Assay for high glucose-mediated islet cell sensitization to apoptosis induced by streptozotocin and cytokines

Pancreatic beta-cell apoptosis is known to participate in the beta-cell destruction process that occurs in diabetes. It has been described that high glucose level induces a hyperfunctional status which could provoke apoptosis. This phenomenon is known as glucotoxicity and has been proposed that it can play a role in type 1 diabetes mellitus pathogenesis. In this study we develop an experimental design to sensitize pancreatic islet cells by high glucose to streptozotocin (STZ) and proinflammatory cytokines [interleukin (IL)-1beta, tumor necrosis factor (TNF)-alpha and interferon (IFN)-gamma]-induced apoptosis. This method is appropriate for subsequent quantification of apoptotic islet cells stained with Tdt-mediated dUTP Nick-End Labeling (TUNEL) and protein expression assays by Western Blotting (WB).

Incretin receptor null mice reveal key role of GLP-1 but not GIP in pancreatic beta cell adaptation to pregnancy.

Islet adaptations to pregnancy were explored in C57BL6/J mice lacking functional receptors for glucagon-like peptide 1 (GLP-1) and gastric inhibitory polypeptide (GIP). Pregnant wild type mice and GIPRKO mice exhibited marked increases in islet and beta cell area, numbers of medium/large sized islets, with positive effects on Ki67/Tunel ratio favouring beta cell growth and enhanced pancreatic insulin content. Alpha cell area and glucagon content were unchanged but prohormone convertases PC2 and PC1/3 together with significant amounts of GLP-1 and GIP were detected in alpha cells. Knockout of GLP-1R abolished these islet adaptations and paradoxically decreased pancreatic insulin, GLP-1 and GIP. This was associated with abolition of normal pregnancy-induced increases in plasma GIP, L-cell numbers, and intestinal GIP and GLP-1 stores. These data indicate that GLP-1 but not GIP is a key mediator of beta cell mass expansion and related adaptations in pregnancy, triggered in part by generation of intra-islet GLP-1.

Effects of triamcinolone acetonide on vessels of the posterior segment of the eye.

PURPOSE: This study investigates the effects of triamcinolone acetonide (TA) on retinal endothelial cells in vitro and explores the potential vascular toxic effect of TA injected into the vitreous cavity of rats in vivo. METHODS: Subconfluent endothelial cells were treated with either 0.1 mg/ml or 1 mg/ml TA in 1% ethanol. Control cells were either untreated or exposed to 1% ethanol. Cell viability was evaluated at 24 h, 72 h, and five days using the tetrazolium 3-(4,5-dimethylthiazol-2-yl)-2,5 phenyltetrazolium bromide test (MTT) and lactate dehydrogenase (LDH) assays. Cell proliferation was evaluated by 5-bromo-2-deoxyuridine (BrdU) test. Apoptosis was evaluated by terminal deoxynucleotidyl transferase dUTP nick end labeling assay (TUNEL assay), annexin-binding, and caspase 3 activation. Caspase-independent cell deaths were investigated by immunohistochemistry using antibodies against apoptosis inducing factor (AIF), cytochrome C, microtubule-associated protein (MAP)-light chain 3 (MAP-LC3), and Leukocyte Elastase Inhibitor/Leukocyte Elastase Inhibitor-derived DNase II (LEI/L-DNase II). In vivo, semithin and ultrathin structure analysis and vascular casts were performed to examine TA-induced changes of the choroidal vasculature. In addition, outer segments phagocytosis assay on primary retinal pigment epithelium (RPE) cells was performed to assess cyclooxygenase (COX-2) and vascular endothelial growth factor (VEGF) mRNAs upregulation with or without TA. RESULTS: The inhibitory effect of TA on cell proliferation could not explain the significant reduction in cell viability. Indeed, TA induced a time-dependent reduction of bovine retinal endothelial cells viability. Annexin-binding positive cells were observed. Cytochrome C was not released from mitochondria. L-DNase II was found translocated to the nucleus, meaning that LEI was changed into L-DNase II. AIF was found nuclearized in some cells. LC3 labeling showed the absence of autophagic vesicles. No autophagy or caspase dependent apoptosis was identified. At 1 mg/ml TA induced necrosis while exposure to lower concentrations for 3 to 5 days induced caspase independent apoptosis involving AIF and LEI/L-DNase II. In vivo, semithin and ultrathin structure analysis and vascular casts revealed that TA mostly affected the choroidal vasculature with a reduction of choroidal thickness and increased the avascular areas of the choriocapillaries. Experiments performed on primary RPE cells showed that TA downregulates the basal expression of COX-2 and VEGF and inhibits the outer segments (OS)-dependent COX-2 induction but not the OS-dependent VEGF induction. CONCLUSIONS: This study demonstrates for the first time that glucocorticoids exert direct toxic effect on endothelial cells through caspase-independent cell death mechanisms. The choroidal changes observed after TA intravitreous injection may have important implications regarding the safety profile of TA use in human eyes.

