The Genotype-Tissue Expression (GTEx) project was established to characterize genetic effects on the transcriptome across human tissues and to link these regulatory mechanisms to trait and disease associations. Here, we present analyses of the version 8 data, examining 15,201 RNA-sequencing samples from 49 tissues of 838 postmortem donors. We comprehensively characterize genetic associations for gene expression and splicing in cis and trans, showing that regulatory associations are found for almost all genes, and describe the underlying molecular mechanisms and their contribution to allelic heterogeneity and pleiotropy of complex traits. Leveraging the large diversity of tissues, we provide insights into the tissue specificity of genetic effects and show that cell type composition is a key factor in understanding gene regulatory mechanisms in human tissues.

Stem cells are a valuable resource for treating disease, but limited access to stem cells from tissues such as brain restricts their utility. Here, we injected marrow stromal cells (MSCs) into the lateral ventricle of neonatal mice and asked whether these multipotential mesenchymal progenitors from bone marrow can adopt neural cell fates when exposed to the brain microenvironment. By 12 days postinjection, MSCs migrated throughout the forebrain and cerebellum without disruption to the host brain architecture. Some MSCs within the striatum and the molecular layer of the hippocampus expressed glial fibrillary acidic protein and, therefore, differentiated into mature astrocytes. MSCs also populated neuron rich regions including the Islands of Calleja, the olfactory bulb, and the internal granular layer of the cerebellum. A large number of MSCs also were found within the external granular layer of the cerebellum. In addition, neurofilament positive donor cells were found within the reticular formation of the brain stem, suggesting that MSCs also may have differentiated into neurons. Therefore, MSCs are capable of producing differentiated progeny of a different dermal origin after implantation into neonatal mouse brains. These results suggest that MSCs are potentially useful as vectors for treating a variety of central nervous system disorders.Stem cells are characterized by a capacity to self-renew and to generate progeny capable of differentiating into multiple yet distinct cell lineages. Although stem cells derived from early embryos can differentiate into all somatic cell types (1-3), those derived from adult tissues are thought to produce only the cell lineages characteristic of the tissues wherein they reside. For example, hematopoietic stem cells resident in bone marrow give rise to only blood elements (4). Stem cells also have been identified in the gut, gonads, skin, and brain of adults (5, 6). Neural stem cells have been proposed as useful vectors for treating diseases of the central nervous system, but their lack of accessibility limits their utility.Recently, several reports demonstrated that some stem cells may have greater plasticity than previously envisioned. Eglitis and Mezey reported that a few donor cells from bone marrow transfused into immunodeficient mice were recovered as macroglia in the host brain (7). However, the nature of the marrow cells that engrafted in brain was not determined. Azizi et al. (8) infused rat brains with human marrow stromal cells (MSCs) that are capable of expansion, self-renewal, and differentiation into multiple mesenchymal cell lineages (9, 10). The human MSCs migrated in brain in a manner similar to paraventricular astrocytes, but whether the cells adopted neural cell fates was not defined. In related experiments, Bjornson et al. showed that neural stem cells differentiated into myeloid and lymphoid cell lineages after transplantation into the hematopoietic system of irradiated hosts (11). Together, these findings suggest that it may be possible to reconsti...

Mesenchymal stem cells (MSCs) are typically enriched from bone marrow via isolation of the plastic adherent, fibroblastoid cell fraction. However, plastic adherent cultures elaborated from murine bone marrow are an admixture of fibroblastoid and hematopoietic cell types. Here we report a reliable method based on immunodepletion to fractionate fibroblastoid cells from hematopoietic cells within plastic adherent murine marrow cultures. The immunodepleted cells expressed the antigens Sca-1, CD29, CD44, CD81, CD106, and the stem cell marker nucleostemin (NST) but not CD11b, CD31, CD34, CD45, CD48, CD90, CD117, CD135, or the transcription factor Oct-4. They were also capable of differentiating into adipocytes, chondrocytes, and osteoblasts in vitro as well as osteoblasts/osteocytes in vivo. Therefore, immunodepletion yields a cell population devoid of hematopoietic and endothelial cells that is phenotypically and functionally equivalent to MSCs. The immunodepleted cells exhibited a population doubling time of approximately 5-7 days in culture. Poor growth was due to the dramatic down regulation of many genes involved in cell proliferation and cell cycle progression as a result of immunodepletion. Exposure of immunodepleted cells to fibroblast growth factor 2 (FGF2) but not insulin-like growth factor (IGF), murine stem cell factor, or leukemia inhibitory factor (LIF) significantly increased their growth rate. Moreover, 82% of the transcripts down regulated by immunodepletion remain unaltered in the presence of FGF2. Exposure to the later also reversibly inhibited the ability of the immunodepleted cells to differentiate into adipocytes, chondrocytes, and osteoblasts in vitro. Therefore, FGF2 appears to function as a mitogen and self-maintenance factor for murine MSCs enriched from bone marrow by negative selection.

