Utilizing machine learning to improve clinical trial design for acute respiratory distress syndrome.

Heterogeneous patient populations, complex pharmacology and low recruitment rates in the Intensive Care Unit (ICU) have led to the failure of many clinical trials. Recently, machine learning (ML) emerged as a new technology to process and identify big data relationships, enabling a new era in clinical trial design. In this study, we designed a ML model for predictively stratifying acute respiratory distress syndrome (ARDS) patients, ultimately reducing the required number of patients by increasing statistical power through cohort homogeneity. From the Philips eICU Research Institute (eRI) database, no less than 51,555 ARDS patients were extracted. We defined three subpopulations by outcome: (1) rapid death, (2) spontaneous recovery, and (3) long-stay patients. A retrospective univariate analysis identified highly predictive variables for each outcome. All 220 variables were used to determine the most accurate and generalizable model to predict long-stay patients. Multiclass gradient boosting was identified as the best-performing ML model. Whereas alterations in pH, bicarbonate or lactate proved to be strong predictors for rapid death in the univariate analysis, only the multivariate ML model was able to reliably differentiate the disease course of the long-stay outcome population (AUC of 0.77). We demonstrate the feasibility of prospective patient stratification using ML algorithms in the by far largest ARDS cohort reported to date. Our algorithm can identify patients with sufficiently long ARDS episodes to allow time for patients to respond to therapy, increasing statistical power. Further, early enrollment alerts may increase recruitment rate.

Virtual reality in preoperative imaging in maxillofacial surgery: implementation of "the next level"?

Not only are current imaging techniques - cone-beam computed tomography (CT), CT, and magnetic resonance imaging (MRI) - becoming more precise in capturing data, but the illustration and interpretation of the acquired images is no longer limited to conventional display screens or projectors. The so-called "virtual reality" (VR) glasses have the potential to engage the viewer in a 3-dimensional space, and ultimately to enable evaluation of the reconstructed anatomical structures from a new perspective. For the first time in the field of oral and maxillofacial surgery (OMFS), a 3-dimensional imaging dataset (cone-beam CT, CT, and MRI) can be evaluated by using VR glasses. A medical student, an OMFS resident, and an OMFS consultant rated the preoperative usability of VR glasses to improve the operative understanding of three cases: a deeply impacted wisdom tooth, a fracture of the lower jaw, and an oncological resection. VR glasses seem to help to simplify operations and give the surgeon a good preoperative overview of the intraoperative findings, particularly in the evaluation of impacted teeth and hard tissue structures. In addition, VR glasses seem to be a promising innovation to help in the training of surgical residents and to teach students. However, the more experienced the surgeon, the smaller is the additional value of VR glasses. Preoperative examination using VR glasses can aid better understanding and planning of the surgical site in the future, and is an innovative piece of advanced technology for displaying CT, cone-beam CT, and MRI anatomical data.

Efficacy of an adapted granzyme B-based anti-CD30 cytolytic fusion protein against PI-9-positive classical Hodgkin lymphoma cells in a murine model.

Tumors develop when infiltrating immune cells contribute growth stimuli, and cancer cells are selected to survive within such a cytotoxic microenvironment. One possible immune-escape mechanism is the upregulation of PI-9 (Serpin B9) within cancer cells. This serine proteinase inhibitor selectively inactivates apoptosis-inducing granzyme B (GrB) from cytotoxic granules of innate immune cells. We demonstrate that most classical Hodgkin lymphoma (cHL)-derived cell lines express PI-9, which protects them against the GrB attack and thereby renders them resistant against GrB-based immunotherapeutics. To circumvent this disadvantage, we developed PI-9-insensitive human GrB mutants as fusion proteins to target the Hodgkin-selective receptor CD30. In contrast to the wild-type GrB, a R201K point-mutated GrB construct most efficiently killed PI-9-positive and -negative cHL cells. This was tested in vitro and also in vivo whereby a novel optical imaging-based tumor model with HL cell line L428 was applied. Therefore, this variant, as part of the next generation immunotherapeutics, also named cytolytic fusion proteins showing reduced immunogenicity, is a promising molecule for (targeted) therapy of patients with relapsing malignancies, such as cHL, and possibly other PI-9-positive malignancies, such as breast or lung carcinoma.

