Modulation of DNA topoisomerase I activity by p53

The tumor suppressor protein p53 plays a central role in the cellular response to genotoxic lesions and has been shown to be activated by most anticancer agents such as mitomycin C. We here show that mitomycin C treatment of human MCF7 breast adenocarcinoma cells results in increased topoisomerase I activity as measured by relaxation of supercoiled DNA and by phosphorylation of SR protein splicing factor. The increase in catalytic activity occurs in parallel with the nuclear accumulation of p53, resulting in detectable activation of topoisomerase I within less than 1 h of drug treatment. Furthermore, topoisomerase I co-immunoprecipitates with nuclear p53, suggesting that the activation of topoisomerase I may be a result of a direct interaction between the two proteins. In vitro experiments with purified recombinant proteins show that p53 increases the catalytic activities of topoisomerase I as measured by relaxation of supercoiled DNA, stabilization of the covalent topoisomerase I-DNA complex (in the presence of camptothecin), and phosphorylation of SR protein splicing factor ASF/SF2. Furthermore, topoisomerase I sediments at a higher molecular weight in the presence of p53 as revealed by sucrose density gradient analysis in the absence of DNA. Finally, p53 modifies the thermal stability of topoisomerase I, protecting it from heat denaturation. Taken together, our results show that topoisomerase I and p53 form molecular complexes in vitro as in vivo, and we suggest that the p53-mediated response to DNA damage may, at least in part, involve activation of topoisomerase I.

Specific phosphorylation of SR proteins by mammalian DNA topoisomerase I

Several metazoan splicing factors are characterized by ribonucleoprotein (RNP) consensus sequences and arginine-serine repeats (RS domain) which are essential for their function in splicing. These include members of the SR-protein family (SC35, SF2/ASF), the U1 small nuclear (sn) RNP protein (U1-70K) and the U2 snRNP auxiliary factor (U2AF). SR proteins are phosphorylated in vivo and the phosphorylation state of U1-70K's RS domain influences its splicing activity. Here we report the purification of a protein kinase that is specific for SR proteins and show that it is DNA topoisomerase I. This enzyme lacks a canonical ATP-binding motif but binds ATP with a dissociation constant of 50 nM. Camptothecin and derivatives, known to be specific inhibitors of DNA topoisomerase I, strongly inhibit the kinase activity in the presence of DNA and affect the phosphorylation state of SR proteins. Thus, DNA topoisomerase I may well be one of the SR protein kinases operating in vivo.

Topographically organized climbing fibre sprouting in the adult rat cerebellum

Adaptive recovery following brain injury requires the topography of projection maps to be restored. In the adult mammalian brain, the regeneration of severed axons does not normally occur and repair mainly relies on collateral reinnervation from uninjured neurons. Although reinnervation can be target specific at the single cell level, it is not known if the new connections are organized correctly. The normal olivocerebellar projection had precise topography in which subnuclei of the inferior olive terminate as climbing fibres on chemically defined bands of cerebellar Purkinje cells. This precision has been exploited to determine the topography of climbing fibre sprouting following an inferior olive lesion in the adult rat. Collateral reinnervation was found to respect the boundaries between the Purkinje cell compartments. Thus, topographical cues are available in the adult during post-lesion plasticity to guide the restoration of the olivocerebellar projection map.

Involvement of nitric oxide in cardiorespiratory regulation in the nucleus tractus solitarii

The aim of this study was to investigate whether nitric oxide (NO) is involved in cardiorespiratory regulation in the nucleus tractus solitarii (NTS). Unilateral microinjections (50 nl) of the NO-donor, sodium nitroprusside (SNP, 40-100-200 mM), into the NTS of anaesthetized rats elicited dose-dependent apnea (7.3 +/- 2.3 sec; 28.6 +/- 5.7 sec; 35.6 +/- 6.4 sec, respectively; n = 6) and a decrease in arterial blood pressure (8.4 +/- 3.1 mmHg; 18.2 +/- 5.8 mmHg; 25.8 +/- 6.7 mmHg, respectively; n = 6). Similarly, unilateral micro-injections (50 nl) of another NO-donor, 3-morpholinosydnonimine (SIN-1, 20-40-100 mM), also induced apnea (5.1 +/- 2.4 sec; 8.7 +/- 4.3 sec; 26.3 +/- 6.4 sec, respectively; n = 6) and a decrease in arterial blood pressure (6.2 +/- 2.3 mmHg; 11.1 +/- 3.3 mmHg; 18.3 +/- 6.1 mmHg, respectively; n = 6). The SNP- and SIN-1-induced apnea and arterial blood pressure decrease were significantly (p < 0.01) blocked by a 3 min pretreatment with two calcium-channel blockers, diltiazem (0.1 mM) and cobalt (10 mM), while lower doses (diltiazem 0.01 and cobalt 1) were ineffective. Microinjections of diltiazem (0.01 mM) and cobalt (1 mM) alone did not induce any change in basal cardiorespiratory values like diltiazem (0.1 mM) and cobalt (10 mM). These data suggest that NO may be involved in NTS cardiorespiratory regulation via calcium-channel activation.

