Discontinuous nanoporous membranes reduce non-specific fouling for immunoaffinity cell capture.

The microfluidic isolation of target cells using adhesion-based surface capture has been widely explored for biology and medicine. However, high-throughput processing can be challenging due to interfacial limitations such as transport, reaction, and non-specific fouling. Here, it is shown that antibody-functionalized capture surfaces with discontinuous permeability enable efficient target cell capture at high flow rates by decreasing fouling. Experimental characterization and theoretical modeling reveal that "wall effects" affect cell-surface interactions and promote excess surface accumulation. These issues are partially circumvented by reducing the transport and deposition of cells near the channel walls. Optimized microfluidic devices can be operated at higher cell concentrations with significant improvements in throughput.

Paradoxical glomerular filtration of carbon nanotubes.

The molecular weight cutoff for glomerular filtration is thought to be 30-50 kDa. Here we report rapid and efficient filtration of molecules 10-20 times that mass and a model for the mechanism of this filtration. We conducted multimodal imaging studies in mice to investigate renal clearance of a single-walled carbon nanotube (SWCNT) construct covalently appended with ligands allowing simultaneous dynamic positron emission tomography, near-infrared fluorescence imaging, and microscopy. These SWCNTs have a length distribution ranging from 100 to 500 nm. The average length was determined to be 200-300 nm, which would yield a functionalized construct with a molecular weight of approximately 350-500 kDa. The construct was rapidly (t(1/2) approximately 6 min) renally cleared intact by glomerular filtration, with partial tubular reabsorption and transient translocation into the proximal tubular cell nuclei. Directional absorption was confirmed in vitro using polarized renal cells. Active secretion via transporters was not involved. Mathematical modeling of the rotational diffusivity showed the tendency of flow to orient SWCNTs of this size to allow clearance via the glomerular pores. Surprisingly, these results raise questions about the rules for renal filtration, given that these large molecules (with aspect ratios ranging from 100:1 to 500:1) were cleared similarly to small molecules. SWCNTs and other novel nanomaterials are being actively investigated for potential biomedical applications, and these observations-that high aspect ratio as well as large molecular size have an impact on glomerular filtration-will allow the design of novel nanoscale-based therapeutics with unusual pharmacologic characteristics.

Antibody-functionalized fluid-permeable surfaces for rolling cell capture at high flow rates.

Adhesion-based cell capture on surfaces in microfluidic devices forms the basis of numerous biomedical diagnostics and in vitro assays. However, the performance of these platforms is partly limited by interfacial phenomena that occur at low Reynolds numbers. In contrast, cell homing to porous vasculature is highly effective in vivo during inflammation, stem cell trafficking, and cancer metastasis. Here, we show that a porous, fluid-permeable surface functionalized with cell-specific antibodies promotes efficient and selective cell capture in vitro. This architecture is advantageous due to enhanced transport as streamlines are diverted toward the surface. Moreover, specific cell-surface interactions are promoted due to reduced shear, allowing gentle cell rolling and arrest. Together, these synergistic effects enable highly effective cell capture at flow rates more than an order of magnitude larger than those provided by existing devices with solid surfaces.

Nitric oxide produced endogenously is responsible for hypoxia-induced HIF-1α stabilization in colon carcinoma cells.

Hypoxia-inducible factor-1α (HIF-1α) is a critical regulator of cellular responses to hypoxia. Under normoxic conditions, the cellular HIF-1α level is regulated by hydroxylation by prolyl hydroxylases (PHDs), ubiquitylation, and proteasomal degradation. During hypoxia, degradation decreases, and its intracellular level is increased. Exogenously administered nitric oxide (NO)-donor drugs stabilize HIF-1α; thus, NO is suggested to mimic hypoxia. However, the role of low levels of endogenously produced NO generated during hypoxia in HIF-1α stabilization has not been defined. Here, we demonstrate that NO and reactive oxygen species (ROS) produced endogenously by human colon carcinoma HCT116 cells are responsible for HIF-1α accumulation in hypoxia. The antioxidant N-acetyl-L-cysteine (NAC) and NO synthase inhibitor N(G)-monomethyl L-arginine (L-NMMA) effectively reduced HIF-1α stabilization and decreased HIF-1α hydroxylation. These effects suggested that endogenous NO and ROS impaired PHD activity, which was confirmed by reversal of L-NMMA- and NAC-mediated effects in the presence of dimethyloxaloylglycine, a PHD inhibitor. Thiol reduction with dithiothreitol decreased HIF-1α stabilization in hypoxic cells, while dinitrochlorobenzene, which stabilizes S-nitrosothiols, favored its accumulation. This suggested that ROS- and NO-mediated HIF-1α stabilization involved S-nitrosation, which was confirmed by demonstrating increased S-nitrosation of PHD2 during hypoxia. Our results support a regulatory mechanism of HIF-1α during hypoxia in which endogenously generated NO and ROS promote inhibition of PHD2 activity, probably by its S-nitrosation.

