AdRoit is an accurate and robust method to infer complex transcriptome composition

Bulk RNA sequencing provides the opportunity to understand biology at the whole transcriptome level without the prohibitive cost of single cell profiling. Advances in spatial transcriptomics enable to dissect tissue organization and function by genome-wide gene expressions. However, the readout of both technologies is the overall gene expression across potentially many cell types without directly providing the information of cell type constitution. Although several in-silico approaches have been proposed to deconvolute RNA-Seq data composed of multiple cell types, many suffer a deterioration of performance in complex tissues. Here we present AdRoit, an accurate and robust method to infer the cell composition from transcriptome data of mixed cell types. AdRoit uses gene expression profiles obtained from single cell RNA sequencing as a reference. It employs an adaptive learning approach to alleviate the sequencing technique difference between the single cell and the bulk (or spatial) transcriptome data, enhancing cross-platform readout comparability. Our systematic benchmarking and applications, which include deconvoluting complex mixtures that encompass 30 cell types, demonstrate its preferable sensitivity and specificity compared to many existing methods as well as its utilities. In addition, AdRoit is computationally efficient and runs orders of magnitude faster than most methods.

Inhibition of calcineurin-mediated endocytosis and alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors prevents amyloid beta oligomer-induced synaptic disruption

Synaptic degeneration, including impairment of synaptic plasticity and loss of synapses, is an important feature of Alzheimer disease pathogenesis. Increasing evidence suggests that these degenerative synaptic changes are associated with an accumulation of soluble oligomeric assemblies of amyloid beta (Abeta) known as ADDLs. In primary hippocampal cultures ADDLs bind to a subpopulation of neurons. However the molecular basis of this cell type-selective interaction is not understood. Here, using siRNA screening technology, we identified alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor subunits and calcineurin as candidate genes potentially involved in ADDL-neuron interactions. Immunocolocalization experiments confirmed that ADDL binding occurs in dendritic spines that express surface AMPA receptors, particularly the calcium-impermeable type II AMPA receptor subunit (GluR2). Pharmacological removal of the surface AMPA receptors or inhibition of AMPA receptors with antagonists reduces ADDL binding. Furthermore, using co-immunoprecipitation and photoreactive amino acid cross-linking, we found that ADDLs interact preferentially with GluR2-containing complexes. We demonstrate that calcineurin mediates an endocytotic process that is responsible for the rapid internalization of bound ADDLs along with surface AMPA receptor subunits, which then both colocalize with cpg2, a molecule localized specifically at the postsynaptic endocytic zone of excitatory synapses that plays an important role in activity-dependent glutamate receptor endocytosis. Both AMPA receptor and calcineurin inhibitors prevent oligomer-induced surface AMPAR and spine loss. These results support a model of disease pathogenesis in which Abeta oligomers interact selectively with neurotransmission pathways at excitatory synapses, resulting in synaptic loss via facilitated endocytosis. Validation of this model in human disease would identify therapeutic targets for Alzheimer disease.

Solution state characterization of amyloid beta-derived diffusible ligands

A growing body of evidence suggests that soluble oligomeric forms of the amyloid beta peptide known as amyloid-derived diffusible ligands (ADDLs) are the toxic species responsible for neurodegeneration associated with Alzheimer's disease. Accurate biophysical characterization of ADDL preparations is hampered by the peptide's strong tendency to self-associate and the effect of factors such as ionic strength, temperature, and pH on its behavior. In addition, amyloid peptides are known to interact with common laboratory excipients, specifically detergents, further complicating the results from standard analytical methods such as denaturing polyacrylamide gel electrophoresis. We have studied the solution behavior of various amyloid peptide preparations using analytical ultracentrifugation and size exclusion chromatography coupled with multiangle laser light scattering. Our results indicate that ADDL preparations exist in solution primarily as a binary mixture of a monomeric peptide and high-molecular mass oligomers. We relate our findings to previously described characterizations utilizing atomic force microscopy and electrophoretic methods and demonstrate that low-molecular mass oligomers identified by gel electrophoresis likely represent artifacts induced by the peptide's interaction with detergent, while atomic force microscopy results are likely skewed by differential binding of monomeric and oligomeric peptide species. Finally, we confirm that only the high-molecular mass oligomeric components of an ADDL preparation are capable of binding to subpopulations of primary hippocampal neurons in vitro.