Endothelin-1 in periaqueductal gray area of mice induces analgesia via glutamatergic receptors

The aim of the study was to examine whether endothelin-1 (ET-1) injected into dorsolateral periaqueductal gray (PAG) area of mice produces antinociception. ET-1, from 1 to 4 pmol/mouse, induced antinociceptive effect in a dose-dependent manner. This antinociceptive effect was prevented by NMDA receptor antagonists (2-APV and MK-801) injected in the same area (2-APV) or by intraperitoneal route (MK-801). CNQX, a non-NMDA receptor antagonist, did not inhibit the ET-1 effects. Prazosin, an alpha 1-adrenergic blocking agent, also prevented the ET-1 antinociceptive effect. We suggest that the activation of NMDA glutamatergic receptors in the PAG area may be a necessary step for ET-1 induced antinociception.

[Heart surgery, cardiopulmonary bypass, and organic inflammatory response. Part II: changes in leukocytes, arachidonic acid derivatives, and hormones]

The use of cardiopulmonary bypass for surgical cardiac procedures is characterized by a whole-body inflammatory reaction due to the contact of blood through nonendothelialized surfaces; this stimulates the organism to recognize the cardiopulmonary bypass system as "nonself" and to activate specific (immune) and nonspecific (inflammatory) responses. These responses are then related with postoperative damage to many body systems of the body, like pulmonary, renal or brain dysfunction, excessive bleeding and postoperative sepsis. In this paper, present knowledge on untoward responses of the patient to cardiopulmonary bypass in cardiac surgery is reviewed and discussed, particularly focusing on the perturbation of the leukocytes, of the hormones and of the products of the arachidonic acid cascade.

Lineage commitment of transformed haematopoietic progenitors is determined by the level of PKC activity

Our previous work showed that haematopoietic precursors transformed by the E26 avian leukemia virus undergo multilineage differentiation in response to the phorbol ester phorbol 12-myristate 13-acetate (PMA). Treatment of the cells with high concentrations of PMA (100 nM) favours myelomonocytic differentiation, while lower concentrations (20 nM) induce predominantly eosinophil differentiation. Here we have investigated the role of protein kinase C (PKC) in this process and found that 100 nM, but not 20 nM, PMA dramatically down-regulates total cellular PKC activity, indicating that high PMA concentrations result in less efficient signalling than lower PMA concentrations. Consistent with these findings is the observation that very low PMA concentrations (1 nM), which presumably only moderately activate PKC, induce myeloid differentiation. This suggests the existence of two PKC thresholds which play a role in lineage commitment. To test the model, alpha- and epsilon-PKC isoforms were expressed in E26-transformed progenitors. These cells exhibited myelomonocytic differentiation even in the absence of PMA, while treatment with concentrations of PMA as high as 100 nM led to the differentiation of predominantly eosinophils and failed to downregulate the exogenous PKC. Our results suggest that different levels of PKC activity result in three different phenotypes: (i) no PKC activity maintains the progenitor phenotype; (ii) low PKC activity favours myelomonocytic differentiation; (iii) high PKC favours eosinophil differentiation.

Tyrosine phosphorylation and subcellular redistribution of p125 ras guanosine triphosphatase-activating protein in human neutrophils stimulated with FMLP

In this paper, we show that the p125 ras guanosine triphosphatase-activating protein (p125 GAP) is present in the cytosol of human neutrophils and is transiently tyrosine phosphorylated and translocated to the membranes upon cell activation with formyl-methionyl-leucyl-phenylalanine (FMLP). When concanavalin A (ConA) or phorbol 12-myristate 13-acetate (PMA), which both induced a long-lasting respiratory burst, were used as stimuli, tyrosine phosphorylation and translocation of p125 GAP did not occur.

