CTLA4-Ig-Based Bifunctional Costimulation Inhibitor Blocks CD28 and ICOS Signaling to Prevent T Cell Priming and Effector Function

CTLA4-Ig/abatacept dampens activation of naive T cells by blocking costimulation via CD28. It is an approved drug for rheumatoid arthritis but failed to deliver efficacy in a number of other autoimmune diseases. One explanation is that activated T cells rely less on CD28 signaling and use alternate coreceptors for effector function. ICOS is critical for activation of T-dependent humoral immune responses, which drives pathophysiology of IgG-mediated autoimmune diseases. In this study, we asked whether CD28 and ICOS play nonredundant roles for maintenance of T-dependent responses in mouse models. Using a hapten-protein immunization model, we show that during an ongoing germinal center response, combination treatment with CTLA4-Ig and ICOS ligand (ICOSL) blocking Ab completely dissolves ongoing germinal center responses, whereas single agents show only partial activity. Next, we took two approaches to engineer a therapeutic molecule that blocks both pathways. First, we engineered CTLA4-Ig to enhance binding to ICOSL while retaining affinity to CD80/CD86. Using a library approach, binding affinity of CTLA4-Ig to human ICOSL was increased significantly from undetectable to 15-42 nM; however, the affinity was still insufficient to completely block binding of ICOSL to ICOS. Second, we designed a bispecific costimulation inhibitor with high-affinity CTLA4 extracellular domains fused to anti-ICOSL Ab termed bifunctional costimulation inhibitor. With this bispecific approach, we achieved complete inhibition of CD80 and CD86 binding to CD28 as well as ICOS binding to ICOSL. Such bispecific molecules may provide greater therapeutic benefit in IgG-mediated inflammatory diseases compared with CTLA4-Ig alone.

Microvesicle induction of prostate specific gene expression in normal human bone marrow cells

PURPOSE

Transfer of genetic material from cancer cells to normal cells occurs via microvesicles. Cell specific phenotypes can be induced in normal cells by the transfer of material in microvesicles, leading to genetic changes. We report the identification and expression of prostate specific genes in normal human marrow cells co-cultured with human prostate cancer cells.

MATERIALS AND METHODS

We harvested prostate tissue from 11 patients with prostate cancer. In 4 cases prostate tissue was co-cultured across from human marrow for 2 or 7 days but separated from it by a 0.4 μM polystyrene membrane. In 5 cases conditioned medium from patient cancer tissue was collected and ultracentrifuged, and microvesicles were collected for co-culture (3) and vesicle characterization (3). Explanted human marrow was harvested from cultures and RNA extracted. Real-time reverse transcriptase-polymerase chain reaction was done for select prostate specific genes.

RESULTS

Marrow exposed to human prostate tumor or isolated microvesicles in culture in 4 and 3 cases, respectively, showed at least 2-fold or greater prostate gene expression than control marrow. In 1 case in which normal prostate was co-cultured there were no prostate gene increases in normal marrow.

CONCLUSIONS

Prostate cancer tumor cells co-cultured with human bone marrow cells induce prostate specific gene expression. The proposed mechanism of transfer of genetic material is via microvesicles. This represents an opportunity for novel therapeutic agents, such as antibodies, to block microvesicle release from cancer cells or for agents that may block cells from accepting microvesicles.

Heterogeneity of non-cycling and cycling synchronized murine hematopoietic stem/progenitor cells

Purified long-term multilineage repopulating marrow stem cells have been considered to be homogenous, but functionally these cells are heterogeneous. Many investigators urge clonal studies to define stem cells but, if stem cells are truly heterogeneous, clonal studies can only define heterogeneity. We have determined the colony growth and differentiation of individual lineage negative, rhodamine low, Hoechst low (LRH) stem cells at various times in cytokine culture, corresponding to specific cell cycle stages. These highly purified and cycle synchronized (98% in S phase at 40 h of culture) stem cells were exposed to two cytokine cocktails for 0, 18, 32, or 40 h and clonal differentiation assessed 14 days later. Total heterogeneity as to gross colony morphology and differentiation stage was demonstrated. This heterogeneity showed patterns of differentiation at different cycle times. These data hearken to previous suggestions that stem cells might be similar to radioactive isotopes; decay rate of a population of radioisotopes being highly predictable, while the decay of individual nuclei is heterogeneous and unpredictable (Till et al., 1964). Marrow stem cells may be most adequately defined on a population basis; stem cells existing in a continuum of reversible change rather than a hierarchy.