Inverse blood pressure rhythm of transgenic hypertensive TGR(mREN2)27 rats: role of norepinephrine and expression of tyrosine-hydroxylase and reuptake1-transporter.

Transgenic hypertensive TGR(mREN2)27 rats (TGR) exhibit an inverse circadian blood pressure profile from the age of 8 to 9 wk. To investigate the role of the sympathetic nervous system in this pathological blood pressure rhythm, we examined postnatal changes in catecholamine concentration, expression of tyrosine-hydroxylase (TH), and norepinephrine (NE) reuptake(1)-transporter (NET) in the heart, adrenal glands, and hypothalamus of non-hypertensive TGR at an age of 4 wk and of hypertensive TGR at an age of 10 wk and compared these to normotensive, age-matched Sprague-Dawley rats. Rats were kept under synchronized light:dark (LD) conditions of 12:12 h. Blood pressure and heart rate were monitored by radiotelemetry, catecholamines by high performance liquid chromatography, expression of TH and NET (mRNA) by RT-PCR, and TH protein by Western blots. In normotensive 4 wk-old Sprague-Dawley rats, cardiac NE concentrations were circadian phase-dependent with lower values at ZT12.5, with no differences observed, in 10-wk-old animals. At both ages however, sympathetic tone was higher during the dark phase, as shown by a higher turnover of NE. This observation confirms earlier data, which indicate that the endogenous amine concentration may not mirror its turnover rate. TGR at either age had lower cardiac NE as well as lower TH expression and did not display a circadian phase-dependency. The increased cardiac NE turnover rate in the dark phase in non-hypertensive TGR was lost in hypertensive rats. Both cardiac NE concentrations and TH expression decreased with age in both strains. In adrenal glands, NE and epinephrine (E) were not circadian phase-dependent in both strains but increased with age. NE concentrations in the hypothalamus were neither circadian phase-dependent nor different in both strains and at both ages. However, sympathetic tone of NE in the hypothalamus, as indicated by the turnover rate, was greater during the dark phase in both strains at an age of 10 wk. Expression of TH and NET were greatly reduced in adrenal glands when compared to Sprague-Dawley rats; whereas, expression of TH in the hypothalamus was significantly increased in hypertensive TGR. These data indicate that the transgene in TGR leads to an increased central stimulation of the sympathetic nervous system and to a consecutive down-regulation in the peripheral organs. It is of interest that rhythmicity in the studied parameters was lost in hypertensive TGR, except in the turnover of NE in the hypothalamus. We concluded that the data on key mechanisms of regulation of the sympathetic system in TGR cannot explain the inverse blood pressure rhythm observed in this transgenic rat strain.

Hydride transfer in liver alcohol dehydrogenase: quantum dynamics, kinetic isotope effects, and role of enzyme motion.