Expression of runtB is modulated during chondrocyte differentiation

The runt locus in Drosophila encodes a nuclear protein involved in embryo segmentation, sex determination/X dosage compensation, and neurogenesis. runt homologues have been identified in higher vertebrates. The encoded proteins share a domain of 128 amino acids called the runt domain. It has been reported that this domain mediates DNA binding and heterodimerization. Here, we analyze runtB expression during chondrocyte differentiation "in vitro" and "in vivo." We have first isolated, from a chondrocyte library, a cDNA clone coding for a runtB chicken homologue and containing a complete open reading frame. The predicted protein product is 84% identical to the mouse PEBP2alphaB2 isoform. By RT-PCR analysis we have also cloned the chicken cDNA fragment coding for delta alphaB2, the exon sequence included in the B1 isoform mRNA. On Northern blot analysis of cultured chondrocytes, runtB mRNA levels increase dramatically with the transition from stage 0 (dedifferentiated) to stages I and II (hypertrophic chondrocytes). Moreover, runt polypeptides were demonstrated in chondrocytes both in vivo and in vitro. These results suggest that runt plays a role in chondrogenic differentiation.

Nef-CD4 physical interaction sensed with the yeast two-hybrid system

HIV-1 Nef protein has been known to induce downmodulation of CD4 receptor. In order to test whether the two proteins physically interact, the yeast two-hybrid system was exploited. A Saccharomyces cerevisiae strain carrying a GAL4-responsive lacZ fusion gene was cotransformed with plasmids in which the Nef and the CD4 cytoplasmic domain (CD4cd) coding sequences were fused to either the DNA binding (DB) or the activation (A) moiety of the GAL4 transcriptional activator. Both the DB-Nef + A-CD4cd and the DB-CD4cd + A-Nef combinations activated the reporter gene, weakly but specifically, as inferred by comparison with a number of controls. Reporter activation was similary observed when DB-Nef was cotransfected with the fusion A-CD4cd(aa 1-23). On the contrary, the combination DB-Nef + A-CD4cd(aa 24-40) was inactive. Also, mutating the CD4cd Leu20-Leu21 motif (known to be essential for both physiological and Nef-induced CD4 endocytosis) to Ala20-Ala21 abolished the GAL4 activity of DB-Nef + A-CD4cd. None of six DB-Nef derivatives in which Nef was partially deleted activated specifically the reporter when coexpressed with A-CD4cd. These findings suggest that CD4cd and Nef directly interact and that a largely complete Nef is required for the interaction. CD4cd aa 1-23 are sufficient for binding; in particular, the Leu20-Leu21 motif is essential. One can infer from these data that: (i) Nef-induced CD4 downmodulation involves a direct CD4-Nef contact and (ii) CD4cd Leu20-Leu21 is required in Nef-induced downmodulation, not simply as an endocytosis signal, but also as an essential component of the Nef-binding moiety.

Mechanisms of NADPH oxidase activation: translocation of p40phox, Rac1 and Rac2 from the cytosol to the membranes in human neutrophils lacking p47phox or p67phox

On neutrophil stimulation, the cytosolic components of NADPH oxidase, p67phox, p47phox, p40phox, as well as the Ras-related G-proteins Rac1 and Rac2, are translocated from the cytosol to cell membranes where they associate with a flavocytochrome b, forming a functional complex responsible for the production of oxygen radicals in phagocytes. In this paper we show that (a) in neutrophils from a patient with a form of chronic granulomatous disease (CGD) in which p67phox is absent, p47phox and Rac2, but not p40phox and Rac1 were translocated from the cytosol to the membrane on stimulation with formylmethionyl-leucylphenylalanine (fMLP) or phorbol 12-myristate 13-acetate (PMA); (b) in neutrophils from a patient with a form of CGD in which p47phox is absent, p67phox, p40phox and Rac1 failed to associate with the membrane on stimulation with fMLP or PMA, whereas Rac2 was translocated as in normal neutrophils. We also show that in neutrophils from a patient lacking p67phox, the amount of cytosolic p40phox was decreased by about 40%. These findings indicate that, on neutrophil stimulation, p67phox mediates the translocation of p40phox and Rac1 from the cytosol to cell membranes and that Rac2 associates with the membranes independently of p47phox and p67phox.