Delivery method, target gene structure, and growth properties of target cells impact mutagenic responses to reactive nitrogen and oxygen species.

Dysregulated production of nitric oxide (NO•) and reactive oxygen species (ROS) by inflammatory cells in vivo may contribute to mutagenesis and carcinogenesis. Here, we compare cytotoxicity and mutagenicity induced by NO• and ROS in TK6 and AS52 cells, delivered by two methods: a well-characterized delivery system and a novel adaptation of a system for coculture. When exposed to preformed NO•, a cumulative dose of 620 µM min reduced the viability of TK6 cells at 24 h to 36% and increased mutation frequencies in the HPRT and TK1 genes to 7.7 × 10⁻6 (p < 0.05) and 24.8 × 10⁻6 (p < 0.01), 2.7- and 3.7-fold higher than background, respectively. In AS52 cells, cumulative doses of 1700 and 3700 µM min reduced viability to 49 and 22%, respectively, and increased the mutation frequency 10.2- and 14.6-fold higher than the argon control (132 × 10⁻6 and 190 × 10⁻6, respectively). These data show that TK6 cells were more sensitive than AS52 cells to killing by NO•. However, the two cell lines were very similar in relative susceptibility to mutagenesis; on the basis of fold increases in MF, average relative sensitivity values [(MF(exp)/MF(control))/cumulative NO• dose] were 5.16 × 10⁻³ and 4.97 × 10⁻³ µM⁻¹ min⁻¹ for TK6 cells and AS52 cells, respectively. When AS52 cells were exposed to reactive species generated by activated macrophages in the coculture system, cell killing was greatly reduced by the addition of NMA to the culture medium and was completely abrogated by combined additions of NMA and the superoxide scavenger Tiron, indicating the relative importance of NO• to loss of viability. Exposure in the coculture system for 48 h increased mutation frequency in the gpt gene by more than 9-fold, and NMA plus Tiron again completely prevented the response. Molecular analysis of gpt mutants induced by preformed NO• or by activated macrophages revealed that both doubled the frequency of gene inactivation (40% in induced vs 20% in spontaneous mutants). Sequencing showed that base-substitution mutations dominated the spectra, with transversions (30-40%) outnumbering transitions (10-20%). Virtually all mutations took place at guanine sites in the gene. G:C to T:A transversions accounted for about 30% of both spontaneous and induced mutations; G:C to A:T transitions amounted to 10-20% of mutants; insertions, small deletions, and multiple mutations were present at frequencies of 0-10%. Taken together, these results indicate that cell type and proximity to generator cells are critical determinants of cytotoxic and genotoxic responses induced by NO• and reactive species produced by activated macrophages.

A system for exposing molecules and cells to biologically relevant and accurately controlled steady-state concentrations of nitric oxide and oxygen.

Nitric oxide (NO) plays key roles in cell signaling and physiology, with diverse functions mediated by NO concentrations varying over three orders-of-magnitude. In spite of this critical concentration dependence, current approaches to NO delivery in vitro result in biologically irrelevant and poorly controlled levels, with hyperoxic conditions imposed by ambient air. To solve these problems, we developed a system for controlled delivery of NO and O(2) over large concentration ranges to mimic biological conditions. Here we describe the fabrication, operation and calibration of the delivery system. We then describe applications for delivery of NO and O(2) into cell culture media, with a comparison of experimental results and predictions from mass transfer models that predict the steady-state levels of various NO-derived reactive species. We also determined that components of culture media do not affect the steady-state levels of NO or O(2) in the device. This system provides critical control of NO delivery for in vitro models of NO biology and chemistry.

