Analysis of clinical flow cytometric immunophenotyping data by clustering on statistical manifolds: treating flow cytometry data as high-dimensional objects.

BACKGROUND: Clinical flow cytometry typically involves the sequential interpretation of two-dimensional histograms, usually culled from six or more cellular characteristics, following initial selection (gating) of cell populations based on a different subset of these characteristics. We examined the feasibility of instead treating gated n-parameter clinical flow cytometry data as objects embedded in n-dimensional space using principles of information geometry via a recently described method known as Fisher Information Non-parametric Embedding (FINE). METHODS: After initial selection of relevant cell populations through an iterative gating strategy, we converted four color (six-parameter) clinical flow cytometry datasets into six-dimensional probability density functions, and calculated differences among these distributions using the Kullback-Leibler divergence (a measurement of relative distributional entropy shown to be an appropriate approximation of Fisher information distance in certain types of statistical manifolds). Neighborhood maps based on Kullback-Leibler divergences were projected onto two dimensional displays for comparison. RESULTS: These methods resulted in the effective unsupervised clustering of cases of acute lymphoblastic leukemia from cases of expansion of physiologic B-cell precursors (hematogones) within a set of 54 patient samples. CONCLUSIONS: The treatment of flow cytometry datasets as objects embedded in high-dimensional space (as opposed to sequential two-dimensional analyses) harbors the potential for use as a decision-support tool in clinical practice or as a means for context-based archiving and searching of clinical flow cytometry data based on high-dimensional distribution patterns contained within stored list mode data. Additional studies will be needed to further test the effectiveness of this approach in clinical practice.

Mechanisms involved in radiation enhancement of intratumoral dendritic cell therapy.

We have previously reported that local tumor irradiation, without inducing cell death, can augment the therapeutic efficacy of intratumoral (IT) dendritic cell (DC) vaccination. This study examined potential mechanisms underlying radiation enhancement of IT DC therapy in this setting. Even though ionizing radiation did not mediate tumor cell killing, bone marrow-derived DCs acquired in vitro tumor antigens from irradiated D5 murine melanoma cells more efficiently than from untreated cells. This radiation-enhanced loading of DCs did not induce DC maturation, but was associated with improved cross-priming of T cells both in vitro and in vivo. Furthermore, in vivo pulsing of DCs with irradiated versus untreated tumor cells resulted in superior presentation of tumor antigens to T cells. In addition, tumor irradiation facilitated homing of IT administered DCs to the draining lymph node, possibly by down-regulating CCL21 expression within the tumor mass. Studies of the tumor microenvironment in irradiated versus untreated tumors did not reveal significant inflammatory changes. Moreover, radiation did not promote accumulation of CD4 or CD8 effector T cells within solid tumors. Our results indicate that, without inducing cytotoxicity, tumor irradiation can enhance the ability of DCs to capture tumor antigens, migrate to the draining lymph node, and present processed antigens to T cells. These findings may prove useful in designing future strategies for human cancer immunotherapy.

Synergistic effects of IL-12 and IL-18 in skewing tumor-reactive T-cell responses towards a type 1 pattern.

We have previously described the antitumor reactivity of tumor-draining lymph node (TDLN) cells after secondary activation with antibodies. In this report, we examined the effects of interleukin (IL)-12 and IL-18 on modulating the immune function of antibody-activated murine TDLN cells. TDLN cells were activated with anti-CD3/anti-CD28 monoclonal antibody followed by stimulation with IL-12 and/or IL-18. IL-18 in combination with IL-12 showed a synergistic effect in augmenting IFNgamma and granulocyte macrophage colony-stimulating factor secretion, whereas IL-18 alone had minimal effect. Concurrently, IL-18 prevented IL-12-stimulated TDLN cells from producing IL-10. The IL-12/IL-18-cultured TDLN cells therefore manifested cytokine responses skewed towards a Th1/Tc1 pattern. IL-12 and IL-18 stimulated CD4(+) TDLN cells and enhanced IFNgamma production by CD4(+) cells to a greater extent than by CD8(+) cells. Use of NF-kappaB p50(-/-) TDLN cells suggested the involvement of NF-kappaB in the IL-12/IL-18 polarization effect. Furthermore, a specific NF-kappaB inhibitor significantly suppressed IL-12/IL-18-induced IFNgamma secretion, thus confirming the requirement for NF-kappaB activation in IL-12/IL-18 signaling. In adoptive immunotherapy, IL-12- and IL-18-cultured TDLN cells infiltrated pulmonary tumor nodules and eradicated established tumor metastases more efficiently than T cells generated with IL-12 or IL-18 alone. Antibody depletion revealed that both CD4(+) and CD8(+) cells were involved in the tumor rejection induced by IL-12/IL-18-cultured TDLN cells. These studies indicate that IL-12 and IL-18 can be used to generate potent CD4(+) and CD8(+) antitumor effector cells by synergistically polarizing antibody-activated TDLN cells towards a Th1 and Tc1 phenotype.

