Characterization of cytokine interactions by flow cytometry and factorial analysis

BACKGROUND: Multiple cytokines are required for the growth and development of hematopoietic cells. The effect of many cytokines depends on the activity of other signaling pathways. These interactions are quantified using factorial experimental design and analysis. METHODS: Human umbilical cord blood (HUCB) CD34+ cells were cultured in fully defined media containing various combinations of recombinant cytokines as defined by resolution IV factorial (2(7-3)(IV)) or full factorial (2(4)) design experiments. The cytokines studied were stem cell factor (SCF), interleukin (IL)-3, megakaryocyte growth and development factor (MGDF), granulocyte-colony stimulating factor (G-CSF), Flt-3 ligand, IL-6, IL-11, and erythropoietin (EPO). In vitro cell divisions were tracked by staining CD34+ cells with 5-(and-6)-carboxyfluorescein diacetate, succinimidyl ester, followed by flow cytometric analysis at 4 days of culture. In separate experiments, lineage commitment and differentiation were determined at 7 days by immunophenotype. RESULTS: In addition to the main effects of single cytokines, cytokine interactions were identified. There was a negative interaction between IL-3 and MGDF that resulted in a less than additive effect of these factors on erythroid and megakaryocytic development. The effect of Flt-3 ligand and SCF factor on CD34+ cell production was also less than additive, although the response to both cytokines was greater than single cytokines. The only positive interaction that was identified was between EPO and SCF, which resulted in the synergistic production of erythroid cells. CONCLUSIONS: Factorial analysis provides a powerful methodology to study the integration of multiple signals at the cellular and molecular level.     

NMR solution structure of the inserted domain of human leukocyte function associated antigen-1

The interaction between the leukocyte function-associated antigen-1 (LFA-1) and the intercellular adhesion molecule is thought to be mediated primarily via the inserted domain (I-domain) in the alpha-subunit. The activation of LFA-1 is an early step in triggering the adhesion of leukocytes to target cells decorated with intercellular adhesion molecules. There is some disagreement in the literature over the respective roles of conformational changes in the I-domain and of divalent cations (Mg(2+), Mn(2+)) in the activation of LFA-1 for intercellular adhesion molecule binding. X-ray crystallographic structures of the I-domains of LFA-1 and Mac-1 in the presence and absence of cations show structural differences in the C-terminal alpha-helix; this change was proposed to represent the active and inactive conformations of the I-domain. However, more recent X-ray results have called this proposal into question. The solution structure of the Mg(2+) complex of the I-domain of LFA-1 has been determined by NMR methods, using a model-based approach to nuclear Overhauser enhancement spectroscopy peak assignment. The protein adopts the same structure in solution as that of the published I-domain X-ray structures, but the C-terminal region, where the X-ray structures are most different from each other, is different again in the solution structures. The secondary structure of this helix is well formed, but NMR relaxation data indicate that there is considerable flexibility present, probably consisting of breathing or segmental motion of the helix. The conformational diversity seen in the various X-ray structures could be explained as a result of the inherent flexibility of this C-terminal region and as a result of crystal contacts. Our NMR data are consistent with a model where the C-terminal helix has the potential flexibility to take up alternative conformations, for example, in the presence and absence of the intercellular adhesion molecule ligand. The role of divalent cations appears from our results not to be as a direct mediator of a conformational change that alters affinity for the ligand. Rather, the presence of the cation appears to be involved in some other way in ligand binding, perhaps by acting as a bridge to the ligand and by modulation of the charge of the binding surface.     

Distributed torsion angle grid search in high dimensions: A systematic approach to NMR structure determination

Two complementary approaches for systematic search in torsion angle space are described for the generation of all conformations of polypeptides which satisfy experimental NMR restraints, hard-sphere van der Waals radii, and rigid covalent geometry. The first procedure is based on a recursive, tree search algorithm for the examination of linear chains of torsion angles, and uses a novel treatment to propagate the search results to neighboring regions so that the structural consequences of the restraints are fully realized. The second procedure is based on a binary combination of torsion vector spaces for connected submolecules, and produces intermediate results in Cartesian space for a more robust restraint analysis. Restraints for NMR applications include bounds on torsion angles and internuclear distances, including relational and degenerate restraints involving equivalent and nonstereoassigned protons. To illustrate these methods, conformation search results are given for the tetrapeptide APGA restrained to an idealized -turn conformation, an alanine octapeptide restrained to a right-handed helical conformation, and the structured region of the peptide SYPFDV.     