Nitric oxide, oxygen, and superoxide formation and consumption in macrophages and colonic epithelial cells.

Knowledge of the rates at which macrophages and epithelial cells synthesize NO is critical for predicting the concentrations of NO and other reactive nitrogen species in colonic crypts during inflammation, and elucidating the linkage between inflammatory bowel disease, NO, and cancer. Macrophage-like RAW264.7 cells, primary bone marrow-derived macrophages (BMDM), and HCT116 colonic epithelial cells were subjected to simulated inflammatory conditions, and rates of formation and consumption were determined for NO, O(2), and O(2)(-). Production rates of NO were determined in either of two ways: continuous monitoring of NO concentrations in a closed chamber with corrections for autoxidation, or NO(2)(-) accumulation measurements in an open system with corrections for diffusional losses of NO. The results obtained using the two methods were in excellent agreement. Rates of NO synthesis (2.3 +/- 0.6 pmol s(-1) 10(6) cells(-1)), NO consumption (1.3 +/- 0.3 s(-1)), and O(2) consumption (59 +/- 17 pmol s(-1) 10(6) cells(-1) when NO is negligible) for activated BMDM were indistinguishable from those of activated RAW264.7 cells. NO production rates calculated from NO(2)(-) accumulation data for HCT116 cells infected with Helicobacter cinaedi (3.9 +/- 0.1 pmol s(-1) 10(6) cells(-1)) were somewhat greater than those of RAW264.7 macrophages infected under similar conditions (2.6 +/- 0.1 pmol s(-1) 10(6) cells(-1)). Thus, RAW264.7 cells have NO kinetics nearly identical to those of primary macrophages, and stimulated epithelial cells are capable of synthesizing NO at rates comparable to those of macrophages. Using these cellular kinetic parameters, simulations of NO diffusion and reaction in a colonic crypt during inflammation predict maximum NO concentrations of about 0.2 microM at the base of a crypt.

Prediction of nitric oxide concentrations in melanomas.

The presence of iNOS and nitrotyrosine in cutaneous melanomas has been correlated with poor survival rates of patients, suggesting that NO plays a role in the tumor pathophysiology. However, the concentrations of NO that melanoma cells are exposed to in vivo have been unknown. To provide cell kinetic data for use in predicting those concentrations, synthesis and consumption of NO was examined in A375 melanoma cells. Nitric oxide synthesis was undetectable. The rate of intracellular NO consumption was determined by continuous monitoring of NO concentrations following injection of NO solutions in a closed chamber. After correcting for autoxidation and consumption from media-generated O(2)(-), the rate constant obtained for cellular consumption was 7.1±1.1 s(-1). This information was combined with previous data on macrophage NO kinetics to develop a mathematical model to predict NO levels in cutaneous melanomas. Synthesis of NO by macrophages in the stroma was found to give a maximum concentration at the tumor periphery of 0.2 µM. Because of the high rates of cellular consumption, the elevation in NO concentration is predicted to be very localized, approximately 90% of the concentration decay occurring within 30 µm of the tumor edge. High NO concentrations at the periphery of a melanoma may contribute to metastasis by stimulating cell proliferation, inhibiting apoptosis, or acting as a lymphangiogenic factor.

Effects of charge on osmotic reflection coefficients of macromolecules in fibrous membranes.

A model based on continuum hydrodynamics and electrostatics was developed to predict the combined effects of molecular charge and size on the osmotic reflection coefficient (sigma(o)) of a macromolecule in a fibrous membrane, such as a biological hydrogel. The macromolecule was represented as a sphere with a constant surface charge density, and the membrane was assumed to consist of an array of parallel fibers of like charge, also with a constant surface charge density. The flow was assumed to be parallel to the fiber axes. The effects of charge were included by computing the electrostatic free energy for a sphere interacting with an array of fibers. It was shown that this energy could be approximated using a pairwise additivity assumption. Results for sigma(o) were obtained for two types of negatively charged fibers, one with properties like those of glycosaminoglycan chains, and the other for thicker fibers having a range of charge densities. Using physiologically reasonable fiber spacings and charge densities, sigma(o) for bovine serum albumin in either type of fiber array was shown to be much larger than that for an uncharged system. Given the close correspondence between sigma(o) and the reflection coefficient for filtration, the results suggest that the negative charge of structures such as the endothelial surface glycocalyx is important in minimizing albumin loss from the circulation.