Donor cell cycling, trafficking, and accumulation during adoptive immunotherapy for murine lung metastases.

Adoptive cellular immunotherapy treats metastatic cancer by infusing cultured T cells derived from resected tumors or primed lymph nodes. The infused cells must accumulate in metastatic lesions to suppress growth; however, this process and the resulting clinical response are dynamic and evolve during the days and weeks following cell infusion. This study used novel experimental techniques to determine the fate of infused, cultured tumor-draining lymph node (TDLN) cells during the treatment of murine pulmonary micrometastases. After infusion, the cultured TDLN cells accumulated in the pulmonary vasculature, systemic lymph nodes, and spleen. Donor cells were initially confined to alveolar capillaries with no movement into metastases. Within 4 h, TDLN cells began migrating across pulmonary postcapillary venules and first appeared within metastases. After 24 h, most donor cells in the lung were associated with tumor nodules. Donor cell proliferation within the lung and lymphoid organs was detected within 24 h of infusion and continued throughout the 5-day period of observation. Furthermore, those proliferating in lymphoid organs trafficked back to the tumor-bearing lungs, accounting for approximately 50% of the donor cells recovered from these sites after 5 days. Finally, donor T cells entering metastases both early (within 1-2 days) and late (after 2 days) suppressed tumor growth, but the early recruits accounted for most of the therapeutic response. Thus, cultured TDLN cells migrate directly into tumor-bearing organs and seed the recirculating pool of lymphocytes after infusion. Small fractions of the later differentiate in lymphoid organs and migrate into the lungs but appear less effective than effector cells in the initial bolus.

Discovery of a potent nanoparticle P-selectin antagonist with anti-inflammatory effects in allergic airway disease.

The severity of allergic asthma is dependent, in part, on the intensity of peribronchial inflammation. P-selectin is known to play a role in the development of allergen-induced peribronchial inflammation and airway hyperreactivity. Selective inhibitors of P-selectin-mediated leukocyte endothelial-cell interactions may therefore attenuate the inflammatory processes associated with allergic airway disease. Novel P-selectin inhibitors were created using a polyvalent polymer nanoparticle capable of displaying multiple synthetic, low molecular weight ligands. By assembling a particle that presents an array of groups, which as monomers interact with only low affinity, we created a construct that binds extremely efficiently to P-selectin. The ligands acted as mimetics of the key binding elements responsible for the high-avidity adhesion of P-selectin to the physiologic ligand, PSGL-1. The inhibitors were initially evaluated using an in vitro shear assay system in which interactions between circulating cells and P-selectin-coated capillary tubes were measured. The nanoparticles were shown to preferentially bind to selectins expressed on activated endothelial cells. We subsequently demonstrated that nanoparticles displaying P-selectin blocking arrays were functionally active in vivo, significantly reducing allergen-induced airway hyperreactivity and peribronchial eosinophilic inflammation in a murine model of asthma.

CD49d overexpression and T cell autoimmunity.

