Kevin P. Weller

Severe pandemic 2009 H1N1 influenza disease due to pathogenic immune complexes

Pandemic influenza viruses often cause severe disease in middle-aged adults without preexisting comorbidities. The mechanism of illness associated with severe disease in this age group is not well understood. Here we find preexisting serum antibodies that cross-react with, but do not protect against, 2009 H1N1 influenza virus in middle-aged adults. Nonprotective antibody is associated with immune complex–mediated disease after infection. We detected high titers of serum antibody of low avidity for H1-2009 antigen, and low-avidity pulmonary immune complexes against the same protein, in severely ill individuals. Moreover, C4d deposition—a marker of complement activation mediated by immune complexes—was present in lung sections of fatal cases. Archived lung sections from middle-aged adults with confirmed fatal influenza 1957 H2N2 infection revealed a similar mechanism of illness. These observations provide a previously unknown biological mechanism for the unusual age distribution of severe cases during influenza pandemics.

Identification of myeloid cell subsets in murine lungs using flow cytometry.

Although the antibody-based recognition of cell-surface markers has been widely used for the identification of immune cells, overlap in the expression of markers by different cell types and the inconsistent use of antibody panels have resulted in a lack of clearly defined signatures for myeloid cell subsets. We developed a 10-fluorochrome flow cytometry panel for the identification and quantitation of myeloid cells in the lungs, including pulmonary monocytes, myeloid dendritic cells, alveolar and interstitial macrophages, and neutrophils. After the initial sorting of viable CD45(+) leukocytes, we detected three leukocyte subpopulations based on CD68 expression: CD68(-), CD68(low), and CD68(hi). Further characterization of the CD68(hi) population revealed CD45(+)/CD68(hi)/F4/80(+)/CD11b(-)/CD11c(+)/Gr1(-) alveolar macrophages and CD45(+)/CD68(hi)/F4/80(-)/CD11c(+)/Gr1(-)/CD103(+)/major histocompatibility complex (MHC) class II(hi) dendritic cells. The CD68(low) population contained primarily CD45(+)/CD68(low)/F4/80(+)/CD11b(+)/CD11c(+)/Gr1(-)/CD14(low) interstitial macrophages and CD45(+)/CD68(low)/F4/80(+)/CD11b(+)/CD11c(-)/Gr1(low)/CD14(hi) monocytes, whereas the CD68(-) population contained neutrophils (CD45(+)/CD68(-)/F4/80(-)/CD11b(+)/Gr1(hi)). The validity of cellular signatures was confirmed by a morphological analysis of FACS-sorted cells, functional studies, and the depletion of specific macrophage subpopulations using liposomal clodronate. We believe our approach provides an accurate and reproducible method for the isolation, quantification, and characterization of myeloid cell subsets in the lungs, which may be useful for studying the roles of myeloid cells during various pathological processes.

Obesity induced by a high-fat diet is associated with increased immune cell entry into the central nervous system.

Obesity is associated with chronic low-grade inflammation in peripheral tissues caused, in part, by the recruitment of inflammatory monocytes into adipose tissue. Studies in rodent models have also shown increased inflammation in the central nervous system (CNS) during obesity. The goal of this study was to determine whether obesity is associated with recruitment of peripheral immune cells into the CNS. To do this we used a bone marrow chimerism model to track the entry of green-fluorescent protein (GFP) labeled peripheral immune cells into the CNS. Flow cytometry was used to quantify the number of GFP(+) immune cells recruited into the CNS of mice fed a high-fat diet compared to standard chow fed controls. High-fat feeding resulted in obesity associated with a 30% increase in the number of GFP(+) cells in the CNS compared to control mice. Greater than 80% of the GFP(+) cells recruited to the CNS were also CD45(+) CD11b(+) indicating that the GFP(+) cells displayed characteristics of microglia/macrophages. Immunohistochemistry further confirmed the increase in GFP(+) cells in the CNS of the high-fat fed group and also indicated that 93% of the recruited cells were found in the parenchyma and had a stellate morphology. These findings indicate that peripheral immune cells can be recruited to the CNS in obesity and may contribute to the inflammatory response.

Genetic background impacts developmental potential of enteric neural crest-derived progenitors in the Sox10Dom model of Hirschsprung disease

Abnormalities in the development of enteric neural crest-derived progenitors (ENPs) that generate the enteric nervous system (ENS) can lead to aganglionosis in a variable portion of the distal gastrointestinal tract. Cumulative evidence suggests that variation of aganglionosis is due to gene interactions that modulate the ability of ENPs to populate the intestine; however, the developmental processes underlying this effect are unknown. We hypothesized that differences in enteric ganglion deficits could be attributable to the effects of genetic background on early developmental processes, including migration, proliferation, or lineage divergence. Developmental processes were investigated in congenic Sox10(Dom) mice, an established Hirschsprung disease (HSCR) model, on distinct inbred backgrounds, C57BL/6J (B6) and C3HeB/FeJ (C3Fe). Immuno-staining on whole-mount fetal gut tissue and dissociated cell suspensions was used to assess migration and proliferation. Flow cytometry utilizing the cell surface markers p75 and HNK-1 was used to isolate live ENPs for analysis of developmental potential. Frequency of ENPs was reduced in Sox10(Dom) embryos relative to wild-type embryos, but was unaffected by genetic background. Both migration and developmental potential of ENPs in Sox10(Dom) embryos were altered by inbred strain background with the most highly significant differences seen for developmental potential between strains and genotypes. In vivo imaging of fetal ENPs and postnatal ganglia demonstrates that altered lineage divergence impacts ganglia in the proximal intestine. Our analysis demonstrates that genetic background alters early ENS development and suggests that abnormalities in lineage diversification can shift the proportions of ENP populations and thus may contribute to ENS deficiencies in vivo.

