent years a number of scientific proposals that do not require the generation and subsequent destruction of human embryos have been put forward in an attempt to fill the gap between ethical questions and potential scientific and medical benefits. In this review we briefly summarize data obtained from the literature related to these different alternative approaches and focus in more details on our experience in the derivation of parthenothes, as a possible alternative source for pluripotent cells, discussing the advantages as well as the limits of these cell lines.

Parthenotes as a source of embryonic stem cells

The derivation and study of human embryonic stem cell lines, despite their potential therapeutic usefulness, raise considerable ethical, religious, legal and political concerns because it inevitably leads to the destruction of viable embryos. In an attempt to bridge the division between ethical questions and potential scientific and medical benefits, considerable efforts have been devoted to the search for alternative sources of pluripotent cell lines. In this review we discuss the use of artificial parthenogenesis as a way to create entities, called parthenotes, that may represent an alternative ethical source for pluripotent cell lines. We describe the biological differences between parthenotes and embryos, in order to provide a rationale for the discussion on whether their use can be acceptable as a source of stem cells. We present data derived from animal models on the extent parthenogenetic stem cells are similar to biparental cell lines and discuss these aspects in the context of their extension to the human species. Finally, we present experiments recently carried out in our laboratory that allowed us to generate human parthenotes through artificial activation of human oocytes and to use them as a source for the derivation of parthenogenetic pluripotent cell lines.

Activation of paternally expressed imprinted genes in newly derived germline-competent mouse parthenogenetic embryonic stem cell lines

Parthenogenetic embryonic stem (pES) cells provide a valuable in vitro model system for studying the molecular mechanisms that underlie genomic imprinting. However, the pluripotency of pES cells and the expression profiles of paternally expressed imprinted genes have not been fully explored. In this study, three mouse pES cell lines were established and the differentiation potential of these cells in extended culture was evaluated. The undifferentiated cells had a normal karyotype and homozygous genome, and expressed ES-cell-specific molecular markers. The cells remained undifferentiated after more than 50 passages and exhibited pluripotent differentiation capacity. All three lines of the established ES cells produced teratomas; two lines of ES cells produced chimeras and germline transmission. Furthermore, activation of the paternally expressed imprinted genes Snrpn, U2af1-rs1, Peg3, Impact, Zfp127, Dlk1 and Mest in these cells was detected. Some paternally expressed imprinted genes were found to be expressed in the blastocyst stage of parthenogenetically activated embryos in vitro and their expression level increased with extended pES cell culture. Furthermore, our data show that the activation of these paternally expressed imprinted genes in pES cells was associated with a change in the methylation of the related differentially methylated regions. These findings provide direct evidence for the pluripotency of pES cells and demonstrate the association between the DNA methylation pattern and the activation of paternally expressed imprinted genes in pES cells. Thus, the established ES cell lines provide a valuable model for studying epigenetic regulation in mammalian development.

High-throughput screen for genes predominantly expressed in the ICM of mouse blastocysts by whole mount in situ hybridization

Mammalian preimplantation embryos provide an excellent opportunity to study temporal and spatial gene expression in whole mount in situ hybridization (WISH). However, large-scale studies are made difficult by the size of the embryos ( approximately 60mum diameter) and their fragility. We have developed a chamber system that allows parallel processing of embryos without the aid of a microscope. We first selected 91 candidate genes that were transcription factors highly expressed in blastocysts, and more highly expressed in embryonic (ES) than in trophoblast (TS) stem cells. We then used the WISH to identify 48 genes expressed predominantly in the inner cell mass (ICM) and to follow several of these genes in all seven preimplantation stages. The ICM-predominant expressions of these genes suggest their involvement in the pluripotency of embryonic cells. This system provides a useful tool to a systematic genome-scale analysis of preimplantation embryos.

