Genome-wide system analysis reveals stable yet flexible network dynamics in yeast.

Recently, important insights into static network topology for biological systems have been obtained, but still global dynamical network properties determining stability and system responsiveness have not been accessible for analysis. Herein, we explore a genome-wide gene-to-gene regulatory network based on expression data from the cell cycle in Saccharomyces cerevisae (budding yeast). We recover static properties like hubs (genes having several out-going connections), network motifs and modules, which have previously been derived from multiple data sources such as whole-genome expression measurements, literature mining, protein-protein and transcription factor binding data. Further, our analysis uncovers some novel dynamical design principles; hubs are both repressed and repressors, and the intra-modular dynamics are either strongly activating or repressing whereas inter-modular couplings are weak. Finally, taking advantage of the inferred strength and direction of all interactions, we perform a global dynamical systems analysis of the network. Our inferred dynamics of hubs, motifs and modules produce a more stable network than what is expected given randomised versions. The main contribution of the repressed hubs is to increase system stability, while higher order dynamic effects (e.g. module dynamics) mainly increase system flexibility. Altogether, the presence of hubs, motifs and modules induce few flexible modes, to which the network is extra sensitive to an external signal. We believe that our approach, and the inferred biological mode of strong flexibility and stability, will also apply to other cellular networks and adaptive systems.

Deconstructing the core dynamics from a complex time-lagged regulatory biological circuit.

Complex regulatory dynamics is ubiquitous in molecular networks composed of genes and proteins. Recent progress in computational biology and its application to molecular data generate a growing number of complex networks. Yet, it has been difficult to understand the governing principles of these networks beyond graphical analysis or extensive numerical simulations. Here the authors exploit several simplifying biological circumstances which thereby enable to directly detect the underlying dynamical regularities driving periodic oscillations in a dynamical nonlinear computational model of a protein-protein network. System analysis is performed using the cell cycle, a mathematically well-described complex regulatory circuit driven by external signals. By introducing an explicit time delay and using a 'tearing-and-zooming' approach the authors reduce the system to a piecewise linear system with two variables that capture the dynamics of this complex network. A key step in the analysis is the identification of functional subsystems by identifying the relations between state-variables within the model. These functional subsystems are referred to as dynamical modules operating as sensitive switches in the original complex model. By using reduced mathematical representations of the subsystems the authors derive explicit conditions on how the cell cycle dynamics depends on system parameters, and can, for the first time, analyse and prove global conditions for system stability. The approach which includes utilising biological simplifying conditions, identification of dynamical modules and mathematical reduction of the model complexity may be applicable to other well-characterised biological regulatory circuits. [Includes supplementary material].

Correlation of serum IGF-I and IGFBP-1 and -3 to cardiovascular risk indicators and early carotid atherosclerosis in healthy middle-aged men.

OBJECTIVES: IGF-I, IGFBP-1 and IGFBP-3 are putative mediators in cardiovascular disease. The present study examined (i) the correlations of circulating IGF-I, IGFBP-1 and IGFBP-3 to established cardiovascular risk factors and signs of early atherosclerosis as reflected by ultrasound measurement of common carotid intima-media thickness (IMT), and (ii) whether serum concentrations of these analytes are modulated during alimentary lipaemia. DESIGN: Cross-sectional clinical study. PATIENTS: A biobank and clinical database based on 96 healthy Caucasian men, aged 50 years, with an apolipoprotein (apo) E3/E3 genotype, who had originally undergone investigations of postprandial lipoprotein metabolism was used for the study. MEASUREMENTS: Total IGF-I, IGFBP-1 and IGFBP-3 were determined in serum by radioimmunoassay (RIA). Free IGF-I was measured by a commercial two-site immunoradiometric assay (IRMA). RESULTS: In multivariate analyses, fasting serum free IGF-I correlated inversely with IMT and accounted for 5% of the variation in multiple R(2). When fasting serum IGFBP-1 was entered in the models instead of IGF-I, IGFBP-1 correlated positively with IMT and accounted for 6% of the variation in IMT. IGFBP-3 and total IGF-I were unrelated to IMT. There were no associations between free IGF-I and cardiovascular risk factors, whereas IGFBP-1 behaved like a component of the insulin resistance syndrome. Serum free IGF-I increased and IGFBP-1 decreased postprandially. CONCLUSION: The data indicate that serum free IGF-I and IGFBP-1 are implicated in early atherosclerosis.

Growing Bayesian network models of gene networks from seed genes.