Microvesicular-mediated gene transfer of prostate tumor markers

e16076

Background: Microvesicles have been a subject of research for many years. Recent work has focused on the potential for cancer vaccines via microvesicles. It has also been demonstrated that various cell-specific phenotypes can be transferred from one cell type to another through microvesicle transfer. Studies in our laboratory have demonstrated that co-culture of murine lung tissue with marrow cells across a cell impermeable membrane can induce elevations in lung-specific mRNA expression in human donor marrow stem cells. Our objective is to determine whether there is transfer of genetic or transcriptional factors via microvesicles from human prostate cancer cells to fresh human marrow cells. Methods: Fresh prostate tissue was harvested from surgical specimens following radical retropubic prostatectomy. Samples were histologically confirmed to contain prostatic adenocarcinoma. Co-cultures were established using a transwell system in which 0.05–0.100 grams of prostate tissue was minced and co-cultured with 1–3 million normal, human donor marrow cells for 2–7 days. Marrow not co-cultured with tumorserved as a control. Target cells were collected and total RNA was analyzed for prostate-specific gene expression byReal Time RT-PCR. Fold differences in expression of the genes were analyzed, using TaqMan®, gene assays (Applied Biosystems) and were expressed in relation to the marrow control. Results: We have observed significant increases in gene expression in marrow cells co-cultured with prostate tumor cells (Gleason grades 6–9). Variable increases in expression were seen in 3 patient samples, as high as 7-fold for ERG, greater than 10-fold for ACPP and greater than 100-fold for STEAP, PART, TMPRSS2, PSCA and ETV1. Conclusions: These studies demonstrate that prostate specific genes are present in fresh human marrow cells after co-culture with tumor tissue. This establishes a base to begin evaluating the significance of microvesicle-mediated genetic transfer, mechanisms of transfer and therapeutic options for blocking or manipulating such transfer to influence the disease process.

No significant financial relationships to disclose.

Marrow cell genetic phenotype change induced by human lung cancer cells

11108

Background: Murine lung-derived microvesicles are capable of inducing lung-specific mRNA in marrow cells, when co-cultured across from these cells, but separated from them by a cell-impermeable (0.4 micron) membrane. These converted murine marrow cells showed mRNA elevations, lung-specific protein production and enhanced capacity to convert to lung epithelial cells after in vivo transplantation into irradiated mice. We examine here whether fresh tissue from lung cancer patients would have the same capacity to genetically alter co-cultured human marrow cells. Methods: Lung cancer samples were collected from 5 patients undergoing surgery. Minced tumor tissue at 50–100 mg was co-cultured in a semi-permeable culture plate insert opposite 3.0 ×106 human marrow cells. The marrow cells were harvested after 2–7 days of co-culture. Marrow cell RNA was analyzed for lung specific mRNA using real time RT-PCR. Relative levels of gene expression was expressed a fold increase compared to level in controls. Results: Lung cancers studied were adenocarcinoma, endobronchial alveolar carcinoma, bronchioloalveolar carcinoma, non-small cell carcinoma and squamous cell carcinoma. mRNAs for aquaporin 1–5, specific for type I pneumocytes and surfactant A-D, specific for type II pneumocytes, were measured. Aquaporin I was elevated in marrow cells from co culture with all lung cancers; elevations ranging from 2.15 to 56.7 fold (mean 23 fold). Similarly surfactant B mRNA was induced in marrow cells by all lung cancers with fold elevations ranging from 7.9 to 2164 (mean fold elevation 668). More variable elevations were also seen with aquaporin 3, 4, and 5, surfactant A, surfactant C, and surfactant D. Ultracentrifugation (28,000 g) of conditioned media from these cancers revealed the presence of microvesicles with diameters of 100–180 nm. Conclusions: These observations indicate that the genetic phenotype of cells in the vicinity of lung cancer cells can be altered and that these alterations might be mediated by microvesicle transfer of genetic information.

No significant financial relationships to disclose.