Translocation of p190rho guanosine triphosphatase-activating protein from cytosol to the membrane in human neutrophils stimulated with different agonists

In this paper, we investigated the subcellular distribution of p190rho guanosine triphosphatase-activating protein (p190 GAP) in human neutrophils stimulated with different agonists. The results show that in neutrophils treated with formyl-methionyl-leucyl-phenylalanine (FMLP) (1) p190 GAP was translocated from the cytosol to the membranes; (2) the translocation of p190 GAP took place only at doses of FMLP that induced the translocation of rac 1 and rac 2 and the activation of the NADPH oxidase; and (3) the kinetic of translocation of p190 GAP paralleled that of rac 1 and rac 2. However, when the agonist was concanavalin A (ConA) or phorbol 12-myristate 13-acetate (PMA), rac 1 and rac 2, but not the p190 GAP, were translocated.

The eosinophil-specific cell surface antigen, EOS47, is a chicken homologue of the oncofetal antigen melanotransferrin

The EOS47 antigen is a 100-kD cell surface glycoprotein selectively expressed by avian retrovirus-transformed eosinophils and their precursors. We have purified the EOS47 protein to homogeneity and used peptide sequence information to clone EOS47-encoding cDNAs. The open reading frames from these cDNAs predict a 738 amino acid protein with homology to human melanotransferrin, a membrane-found, transferrin-like protein that is expressed at high levels by a subset of melanomas, tumor cell lines, fetal intestine, and liver, but not by most normal adult tissues. The predicted protein sequence of EOS47 displays a 61% sequence identity with melanotransferrin and conservation of all 28 cysteine residues, indicating a similar tertiary structure. The finding that EOS47 lacks several of the iron-coordinating amino acids present in all transferrins suggests that it may be impaired in its ability to bind iron. In nonhematopoietic tissues, EOS47 is expressed at high levels by epithelial brush borders of small intestine and kidney and at lower levels by cells lining the sinusoids of the liver. Within hematopoietic tissues, EOS47 is restricted to a subpopulation of cells (1% to 5%) in bone marrow and early spleen and fluorescence-activated cell sorting of EOS47+ cells leads to a dramatic ( > 30-fold) enrichment of peroxidase+ eosinophils. In contrast, peripheral blood eosinophils are EOS47-, suggesting that the antigen is expressed by newly formed eosinophils and that expression ceases shortly before these cells emigrate from the bone marrow into the peripheral blood. Our results show that melanotransferrin is a stage-specific marker of eosinophils and should be useful for their isolation and further characterization.

Pain pressure threshold in cluster headache patients

Pain perception threshold (PPT) in the head was assessed with a pressure algometer in 58 cluster headache (CH) patients (52M, 6F; 41 episodic and 17 chronic). Fourteen patients in cluster period were retested in remission. Thresholds were assessed at 10 symmetrical points on each side of the head and at the deltoid. Compared with controls (n = 50), CH patients had lower PPT in the head and in the deltoid. PPT was lower on the symptomatic side than on the non-symptomatic side in patients with episodic CH during a cluster period (p < 0.001) and in patients with chronic CH (p < 0.05). This pattern was more evident during a cluster period than during remission (p < 0.05). A reduced PPT did not correlate with illness duration and pain side. The lowest PPT mean values were found at the anterior and intermediate levels of the temporal muscle on the symptomatic side. These results imply a central mechanism underlying the pathogenesis of CH.

Synthesis and pharmacological profile of novel N-substituted N-[5H-[1]benzopyrano[4,3-d]pyrimidin-2-YL]-N-methylglycinamides

The title compounds were synthesized by cyclizing 3-dimethylaminomethylene-4-chromanone with creatine to give the acid 2 that was successively converted into the corresponding amides 3a-f via diphenylphosphorylazide and the relevant amines. The free acid 2 and some of 3a-f showed antiinflammatory, hypotensive and antiarrhythmic properties.