MOTIVATION: For the last few years, Bayesian networks (BNs) have received increasing attention from the computational biology community as models of gene networks, though learning them from gene-expression data is problematic. Most gene-expression databases contain measurements for thousands of genes, but the existing algorithms for learning BNs from data do not scale to such high-dimensional databases. This means that the user has to decide in advance which genes are included in the learning process, typically no more than a few hundreds, and which genes are excluded from it. This is not a trivial decision. We propose an alternative approach to overcome this problem. RESULTS: We propose a new algorithm for learning BN models of gene networks from gene-expression data. Our algorithm receives a seed gene S and a positive integer R from the user, and returns a BN for the genes that depend on S such that less than R other genes mediate the dependency. Our algorithm grows the BN, which initially only contains S, by repeating the following step R + 1 times and, then, pruning some genes; find the parents and children of all the genes in the BN and add them to it. Intuitively, our algorithm provides the user with a window of radius R around S to look at the BN model of a gene network without having to exclude any gene in advance. We prove that our algorithm is correct under the faithfulness assumption. We evaluate our algorithm on simulated and biological data (Rosetta compendium) with satisfactory results.

Lipoprotein secretion and triglyceride stores in the heart.

The genes for apolipoprotein B and microsomal triglyceride transfer protein are expressed in mouse and human heart tissue. Why the heart would express these "lipoprotein assembly" genes has been unclear. Here we demonstrate that the beating mouse heart actually secretes spherical lipoproteins. Moreover, increased cardiac production of lipoproteins (e.g., in mice that express a human apolipoprotein B transgene) was associated with increased triglyceride secretion from the heart and decreased stores of triglycerides within the heart. Increased cardiac production of lipoproteins also reduced the pathological accumulation of triglycerides that occurs in the hearts of mice lacking long-chain acyl coenzyme A dehydrogenase. In contrast, blocking heart lipoprotein secretion (e.g., in heart-specific microsomal triglyceride transfer protein knockout mice) increased cardiac triglyceride stores. Thus, heart lipoprotein secretion helps regulate cardiac triglyceride stores and may protect the heart from the detrimental effects of surplus lipids.

A deficiency of microsomal triglyceride transfer protein reduces apolipoprotein B secretion.

Microsomal triglyceride transfer protein (MTP) transfers lipids to apolipoprotein B (apoB) within the endoplasmic reticulum, a process that involves direct interactions between apoB and the large subunit of MTP. Recent studies with heterozygous MTP knockout mice have suggested that half-normal levels of MTP in the liver reduce apoB secretion. We hypothesized that reduced apoB secretion in the setting of half-normal MTP levels might be caused by a reduced MTP:apoB ratio in the endoplasmic reticulum, which would reduce the number of apoB-MTP interactions. If this hypothesis were true, half-normal levels of MTP might have little impact on lipoprotein secretion in the setting of half-normal levels of apoB synthesis (since the ratio of MTP to apoB would not be abnormally low) and might cause an exaggerated reduction in lipoprotein secretion in the setting of apoB overexpression (since the MTP:apoB ratio would be even lower). To test this hypothesis, we examined the effects of heterozygous MTP deficiency on apoB metabolism in the setting of normal levels of apoB synthesis, half-normal levels of apoB synthesis (heterozygous Apob deficiency), and increased levels of apoB synthesis (transgenic overexpression of human apoB). Contrary to our expectations, half-normal levels of MTP reduced the plasma apoB100 levels to the same extent ( approximately 25-35%) at each level of apoB synthesis. In addition, apoB secretion from primary hepatocytes was reduced to a comparable extent at each level of apoB synthesis. Thus, these results indicate that the concentration of MTP within the endoplasmic reticulum rather than the MTP:apoB ratio is the critical determinant of lipoprotein secretion. Finally, we found that heterozygosity for an apoB knockout mutation lowered plasma apoB100 levels more than heterozygosity for an MTP knockout allele. Consistent with that result, hepatic triglyceride accumulation was greater in heterozygous apoB knockout mice than in heterozygous MTP knockout mice.

Accumulation of apolipoprotein C-I-rich and cholesterol-rich VLDL remnants during exaggerated postprandial triglyceridemia in normolipidemic patients with coronary artery disease.