Role of Endogenous GLP-1 and GIP in Beta Cell Compensatory Responses to Insulin Resistance and Cellular Stress.

Role of GLP-1 and GIP in beta cell compensatory responses to beta cell attack and insulin resistance were examined in C57BL/6 mice lacking functional receptors for GLP-1 and GIP. Mice were treated with multiple low dose streptozotocin or hydrocortisone. Islet parameters were assessed by immunohistochemistry and hormone measurements were determined by specific enzyme linked immunoassays. Wild-type streptozotocin controls exhibited severe diabetes, irregularly shaped islets with lymphocytic infiltration, decreased Ki67/TUNEL ratio with decreased beta cell and increased alpha cell areas. GLP-1 and GIP were co-expressed with glucagon and numbers of alpha cells mainly expressing GLP-1 were increased. In contrast, hydrocortisone treatment and induction of insulin resistance increased islet numbers and area, with enhanced beta cell replication, elevated mass of beta and alpha cells, together with co-expression of GLP-1 and GIP with glucagon in islets. The metabolic responses to streptozotocin in GLP-1RKO and GIPRKO mice were broadly similar to C57BL/6 controls, although decreases in islet numbers and size were more severe. In contrast, both groups of mice lacking functional incretin receptors displayed substantially impaired islet adaptations to insulin resistance induced by hydrocortisone, including marked curtailment of expansion of islet area, beta cell mass and islet number. Our observations cannot be explained by simple changes in circulating incretin concentrations, suggesting that intra-islet GLP-1 and GIP make a significant contribution to islet adaptation, particularly expansion of beta cell mass and compensatory islet compensation to hydrocortisone and insulin resistance.

Dying neurons in thalamus of asphyxiated term newborns and rats are autophagic.

OBJECTIVE: Neonatal hypoxic-ischemic encephalopathy (HIE) still carries a high burden by its mortality and long-term neurological morbidity in survivors. Apart from hypothermia, there is no acknowledged therapy for HIE, reflecting the lack of mechanistic understanding of its pathophysiology. (Macro)autophagy, a physiological intracellular process of lysosomal degradation, has been proposed to be excessively activated in excitotoxic conditions such as HIE. The present study examines whether neuronal autophagy in the thalamus of asphyxiated human newborns or P7 rats is enhanced and related to neuronal death processes. METHODS: Neuronal autophagy and cell death were evaluated in the thalamus (frequently injured in severe HIE) of both human newborns who died after severe HIE (n = 5) and P7 hypoxic-ischemic rats (Rice-Vannuci model). Autophagic (LC3, p62), lysosomal (LAMP1, cathepsins), and cell death (TUNEL, caspase-3) markers were studied by immunohistochemistry in human and rat brain sections, and by additional methods in rats (immunoblotting, histochemistry, and electron microscopy). RESULTS: Following severe perinatal asphyxia in both humans and rats, thalamic neurons displayed up to 10-fold (p < 0.001) higher numbers of autophagosomes and lysosomes, implying an enhanced autophagic flux. The highly autophagic neurons presented strong features of apoptosis. These findings were confirmed and elucidated in more detail in rats. INTERPRETATION: These results show for the first time that autophagy is enhanced in severe HIE in dying thalamic neurons of human newborns, as in rats. Experimental neuroprotective strategies targeting autophagy could thus be a promising lead to follow for the development of future therapeutic approaches. Ann Neurol 2014;76:695-711.

Krypton laser photocoagulation induces retinal vascular remodeling rather than choroidal neovascularization.