Generation and characterization of pluripotent stem cells from cloned bovine embryos

Bovine embryonic stem (ES) cell lines reported to date vary in morphology and marker expression (e.g., alkaline phosphatase [ALPL], stage-specific embryonic antigen 4 [SSEA4], and OCT4) that normally are associated with the undifferentiated, pluripotent state. These observations suggest that the proper experimental conditions for consistently producing bovine ES cells have not been identified. Here, we report three bovine ES cell lines, one from in vitro-fertilized and two from nuclear transfer embryos. These bovine ES cells grew in large, multicellular colonies resembling the mouse ES and embryonic germ (EG) cells and human EG cells. Throughout the culture period, most of the cells within the colonies stained positive for ALPL and the cell surface markers SSEA4 and OCT4. The staining patterns of nuclear transfer ES cells were identical to those of the blastocysts generated in vitro yet different from most previously reported bovine ES cell lines, which were either negative or not detected. After undifferentiated culture for more than 1 yr, these cells maintained the ability to differentiate into embryoid bodies and derivatives of all three EG layers, thus demonstrating their pluripotency. However, unlike the mouse and human ES cells, following treatment with trypsin, type IV collagenase, or protease E, our bovine ES cells failed to self-renew and became spontaneously differentiated. Presumably, this resulted from an interruption of the self-renewal pathway. In summary, we generated pluripotent bovine ES cells with morphology similar to those of established ES cells in humans and mice as well as marker-staining patterns identical to those of the bovine blastocysts.

Effect of ploidy and parental genome composition on expression of Oct-4 protein in mouse embryos

The transcription factor Oct-4 is expressed in germ cells and also is considered as a marker for pluripotency of stem cells. We first examined dynamics of Oct-4 protein expression during preimplantation development using both Western blot analysis, and immunofluorescence staining. We show that intact Oct-4 protein is not detected in either ovulated mature oocytes, or in zygotes and 2-4-cell embryos, which are the only known totipotent cell types in mammals. This finding is unexpected, since Oct-4 has been proposed to play a role in the control of totipotency. The results suggest that Oct-4 is not indispensable for fertilization and early cleavage. Rather, expression of Oct-4 protein is first detected in the nuclei of 8-16 cell morula, increases in early blastocysts, and declines in late blastocysts, in which most Oct-4 protein is confined to the inner cell mass (ICM) region, consistent with previous findings. We further compared Oct-4 protein expression in diploid and tetraploid blastocysts derived from normal fertilization or parthenogenesis, as well as expression in diploid androgenetic blastocysts. Expression levels and localization of Oct-4 protein are similar in both diploid and tetraploid early blastocysts, regardless of whether blastocysts are derived from fertilization or parthenogenesis. Androgenetic diploid blastocysts also express similar levels of Oct-4. Late blastocysts generated by both fertilization and parthenogenesis show a similar pattern of Oct-4 expression, suggesting that paternal genome activation is not required for Oct-4 expression. Expression of Oct-4 protein does not differ between diploid and tetraploid embryos, indicating that tetraploidy does not influence Oct-4 expression. Thus, expression of Oct-4 protein is initiated at morula stage in preimplantation embryos and completely controlled by a mechanism activated in oocytes. Downregulation of Oct-4 expression coincides with differentiation of trophectoderm. Similar profiles of Oct-4 expression observed in embryos with different ploidy and genome composition, are suggestive of Oct-4 being necessary but not sufficient for developmental potency.

Insulin-like growth factor 2 expressed in a novel fetal liver cell population is a growth factor for hematopoietic stem cells

Hematopoietic stem cells (HSCs) undergo dramatic expansion during fetal liver development, but attempts to expand their numbers ex vivo have failed. We hypothesized that unidentified fetal liver cells produce growth factors that support HSC proliferation. Here we describe a novel population of CD3+ and Ter119- day-15 fetal liver cells that support HSC expansion in culture, as determined by limiting dilution mouse reconstitution analyses. DNA array experiments showed that, among other proteins, insulin-like growth factor 2 (IGF-2) is specifically expressed in fetal liver CD3+ cells but not in several cells that do not support HSCs. Treatment of fetal liver CD3+Ter119- cells with anti-IGF-2 abrogated their HSC supportive activity, suggesting that IGF-2 is the key molecule produced by these cells that stimulates HSC expansion. All mouse fetal liver and adult bone marrow HSCs express receptors for IGF-2. Indeed, when combined with other growth factors, IGF-2 supports a 2-fold expansion of day-15 fetal liver Lin-Sca-1+c-Kit+ long-term (LT)-HSC numbers. Thus, fetal liver CD3+Ter119- cells are a novel stromal population that is capable of supporting HSC expansion, and IGF-2, produced by these cells, is an important growth factor for fetal liver and, as we show, adult bone marrow HSCs.

