Using machine learning models to explore the solution space of large nonlinear systems underlying flowsheet simulations with constraints

Flowsheet simulations of chemical processes on an industrial scale require the solution of large systems of nonlinear equations,so that solvability becomes a practical issue.Additional constraints from technical,economic,environmental,and safety considerations may further limit the feasible solution space beyond the convergence requirement.A priori,the design variable domains for which a simulation converges and fulfills the imposed constraints are usually unknown and it can become very time-consuming to distinguish feasible from infeasible design variable choices by simply running the simulation for each choice.To support the exploration of the design variable space for such scenarios,an adaptive sampling technique based on machine learning models has recently been proposed.However,that approach only considers the exploration of the convergent domain and ignores additional constraints.In this paper,we present an improvement which particularly takes the fulfillment of constraints into account.We successfully apply the proposed algorithm to a toy example in up to 20 dimensions and to an industrially relevant flowsheet simulation.     

Gut Microbiota and Short Chain Fatty Acids: Implications in Glucose Homeostasis

Gut microbiota encompasses a wide variety of commensal microorganisms consisting of trillions of bacteria, fungi, and viruses. This microbial population coexists in symbiosis with the host, and related metabolites have profound effects on human health. In this respect, gut microbiota plays a pivotal role in the regulation of metabolic, endocrine, and immune functions. Bacterial metabolites include the short chain fatty acids (SCFAs) acetate (C2), propionate (C3), and butyrate (C4), which are the most abundant SCFAs in the human body and the most abundant anions in the colon. SCFAs are made from fermentation of dietary fiber and resistant starch in the gut. They modulate several metabolic pathways and are involved in obesity, insulin resistance, and type 2 diabetes. Thus, diet might influence gut microbiota composition and activity, SCFAs production, and metabolic effects. In this narrative review, we discuss the relevant research focusing on the relationship between gut microbiota, SCFAs, and glucose metabolism.     

High Iron and Iron Household Protein Contents in Perineuronal Net-Ensheathed Neurons Ensure Energy Metabolism with Safe Iron Handling

A subpopulation of neurons is less vulnerable against iron-induced oxidative stress and neurodegeneration. A key feature of these neurons is a special extracellular matrix composition that forms a perineuronal net (PN). The PN has a high affinity to iron, which suggests an adapted iron sequestration and metabolism of the ensheathed neurons. Highly active, fast-firing neurons—which are often ensheathed by a PN—have a particular high metabolic demand, and therefore may have a higher need in iron. We hypothesize that PN-ensheathed neurons have a higher intracellular iron concentration and increased levels of iron proteins. Thus, analyses of cellular and regional iron and the iron proteins transferrin (Tf), Tf receptor 1 (TfR), ferritin H/L (FtH/FtL), metal transport protein 1 (MTP1 aka ferroportin), and divalent metal transporter 1 (DMT1) were performed on Wistar rats in the parietal cortex (PC), subiculum (SUB), red nucleus (RN), and substantia nigra (SNpr/SNpc). Neurons with a PN (PN⁺) have higher iron concentrations than neurons without a PN: PC 0.69 mM vs. 0.51 mM, SUB 0.84 mM vs. 0.69 mM, SN 0.71 mM vs. 0.63 mM (SNpr)/0.45 mM (SNpc). Intracellular Tf, TfR and MTP1 contents of PN⁺ neurons were consistently increased. The iron concentration of the PN itself is not increased. We also determined the percentage of PN⁺ neurons: PC 4%, SUB 5%, SNpr 45%, RN 86%. We conclude that PN⁺ neurons constitute a subpopulation of resilient pacemaker neurons characterized by a bustling iron metabolism and outstanding iron handling capabilities. These properties could contribute to the low vulnerability of PN⁺ neurons against iron-induced oxidative stress and degeneration.     