The purpose of this study is to analyze the retina and choroid response following krypton laser photocoagulation. Ninety-two C57BL6/Sev129 and 32 C57BL/6J, 5-6-week-old mice received one single krypton (630 nm) laser lesion: 50 microm, 0.05 s, 400 mW. On the following day, every day thereafter for 1 week and every 2-3 days for the following 3 weeks, serial sections throughout the lesion were systematically collected and studied. Immunohistology using specific markers or antibodies for glial fibrillary acidic protein (GFAP) (astrocytes, glia and Muller's cells), von Willebrand (vW) (vascular endothelial cells), TUNEL (cells undergoing caspase dependent apoptosis), PCNA (proliferating cell nuclear antigen) p36, CD4 and F4/80 (infiltrating inflammatory and T cells), DAPI (cell nuclei) and routine histology were carried out. Laser confocal microscopy was also performed on flat mounts. Temporal and spatial observations of the created photocoagulation lesions demonstrate that, after a few hours, activated glial cells within the retinal path of the laser beam express GFAP. After 48 h, GFAP-positive staining was also detected within the choroid lesion center. "Movement" of this GFAP-positive expression towards the lasered choroid was preceded by a well-demarcated and localized apoptosis of the retina outer nuclear layer cells within the laser beam path. Later, death of retinal outer nuclear cells and layer thinning at this site was followed by evagination of the inner nuclear retinal layer. Funneling of the entire inner nuclear and the thinned outer nuclear layers into the choroid lesion center was accompanied by "dragging" of the retinal capillaries. Thus, from days 10 to 14 after krypton laser photocoagulation onward, well-formed blood capillaries (of retinal origin) were observed within the lesion. Only a few of the vW-positive capillary endothelial cells stained also for PCNA p36. In the choroid, dilatation of the vascular bed occurred at the vicinity of the photocoagulation site and around it. Confocal microscopy demonstrates that the vessels throughout the path lesion are located within the neuroretina while in the choroid (after separation of the neural retina) only GFAP-positive but no lectin-positive cells can be seen. The involvement of infiltrating inflammatory cells in these remodeling and healing processes remained minimal throughout the study period. During the 4 weeks following krypton laser photocoagulation in the mouse eye, processes of wound healing and remodeling appear to be driven by cells (and vessels) originating from the retina.

Myocardial impairment in chronic hypoxia is abolished by short aeration episodes: involvement of K+ATP channels.

In vivo exposure to chronic hypoxia is considered to be a cause of myocardial dysfunction, thereby representing a deleterious condition, but repeated aeration episodes may exert some cardioprotection. We investigated the possible role of ATP-sensitive potassium channels in these mechanisms. First, rats (n = 8/group) were exposed for 14 days to either chronic hypoxia (CH; 10% O(2)) or chronic hypoxia with one episode/day of 1-hr normoxic aeration (CH+A), with normoxia (N) as the control. Second, isolated hearts were Langendorff perfused under hypoxia (10% O(2), 30 min) and reoxygenated (94% O(2), 30 min) with or without 3 microM glibenclamide (nonselective K(+)(ATP) channel-blocker) or 100 microM diazoxide (selective mitochondrial K(+)(ATP) channel-opener). Blood gasses, hemoglobin concentration, and plasma malondialdehyde were similar in CH and CH+A and in both different from normoxic (P < 0.01), body weight gain and plasma nitrate/nitrite were higher in CH+A than CH (P < 0.01), whereas apoptosis (number of TUNEL-positive nuclei) was less in CH+A than CH (P < 0.05). During in vitro hypoxia, the efficiency (ratio of ATP production/pressure x rate product) was the same in all groups and diazoxide had no measurable effects on myocardial performance, whereas glibenclamide increased end-diastolic pressure more in N and CH than in CH+A hearts (P < 0.05). During reoxgenation, efficiency was markedly less in CH with respect to N and CH+A (P < 0.0001), and ratex pressure product remained lower in CH than N and CH+A hearts (P < 0.001), but glibenclamide or diazoxide abolished this difference. Glibenclamide, but not diazoxide, decreased vascular resistance in N and CH (P < 0.005 and < 0.001) without changes in CH+A. We hypothesize that cardioprotection in chronically hypoxic hearts derive from cell depolarization by sarcolemmal K(+)(ATP) blockade or from preservation of oxidative phosphorylation efficiency (ATP turnover/myocardial performance) by mitochondrial K(+)(ATP) opening. Therefore K(+)(ATP) channels are involved in the deleterious effects of chronic hypoxia and in the cardioprotection elicited when chronic hypoxia is interrupted with short normoxic aeration episodes.