Generation and characterization of a humanized bradykinin B1 receptor mouse

Antagonists of the B1 bradykinin receptor (B1R), encoded by the BDKRB1 gene, offer the promise of novel therapeutic agents for inflammatory and neuropathic pain. However, the in vivo characterization of the pharmacodynamics of B1R antagonists is hindered by the low level of B1R expression in healthy tissue and the profound species selectivity exhibited by many compounds for the B1R. To circumvent these issues we generated two genetically engineered rodent models. The first is a transgenic rat over-expressing the human B1R under the control of the neuronal-specific enolase promoter; we previously reported the utility of this model in assessing human B1R receptor occupancy in the central nervous system of the rat. The second model, reported here, utilized gene-targeting by homologous recombination to replace the genomic coding sequence for the endogenous mouse B1R with that of the human B1R. The mRNA expression profile of the humanized Bdkrb1 (hBkdrb1) allele is similar to that of the mouse Bdkrb1 (mBkdrb1) in the wild-type animal. Furthermore, in vitro assays indicate that tissues isolated from the humanized mouse possess pharmacological properties characteristic of the human B1R. Therefore, we have generated a humanized B1R mouse model that is suitable for testing the efficacy of human B1R-selective compounds.

Engraftment potential of human fetal hematopoietic cells in NOD/SCID mice is not restricted to mitotically quiescent cells

During fetal development, there is a continued demand for large numbers of primitive and mature hematopoietic cells. This demand may require that all potential hematopoietic stem cells (HSCs) migrate effectively to emerging hematopoietic sites and subsequently contribute to blood cell production, regardless of their cell cycle status. We recently established that umbilical cord blood cells in the G(1) phase of the cell cycle have a repopulating potential similar to cells in G(0), suggesting that cycling prenatal and neonatal HSCs may have the same functional capabilities described for quiescent, but not cycling, cells from adult sources. To establish the relationship between cell cycle status and hematopoietic potential at early stages of human ontogeny, the in vivo engraftment potential of mitotically defined fetal liver (FL) and fetal bone marrow (FBM) cells were examined in NOD/SCID recipients. Following transplantation of the same numbers of G(0), G(1), or S/G(2)+M CD34(+) cells from FL, equivalent percentages of recipient mice were chimeric (55%, 60%, and 60%, respectively). FBM-derived CD34(+) cells in all phases of the cell cycle engrafted in conditioned recipients and sustained human hematopoiesis, albeit at lower levels than their FL-derived counterparts. Multilineage differentiation was evident in all transplanted mice independent of the source or cell cycle status of graft cells. In addition, levels of chimerism in mice transplanted with fetal blood-derived G(0) or G(1) CD34(+) lineage-depleted cells were similar. These results support the assertion that mitotically quiescent and cycling fetal hematopoietic cells contain marrow-repopulating stem cells capable of multilineage engraftment in NOD/SCID mouse recipients.

Homing efficiency, cell cycle kinetics, and survival of quiescent and cycling human CD34(+) cells transplanted into conditioned NOD/SCID recipients

Differences in engraftment potential of hematopoietic stem cells (HSCs) in distinct phases of cell cycle may result from the inability of cycling cells to home to the bone marrow (BM) and may be influenced by the rate of entry of BM-homed HSCs into cell cycle. Alternatively, preferential apoptosis of cycling cells may contribute to their low engraftment potential. This study examined homing, cell cycle progression, and survival of human hematopoietic cells transplanted into nonobese diabetic severe combined immunodeficient (NOD/SCID) recipients. At 40 hours after transplantation (AT), only 1% of CD34(+) cells, or their G(0) (G(0)CD34(+)) or G(1) (G(1)CD34(+)) subfractions, was detected in the BM of recipient mice, suggesting that homing of engrafting cells to the BM was not specific. BM of NOD/SCID mice receiving grafts containing approximately 50% CD34(+) cells harbored similar numbers of CD34(+) and CD34(-) cells, indicating that CD34(+) cells did not preferentially traffic to the BM. Although more than 64% of human hematopoietic cells cycled in culture at 40 hours, more than 92% of cells recovered from NOD/SCID marrow were quiescent. Interestingly, more apoptotic human cells were detected at 40 hours AT in the BM of mice that received xenografts of expanded cells in S/G(2)+M than in recipients of G(0)/G(1) cells (34.6% +/- 5.9% and 17.1% +/- 6.3%, respectively; P <.01). These results suggest that active proliferation inhibition in the BM of irradiated recipients maintains mitotic quiescence of transplanted HSCs early AT and may trigger apoptosis of cycling cells. These data also illustrate that trafficking of transplanted cells to the BM is not selective, but lodgment of BM-homed cells may be specific.