BACKGROUND: Exaggerated postprandial triglyceridemia is common in normolipidemic patients with coronary artery disease (CAD). Alterations in the composition of triglyceride-rich lipoproteins (TRLs) are likely to underlie this metabolic disturbance. However, the composition of very-low-density lipoproteins (VLDLs), which are the most abundant postprandial TRLs, has never been defined in CAD patients. METHODS AND RESULTS: We examined postprandial changes in the number and composition of VLDLs in middle-aged, normolipidemic CAD patients and control subjects. TRLs from 14 patients and 14 control subjects aged 45 to 55 years were subfractionated by density gradient ultracentrifugation into Svedberg flotation rate (Sf) fractions >400, 60 to 400, and 20 to 60. The VLDLs were separated from chylomicron remnants by immunoaffinity chromatography. In CAD patients, the postprandial concentrations of triglycerides and large (Sf 60 to 400) VLDL particles were elevated. In addition, their postprandial large VLDLs were enriched in apolipoprotein (apo) C-I and their postprandial small (Sf 20 to 60) VLDL remnants were enriched with apo C-I and cholesterol. CONCLUSIONS: Perturbed handling of postprandial triglycerides in normolipidemic CAD patients involves the accumulation of apo C-I-rich large VLDL particles and the generation of small, apo C-I- and cholesterol-rich VLDL remnants.

Alimentary lipemia, postprandial triglyceride-rich lipoproteins, and common carotid intima-media thickness in healthy, middle-aged men.

BACKGROUND: Alimentary lipemia has been associated with coronary heart disease and common carotid artery intima-media thickness (IMT). This study was designed to investigate the relations of subclasses of postprandial triglyceride-rich lipoproteins (TRLs) with IMT. METHODS AND RESULTS: Ninety-six healthy 50-year-old men with an apolipoprotein (apo) E3/E3 genotype underwent an oral fat tolerance test and B-mode carotid ultrasound examination. The apo B-48 and apo B-100 contents of each fraction of TRLs were determined as a measure of chylomicron remnant and VLDL particle concentrations. In the fasting state, LDL cholesterol (P<0.05) and basal proinsulin (P<0. 05) were significantly related to IMT, whereas HDL cholesterol, plasma triglycerides, and insulin were not. In the postprandial state, plasma triglycerides at 1 to 4 hours (P<0.01 at 2 hours), total triglyceride area under the curve (AUC) (P<0.05), incremental triglyceride AUC (P<0.01), and the large VLDL (Sf 60 to 400 apo B-100) concentration at 3 hours (P<0.05) were significantly related to IMT. Multivariate analyses showed that plasma triglycerides at 2 hours, LDL cholesterol, and basal proinsulin were consistently and independently related to IMT when cumulative tobacco consumption, alcohol intake, waist-to-hip circumference ratio, and systolic blood pressure were included as confounders. CONCLUSIONS: These results provide further evidence for postprandial triglyceridemia as an independent risk factor for early atherosclerosis and also suggest that the postprandial triglyceridemia is a better predictor of IMT than particle concentrations of individual TRLs.

Analysis of the role of microsomal triglyceride transfer protein in the liver of tissue-specific knockout mice.

A deficiency in microsomal triglyceride transfer protein (MTP) causes the human lipoprotein deficiency syndrome abetalipoproteinemia. However, the role of MTP in the assembly and secretion of VLDL in the liver is not precisely understood. It is not clear, for instance, whether MTP is required to move the bulk of triglycerides into the lumen of the endoplasmic reticulum (ER) during the assembly of VLDL particles. To define MTP's role in hepatic lipoprotein assembly, we recently knocked out the mouse MTP gene (Mttp). Unfortunately, achieving our objective was thwarted by a lethal embryonic phenotype. In this study, we produced mice harboring a "floxed" Mttp allele and then used Cre-mediated recombination to generate liver-specific Mttp knockout mice. Inactivating the Mttp gene in the liver caused a striking reduction in VLDL triglycerides and large reductions in both VLDL/LDL and HDL cholesterol levels. The Mttp inactivation lowered apo B-100 levels in the plasma by >95% but reduced plasma apo B-48 levels by only approximately 20%. Histologic studies in liver-specific knockout mice revealed moderate hepatic steatosis. Ultrastructural studies of wild-type mouse livers revealed numerous VLDL-sized lipid-staining particles within membrane-bound compartments of the secretory pathway (ER and Golgi apparatus) and few cytosolic lipid droplets. In contrast, VLDL-sized lipid-staining particles were not observed in MTP-deficient hepatocytes, either in the ER or in the Golgi apparatus, and there were numerous cytosolic fat droplets. We conclude that MTP is essential for transferring the bulk of triglycerides into the lumen of the ER for VLDL assembly and is required for the secretion of apo B-100 from the liver.