Cornea graft endothelial cells undergo apoptosis by way of an alternate (caspase-independent) pathway.

PURPOSE: To look for apoptosis pathways involved in corneal endothelial cell death during acute graft rejection and to evaluate the potential role of nitric oxide in this process. MATERIALS AND METHODS: Corneal buttons from Brown-Norway rats were transplanted into Lewis rat corneas. At different time intervals after transplantation, apoptosis was assessed by diamino-2-phenylindol staining and annexin-V binding on flat-mount corneas, and by terminal transferase dUTP nick end labeling (TUNEL), caspase-3 dependent and leukocyte elastase inhibitor (LEI)/LDNase II caspase-independent pathways on sections. Inducible nitric oxide synthase (NOS-II) expression and the presence of nitrotyrosine were assayed by immunohistochemistry. RESULTS: Graft endothelial cells demonstrated nuclear fragmentation and LEI nuclear translocation, annexin-V binding, and membranes bleb formation. Apoptosis associated with caspase-3 activity or TUNEL-positive reaction was not observed at any time either in the graft or in the recipient corneal endothelial cells. During 14 days posttransplantation, the recipient corneal endothelial cells remained unaltered and their number unchanged in all studied corneas. NOS-II was expressed in infiltrating cells present within the graft. This expression was closely associated with the presence of nitrotyrosine in endothelial and infiltrating cells. CONCLUSION: During the time course of corneal graft rejection, graft endothelial cells undergo apoptosis. Apoptosis is caspase 3 independent and TUNEL negative and is, probably, carried out by an alternative pathway driven by an LEI/L-Dnase II. Peroxynitrite formation may be an additional mechanism for cell toxicity and programmed cell death of the graft endothelial cells during the rejection process in this model.

Protein kinase C-zeta mediates retinal degeneration in response to TNF.

Tumor necrosis factor-alpha (TNF) has been implicated in retinal ganglion cells (RGC) degeneration in glaucoma. Atypical protein kinase C (PKC) zeta is involved in cell protection against various stresses. The aim of this study was to investigate the potential proapoptotic effects of intravitreal injections of TNF with or without PKCzeta specific inhibitor on the rat retina. TNF was injected in the vitreous of rat eyes alone or in combination with specific PKCzeta inhibitor. PKCzeta and NF-kappaB were studied by immunohistochemistry and western-blotting analysis on retina, and apoptosis quantified by the TUNEL assay. While low basal PKCzeta was observed in the control eyes, TNF induced intense expression of PKCzeta mostly in bipolar cells processes. PKCzeta staining became nuclear when TNF was coinjected with PKCzeta inhibitor. TNF alone did not induce apoptosis in the retina. Coinjection of the PKCzeta-specific inhibitor and TNF, however, induced apoptosis in the inner nuclear and ganglion cell layers. The PKCzeta-specific inhibitor unmasks retinal cells to TNF cytotoxicity showing a link between the proapoptotic effects of TNF and the antiapoptotic PKCzeta signaling pathway.

Bax-induced apoptosis in Leber's congenital amaurosis : a dual role in rod and cone degeneration

Purpose:We previously observed that anti- and pro-apoptotic genes of the Bcl-2 family were differentially expressed during the development of LCA in the Rpe65-/- mouse model (Cottet et al. 2006). Moreover, we reported that activation and translocation of pro-apoptotic Bax to the mitochondria was associated with apoptosis of rod photoreceptors as the disease progressed (Cottet et al. 2008). In this study we challenged whether disruption of the pro-apoptotic pro-apoptotic Bax protein is sufficient to protect photoreceptor cells against apoptosis. Methods:Apoptosis of photoreceptor cells was addressed by TUNEL assay on flatmounted retinas. Counting of the rod nuclei within the ONL was performed following hematoxylin/eosin histological staining of retina sections. Expression level and localization of photoreceptor gene markers were assessed by quantitative PCR and immunohistological analyses. Results:While expression of rod photoreceptor genes was decreased in Rpe65-deficient retina, expression level remained unchanged in Rpe65-/- / Bax-/- mice. Moreover, OS dysorganization and shortening as well as decrease in ONL thickness observed in diseased retina were prevented in mice lacking functional Bax protein. TUNEL assay confirmed that Bax-dependent rod photoreceptor apoptosis was abolished in Rpe65-/- / Bax-/- mice. However, early and fast degeneration of cone cells was not prevented in Rpe65-/- / Bax-/- mice, indicating that Bax-induced apoptotic pathway was not involved in the degenerating process of cones in Rpe65-deficient retina. Conclusions:Altogether, these data show for the first time that a single genetic mutation can trigger two independent apoptotic pathways in rod and cone photoreceptors in LCA disease. While pro-apoptotic Bax is essential to trigger rod photoreceptor apoptosis, early degeneration of cones is not dependent on Bax-mediated apoptotic pathway in Rpe65-deficientmice.