D10.G4.1 (D10) cells, a murine conalbumin-reactive Th2 cell line, made to overexpress the beta(2) integrin LFA-1 by pharmacological manipulation or by transfection become autoreactive and are capable of inducing in vivo autoimmunity. However, whether this is specific to LFA-1 and whether overexpression of other T cell integrin molecules has the same effect are unknown. We examined the functional consequences of T cell CD49d (alpha(4) integrin) overexpression by transfecting murine CD49d cDNA into D10 cells. Similar to the LFA-1-transfected cells, the CD49d-overexpressing T cells are autoreactive and proliferate in response to APCs in an MHC class II-dependent manner in the absence of nominal Ag. Additionally, CD49d overexpression is associated with increased in vitro adhesion to endothelial cells and increased in vivo splenic homing. However, in contrast to LFA-1 overexpression, increased T cell CD49d expression is not associated with autoreactive cytotoxicity or the ability to induce in vivo autoimmunity. In addition to the novel observation that CD49d overexpression is sufficient to induce T cell autoreactivity, our results also support the hypothesis that the ability to induce in vivo autoimmunity is related to T cell cytotoxicity and not to T cell proliferation function in the D10 murine adoptive transfer model of autoimmunity.

Sustained high-yield production of recombinant proteins in transiently transfected COS-7 cells grown on trimethylamine-coated (hillex) microcarrier beads.

The present study shows that COS-7 cells transiently transfected and maintained on positively charged (trimethylamine-coated) microcarrier beads synthesize recombinant protein at higher levels and for longer periods of time than cells transfected and maintained on polystyrene flasks in monolayer culture. Sustained, high-level synthesis was observed with secreted chimeric proteins (murine E-selectin- and P-selectin-human IgM chimeras) and a secreted hematopoietic growth factor (granulocyte-macrophage colony-stimulating factor). Studies with green fluorescent protein indicated that the transfected cells attached more firmly to the trimethylamine-coated microcarriers than to polystyrene flasks. After 10-14 days in culture, most of the transfected cells detached from the surface of the polystyrene flasks, whereas most transfected cells remained attached to the microcarriers. The transiently transfected microcarrier cultures produced higher levels of protein per transfected cell due to this prolonged attachment. The prolonged attachment and higher output of transfected cells on microcarriers resulted in a 5-fold increase in protein production from a single transfection over two weeks. Thus, microcarrier-based transient transfection yields quantities of recombinant proteins with a significant savings of time and reagents over monolayer culture.

CCR6 colocalizes with CD18 and enhances adhesion to activated endothelial cells in CCR6-transduced Jurkat T cells.

CCR6 is expressed by memory T cells (mTC) and is a requirement for efficient arrest of a subset of mTC to activated human dermal microvascular endothelial cells (HDMEC) under physiologic shear stress. We now address whether CCR6 alone is sufficient to induce arrest of a model T cell line (Jurkat) that shows low expression of all CCRs tested (CCR1-10). Herein, we transduced Jurkat (JK) T cells expressing fucosyltransferase VII with a chimeric chemokine receptor consisting of CCR6 fused to enhanced green fluorescent protein. In contrast to the starting JK lines, the resulting cell line (JK fucosyltransferase VII-CCR6) migrated 6-fold better to CCL20 in chemotaxis assays, arrested in response to CCL20 that was immobilized to plastic, and demonstrated a 2.5-fold increase in adhesion to activated HDMEC (p = 0.001). Adhesion was blocked by anti-CD18 mAb (p = 0.005) but not by anti-CD49d mAb (p = 0.3). After arrest on recombinant substrates, CCR6 clustered on the surface as detected by real-time observation of enhanced green fluorescent protein fluorescence. Dual-label confocal microscopy revealed that LFA-1 (CD18 and CD11a), but not CXCR4, colocalized with clustered CCR6 in the presence of immobilized CCL20. Thus, the functional expression of CCR6 is sufficient to provide the chemokine signaling necessary to induce arrest of a JK T cell line to activated HDMEC. Clustering of CCR6 and coassociation with critical integrins may serve to strengthen adhesion between T cells and activated endothelial cells.

Subset-specific reductions in lung lymphocyte accumulation following intratracheal antigen challenge in endothelial selectin-deficient mice.