Differences in apolipoprotein and lipid composition between human chylomicron remnants and very low density lipoproteins isolated from fasting and postprandial plasma.

Triglyceride-rich lipoproteins (TRLs) that are modified during alimentary lipemia and their remnants are indicated to play an important role in the development of atherosclerosis. Although recent studies in transgenic and gene knock-out animal models have shed new light on the function of different apolipoproteins (apos) in the metabolism of TRLs and on their respective role in atherogenesis in these models, little is known about the compositional properties of human chylomicron remnants and very low density lipoprotein (VLDL). To address this issue, apos E, C-I, C-II, and C-III and lipids (triglycerides, phospholipids and cholesterol) were measured in Svedberg flotation rate (Sf) 60-400 and Sf 20-60 subfractions of VLDL and chylomicron remnants isolated from fasting and postprandial plasma samples in ten normotriglyceridemic men. VLDL was separated from chylomicron remnants by immunoaffinity chromatography using monoclonal antibodies (4G3 and 5E11) recognizing apoB-100 but not apoB-48 epitopes. The triglyceride, cholesterol and apoC-II contents of large (Sf 60-400) chylomicron remnants were significantly higher compared with large VLDL particles, while the small (Sf 20-60) chylomicron remnants contained significantly more apoC-II molecules but fewer apoC-I molecules than small VLDL. Whereas the apoC-III contents of large chylomicrons decreased, the apoC-III contents of large VLDL increased postprandially. The cholesterol to triglyceride ratio of large VLDL particles increased transiently by 50% in response to the oral fat load, whereas the cholesterol to triglyceride ratio of large chylomicron remnant particles and small TRL remnants increased 50-100% throughout the entire postprandial period. The specific alterations of the apolipoprotein and lipid composition of chylomicron remnants and VLDL particles observed during alimentary lipemia are likely to target these lipoprotein species differently to metabolic routes and to confer both endogenous and exogenous remnant lipoprotein roles in atherogenesis.

Transient triglyceridemia in healthy normolipidemic men increases cellular processing of large very low density lipoproteins by fibroblasts in vitro.

Exaggerated and prolonged postprandial triglyceridemia is a characteristic of patients with precocious coronary heart disease. Although large very low density lipoprotein (VLDL) particles accumulate during alimentary lipemia, the biological properties of the postprandial VLDL remain unknown. In the present study, an intravenous infusion of a chylomicron-like emulsion was given to healthy normolipidemic men to examine the effects of transient triglyceridemia in vivo on compositional and cell biological characteristics of VLDL. The postinfusion large(Svedberg flotation rate (Sf) (60-400) VLDL was found to have increased capacity to inhibit low density lipoprotein (LDL) binding to the LDL-receptor and a greater ability to suppress the 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase activity of cultured fibroblasts compared to VLDL isolated from fasting plasma. These alterations in cellular interactions were accompanied by increases in the number of apolipoprotein (apo) E, C-I, and C-III molecules per large VLDL particle and loss of apoC-II, compositional changes similar to those observed after an oral fat load. The increase in number of apoE molecules per large VLDL particle correlated positively and significantly with the increase in the capacity of large VLDL to inhibit LDL binding to the LDL receptor (r = 0.76, P = 0.01, n = 10). In contrast, the composition of the small (Sf 20-60) VLDL particles did not change significantly, nor was the LDL receptor-mediated processing of these particles altered consistently. These observations indicate that large VLDL particles that accumulate during alimentary lipemia undergo compositional changes that render them more prone to cellular binding and uptake.

Alterations of VLDL composition during alimentary lipemia.