Resolved: normal glomeruli filter nephrotic levels of albumin.

The glomerular filtration barrier has long been thought largely impermeable to albumin. Startling new data suggest it may not be especially important in this process. Much of the argument hinges on whether charge selectivity really exists, how feasible it is for enormous quantities of albumin to instead be reclaimed by the proximal tubule, and the technical merits of previous experiments. Three experts in the field debate the merits of this argument.

Darcy permeability of agarose-glycosaminoglycan gels analyzed using fiber-mixture and donnan models.

Agarose-glycosaminoglycan (GAG) membranes were synthesized to provide a model system in which the factors controlling the Darcy (or hydraulic) permeability could be assessed in composite gels of biological relevance. The membranes contained a GAG (chondroitin sulfate) that was covalently bound to agarose via terminal amine groups, and the variables examined were GAG concentration and solution ionic strength. The addition of even small amounts of GAG (0.4 vol/vol %) resulted in a twofold reduction in the Darcy permeability of 3 vol/vol % agarose gels. Electrokinetic coupling, caused by the negative charge of the GAG, resulted in an additional twofold reduction in the open-circuit permeability when the ionic strength was decreased from 1 M to 0.01 M. A microstructural hydrodynamic model was developed, based on a mixture of neutral, coarse fibers (agarose fibrils), and fine, charged fibers (GAG chains). Heterogeneity within agarose gels was modeled by assuming that fiber-rich, spherical inclusions were distributed throughout a fiber-poor matrix. That model accurately predicted the Darcy permeability when the ionic strength was high enough to suppress the effects of charge, but underestimated the influence of ionic strength. A more macroscopic approach, based on Donnan equilibria, better captured the reductions in Darcy permeability caused by GAG charge.

Kinetic analysis of intracellular concentrations of reactive nitrogen species.

Reactive nitrogen species derived from NO have been implicated in cancer and other diseases, but their intracellular concentrations are largely unknown. To estimate them under steady-state conditions representative of inflamed tissues, a kinetic model was developed that included the effects of cellular antioxidants, amino acids, proteins, and lipids. For an NO concentration of 1 microM, total peroxynitrite (Per, the sum of ONOO(-) and ONOOH), NO(2)(*), and N(2)O(3) were calculated to have concentrations in the nanomolar, picomolar, and femtomolar ranges, respectively. The concentrations of NO(2)(*) and N(2)O(3) were predicted to decrease markedly with increases in glutathione (GSH) levels, due to the scavenging of each by GSH. Although lipids accelerate the oxidation of NO by O(2) (because of the high solubility of each in hydrophobic media), lipid-phase reactions were calculated to have little effect on NO(2)(*) or N(2)O(3) concentrations. The major sources of intracellular NO(2)(*) were found to be the reaction of Per with metals and with CO(2), whereas the major sinks were its reactions with GSH and ascorbate (AH(-)). The radical-scavenging ability of GSH and AH(-) caused 3-nitrotyrosine to be the only tyrosine derivative predicted to be formed at a significant rate. The major GSH reaction product was S-nitrosoglutathione. Analytical (algebraic) expressions are provided for the concentrations of the key reactive intermediates, allowing the calculations to be extended readily.

Prediction of nitric oxide concentrations in colonic crypts during inflammation.