We previously demonstrated induction and expression of CD62E and CD62P in the lungs of mice primed and then challenged with intratracheal (i.t.) SRBC. The current study examined accumulation of endogenous lymphocytes in the lungs of endothelial E- and P-selectin-deficient (E(-)P(-)) mice after i.t. SRBC challenge. Compared with syngeneic wild-type (wt) mice, E(-)P(-) mice showed an 85-95% decrease in CD8(+) T cells and B cells in the lungs at both early and late time points. In contrast, CD4(+) T cell accumulation was reduced by approximately 60% early, but equivalent to wt levels later. Surprisingly, many gammadelta T cells were found in lungs and blood of E(-)P(-) mice but were undetectable in the lungs and blood of wt mice. Absolute numbers of peripheral blood CD4, CD8, and B lymphocytes in E(-)P(-) mice equaled or exceeded the levels in wt mice, particularly after challenge. Trafficking studies using alphabeta T lymphoblasts confirmed that the recruitment of circulating cells after challenge was markedly reduced in E(-)P(-) mice. Furthermore, Ag priming occurred normally in both the selectin-deficient and wt mice, because primed lymphocytes from both groups transferred Ag sensitivity into naive wt mice. Lung production of mRNA for six CC and two CXC chemokines after challenge was equivalent by RT-PCR analysis in wt and E(-)P(-) mice. Therefore, reduced lung accumulation of alphabeta T cells and B cells in E(-)P(-) mice did not result from reduced delivery of circulating lymphocytes to the lungs, unsuccessful Ag priming, or defective pulmonary chemokine production. Selectin-dependent lymphocyte recruitment into the lungs following i.t.-SRBC challenge is subset specific and time dependent.

L-selectin serves as an E-selectin ligand on cultured human T lymphoblasts.

Previous studies reported that L-selectin (CD62L) on human peripheral blood neutrophils serves as an E-selectin ligand. This study shows that CD62L acquired E-selectin-binding activity following phorbol ester (PMA) treatment of the Jurkat T cell line and anti-CD3/IL-2-driven proliferation of human T lymphocytes in vitro. The recombinant porcine E-selectin/human Ig chimera P11.4 showed neuraminidase-sensitive and calcium-dependent attachment to PMA-stimulated human Jurkat T cells in a flow cytometry assay. The anti-CD62L mAb (DREG 56) blocked this binding interaction by approximately 60% and P11.4 precipitated CD62L from detergent lysates of PMA-activated Jurkat cells. In contrast, P11.4 precipitated minimal amounts of CD62L from detergent lysates of nonactivated human PBL. As reported previously, P-selectin glycoprotein ligand 1 and a distinct 130-kDa glycoprotein were the major species in these precipitates. However, T cell activation on plate-immobilized anti-CD3 and growth in low-dose IL-2 increased the percentage of CD62L molecules with E-selectin-binding activity. After two cycles of activation and culture, approximately 60-70% of the CD62L was precipitated with the P11.4 chimera. These cultured T lymphoblasts rolled avidly on both E-selectin and P-selectin at physiologic levels of linear shear stress. The DREG 56 Ab partially blocked rolling on the E-selectin substrate, whereas no effect was seen on P-selectin. Thus, CD62L on human cultured T lymphoblasts is one of several glycoproteins that interacts directly with E-selectin and contributes to rolling under flow.

E- and P-selectins are essential for the development of cockroach allergen-induced airway responses.

Peribronchial inflammation contributes to the pathophysiology of allergic asthma. In many vascular beds, adhesive interactions between leukocytes and the endothelial surface initiate the recruitment of circulating cells. Previous studies using OVA-induced airway hyperreactivity indicated that P-selectin, a member of the selectin family expressed by activated platelets and endothelium, contributed to both inflammation and bronchoconstriction. The current study used cockroach allergen (CRA), an allergen that induces asthmatic responses in both humans and mice, to further investigate the role of selectins in the development of peribronchial inflammation and airway hyperreactivity. P- and E-selectin mRNAs were detected in extracts of CRA-sensitized animals beginning shortly after intratracheal challenge with CRA. The P-selectin mRNA was transiently induced at early time points while up-regulation of the E-selectin mRNA was more prolonged. Mice with targeted deletions in E-selectin (E(-)), P-selectin (P(-)), and both genes (E(-)/P(-)) showed 70-85% reductions in airway hyperreactivity, peribronchial inflammation, and eosinophil accumulation. The P(-) and E(-)/P(-) groups showed the most profound reductions. The transfer of splenic lymphocytes from CRA-primed E(-)/P(-) into naive wild-type (WT) mice produced the same level of airway hyperreactivity as transfers from CRA-primed WT into naive WT hosts, indicating that peripheral immunization was similar. The observed changes in the selectin-deficient animals were not related to inadequate sensitization, because CRA priming and challenge increased serum IgE levels. Furthermore, pulmonary Th2-type cytokines and chemokines in the E-selectin(-/-) and WT animals were similar. The findings indicate that both P- and E-selectin contribute to CRA-induced peribronchial inflammation and airway hyperreactivity.