Multilineage potential of homozygous stem cells derived from metaphase II oocytes

Human stem cells derived from human fertilized oocytes, fetal primordial germ cells, umbilical cord blood, and adult tissues provide potential cell-based therapies for repair of degenerating or damaged tissues. However, the diversity of major histocompatibility complex (MHC) antigens in the general population and the resultant risk of immune-mediated rejection complicates the allogenic use of established stem cells. We assessed an alternative approach, employing chemical activation of nonfertilized metaphase II oocytes for producing stem cells homozygous for MHC. By using F1 hybrid mice (H-2-B/D), we established stem cell lines homozygous for H-2-B and H-2-D, respectively. The undifferentiated cells retained a normal karyotype, expressed stage-specific embryonic antigen-1 and Oct4, and were positive for alkaline phosphatase and telomerase. Teratomatous growth of these cells displayed the development of a variety of tissue types encompassing all three germ layers. In addition, these cells demonstrated the potential for in vitro differentiation into endoderm, neuronal, and hematopoietic lineages. We also evaluated this homozygous stem cell approach in human tissue. Five unfertilized blastocysts were derived from a total of 25 human oocytes, and cells from one of the five hatched blastocysts proliferated and survived beyond two passages. Our studies demonstrate a plausible "homozygous stem cell" approach for deriving pluripotent stem cells that can overcome the immune-mediated rejection response common in allotransplantation, while decreasing the ethical concerns surrounding human embryonic stem cell research.

Effects of fibroblast growth factor 2 and insulin-like growth factor II on the development of parthenogenetic mouse embryos in vitro

Most parthenogenetic embryos (PEs) in mammals die shortly after implantation, and this failure to develop is associated with genomic imprinting. We have examined the influence of human recombinant basic fibroblast growth factor 2 (FGF-2) and human recombinant insulin-like growth factor II (ICF-II) on the development of (CBA x C57BL/6)F1 parthenogenetic mouse embryos. Embryos were treated in vitro at the morula stage with different doses of FGF-2 and, after their development to blastocysts, transferred to pseudopregnant recipients. The optimal doses of FGF-2 did not affect the number of forming and implanting blastocysts, but increased, from 20 to 42%, the number of embryos developing to somite stages. PEs (18-21 somites) treated with an optimal dose of FGF-2 were explanted for further development in culture by treatment with the second growth factor, IGF-II. Eighty-three percent of those embryos cultured with IGF-II (2.5 microg/ml) developed to 35 or more somites, as compared with 36% of embryos cultured without any growth factors (P < 0.01). Also, a significantly higher proportion of PEs developed to 40-50 somites in this case. These results show that the in vitro treatment of PEs with FGF-2 at the morula stage increases the number of somite embryos, and the second treatment of somite PEs with IGF-II in culture medium prolongs their development significantly.