Role of superoxide, nitric oxide, and peroxynitrite in doxorubicin-induced cell death in vivo and in vitro.

Doxorubicin (DOX) is a potent available antitumor agent; however, its clinical use is limited because of its cardiotoxicity. Cell death is a key component in DOX-induced cardiotoxicity, but its mechanisms are elusive. Here, we explore the role of superoxide, nitric oxide (NO), and peroxynitrite in DOX-induced cell death using both in vivo and in vitro models of cardiotoxicity. Western blot analysis, real-time PCR, immunohistochemistry, flow cytometry, fluorescent microscopy, and biochemical assays were used to determine the markers of apoptosis/necrosis and sources of NO and superoxide and their production. Left ventricular function was measured by a pressure-volume system. We demonstrated increases in myocardial apoptosis (caspase-3 cleavage/activity, cytochrome c release, and TUNEL), inducible NO synthase (iNOS) expression, mitochondrial superoxide generation, 3-nitrotyrosine (NT) formation, matrix metalloproteinase (MMP)-2/MMP-9 gene expression, poly(ADP-ribose) polymerase activation [without major changes in NAD(P)H oxidase isoform 1, NAD(P)H oxidase isoform 2, p22(phox), p40(phox), p47(phox), p67(phox), xanthine oxidase, endothelial NOS, and neuronal NOS expression] and decreases in myocardial contractility, catalase, and glutathione peroxidase activities 5 days after DOX treatment to mice. All these effects of DOX were markedly attenuated by peroxynitrite scavengers. Doxorubicin dose dependently increased mitochondrial superoxide and NT generation and apoptosis/necrosis in cardiac-derived H9c2 cells. DOX- or peroxynitrite-induced apoptosis/necrosis positively correlated with intracellular NT formation and could be abolished by peroxynitrite scavengers. DOX-induced cell death and NT formation were also attenuated by selective iNOS inhibitors or in iNOS knockout mice. Various NO donors when coadministered with DOX but not alone dramatically enhanced DOX-induced cell death with concomitant increased NT formation. DOX-induced cell death was also attenuated by cell-permeable SOD but not by cell-permeable catalase, the xanthine oxidase inhibitor allopurinol, or the NADPH oxidase inhibitors apocynine or diphenylene iodonium. Thus, peroxynitrite is a major trigger of DOX-induced cell death both in vivo and in vivo, and the modulation of the pathways leading to its generation or its effective neutralization can be of significant therapeutic benefit.

Sleep deprivation (SD) on focal brain ischemia in the rat : effects of different SD protocols