Apoliprotein (apo) B-100-containing very low density lipoprotein (VLDL) particles secreted from the liver accumulate in plasma during alimentary lipemia. To determine whether changes of VLDL composition occur in the postprandial state that may render these lipoproteins more atherogenic, apoE, C-I, C-II, and C-III, and lipids (triglycerides, phospholipids, and cholesterol) were measured in Svedberg flotation (Sf) 60-400 (large) and Sf 20-60 (small) VLDL before and after an oral fat load. Ten normotriglyceridemic (NTG) and three hypertriglyceridemic (HTG) healthy men were given a fat-rich mixed meal (1,000 kCal with 60.2 E% from fat). Triglyceride-rich lipoproteins were isolated by density gradient ultracentrifugation from plasma samples obtained before (fasting) and at 2-h intervals after the meal. VLDL was then separated from chylomicrons and their remnants by immunoaffinity chromatography using monoclonal antibodies 4G3 and 5E11, recognizing apoB-100, but not apoB-48 epitopes. Large and small VLDL isolated from the NTG group were enriched with apoE and C-I, and cholesterol, but depleted of apoC-II in the postprandial state, whereas the apoC-III, triglyceride, and phospholipid contents were essentially unchanged. The compositional changes of VLDL in HTG subjects were similar but more pronounced compared with NTG subjects. We conclude that postprandial lipemia in healthy men induces transient compositional alterations of VLDL that link these lipoprotein species to the formation of atherosclerosis.

Accumulation of large very low density lipoprotein in plasma during intravenous infusion of a chylomicron-like triglyceride emulsion reflects competition for a common lipolytic pathway.

Very low density lipoproteins (VLDL) are produced in the liver and contain apolipoprotein (apo) B-100 and endogenous lipids. By contrast, ingestion of fat leads to formation of chylomicrons containing apoB-48 secreted from the intestine. In this study, a 60-min intravenous infusion of a chylomicron-like triglyceride emulsion was given to healthy young men to examine whether competition between chylomicrons and VLDL for the same lipolytic pathway explains the increase in VLDL seen after meals. The responses of two major VLDL subfractions were determined by measuring the concentrations of apoB-100 in fractions of triglyceride-rich lipoproteins with Svedberg flotation rates of 60-400 (large VLDL) and 20-60 (small VLDL) that were separated from plasma by density gradient ultracentrifugation. A threefold elevation in plasma triglycerides was observed during the infusion together with a consistent linear increase of large VLDL. The rate at which large VLDL accumulated in plasma differed markedly among individuals and was not enhanced by doubling of the infusion rate. The response of small VLDL was more heterogeneous; however, a decrease was seen in most subjects. The combined pattern for the two VLDL species is what would be expected if large VLDL particles are the precursors of smaller VLDL species and if lipolysis of large VLDL is inhibited through competition from the triglyceride emulsion. The extent to which the triglyceride emulsion inhibited the lipolysis of VLDL and/or influenced the synthesis rate of large VLDL was estimated from simultaneous stable isotope studies. The emulsion caused a 75-90% block of the conversion of large VLDL apoB to small VLDL apoB and there was no sign of enhanced synthesis of large VLDL after infusion of the triglyceride emulsion. The corollary of these findings is that chylomicrons and their remnants impede the normal lipolytic degradation of VLDL and could thereby be indirectly implicated in the generation of atherogenic remnant lipoproteins.

Quantification of postprandial triglyceride-rich lipoproteins in healthy men by retinyl ester labeling and simultaneous measurement of apolipoproteins B-48 and B-100.

The metabolism of chylomicrons, very-low-density lipoprotein (VLDL), and their remnants in the postprandial state was studied in normolipidemic healthy men by measuring apoB-48 and apoB-100 and retinyl palmitate (RP) in fractions of triglyceride-rich lipoproteins after a mixed meal type of oral fat load supplemented with vitamin A. ApoB-48 was present at low concentrations in the fasting plasma samples in most subjects and increased in response to the test meal in Svedberg's flotation rate (Sf) > 20 lipoprotein fractions. Concomitantly, the level of Sf 60 to 400 apoB-100 (large VLDL) had doubled at 3 hours and returned to baseline at 9 hours. The number of apoB-48-containing lipoprotein particles did not exceed 20% of the total number of apoB-containing lipoproteins contained in Sf 12 to 400 fractions at any time point after fat intake. The peak plasma level of RP was delayed compared with the peak plasma concentration of apoB-48, suggesting that retinyl ester labeling of chylomicrons is questionable as a means of quantifying postprandial triglyceride-rich lipoproteins of intestinal origin. Approximately 2000 and 4000 RP molecules were carried in each chylomicron particle in the 3- and 6-hour samples, respectively, in contrast to the remnant fractions in which 100 to 600 RP molecules were found for each lipoprotein particle. The limited RP exchange between lipoprotein particles indicates that the smaller intestinal lipoproteins do not originate primarily from larger Sf > 400 chylomicron particles but instead are secreted directly into the Sf 20 to 400 fraction and subsequently converted to smaller chylomicron remnants.(ABSTRACT TRUNCATED AT 250 WORDS)