Nitric oxide production in the colon has been linked to inflammatory bowel disease (IBD) and increased risk for colon cancer. However, measurements of NO concentration in the inflamed colon have not been available and it is not known what NO levels are pathophysiological. A computational model, based on anatomical length scales and rates of NO production measured in cell cultures, was used to predict spatially varying NO concentrations within a colonic crypt under inflammatory conditions. A variety of scenarios were considered, including different spatial distributions of macrophages and a range of possible macrophage and epithelial synthesis rates for NO. Activated macrophages arranged as a monolayer at the base of the crypt elicited maximum NO concentrations of approximately 0.3 microM. The epithelial contribution to NO synthesis was calculated to be negligible. Assuming a uniform macrophage layer, NO synthesis rates greater than 20 microM/s, or more than three times that measured in vitro, would be necessary to achieve maximum NO concentrations of 1 microM in the crypt. Thus, unless NO synthesis rates in macrophages and/or epithelial cells greatly exceed those measured in cell cultures, NO concentrations will remain submicromolar in the crypt during inflammation. Additionally, the results were used to predict the range of NO concentrations (<0.3 microM) and cumulative NO dose (560 microM min) experienced by a given epithelial cell migrating from the base to the top of the crypt. These estimates of NO concentrations in inflamed crypts should facilitate efforts to elucidate the molecular biological linkage between NO exposure and carcinogenesis in IBD.

A mathematical analysis of Prx2-STAT3 disulfide exchange rate constants for a bimolecular reaction mechanism.

Appreciation of peroxiredoxins as the major regulators of H2O2 concentrations in human cells has led to a new understanding of redox signaling. In addition to their status as the primary reducers of H2O2 to water, the oxidized peroxiredoxin byproduct of this reaction has recently been shown capable of participation in H2O2-mediated signaling pathways through disulfide exchange reactions with the transcription factor STAT3. The dynamics of peroxidase-transcription factor disulfide exchange reactions have not yet been considered in detail with respect to how these reactions fit into the larger network of competing reactions in human cells. In this study, we used a kinetic model of oxidation and reduction reactions related to H2O2 metabolism in the cytosol of human cells to study the dynamics of peroxiredoxin-2 mediated oxidation of the redox-regulated transcription factor STAT3. In combination with previously reported experimental data, the model was used to estimate the rate coefficient of a biomolecular reaction between Prx2 and STAT3 for two sets of assumptions that constitute lower and upper bound cases. Using these estimates, we calculated the relative rates of the reaction of oxidized peroxiredoxin-2 and STAT3 and other competing reactions in the cytosol. These calculations revealed that peroxiredoxin-2-mediated oxidation of STAT3 likely occurs at a much slower rate than competing reactions in the cytosol. This analysis suggests the existence of more complex mechanisms, potentially involving currently unknown protein-protein recognition partners, which facilitate disulfide exchange reactions between peroxiredoxin-2 and STAT3.

What determines glomerular capillary permeability?

There have been exciting recent advances in our understanding of the structural and molecular biology of the glomerular slit diaphragm, as described in a report in this issue of the JCI. These findings, combined with data on the permeability of the basement membrane and evidence that the endothelium may be a more important barrier than often supposed, are allowing a clearer understanding to emerge of how the 3 parts of the glomerular capillary wall jointly determine its functional properties.

Binding of glycosaminoglycans to cyano-activated agarose membranes: kinetic and diffusional effects on yield and homogeneity.

Methods were developed for binding a glycosaminoglycan (GAG, a 50 kDa chondroitin sulfate) to thin agarose membranes using 1-cyano-4-(dimethylamino)pyridinium tetrafluoroborate (CDAP) as the activating agent. Process conditions were optimized to achieve high yields and spatially uniform concentrations of bound ligand. Yields were varied mainly by manipulating the duration and temperature of the aqueous washes prior to coupling, which affected the concentration of active sites available for subsequent GAG binding. The rate constants for degradation of the active cyanate esters in 0.1M bicarbonate solutions were 0.24+/-0.02 h(-1) at 4 degrees C and 0.08+/-0.03 h(-1) at 0 degrees C. Steric limitations in the 3% agarose gels severely restricted binding, with only about 0.1% of active sites being accessible to GAG molecules. The GAG binding occurred primarily in the outer 50-70 microm of the membranes, so that coupling was homogeneous only for thin gels. A model of GAG diffusion and reaction in the coupling step was developed to explain the observed effects of parameters such as the GAG concentration in solution and the membrane thickness. An analysis of the key time scales in the synthesis provides design principles that should be useful also for other cyanylating agents, other ligands, and for beads as well as membranes.

A reaction-diffusion model of cytosolic hydrogen peroxide.