The quantum dynamics of the hydride transfer reaction catalyzed by liver alcohol dehydrogenase (LADH) are studied with real-time dynamical simulations including the motion of the entire solvated enzyme. The electronic quantum effects are incorporated with an empirical valence bond potential, and the nuclear quantum effects of the transferring hydrogen are incorporated with a mixed quantum/classical molecular dynamics method in which the transferring hydrogen nucleus is represented by a three-dimensional vibrational wave function. The equilibrium transition state theory rate constants are determined from the adiabatic quantum free energy profiles, which include the free energy of the zero point motion for the transferring nucleus. The nonequilibrium dynamical effects are determined by calculating the transmission coefficients with a reactive flux scheme based on real-time molecular dynamics with quantum transitions (MDQT) surface hopping trajectories. The values of nearly unity for these transmission coefficients imply that nonequilibrium dynamical effects such as barrier recrossings are not dominant for this reaction. The calculated deuterium and tritium kinetic isotope effects for the overall rate agree with experimental results. These simulations elucidate the fundamental nature of the nuclear quantum effects and provide evidence of hydrogen tunneling in the direction along the donor-acceptor axis. An analysis of the geometrical parameters during the equilibrium and nonequilibrium simulations provides insight into the relation between specific enzyme motions and enzyme activity. The donor-acceptor distance, the catalytic zinc-substrate oxygen distance, and the coenzyme (NAD(+)/NADH) ring angles are found to strongly impact the activation free energy barrier, while the donor-acceptor distance and one of the coenzyme ring angles are found to be correlated to the degree of barrier recrossing. The distance between VAL-203 and the reactive center is found to significantly impact the activation free energy but not the degree of barrier recrossing. This result indicates that the experimentally observed effect of mutating VAL-203 on the enzyme activity is due to the alteration of the equilibrium free energy difference between the transition state and the reactant rather than nonequilibrium dynamical factors. The promoting motion of VAL-203 is characterized in terms of steric interactions involving THR-178 and the coenzyme.

This is the end: processing, editing and repair at the tRNA 3'-terminus.

The generation of a mature tRNA 3'-end is an important step in the processing pathways leading to functional tRNA molecules. While 5'-end processing by RNase P is similar in all organisms, generation of the mature 3'-terminus seems to be more variable and complex. The first step in this reaction is the removal of 3'-trailer sequences. In bacteria, this is a multistep process performed by endo- and exonucleases. In contrast, the majority of eukaryotes generate the mature tRNA 3'-end in a single step reaction, which consists of an endonucleolytic cut at the tRNA terminus. After removal of the 3'-trailer, a terminal CCA triplet has to be added to allow charging of the tRNA with its cognate amino acid. The enzyme catalyzing this reaction is tRNA nucleotidyltransferase, homologs of which have been found in representatives of all three kingdoms. Furthermore, in metazoan mitochondria, some genes encode 3'-terminally truncated tRNAs, which are restored in an editing reaction in order to yield functional tRNAs. Interestingly, this reaction is not restricted to distinct tRNAs, but seems to act on a variety of tRNA molecules and represents therefore a more general tRNA repair mechanism than a specialized editing reaction. In this review, the current knowledge about these crucial reactions is summarized.

Fucosylation of Cripto is required for its ability to facilitate nodal signaling.

O-linked fucose modification is rare and has been shown to occur almost exclusively within epidermal growth factor (EGF)-like modules. We have found that the EGF-CFC family member human Cripto-1 (CR) is modified with fucose and through a combination of peptide mapping, mass spectrometry, and sequence analysis localized the site of attachment to Thr-88. The identification of a fucose modification on human CR within its EGF-like domain and the presence of a consensus fucosylation site within all EGF-CFC family members suggest that this is a biologically important modification in CR, which functionally distinguishes it from the EGF ligands that bind the type 1 erbB growth factor receptors. A single CR point mutation, Thr-88 --> Ala, results in a form of the protein that is not fucosylated and has substantially weaker activity in cell-based CR/Nodal signaling assays, indicating that fucosylation is functionally important for CR to facilitate Nodal signaling.

Cardiovascular regulation in TGR(mREN2)27 rats: 24h variation in plasma catecholamines, angiotensin peptides, and telemetric heart rate variability.