Opposing effects on blood pressure following the activation of metabotropic and ionotropic glutamate receptors in raphe obscurus in the anaesthetized rat

The microinjection of L-glutamate (1-6 nmol/rat) and N-methyl-D-aspartate (NMDA 1-10 nmol/rat), ionotropic glutamate receptor (iGluR) agonists, into the nucleus raphe obscurus caused a concentration -dependent increase of arterial blood pressure. In contrast, (+/-)-1-aminocyclopentane-trans-1,3-dicarboxylic acid (t-ACPD, 14-42 nmol/rat), a metabotropic glutamate receptor (mGluRs) agonist, caused a concentration-dependent decrease in blood pressure. Pretreatment with D,L-2-amino-phosphono valeric acid (2-APV, 5 nmol/rat) a selective NMDA iGluR antagonist, and (+)-5-methyl-10,11-dihydro-5H-dibenzo[a,b] cyclohepten-5,10-imine hydrogen maleate (MK801, 0.9 nmol/rat), a noncompetitive NMDA iGluR antagonist, blocked both the glutamate and NMDA pressor responses, while pretreatment with (+)-alpha-methyl-4-carboxyphenylglycine (MCPG, 0.05 nmol/rat), a mGluR1 antagonist, increased the glutamate-induced pressor effects and blocked the fall in blood pressure induced by t-ACPD. 6-Cyano-7-nitroquinoxaline-2,3-dione (CNQX, 0.4 nmol/rat) a non-NMDA iGluR antagonist, did not affected the glutamate-induced hypertension. These observations indicate opposing roles for ionotropic and metabotropic receptors in the glutamate-induced blood pressure changes elicited from the nucleus raphe obscurus. Moreover, we suggest that the glutamate-induced hypertension may be due to the activation of NMDA ionotropic receptor subtypes and the metabotropic receptors may influence this activation through a reduction of excitability at level of synapses.

Automatic external defibrillators in the hospital as well?

When a cardiac arrest occurs in a non-intensive area of the hospital, the emergency response is not always adequate from the point of view of timeliness and technical quality. The aims of this study were evaluate an experimental programme to improve the CPR skills of staff operating in non-intensive areas of our general hospital and to test the usefulness of placing automatic external defibrillators (AEDs) within these areas. In the experimental phase, two AEDs were placed in 2 non-intensive wards of our hospital for 8 months. The staff of these wards received specific training in CPR and early defibrillation (CPR-D). The devices were used in 19 cases; for defibrillation in four cases of ventricular fibrillation (VF) (three patients were discharged alive from hospital), and for monitoring three supraventricular arrhythmias, one bradyarrhythmia and 11 cardiac rhythms during critical situations. In the implementation phase, four AEDs were indefinitely assigned to as many non-intensive awards. Periodical CPR-D courses and refresher exercises were run; the cardiology staff co-operated in the maintenance of the AEDs and in the registration of technical and clinical data. In the first period of this phase (9 months), AEDs were utilized in 24 cases by the ward-staff: in nine cases for VF (three patients were discharged alive from hospital) and in 15 cases for other rhythm detection in critical conditions. The number and the quality of these uses seem to confirm the favourable impact of the adoption of a more user-friendly defibrillator, such as an AED. The active co-operation between intensive and non-intensive staff was important to facilitate a quick activation of the chain of survival outside the intensive care units. We conclude that AEDs, which were developed for out-of-hospital use by non-physician operators, are suitable for use inside the hospital as well.

Selective and non-selective ET antagonists reveal an ETB receptors mediated ET-1-induced behavioural effect in conscious rats

The injection of endothelin-1 (ET-1) (1 pmol/rat) into the dorsolateral periaqueductal gray (PAG) area of freely moving rats induced rotation along the long axis of the body (barrel-rolling). Preinjection (10 min before) of BQ-788 (an ETB receptor selective antagonist; 5 nmol) or bosentan (an ETA/ETB receptor non-selective antagonist; 10 nmol) to the PAG reduced the behavioral response to ET-1. In contrast, pretreatment with FR139317 (an ETA receptor selective antagonist; 5 nmol) did not affect the ET-1-induced barrel-rolling. These results suggest that barrel-rolling induced by microinjection of ET-1 into the PAG area is predominantly mediated via ETB-like receptors.