Functional analysis of trophoblast giant cells in parthenogenetic mouse embryos

Diploid mouse embryos containing only maternal DNA (parthenotes) fail, in part, because the inner cell mass does not induce the trophoblast to grow. In this study, we asked whether any of the defects in parthenotes may arise from alterations in trophoblast function. We examined the expression of genes important for normal trophoblast function and found several trophoblast genes that were expressed at normal levels in the primary trophoblast cells of parthenotes: E-cadherin, a cell adhesion molecule, was expressed normally in both the ICM and trophectoderm of parthenogenetic blastocysts and blastocyst outgrowths; the gene for Hxt, a basic helix-loop-helix factor that regulates trophoblast development, was expressed in both zygotic and parthenogenetic giant cells; placental lactogen-1, a hormone that is normally secreted by trophoblast giant cells, was expressed in most of both parthenogenetic and normal trophoblast cells; and the 92 kDa matrix metalloproteinase, gelatinase B, also known as MMP-9, was secreted at equivalent levels by both zygotic and parthenogenetic blastocyst outgrowths. However, once the outgrowths had developed, a subpopulation of trophoblast cells in parthenogenetic embryos had decreased DNA replication and significantly fewer nucleoli per nucleus than did zygotic embryos. Moreover, the parthenogenetic trophoblast cells growing out from blastocysts had a decreased viability in culture. These data suggest that, although parthenogenetic embryos are able to initiate primary trophoblast differentiation, the stability and continued differentiation of trophoblast giant cells may be abnormal. Our data support the hypothesis that the deficiency of secondary trophoblast giant cells may contribute to the decline of parthenogenetic embryos and suggest that the factors controlling this subset of trophoblast are distinct from those for primary trophoblast.

The mouse homeobox gene, Gbx2: genomic organization and expression in pluripotent cells in vitro and in vivo

The Gbx2 homeodomain is widely conserved in metazoans. We investigated the mouse Gbx2 locus by isolation and characterization of genomic clones and by physical localization to the genome. The Gbx2 gene contained a single intron that separated the proposed functional protein domains. This organization was conserved with human GBX2. Physical localization of Gbx2 to Chromosome 1C5-E1 indicated that the genomic relationship between the linked Gbx2 and En1 genes differs between mouse and human, making it unlikely to be of functional significance. We also extended the known expression pattern of Gbx2 beyond the gastrulation stage embryo and the developing CNS to pluripotent cells in vitro and in vivo. Gbx2 expression was demonstrated in undifferentiated embryonic stem cells but was downregulated in differentiated cell populations. In the embryo, Gbx2 expression was detected before primitive streak formation, in the inner cell mass of the preimplantation embryo. Gbx2 is therefore a candidate control gene for cell pluripotency and differentiation in the embryo.

Cooperative interactions between extracellular matrix, integrins and parathyroid hormone-related peptide regulate parietal endoderm differentiation in mouse embryos

The outgrowth of parietal endoderm (PE) cells from precursor endodermal cells is one of the first differentiation events that occur in mouse embryos. We have analyzed the molecular determinants of this process by placing isolated inner cell masses (ICMs) on defined extracellular matrix substrata in microdrop cultures. Differentiation and outgrowth of PE required a fibronectin substratum. Laminin supported the adhesion and outgrowth of visceral endoderm (VE) and actively suppressed the differentiation of PE in mixtures of fibronectin and laminin. Collagen type IV, gelatin, vitronectin or entactin supported little or no endodermal outgrowth. Trophectoderm (TE) cells have been implied to be important in PE induction in vivo. We found that recombination of ICMs in culture with TE cells, or with medium conditioned by TE cells, greatly increased the differentiation of PE. TE cells stimulated PE outgrowth on substrata other than fibronectin. One cytokine secreted by trophoblast and endodermal cells, parathyroid hormone-related peptide (PTHrP), was critical for outgrowth on any substratum. A function-perturbing antibody to PTHrP reduced the number of PE cells, whereas the addition of PTHrP increased that number. Furthermore, addition of PTHrP changed the substratum requirements for outgrowth, making laminin, vitronectin and low concentrations of fibronectin permissive for PE outgrowth. Immunostaining with anti-integrin antibodies showed that fully differentiated PE cells outgrowing on fibronectin expressed alpha 5, alpha 6 and alpha v beta 3 integrins. However, analysis of outgrowths in the presence of function-perturbing antibodies to alpha 5, alpha 6 and alpha v beta 3 integrins showed that these integrins directed PE outgrowth only on fibronectin, laminin and vitronectin substrata, respectively. We have shown that there is a cooperative interplay of extracellular matrix, integrins and PTHrP that modulates PE outgrowth.