The stem cell continuum: cell cycle, injury, and phenotype lability

The phenotype of the hematopoietic stem cell is intrinsically labile and impacted by cell cycle and the effects of tissue injury. In published studies we have shown that there are changes in short- and long-term engraftment, progenitor numbers, gene expression, and differentiation potential with cytokine-induced cell cycle transit. Critical points here are that these changes are reversible and not unidirectional weighing, heavily against a hierarchical model of stem cell regulation. Furthermore, a number of studies have now established that stem cells separated by lineage depletion and selection for Sca-1 or c-kit or low rhodamine and Hoechst staining are in fact a cycling population. Last, studies on Hoechst separated "cycling" stem cells indicates that the observed phenotype shifts relate to phase of cell cycle and are not due to in vitro exposure to cytokines. These data suggest a continuum model of stem cell regulation and further indicate that this model holds for in vivo situations. Observations that marrow cells can convert to various tissue cells under different injury conditions continue to be published despite a small, but influential, number of negative studies. Our studies and those of others indicate that conversions of marrow-derived cells to different tissue cells, such as skeletal muscle and lung, is critically dependent upon multiple variables, the most important of which is the presence of tissue injury. Variables which affect conversion of marrow cells to nonhematopoietic cells after in vivo transplantation include the nature and timing of the injury; marrow mobilization; the marrow cell type infused; the timing of cell infusion and the number of cells infused; the cell cycle state of the marrow cells, and other functional alterations in the marrow cells the treatment of the host mouse separate from specific injury; the mode of cell delivery; and possibly the presence of microvesicles from injured tissue. At least some of the highlighted negative reports on stem cell plasticity appear to be due to a failure to address these variables. Recently, we have observed that irradiated lung releases microvesicles which can enter marrow cells and lead to the marrow cells expressing lung-specific mRNA and protein. This could provide an underlying mechanism for many of the plasticity phenomena. Altogether, marrow appears to represent a highly flexible ever-changing cell system with the capacity to respond to products of injured cells and top repair a broad range of tissues.

The stem cell continuum: considerations on the heterogeneity and plasticity of marrow stem cells

Traditional models of hematopoiesis have been hierarchical. Recent evidence showing that marrow stem cells are a cycling population and that the hematopoietic phenotype of these cells reversibly changes with cycle transit have suggested a continuum model of stem cell regulators. Studies on marrow cell conversion to lung cells have extended this continuum to cycle-related differentiation into nonhematopoietic stem cells. We postulate that stem cells transiting cell cycle continually change their chromatin structure, thus providing different windows of transcriptional opportunity and a continually changing phenotype. Final outcomes with this continuum model would be determined by the specific chromatin state of the cell and the presence of specific differentiation inducers.

The stem cell continuum

Hematopoietic stem cells have been felt to exist in a hierarchical structure with a relatively fixed phenotype at each stage of differentiation. Recent studies on the phenotype of the marrow hematopoietic stem cell indicate that it is not a fixed entity, but rather that it fluctuates and shows marked heterogeneity. Past studies have shown that stem cell engraftment characteristics, adhesion protein, and gene expression varies with the phase of the cell cycle. More recently, we demonstrated that progenitor numbers and differentiation potential also vary reversibly during one cytokine-induced cell cycle transit. We have also shown high levels of conversion of marrow cells to skeletal muscle and lung cells, indicating a different level of plasticity. Recently, we demonstrated that homing to lung and conversion to lung cells in a mouse transplant model also fluctuates reversibly with cell cycle transit. This could be considered plasticity squared. These data indicate that marrow stem cells are regulated on a continuum related to the cell cycle both as to hematopoietic and to nonhematopoietic differentiation.

Bone marrow production of lung cells: the impact of G-CSF, cardiotoxin, graded doses of irradiation, and subpopulation phenotype

OBJECTIVE

Previous studies have demonstrated the production of various types of lung cells from marrow cells under diverse experimental conditions. Our aim was to identify some of the variables that influence conversion in the lung.

METHODS

In separate experiments, mice received various doses of total-body irradiation followed by transplantation with whole bone marrow or various subpopulations of marrow cells (Lin(-/+), c-kit(-/+), Sca-1(-/+)) from GFP(+) (C57BL/6-TgN[ACTbEGFP]1Osb) mice. Some were given intramuscular cardiotoxin and/or mobilized with granulocyte colony-stimulating factor (G-CSF).