Nuclear magnetic resonance structural studies and molecular modeling of duplex DNA containing normal and 4'-oxidized abasic sites

A 4'-oxidized abasic site (X) has been synthesized in a defined duplex DNA sequence, 5'-d(CCAAAGXACCGGG)-3'/3'-d(GGTTTCATGGCCC)-5' (1). Its structure has been determined by two-dimensional NMR methods, molecular modeling, and molecular dynamics simulations. 1 is globally B-form with the base (A) opposite X intrahelical and well-stacked. Only the alpha anomer of X is observed, and the abasic site deoxyribose is largely intrahelical. These results are compared with a normal abasic site (Y) in the same sequence context (2). Y is composed of a 60:40 mixture of alpha and beta anomers (2alpha and 2beta). In both 2alpha and 2beta, the base (A) opposite Y is intrahelical and well-stacked and the abasic site deoxyribose is predominantly extrahelical, consistent with the reported structures of the normal abasic site in a similar sequence context [Hoehn, S. T., Turner, C. J., and Stubbe, J. (2001) Nucleic Acids Res. 29, 3413-3423]. Molecular dynamics simulations reveal that the normal abasic site appears to be conformationally more flexible than the 4'-oxidized abasic site. The importance of the structure and flexibility of the abasic site in the recognition by the DNA repair enzyme Ape1 is discussed.     

Phosphopantetheine adenylyltransferase from Escherichia coli: investigation of the kinetic mechanism and role in regulation of coenzyme A biosynthesis

Phosphopantetheine adenylyltransferase (PPAT) from Escherichia coli is an essential hexameric enzyme that catalyzes the penultimate step in coenzyme A (CoA) biosynthesis and is a target for antibacterial drug discovery. The enzyme utilizes Mg-ATP and phosphopantetheine (PhP) to generate dephospho-CoA (dPCoA) and pyrophosphate. When overexpressed in E. coli, PPAT copurifies with tightly bound CoA, suggesting a feedback inhibitory role for this cofactor. Using an enzyme-coupled assay for the forward-direction reaction (dPCoA-generating) and isothermal titration calorimetry, we investigated the steady-state kinetics and ligand binding properties of PPAT. All substrates and products bind the free enzyme, and product inhibition studies are consistent with a random bi-bi kinetic mechanism. CoA inhibits PPAT and is competitive with ATP, PhP, and dPCoA. Previously published structures of PPAT crystallized at pH 5.0 show half-the-sites reactivity for PhP and dPCoA and full occupancy by ATP and CoA. Ligand-binding studies at pH 8.0 show that ATP, PhP, dPCoA, and CoA occupy all six monomers of the PPAT hexamer, although CoA exhibits two thermodynamically distinct binding modes. These results suggest that the half-the-sites reactivity observed in PPAT crystal structures may be pH dependent. In light of previous studies on the regulation of CoA biosynthesis, the PPAT kinetic and ligand binding data suggest that intracellular PhP concentrations modulate the distribution of PPAT monomers between high- and low-affinity CoA binding modes. This model is consistent with PPAT serving as a “backup” regulator of pathway flux relative to pantothenate kinase.     

A new model for chemical shifts of amide hydrogens in proteins

We propose a new computational model to predict amide proton chemical shifts in proteins. In addition to the ring-current, susceptibility and electrostatic effects of earlier models, we add a hydrogen-bonding term based on density functional calculations of model peptide–peptide and peptide–water systems. Both distance and angular terms are included, and the results are rationalized in terms of natural bond orbital analysis of the interactions. Comparison to observed shifts for 15 proteins shows a significant improvement over existing structure-shift correlations. These additions are incorporated in a new version of the SHIFTS program.     

Frequency‐dependent fitness in gynodioecious Lobelia siphilitica

Selection is frequency dependent when an individual's fitness depends on the frequency of its phenotype. Frequency‐dependent selection should be common in gynodioecious plants, where individuals are female or hermaphroditic; if the fitness of females is limited by the availability of pollen to fertilize their ovules, then they should have higher fitness when rare than when common. To test whether the fitness of females is frequency dependent, we manipulated the sex ratio in arrays of gynodioecious Lobelia siphilitica. To test whether fitness was frequency dependent because of variation in pollen availability, we compared open‐pollinated and supplemental hand‐pollinated plants. Open‐pollinated females produced more seeds when they were rare than when they were common, as expected if fitness is negatively frequency dependent. However, hand‐pollinated females also produced more seeds when they were rare, indicating that variation in pollen availability was not the cause of frequency‐dependent fitness. Instead, fitness was frequency dependent because both hand‐ and open‐pollinated females opened more flowers when they were rare than when they were common. This plasticity in the rate of anthesis could cause fitness to be frequency dependent even when reproduction is not pollen limited, and thus expand the conditions under which frequency‐dependent selection operates in gynodioecious species.