Sleep-wake disturbances are frequently observed in stroke patients and are associated with poorer functional outcome. Until now the effects of sleep on stroke evolution are unknown. The purpose of the present study was to evaluate the effects of three sleep deprivation (SD) protocols on brain damages after focal cerebral ischemia in a rat model. Permanent occlusion of distal branches of the middle cerebral artery was induced in adult rats. The animals were then subjected to 6h SD, 12h SD or sleep disturbances (SDis) in which 3 x 12h sleep deprivation were performed by gentle handling. Infarct size and brain swelling were assessed by Cresyl violet staining, and the number of damaged cells was measured by terminal deoxynucleotidyl transferase mediated dUTP nick end labeling (TUNEL) staining. Behavioral tests, namely tape removal and cylinder tests, were performed for assessing sensorimotor function. In the 6h SD protocol, no significant difference (P > 0.05) was found either in infarct size (42.5 ± 30.4 mm3 in sleep deprived animals vs. 44.5 ± 20.5 mm3 in controls, mean ± s.d.), in brain swelling (10.2 ± 3.8 % in sleep deprived animals vs. 11.3 ± 2.0 % in controls) or in number of TUNEL-positive cells (21.7 ± 2.0/mm2 in sleep deprived animals vs. 23.0 ± 1.1/mm2 in controls). In contrast, 12h sleep deprivation increased infarct size by 40 % (82.8 ± 10.9 mm3 in SD group vs. 59.2 ± 13.9 mm3 in control group, P = 0.008) and number of TUNEL-positive cells by 137 % (46.8 ± 15/mm in SD group vs. 19.7 ± 7.7/mm2 in control group, P = 0.003). There was no significant difference (P > 0.05) in brain swelling (12.9 ± 6.3 % in sleep deprived animals vs. 11.6 ± 6.0 % in controls). The SDis protocol also increased infarct size by 76 % (3 x 12h SD 58.8 ± 20.4 mm3 vs. no SD 33.8 ± 6.3 mm3, P = 0.017) and number of TUNEL-positive cells by 219 % (32.9 ± 13.2/mm2 vs. 10.3 ± 2.5/mm2, P = 0.008). Brain swelling did not show any difference between the two groups (24.5 ± 8.4 % in SD group vs. 16.7 ± 8.9 % in control group, p > 0.05). Both behavioral tests did not show any concluding results. In summary, we demonstrate that sleep deprivation aggravates brain damages in a rat model of stroke. Further experiments are needed to unveil the mechanisms underlying these effects.

In defense of the "tunneling" wave function of the universe

The tunneling approach to the wave function of the Universe has been recently criticized by Bousso and Hawking who claim that it predicts a catastrophic instability of de Sitter space with respect to pair production of black holes. We show that this claim is unfounded. First, we argue that different horizon size regions in de Sitter space cannot be treated as independently created, as they contend. And second, the WKB tunneling wave function is not simply the inverse of the Hartle-Hawking one, except in very special cases. Applied to the related problem of pair production of massive particles, we argue that the tunneling wave function leads to a small constant production rate, and not to a catastrophe as the argument of Bousso and Hawking would suggest.

Cell death in Leishmania induced by stress and differentiation: programmed cell death or necrosis?

Unicellular organisms, such as the protozoan parasite Leishmania, can be stimulated to show some morphological and biochemical features characteristic of mammalian apoptosis. This study demonstrates that under a variety of stress conditions such as serum deprivation, heat shock and nitric oxide, cell death can be induced leading to genomic DNA fragmentation into oligonucleosomes. DNA fragmentation was observed, without induction, in the infectious stages of the parasite, and correlated with the presence of internucleosomal nuclease activity, visualisation of 45 to 59 kDa nucleases and detection of TUNEL-positive nuclei. DNA fragmentation was not dependent on active effector downstream caspases nor on the lysosomal cathepsin L-like enzymes CPA and CPB. These data are consistent with the presence of a caspase-independent cell death mechanism in Leishmania, induced by stress and differentiation that differs significantly from metazoa.

Induction of apoptosis by mercury compounds depends on maturation and is not associated with microglial activation

The earliest sign of neurotoxicity observed after exposure of three-dimensional brain cell cultures to low concentrations of mercury compounds is a microglial reaction. We hypothesized that an induction of apoptosis by mercury compounds could be an activating signal of the microglial reaction. Aggregating brain cell cultures of fetal rat telencephalon were treated for 10 days with either mercury chloride or monomethylmercury chloride at noncytotoxic concentrations during two developmental periods: from day 5 to 15, corresponding to an immature stage, and from day 25 to 35 corresponding to a mature stage. Apoptosis was evaluated by the TUNEL technique. It was found that both mercury compounds caused a significant increase in the number of apoptotic cells, but exclusively in immature cultures exhibiting also spontaneous apoptosis. Double staining by the TUNEL technique combined with either neuronal or astroglial markers revealed that the proportion of cells undergoing apoptosis was highest for astrocytes. Furthermore neither an association nor a colocalization was found between apoptotic cells and microglial cells. In conclusion, it appears that the induction of apoptosis by mercury compounds in immature cells is only an acceleration of a spontaneously occurring process, and that it is not a directly related to the early microglial reaction.