Mdm2 and Aurora Kinase A Inhibitors Synergize to Block Melanoma Growth by Driving Apoptosis and Immune Clearance of Tumor Cells

Therapeutics that induce cancer cell senescence can block cell proliferation and promote immune rejection. However, the risk of tumor relapse due to senescence escape may remain high due to the long lifespan of senescent cells that are not cleared. Here, we show how combining a senescence-inducing inhibitor of the mitotic kinase Aurora A (AURKA) with an MDM2 antagonist activates p53 in senescent tumors harboring wild-type 53. In the model studied, this effect is accompanied by proliferation arrest, mitochondrial depolarization, apoptosis, and immune clearance of cancer cells by antitumor leukocytes in a manner reliant upon Ccl5, Ccl1, and Cxcl9. The AURKA/MDM2 combination therapy shows adequate bioavailability and low toxicity to the host. Moreover, the prominent response of patient-derived melanoma tumors to coadministered MDM2 and AURKA inhibitors offers a sound rationale for clinical evaluation. Taken together, our work provides a preclinical proof of concept for a combination treatment that leverages both senescence and immune surveillance to therapeutic ends.

Abstract B12: Synergistic anticancer activity of Aurora A kinase and MDM2 antagonists in melanoma

Senescence-inducing therapies can block proliferation of malignant cells and promote anti-tumor immune activity. However, the risk of tumor relapse remains high due to the long lifespan of senescence cells with potential to escape senescence. Here our preclinical studies demonstrate that combining a senescent-inducing aurora kinase A (AURKA) inhibitor alisertib (MLN8237) with an MDM2 antagonist [(-)-nutlin 3a] effectively induces robust p53 activation in senescent Tp53WT tumors accompanied by: 1) tumor cell proliferation arrest; 2) mitochondrial depolarization and tumor cell apoptosis; and 3) tumor cell clearance via CCL5-, CCL1- and CCL9-mediated recruitment of anti-tumor leukocytes. This combined therapy shows adequate bioavailability and low toxicity to the host in the mouse model. Moreover, the prominent preclinical response of patient-derived melanoma tumors to the co-targeting of MDM2 and AURKA provides rationale for further investigation of alisertib and MDM2 inhibitors. Citation Format: Anna E. Vilgelm, Jeff S. Pawlikowski, Yan Liu, E. Hawkins Oriana, A. Davis Tyler, Kevin P. Weller, Linda W. Horton, Colt M. McClain, Gregory D. Ayers, David Turner, David C. Essaka, Clinton F. Stewart, Jeffrey A. Sosman, Mark C. Kelley, Jeffrey A. Ecsedy, Jeffrey N. Johnston, Ann Richmond. Synergistic anticancer activity of Aurora A kinase and MDM2 antagonists in melanoma. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Melanoma: From Biology to Therapy; Sep 20-23, 2014; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(14 Suppl):Abstract nr B12.

Studies of quiescence of hematopoietic stem cells in non-human primates

Abstract Well documented differences exist between murine and primate hematopoiesis, therefore we examined the quiescence of PHSC in vivo by continuous oral bromodeoxyuridine (BrdU) administration to baboons in unperturbed and perturbed states. CD34+ cells from bone marrow (BM) or peripheral blood (PB) were fractionated based upon relative fluorescence of Hoechst (Ho) and Rhodamine (Rho). HoLow/RhoLow cells contained the highest frequency of cobblestone area-forming cells (CAFCs). Within 1 week of BrdU feeding more than 90% of granulocytes but only 2% of HoLow/RhoLow cells were BrdU+. The BrdU labeling HoLow/RhoLow cells never exceeded 26% and 44% despite 52 and 61 weeks of BrdU feeding respectively during steady state. Granulocytes (>90%) from one of these animals remained BrdU labeled despite withdrawal of BrdU for more than 1 month, confirming that indeed granulocytes are derived from previously labeled progenitors. To determine the cell cycle status of HSC during stress, two additional baboons were given G-CSF alone (100μg/kg × 5 days), one baboon was given SCF (25 μg/kg/day, day 1–8) and G-CSF (100 μg/kg/day day 4–8), and one baboon was given thrombopoietin alone (TPO, 2.5 μg/kg/day × 14 days). During this period these animals were also fed with BrdU continuously. Following administration of G-CSF alone 18% of the HoLow/RhoLow cells were BrdU+ in the mobilized peripheral blood (MPB). SCF alone resulted in only 3% of HoLow/RhoLow cells undergoing cell division, however, when G-CSF was combined with SCF, 34% of HoLow/RhoLow cells acquired BrdU labeling. Although TPO did not effectively mobilize HSC, 27% of BM HoLow/RhoLow cells were labeled with BrdU after 6 days. Since the majority of HSCs in steady state BM were quiescent, these findings are consistent with the clonal succession hypothesis. Our studies provide a direct evidence in a large animal model that hematopoietic stress induced by growth factors leads to a rapid increase in the proportion of cycling HSCs within few days. The presence of BrdU+ cells in MPB suggests that PHSCs cycle in the BM following cytokine administration and then mobilize into the PB where they re-enter the quiescent phase. Indeed, a significant proportion of HoLow/RhoLow cells present in MPB were not labeled with BrdU, which suggest that HSC mobilization is independent of cell cycle status. The quiescence of PHSC in baboons and the potential of growth factors to perturb the quiescence may provide insights regarding human clinical therapies involving HSC. These studies likely more closely portray the behavior of human PHSCs than studies reported in mice.