RESULTS

The production of pulmonary epithelial cells from engrafted bone marrow was established utilizing green fluorescent protein (GFP) antibody labeling to rule out autofluorescence and deconvolution microscopy to establish the colocaliztion of GFP and cytokeratin and the absence of CD45 in lung samples after transplantation. More donor-derived lung cells (GFP(+)/CD45(-)) were seen with increasing doses of radiation (5.43% of all lung cells, 1200 cGy). In the 900-cGy group, 61.43% of GFP(+)/CD45(-) cells were also cytokeratin(+). Mobilization further increased GFP(+)/CD45(-) cells to 7.88% in radiation-injured mice. Up to 1.67% of lung cells were GFP(+)/CD45(-) in radiation-injured mice transplanted with Lin(-), c-kit(+), or Sca-1(+) marrow cells. Lin(+), c-kit(-), and Sca-1(-) subpopulations did not significantly engraft the lung.

CONCLUSIONS

We have established that marrow cells are capable of producing pulmonary epithelial cells and identified radiation dose and G-CSF mobilization as variables influencing the production of lung cells from marrow cells. Furthermore, the putative lung cell-producing marrow cell has the phenotype of a hematopoietic stem cell.

Stem cell biology and the plasticity polemic

Characterization of a cord blood derived unrestricted somatic stem cell (USSC) with capacity to differentiate into hematopoietic and nonhematopoietic tissues in the absence of cell fusion has highlighted the great potential of stem cell plasticity. A great variety of stem cell types have been defined and even the most pure marrow stem cells are highly heterogeneous. Data suggest that stem cells may exist in a continuum with continually and reversibly changing phenotype. These cells also possess a capacity to produce lung, liver, skin, and skeletal muscle under conditions of tissue injury. Arguments raised against the significance of adult marrow to nonmarrow conversions including the importance of cell fusion appear fallacious. We are at the beginning of an exciting and burgeoning field of research with great clinical potential.

The marrow cell continuum: stochastic determinism

Traditional models of hematopoiesis have been hierarchical in nature. Over the past 10 years, we have developed data indicating that hematopoiesis is regulated in a continuum with deterministic and stochastic components. We have shown that the most primitive stem cells, as represented by lineage negative rhodamine(low) Hoechst(low) murine marrow cells are continuously or intermittently cycling as determined by in vivo BrdU labeling. When marrow stem cells are induced to transit cell cycle by in vitro exposure to cytokines, either IL-3, IL-6, IL-11, and steel factor or thrombopoietin, FLT3 ligand, and steel factor, they progress through cycle in a highly synchronized fashion. We have determined that when the stem cells progress through a cytokine stimulated cell cycle the homing, engraftment, adhesion protein, global gene expression, and hematopoietic differentiation phenotypes all change in a reversible fashion. This has led to the continuum model, in which, with cycle transit, chromatin is continually changing altering open transcription areas and providing a continually changing landscape of transcriptional opportunity. More recently, we have extended the changing differentiation profiles to differentiation into lung cells and found that non-hematopoietic differentiation also shows cycle related reversibly modulation. These observations all together support a continuum model of stem cell regulation in which the phenotype of the marrow stem cells is continually and reversibly changing over time.

Effect of ex vivo cytokine treatment on human cord blood engraftment in NOD-scid mice

Umbilical cord blood transplantation is considered an alternative to traditional bone marrow transplantation for patients who do not have matched sibling donors. In this study, we examined the effects of ex vivo treatment of human cord blood cells with cytokine mixtures and assessed the ability of treated cells to engraft in NOD-scid mice. We incubated the cord blood with a four-factor cytokine mixture of interleukin (IL)-3, IL-6, IL-11 and stem cell factor, or with a two-factor cytokine mixture of thrombopoietin and flt-3. Incubation of cord blood for 48 h with either cytokine mixture did not affect progenitor cell number or proliferative potential as measured by the high proliferative potential (HPP) assay. Cytokine-treated cord blood injected into irradiated NOD-scid mice resulted in multilineage human engraftment. Overall, incubation with cytokines resulted in variable levels of engraftment with different cord blood samples. Incubation of cord blood with the four-factor cytokine mixture resulted in increased survival of irradiated NOD-scid recipients. These results demonstrate that short-term ex vivo treatment of human progenitor cells gives variable results on in vivo multipotential capabilities.