Prodynorphin and Proenkephalin in Cerebrospinal Fluid of Sporadic Creutzfeldt–Jakob Disease

Proenkephalin (PENK) and prodynorphin (PDYN) are endogenous opioid peptides mainly produced in the striatum and, to a lesser extent, in the cerebral cortex. Dysregulated metabolism and altered cerebrospinal fluid (CSF) levels of PENK and PDYN have been described in several neurodegenerative diseases. However, no study to date investigated these peptides in the CSF of sporadic Creutzfeldt–Jakob disease (sCJD). Using liquid chromatography-multiple reaction monitoring mass spectrometry, we evaluated the CSF PDYN- and PENK-derived peptide levels in 25 controls and 63 patients with sCJD belonging to the most prevalent molecular subtypes (MM(V)1, VV2 and MV2K). One of the PENK-derived peptides was significantly decreased in each sCJD subtype compared to the controls without a difference among subtypes. Conversely, PDYN-derived peptides were selectively decreased in the CSF of sCJD MV2K, a subtype with a more widespread overall pathology compared to the sCJD MM(V)1 and the VV2 subtypes, which we confirmed by semiquantitative analysis of cortical and striatal neuronal loss and astrocytosis. In sCJD CSF PENK and PDYN were associated with CSF biomarkers of neurodegeneration but not with clinical variables and showed a poor diagnostic performance. CSF PDYN and PENK-derived peptides had no significant diagnostic and prognostic values in sCJD; however, the distinct marker levels between molecular subtypes might help to better understand the basis of phenotypic heterogeneity determined by divergent neuronal targeting.     

An Iterative O‐Methyltransferase Catalyzes 1,11‐Dimethylation of Aspergillus fumigatus Fumaric Acid Amides

S‐adenosyl‐l ‐methionine (SAM)‐dependent methyltransfer is a common biosynthetic strategy to modify natural products. We investigated the previously uncharacterized Aspergillus fumigatus methyltransferase FtpM, which is encoded next to the bimodular fumaric acid amide synthetase FtpA. Structure elucidation of two new A. fumigatus natural products, the 1,11‐dimethyl esters of fumaryl‐l ‐tyrosine and fumaryl‐l ‐phenylalanine, together with ftpM gene disruption suggested that FtpM catalyzes iterative methylation. Final evidence that a single enzyme repeatedly acts on fumaric acid amides came from an in vitro biochemical investigation with recombinantly produced FtpM. Size‐exclusion chromatography indicated that this methyltransferase is active as a dimer. As ftpA and ftpM homologues are found clustered in other fungi, we expect our work will help to identify and annotate natural product biosynthesis genes in various species.     

4‐Biphenylalanine‐ and 3‐Phenyltyrosine‐Derived Hydroxamic Acids as Inhibitors of the JumonjiC‐Domain‐Containing Histone Demethylase KDM4A

Overexpression of the histone lysine demethylase KDM4A, which regulates H3K9 and H3K36 methylation states, has been related to the pathology of several human cancers. We found that a previously reported hydroxamate‐based histone deacetylase (HDAC) inhibitor (SW55) was also able to weakly inhibit this demethylase with an IC₅₀ value of 25.4 μm . Herein we report the synthesis and biochemical evaluations, with two orthogonal in vitro assays, of a series of derivatives of this lead structure. With extensive chemical modifications on the lead structure, also by exploiting the versatility of the radical arylation with aryldiazonium salts, we were able to increase the potency of the derivatives against KDM4A to the low‐micromolar range and, more importantly, to obtain demethylase selectivity with respect to HDACs. Cell‐permeable derivatives clearly showed a demethylase‐inhibition‐dependent antiproliferative effect against HL‐60 human promyelocytic leukemia cells.     

Transcriptome Analyses of Adipose Tissue Samples Identify EGFL6 as a Candidate Gene Involved in Obesity-Related Adipose Tissue Dysfunction in Children