Increase of B-50/GAP-43 immunoreactivity in uninjured muscle nerves of MDX mice

Lack of dystrophin in mdx mice leads to muscle fibre degeneration followed by the formation of new myofibres. This degeneration-regeneration event occurs in clusters. It is accompanied by inflammation and remodelling of the intramuscular terminal nerve fibres. Since the growth-associated protein B-50/GAP-43 has been shown to be involved in axonal outgrowth and synaptic remodelling following neuronal injury, we have investigated the presence of B-50 in gastrocnemius and quadriceps muscles of mdx mice. Using immunocytochemistry we demonstrate increased presence of B-50 in terminal nerve branches at motor endplates of mdx mice, particularly in the clusters of de- and regenerating myofibres. In comparison, the control mice displayed no B-50 immunoreactivity in nerve fibres contacting motor endplates. Our findings indicate that during axonal remodelling and collateral sprouting the B-50 level in the terminal axon arbours is increased although there is no direct injury to the motoneurons. We suggest that the degenerating target and/or the inflammatory reaction induces the increased B-50 level in the motoaxons. The increased B-50 may be important for sprouting of the nerve fibres and re-establishment of synaptic contacts, and in addition, for maturation and survival of the newly formed myofibres.

Hypothalamic paraventricular nucleus involvement in the pressor response to N-methyl-d-aspartic acid in the periaqueductal grey matter

The aim of this study was to demonstrate paraventricular hypothalamic nucleus (PVN) involvement in the cardiovascular changes induced by N-methyl-d-aspartic acid (NMDA) microinjections at the level of periaqueductal grey (PAG) matter. The study was carried out in anaesthetized rats and the arterial blood pressure monitored by a polygraph. NMDA injections (0.68-6.8 nmol/rat) into the PAG area induced a significant increase in blood pressure. After pretreatment by injection of the NMDA receptor antagonist dl-2-amino-5-phosphonovaleric acid (2-APV, 0.05-5 nmol/rat) into the PVN, administration of NMDA (0.68 nmol/rat) into the PAG area elicited a decrease, rather than an increase, of blood pressure. We observed a significant reduction of the pressor effect induced by 6.8 nmol/rat NMDA after 2-APV injection into the PVN. 2-APV injection into the dorsomedial hypothalamic nucleus, an area near the PVN, did not modify the increase in blood pressure induced by NMDA in the PAG area. We suggest the existence of a glutamatergic connection between the PAG area and the PVN in the cardiovascular effects of NMDA.

Structural features of U6 snRNA and dynamic interactions with other spliceosomal components leading to pre-mRNA splicing

In the spliceosome, the pre-mRNA, U2 and U6 snRNAs fold into a catalytic structure exhibiting striking similarities with domain V and VI of group II introns. Building of this tripartite structure implies that an evolutionary conserved base pairing between U4 and U6 snRNAs should be disrupted to allow potentially U6 catalytic residue to interact with U2 snRNAs and the pre-mRNA. The steps leading to U4/U6 disruption have been recently discovered and have been shown to involve a modification of the 3' end of U6 snRNA and the hnRNP C protein.

A multinational economic evaluation of rhDNase in the treatment of cystic fibrosis

Economic evaluations of pharmaceuticals are increasingly being conducted in conjunction with randomized phase III clinical trials to meet the demand for pharmacoeconomic data when new products are launched. While the need for such data is often global, the trials in which relevant information may be collected are often conducted in only one or a limited number of countries. A critical issue is how data from pivotal clinical trials in one setting can serve as the basis for pharmacoeconomic evaluations in others. We address this issue and report on four economic evaluations that we undertook in conjunction with a recent U.S. phase III clinical trial of recombinant human deoxyribonuclease (rhDNase), which is used to improve pulmonary function in patients with cystic fibrosis (CF). The objective of these evaluations was to estimate the potential impact of rhDNase therapy in France, Germany, Italy, and the United Kingdom on the direct costs of medical care for the treatment of respiratory tract infections (RTIs) in patients with CF. Analyses of economic impact were undertaken both with and without adjustment for differences in practice patterns between the United States and the countries of interest. Our findings suggest that rhDNase therapy may reduce the cost of RTI-related care by between US$600 and US$1,100 over a 24-week period; the cost of rhDNase is not included in these figures, as a price was unavailable when our analyses were undertaken. Despite methodologic challenges, economic evaluations that meet the information needs of decision makers in diverse countries can nonetheless be undertaken in conjunction with phase III clinical trials.