Human marrow stromal cells (MSCs) were isolated from posterior illiac crest marrow aspirates obtained from 17 healthy donors, ages 19-45 years, with no apparent physical disability. First passage hMSCs exhibited growth rates in vitro that varied up to 12-fold between donors. No correlation between growth rate and the age or gender of the donor was evident (P

Bone marrow stroma contains a unique cell population, referred to as marrow stromal cells (MSCs), capable of differentiating along multiple mesenchymal cell lineages. A standard liquid culture system has been developed to isolate MSCs from whole marrow by their adherence to plastic wherein the cells grow as clonal populations derived from a single precursor termed the colony-forming-unit fibroblast (CFU-F). Using this liquid culture system, we demonstrate that the relative abundance of MSCs in the bone marrow of five commonly used inbred strains of mice varies as much as 10-fold, and that the cells also exhibit markedly disparate levels of alkaline phosphatase expression, an early marker of osteoblast differentiation. For each strain examined, the method of isolating MSCs by plastic adherence yields a heterogeneous cell population. These plastic adherent cells also exhibit widely varying growth kinetics between the different strains. Importantly, of three inbred strains commonly used to prepare transgenic mice that we examined, only cells derived from FVB/N marrow readily expand in culture. Further analysis of cultures derived from FVB/N marrow showed that most plastic adherent cells express CD11b and CD45, epitopes of lymphohematopoietic cells. The later consists of both pre-B-cell progenitors, granulocytic and monocytic precursors, and macrophages. However, a subpopulation of the MSCs appear to represent bona fide mesenchymal progenitors, as cells can be induced to differentiate into osteoblasts and adipocytes after exposure to dexamethasone and into myoblasts after exposure to amphotericin B. Our results point to significant strain differences in the properties of MSCs and indicate that standard methods cannot be applied to murine bone marrow to isolate relatively pure populations of MSCs.

This study evaluates functional recovery after transplanting human bone marrow-derived stromal cells (BMSCs) into contusion models of spinal cord injury (SCI). The authors used a high-throughput process to expand BMSCs and characterized them by flow cytometry, ELISA, and gene expression. They found that BMSCs secrete neurotrophic factors and cytokines with therapeutic potential for cell survival and axon growth. In adult immune-suppressed rats, mild, moderate, or severe contusions were generated using the MASCIS impactor. One week following injury, 0.5 to 1 x 106 BMSCs were injected into the lesioned spinal cord; control animals received vehicle injection. Biweekly behavioral tests included the Basso, Beattie, and Bresnahan Locomotor Rating Scale (BBB), exploratory rearing, grid walking, and thermal sensitivity. Animals receiving moderate contusions followed by BMSC grafts showed significant behavioral recovery in BBB and rearing tests when compared to controls. Animals receiving BMSC grafts after mild or severe contusion showed trends toward improved recovery. Immunocytochemistry identified numerous axons passing through the injury in animals with BMSC grafts but few in controls. BMSCS were detected at 2 weeks after transplantation; however, at 11 weeks very few grafted cells remained. The authors conclude that BMSCs show potential for repairing SCI. However, the use of carefully characterized BMSCs improved transplantation protocols ensuring BMSC, survival, and systematic motor and sensory behavioral testing to identify robust recovery is imperative for further improvement.

Mesenchymal stem cells (MSCs) isolated from the bone marrow of adult organisms are capable of differentiating into adipocytes, chondrocytes, myoblasts, osteoblasts, and hematopoiesis-supporting stroma. We recently demonstrated that MSCs also adopt glial cell fates when transplanted into the developing central nervous system and hence can produce tissue elements derived from a separate embryonic layer. Despite these remarkable properties, it has been difficult to establish specific criteria to characterize MSCs. Using a modified protocol for microserial analysis of gene expression, we cataloged 2,353 unique genes expressed by a single cell-derived colony of undifferentiated human MSCs. This analysis revealed that the MSC colony simultaneously expressed transcripts characteristic of various mesenchymal cell lineages including chondrocytes, myoblasts, osteoblasts, and hematopoiesis-supporting stroma. Therefore, the profile of expressed transcripts reflects the developmental potential of the cells. Additionally, the MSC colony expressed mRNAs characteristic of endothelial, epithelial and neuronal cell lineages, a combination that provides a unique molecular signature for the cells. Other expressed transcripts included various products involved in wound repair as well as several neurotrophic factors. A total of 268 novel transcripts were also identified, one of which is the most abundantly expressed mRNA in MSCs. This study represents the first extensive gene expression analysis of MSCs and as such reveals new insight into the biology, ontogeny, and in vivo function of the cells.

Gene co-expression networks capture biologically important patterns in gene expression data, enabling functional analyses of genes, discovery of biomarkers, and interpretation of genetic variants. Most network analyses to date have been limited to assessing correlation between total gene expression levels in a single tissue or small sets of tissues. Here, we built networks that additionally capture the regulation of relative isoform abundance and splicing, along with tissue-specific connections unique to each of a diverse set of tissues. We used the Genotype-Tissue Expression (GTEx) project v6 RNA sequencing data across 50 tissues and 449 individuals. First, we developed a framework called Transcriptome-Wide Networks (TWNs) for combining total expression and relative isoform levels into a single sparse network, capturing the interplay between the regulation of splicing and transcription. We built TWNs for 16 tissues and found that hubs in these networks were strongly enriched for splicing and RNA binding genes, demonstrating their utility in unraveling regulation of splicing in the human transcriptome. Next, we used a Bayesian biclustering model that identifies network edges unique to a single tissue to reconstruct Tissue-Specific Networks (TSNs) for 26 distinct tissues and 10 groups of related tissues. Finally, we found genetic variants associated with pairs of adjacent nodes in our networks, supporting the estimated network structures and identifying 20 genetic variants with distant regulatory impact on transcription and splicing. Our networks provide an improved understanding of the complex relationships of the human transcriptome across tissues.