Dysfunction of the sympathetic nervous system might play an important role in disturbed 24h blood pressure regulation in transgenic hypertensive TGR (mREN2)27 (TGR) rats. Our study was performed to determine possible differences in activity of the sympathetic nervous system in TGR rats in comparison to their normotensive Sprague-Dawley (SPRD) controls; we measured plasma catecholamine and angiotensin concentrations throughout 24h under synchronized light-dark 12h:12H (LD 12:12) conditions. In the TGR rat strain, rhythms of plasma catecholamines were blunted, and the concentrations were significantly decreased. In addition, TGR rats showed increased plasma angiotensin I and II concentrations without any significant rhythm. An impaired autonomic regulation was confirmed by monitoring heart rate variability in TGR rats. Data showed that the TGR rat strain is characterized by a reduction in plasma catecholamines and an increase in angiotensin peptides. At present, it is not clear whether the reduction in catecholamines represents a decrease in sympathetic tone mediated by baroreflex activation or an increased catecholamine turnover induced by elevated angiotensin II. However, the blunted, but normally phased, rhythms in plasma catecholamines in TGR rats make it unlikely that the sympathetic nervous system is mainly responsible for the inverse circadian blood pressure rhythm in the transgenic strain.

The plant tRNA 3' processing enzyme has a broad substrate spectrum.

To elucidate the minimal substrate for the plant nuclear tRNA 3' processing enzyme, we synthesized a set of tRNA variants, which were subsequently incubated with the nuclear tRNA 3' processing enzyme. Our experiments show that the minimal substrate for the nuclear RNase Z consists of the acceptor stem and T arm. The broad substrate spectrum of the nuclear RNase Z raises the possibility that this enzyme might have additional functions in the nucleus besides tRNA 3' processing. Incubation of tRNA variants with the plant mitochondrial enzyme revealed that the organellar counterpart of the nuclear enzyme has a much narrower substrate spectrum. The mitochondrial RNase Z only tolerates deletion of anticodon and variable arms and only with a drastic reduction in cleavage efficiency, indicating that the mitochondrial activity can only cleave bona fide tRNA substrates efficiently. Both enzymes prefer precursors containing short 3' trailers over extended 3' additional sequences. Determination of cleavage sites showed that the cleavage site is not shifted in any of the tRNA variant precursors.

Theoretical study of electron, proton, and proton-coupled electron transfer in iron bi-imidazoline complexes.

A comparative theoretical investigation of single electron transfer (ET), single proton transfer (PT), and proton-coupled electron transfer (PCET) reactions in iron bi-imidazoline complexes is presented. These calculations are motivated by experimental studies showing that the rates of ET and PCET are similar and are both slower than the rate of PT for these systems (Roth, J. P.; Lovel, S.; Mayer, J. M. J. Am. Chem. Soc. 2000, 122, 5486). The theoretical calculations are based on a multistate continuum theory, in which the solute is described by a multistate valence bond model, the transferring hydrogen nucleus is treated quantum mechanically, and the solvent is represented as a dielectric continuum. For electronically nonadiabatic electron transfer, the rate expressions for ET and PCET depend on the inner-sphere (solute) and outer-sphere (solvent) reorganization energies and on the electronic coupling, which is averaged over the reactant and product proton vibrational wave functions for PCET. The small overlap of the proton vibrational wave functions localized on opposite sides of the proton transfer interface decreases the coupling for PCET relative to ET. The theory accurately reproduces the experimentally measured rates and deuterium kinetic isotope effects for ET and PCET. The calculations indicate that the similarity of the rates for ET and PCET is due mainly to the compensation of the smaller outer-sphere solvent reorganization energy for PCET by the larger coupling for ET. The moderate kinetic isotope effect for PCET arises from the relatively short proton transfer distance. The PT reaction is found to be dominated by solute reorganization (with very small solvent reorganization energy) and to be electronically adiabatic, leading to a fundamentally different mechanism that accounts for the faster rate.

Theoretical perspectives on proton-coupled electron transfer reactions.

This Account presents a theoretical formulation for proton-coupled electron transfer reactions. The active electrons and transferring protons are treated quantum mechanically, and the free energy surfaces are obtained as functions of collective solvent coordinates corresponding to the proton and electron transfer reactions. Rate expressions have been derived in the relevant limits, and methodology for including the dynamical effects of the solvent and protein has been developed. This theoretical framework allows predictions of rates, mechanisms, and kinetic isotope effects for proton-coupled electron transfer reactions.