As a signaling molecule in mammalian cells, hydrogen peroxide (H2O2) determines the thiol/disulfide oxidation state of several key proteins in the cytosol. Localization is a key concept in redox signaling; the concentrations of signaling molecules within the cell are expected to vary in time and in space in manner that is essential for function. However, as a simplification, all theoretical studies of intracellular hydrogen peroxide and many experimental studies to date have treated the cytosol as a well-mixed compartment. In this work, we incorporate our previously reported reduced kinetic model of the network of reactions that metabolize hydrogen peroxide in the cytosol into a model that explicitly treats diffusion along with reaction. We modeled a bolus addition experiment, solved the model analytically, and used the resulting equations to quantify the spatiotemporal variations in intracellular H2O2 that result from this kind of perturbation to the extracellular H2O2 concentration. We predict that micromolar bolus additions of H2O2 to suspensions of HeLa cells (0.8 × 10(9)cells/l) result in increases in the intracellular concentration that are localized near the membrane. These findings challenge the assumption that intracellular concentrations of H2O2 are increased uniformly throughout the cell during bolus addition experiments and provide a theoretical basis for differing phenotypic responses of cells to intracellular versus extracellular perturbations to H2O2 levels.

Analysis of the lifetime and spatial localization of hydrogen peroxide generated in the cytosol using a reduced kinetic model.

Hydrogen peroxide (H2O2) acts as a signaling molecule via its reactions with particular cysteine residues of certain proteins. Determining the roles of direct oxidation by H2O2 versus disulfide exchange reactions (i.e. relay reactions) between oxidized and reduced proteins of different identities is a current focus. Here, we use kinetic modeling to estimate the spatial and temporal localization of H2O2 and its most likely oxidation targets during a sudden increase in H2O2 above the basal level in the cytosol. We updated a previous redox kinetic model with recently measured parameters for HeLa cells and used the model to estimate the length and time scales of H2O2 diffusion through the cytosol before it is consumed by reaction. These estimates were on the order of one micron and one millisecond, respectively. We found oxidation of peroxiredoxin by H2O2 to be the dominant reaction in the network and that the overall concentration of reduced peroxiredoxin is not significantly affected by physiological increases in intracellular H2O2 concentration. We used this information to reduce the model from 22 parameters and reactions and 21 species to a single analytical equation with only one dependent variable, i.e. the concentration of H2O2, and reproduced results from the complete model. The reduced kinetic model will facilitate future efforts to progress beyond estimates and precisely quantify how reactions and diffusion jointly influence the distribution of H2O2 within cells.

Glomerular filtration of albumin: how small is the sieving coefficient?

A model was developed to describe how the concentration of albumin in proximal tubule fluid will vary with axial position, including the effects of luminal flow and water and albumin reabsorption. The results show that the high albumin sieving coefficients proposed recently cannot be reconciled with micropuncture data in rats.

Effects of concentration on the partitioning of macromolecule mixtures in agarose gels.

To test the effects of solute concentration on the equilibrium partitioning of single macromolecules and macromolecule mixtures between bulk solutions and gels, the partition coefficient in agarose was measured for BSA and for four narrow fractions of Ficoll with Stokes radii of 30-59 A. Solutions of each test macromolecule were equilibrated with a known volume of gel, final liquid concentrations measured, and partition coefficients (gel concentration divided by bulk concentration) calculated by applying a material balance. The partition coefficient of each macromolecule was measured in 4 and 6% gels under dilute conditions and with BSA present at initial concentrations up to 13.5 g/dl. As expected, the partition coefficients decreased with increasing agarose concentration and with increasing macromolecular size. Moreover, increasing the BSA concentration increased the partition coefficient of BSA itself and that of all four Ficolls. This effect was most pronounced for the largest test solutes. Measurements at two ionic strengths confirmed that electrostatic interactions were negligible under the conditions used. The experimental results were compared with predictions from a previously developed excluded volume theory for the partitioning of mixtures of rigid, spheroidal macromolecules in fibrous media. Agarose was represented as a randomly oriented array of cylindrical fibers, BSA as a prolate spheroid, and Ficoll as a sphere. The quantitative agreement between the model predictions and the data was generally quite good, indicating that steric interactions among solute molecules and between solute molecules and gel fibers could explain the partitioning results. The theory is simple enough computationally to be applied to a variety of processes that are influenced by the equilibrium partitioning of macromolecules.