The monoclonal antibody CHO-131 binds to a core 2 O-glycan terminated with sialyl-Lewis x, which is a functional glycan ligand for P-selectin.

Core 2 O-glycans terminated with sialyl-Lewis x (sLe(X)) are functionally important oligosaccharides that endow particular macromolecules with high-affinity glycan ligands for the selectin family. To date, antibodies that recognize these structures on leukocytes have not been described. We characterize such a monoclonal antibody (mAb) here (CHO-131). The binding specificity of CHO-131 was directly examined by means of synthetic glycopeptides containing precise O-glycan structures. CHO-131 bound to sLe(X) extended from a core 2 branch (C2-O-sLe(X)), but CHO-131 demonstrated no reactivity if this oligosaccharide lacked fucose or if sLe(X) was extended from a core 1 branch. Using transfected cell lines, we found that CHO-131 binding required the functional activity of the glycosyltransferases alpha2,3-sialyltransferase, alpha1,3-fucosyltransferase-VII, and core 2 beta1,6 N-acetylglucosaminyltransferase (C2GnT). The C2-O-sLe(X) motif occurs primarily on sialomucins and has been directly shown to contribute to high-affinity P-selectin glycoprotein ligand-1 binding by P-selectin. Indeed, CHO-131 staining of neutrophils was diminished following sialomucin removal by O-glycoprotease, and its reactivity with transfected hematopoietic cell lines correlated with the expression of P-selectin ligands. CHO-131 also stained a small population of lymphocytes that were primarily CD3(+), CD4(+), and CD45RO(+) and represented a subset (37.8% +/- 18.3%) of cutaneous lymphocyte-associated antigen (CLA) T cells, distinguished by the mAb HECA-452, which detects sLe(X)-related glycans. Unlike anti-sLe(X) mAbs, CHO-131 binding also indicates C2GnT activity and demonstrates that CLA T cells are heterogeneous based on the glycan structures they synthesize. These findings support evidence that differential C2GnT activity results in T-cell subsets that express ligands for E-selectin, P-selectin, or both.

Mice genetically lacking endothelial selectins are resistant to the lethality in septic peritonitis.

Leukocyte interactions with vascular endothelium are an initial step for leukocyte entry into infectious foci where endothelial selectins may play a key role. Infiltrating leukocyte is essential for bacterial clearance, suggesting that endothelial selectins would be important in host defense against microorganisms. To address this, E-, P-, and E/P-selectin-deficient mice (E(-/-), P(-/-), E/P(-/-)) and wild-type (WT) mice underwent cecal ligation and puncture (CLP). Neither leukocyte infiltration nor bacterial load in the peritoneum was altered in E(-/-), P(-/-), and E/P(-/-) mice compared to WT mice. However, E(-/-), P(-/-), and E/P(-/-) mice were resistant to the lethality induced by CLP. At the mechanistic level, E(-/-), P(-/-), and E/P(-/-) mice did not develop renal dysfunction, a possible cause of death during sepsis. The serum level of interleukin-13 in E(-/-), P(-/-), and E/P(-/-) mice that had undergone CLP was higher than that in WT mice, whereas levels of macrophage inflammatory protein-2, KC in serum, and KC in kidney were lower than those in WT mice. These experiments demonstrate that endothelial selectin-mediated leukocyte rolling is not required for leukocyte entry in septic peritonitis and that endothelial selectins may affect mice survival during sepsis by influencing the cytokine profiles.