Normalisation of blood pressure in hypertensive TGR(mREN2)27 rats by amlodipine vs. enalapril: effects on cardiac hypertrophy and signal transduction pathways.

It is still a controversial issue whether different classes of antihypertensive drugs are equally effective in the regression of cardiac hypertrophy and associated complications. The present study compared the effects of prolonged treatment with the Ca2+-channel blocker amlodipine and the ACE inhibitor enalapril, respectively, in TGR(mREN2)27 rats (TGR), an animal model of renin-dependent hypertension. TGR were divided into three groups and received either amlodipine, enalapril or drinking water without addition, Sprague-Dawley rats (SPRD) served as normotensive control group. Cardiovascular parameters were monitored by radiotelemetry, and drug doses were titrated until 24-h blood pressure was reduced to approximately 140/90 mmHg in both active treatment groups. After 8 weeks of treatment left ventricular (LV) hypertrophy was completely reversed in both treatment groups despite a tenfold increase in plasma angiotensin II in amlodipine-treated TGR. In untreated TGR LV catecholamines were depleted, and beta1-adrenergic stimulation of adenylyl cyclase was blunted. Treatment of TGR with enalapril prevented both the depletion of tissue catecholamines and the desensitisation of LV beta1-adrenoceptors. Amlodipine had no effect on cardiac adrenergic signal transduction. Basal activity of LV soluble guanylyl cyclase was not different between TGR and SPRD, but its sensitivity to stimulation by nitric oxide was slightly reduced in TGR. Treatment had no effect on basal and stimulated guanylyl cyclase activity. The present study in an animal model of renin-dependent hypertension suggests that blood pressure reduction per se is sufficient for a regression of cardiac hypertrophy. However, beta-adrenergic desensitisation was prevented only in the enalapril-treated group, supporting a blood pressure-independent contribution of the renin-angiotensin system to the regulation of beta-adrenergic signal transduction.

tRNA 3' processing in plants: nuclear and mitochondrial activities differ.

The nuclear tRNA 3' processing activity from wheat has been characterized and partially purified. Several characteristics of the wheat nuclear 3' processing enzyme now allow this activity to be distinguished from its mitochondrial counterpart. The nuclear enzyme is an endonuclease, which we termed nuclear RNase Z. The enzyme cleaves at the discriminator base and seems to consist only of protein subunits, since essential RNA subunits could not be detected. RNase Z leaves 5' terminal phosphoryl and 3' terminal hydroxyl groups at the processing products. It is a stable enzyme being active over broad temperature and pH ranges, with the highest activity at 35 degrees C and pH 8.4. The apparent molecular mass according to gel filtration chromatography is 122 kDa. The nuclear RNase Z does process 5' extended pretRNAs but with a much lower efficiency than 5' matured pretRNAs. Nuclear intron-containing precursor tRNAs as well as mitochondrial precursor tRNAs are efficiently cleaved by the nuclear RNase Z. Mitochondrial pretRNA(His) is processed by the nuclear RNase Z, generating a mature tRNA(His) containing an 8 base pair acceptor stem. The edited mitochondrial pretRNA(Phe) is cleaved easily, while the unedited version having a mismatch in the acceptor stem is not cleaved. Thus, an intact acceptor stem seems to be required for processing. Experiments with precursors containing mutated tRNAs showed that a completely intact anticodon arm is not necessary for processing by RNase Z. Comparison of the plant nuclear tRNA 3' processing enzyme with the plant mitochondrial one suggests that both activities are different enzymes.

Animal welfare and colony management in cancer research.