Differential transduction efficiency of SCID-repopulating cells derived from umbilical cord blood and granulocyte colony-stimulating factor-mobilized peripheral blood

The gene transfer efficiency into nonobese diabetic/severe combined immunodeficient (NOD/SCID)-repopulating cells (SRCs) derived from umbilical cord blood (UCB) (n = 11 NOD/SCID mice) and granulocyte-colony stimulating factor (G-CSF)-mobilized peripheral blood (MPB) (n = 64 NOD/SCID mice) was compared using a clinically relevant protocol and a retrovirus vector expressing the enhanced green fluorescent protein (EGFP). At 6-9 weeks after transplantation, the frequency of transduced human cells in the bone marrow (BM) (40.5% +/- 2.4% [mean +/- SE]) and spleen (SPL) (36.4% +/- 3.2%) in recipients of UCB cells was significantly higher (p < 0.001) than that observed in the BM (2.2% +/- 1.8%) and SPL (2.0% +/- 2.6%) in recipients of MPB. In subsequent studies, MPB was cultured for 2-8 days in cytokines prior to transduction to determine if longer prestimulation was required for optimal gene transfer. A significant increase in gene transfer into CD45(+) human cells and clonogenic cells derived from MPB SRCs was observed when cells were prestimulated for 6 days compared to 2 days prior to transduction (p = 0.019). However, even after 6 days of prestimulation, transduction was still significantly less than UCB. A substantial discrepancy exists in the ability to introduce genes effectively via retrovirus vectors into SRCs derived from MPB as compared to UCB.

Similar repopulating capacity of mitotically active and resting umbilical cord blood CD34(+) cells in NOD/SCID mice

It was hypothesized that during mammalian development, the extensive need for hematopoietic cells requires equal contribution to blood cell production from both quiescent and cycling hematopoietic stem cells (HSCs) while maintaining the stem cell pool. To investigate this hypothesis, the engraftment potential of umbilical cord blood (UCB) CD34(+) cells residing in either G(0) (G(0)CD34(+) cells) or G(1) (G(1)CD34(+) cells) phases of the cell cycle was assessed in nonobese diabetic/severe combined immune-deficient (NOD/SCID) mice. Whereas the level of chimerism in mice transplanted with UCB G(0)CD34(+) cells was 69.9% +/- 24.0%, mice receiving equal numbers of G(1)CD34(+) cells harbored 46.7% +/- 21.3% human cells 8 weeks posttransplantation. Both groups of cells sustained multilineage differentiation and the production of CD34(+) cells in recipient animals. The relationship between the number of transplanted G(0)CD34(+) or G(1)CD34(+) cells and the level of chimerism was analyzed by a general linear models procedure. Although the initial level of chimerism following transplantation of G(0)CD34(+) cells was higher than that sustained by G(1)CD34(+) cells, the increment in the degree of chimerism obtained with each additional 10(3) cells of either phenotype was identical, suggesting that the reconstitution potential of these 2 types of cells was similar. Of interest is that human cells recovered from primary recipients of both G(0)CD34(+) and G(1)CD34(+) cells engrafted in secondary NOD/SCID recipients, albeit at a substantially lower level, confirming the primitive nature of UCB CD34(+) cells residing in G(1).

Use of enhanced green fluorescent protein to monitor retroviral-mediated gene therapy in human keratinocytes

Keratinocytes have great promise as targets for gene therapy involving both skin as well as for systemic disorders due to their availability and potential long life span. Improvement of gene transfer into keratinocytes will be greatly facilitated by markers that will allow both rapid detection and efficient selection of transduced cells. For these purposes, a recombinant version of the Aequorea victoria green fluorescent protein that is enhanced for high-level expression in mammalian cells (EGFP) was placed into a replication-deficient retroviral vector. High-titer retrovirus was used to transduce both primary cultures of neonatal foreskin-derived human keratinocytes (HK) as well as the immortalized keratinocyte-derived cell line HaCaT. Both cell types stably expressed the EGFP, and this marker allowed rapid purification of transduced cells by fluorescence-activated cell sorting. EGFP expression was seen in HaCaT keratinocytes for at least 40 passages, and the presence of this construct did not effect cell growth, or apoptosis in response to UVB or etoposide. Transduced populations of HK were grafted into SCID mice, resulting in a functional epidermis. EGFP expression was readily seen in vivo by exposing the xenografts to an ultraviolet light source. These studies demonstrate the feasibility of using EGFP as a convenient and rapid marker to monitor keratinocyte gene transfer both in vitro and in vivo.

Rac2 stimulates Akt activation affecting BAD/Bcl-XL expression while mediating survival and actin function in primary mast cells

Mast cells generated from Rac2-deficient (-/-) mice demonstrated defective actin-based functions, including adhesion, migration, and degranulation. Rac2(-/-) mast cells generated lower numbers and less mast cell colonies in response to growth factors and were deficient in vivo. Rac2(-/-) mast cells demonstrated a significant reduction in growth factor-induced survival, which correlated with the lack of activation of Akt and significant changes in the expression of the Bcl-2 family members BAD and Bcl-XL, in spite of a 3-fold induction of Rac1 protein. These results suggest that Rac2 plays a unique role in multiple cellular functions and describe an essential role for Rac2 in growth factor-dependent survival and expression of BAD/Bcl-XL.