Obesity develops early in childhood and is accompanied by early signs of adipose tissue (AT) dysfunction and metabolic disease in children. In order to analyse the molecular processes during obesity-related AT accumulation in children, we investigated genome-wide expression profiles in AT samples, isolated adipocytes, and stromal vascular fraction (SVF) cells and assessed their relation to obesity as well as biological and functional AT parameters. We detected alterations in gene expression associated with obesity and related parameters, i.e., BMI SDS, adipocyte size, macrophage infiltration, adiponectin, and/or leptin. While differential gene expression in AT and adipocytes shared an enrichment in metabolic pathways and pathways related to extracellular structural organisation, SVF cells showed an overrepresentation in inflammatory pathways. In adipocytes, we found the strongest positive association for epidermal growth factor-like protein 6 (EGFL6) with adipocyte hypertrophy. EGFL6 was also upregulated during in vitro adipocyte differentiation. In children, EGFL6 expression was positively correlated to parameters of AT dysfunction and metabolic disease such as macrophage infiltration into AT, hs-CRP, leptin levels, and HOMA-IR. In conclusion, we provide evidence for early alterations in AT gene expression related to AT dysfunction in children and identified EGFL6 as potentially being involved in processes underlying the pathogenesis of metabolic disease.     

High Versatility of IPP and DMAPP Methyltransferases Enables Synthesis of C6, C7 and C8 Terpenoid Building Blocks

The natural substance class of terpenoids covers an extremely wide range of different structures, although their building block repertoire is limited to the C₅ compounds DMAPP and IPP. This study aims at the characterization of methyltransferases (MTases) that modify these terpene precursors and the demonstration of their suitability for biotechnological purposes. All seven enzymes tested accepted IPP as substrate and altogether five C₆ compounds and six C₇ compounds were formed within the reactions. A high selectivity for the deprotonation site as well as high stereoselectivity could be observed for most of the biocatalysts. Only the enzyme from Micromonospora humi also accepted DMAPP as substrate, converting it into (2R)-2-methyl-IPP in vitro. In vivo studies demonstrated the production of a C₈ compound and a hydride shift step within the MTase-catalyzed reaction. Our study presents IPP/DMAPP MTases with very different catalytic properties, which provide biosynthetic access to many novel terpene-derived structures.     

Computational spectroscopy and reaction dynamics

Physico- and bio-chemical processes on the femto- to picosecond time scale are ideally suited to be investigated with all-atom simulations. They include, amongst others, vibrational relaxation, ligand migration in sterically demanding environments (proteins, ices), or vibrational spectra. By comparing with experimental data, the results can be used to obtain an understanding of the mechanisms underlying the observations. Furthermore, most of these processes are sensitive to the intermolecular interactions. Therefore, detailed refinement of such interaction potentials is possible.     

Spacer effects on in vivo properties of DOTA-conjugated dimeric [Tyr3]octreotate peptides synthesized by a "Cu(I)-click" and "sulfo-click" ligation method

We report on the SSTR2-binding properties of a series of four dimeric [Tyr3]octreotate analogues with different spacer lengths (nine, 19, 41, and 57 atoms) between the peptides. Two analogues (9 and 57 atoms) were selected as precursors for the design, synthesis, and biological evaluation of DOTA-conjugated dimeric [Tyr3]octreotate analogues for tumor targeting. These compounds were synthesized by using a two-stage click ligation procedure: a Cu(I) -catalyzed 1,3-dipolar cycloaddition ("copper-click" reaction) and a thio acid/sulfonyl azide amidation ("sulfo-click" reaction). The IC(50) values of these DOTA-conjugated [Tyr3]octreotate analogues were comparable, and internalization studies showed that the nine-atom (111) In-DOTA-labeled [Tyr3]octreotate dimer had rapid and high receptor binding. Biodistribution studies with BALB/c nude mice bearing subcutaneous AR42J tumors showed that the (111) In-labeled [Tyr3]octreotate dimer (nine atoms) had a high tumor uptake at 1 h p.i. (38.8 ± 8.3 % ID g(-1) ), and excellent tumor retention at 4 h p.i. (40.9 ± 2.5 % ID g(-1) ). However, the introduction of the extended hydrophilic 57 atoms spacer led to rapid clearance from the circulation; this limited tumor accumulation of the radiotracer (21.4 ± 4.9 % ID g(-1) at 1 h p.i.). These findings provide important insight on dimerization and spacer effects on the in vivo properties of DOTA-conjugated [Tyr3]octreotate dimers.