The use of animals in cancer research continues to be important for the study of tumor biology, the development and testing of new therapies, and risk assessment. The new knowledge generated from this research contributes to the health and welfare not only of human beings, but also of animals which develop cancer. However, the use of animals for cancer research is a privilege which carries with it scientific, professional, and moral obligations. The three tenets of a sound animal research program include good science, humane care, and regulatory compliance. Recognizing the complex interactions in the tumor-animal model, the investigator needs to address a wide range of issues during experimental design and implementation including animal welfare concerns, complicating factors in colony management, and compliance with laws, regulations, and policies. Therefore, both practical and philosophical considerations enjoin researchers to maintain the highest standards of animal care.

An alternately spliced mRNA encoding functional domains of murine MAdCAM-1.

cDNA clones representing a small (0.8 kb) form of murine mucosal addressin cell adhesion molecule-1 (MAdCAM-1) mRNA were obtained and sequenced. The sequence was identical to the published 1.6 kb murine MAdCAM-1 cDNA sequence, except that 432 nucleotides encoding the mucin-like and IgA-homologous portions were deleted. This cDNA most likely represents an alternately spliced mRNA. Substantial amounts of both the short and long MAdCAM-1 mRNAs are present in murine mesenteric lymph node. Ig fusion proteins displaying either the short or long forms of MAdCAM-1 can bind Mn(++)-activated JY cells bearing human alpha 4 beta 7 integrin, indicating that the two N-terminal Ig-like domains of MAdCAM-1 are sufficient to bind its integrin counter-receptor.

Expression of vascular cell adhesion molecule-1 by follicular dendritic cells.

Follicular dendritic cells are the major supporting cell of the germinal center microenvironment. The major function of follicular dendritic cells is to present antigen to B cells in secondary lymphoid tissues. Through cell-cell interactions, FDCs are hypothesized to be central to the regulation of normal B cell growth and differentiation. The major receptor-ligand pair which mediates B cell-FDC adhesion is the beta 1 integrin VLA-4, present on B cells and VCAM-1 expressed on FDCs. Follicular non-Hodgkin's lymphomas similarly employ this mechanism to bind to neoplastic germinal centers. The VCAM-1 molecule can exist as a 6 or 7 immunoglobulin domain form. The major form of VCAM-1 on activated endothelium is the 7 domain form. In this report we have determined by polymerase chain reaction of purified FDCs that they express predominantly mRNA for 7 domain VCAM-1. It is likely that the two forms of VCAM-1 are associated with distinct functions, therefore the expression of 7 domain VCAM-1 may be important in normal and neoplastic B cell-FDC interactions.

Molecular mapping of functional antibody binding sites of alpha 4 integrin.

Integrin alpha 4 beta 1 is a leukocyte receptor for fibronectin and vascular cell adhesion molecule-1 (VCAM-1). It is important in inflammatory recruitment of leukocytes, lymphopoiesis, and a number of development events. Here we have mapped a panel of functional monoclonal antibodies (mAbs) recognizing the integrin alpha 4 chain, using murine/human chimeric constructs expressed in COS7 cells. We find that: 1) mAbs that induce homotypic aggregation (epitope A mAbs) map to the most N-terminal 100 amino acids of the human alpha 4 chain; 2) mAbs that block adhesion of alpha 4 beta 1 to VCAM-1 and fibronectin (epitope B mAbs) map to a 52-amino-acid region between residues 152 and 203 of human alpha 4; 3) epitope B mAbs that do or do not induce aggregation (epitope B2 and B1 mAbs, respectively) map to the same regions and are therefore indistinguishable by this analysis; 4) mAbs that neither induce homotypic aggregation nor block adhesion (epitope C mAbs) map to a distinct region of the molecule comprising amino acids 422-606. The N-terminal region of the alpha 4 chain identified by functional A and B epitope mAbs does not correspond to ligand binding sites identified in other alpha subunits, such as cation binding sites or the "I-domain," which alpha 4 lacks, and thus represents a novel site for epitope functionality among the integrins.