TRH-like immunoreactivity in rat pancreas and eye, bovine and sheep pineals, and human placenta: Non-identity with synthetic Pyroglu-His-Pro-NH2 (TRH)

Extracts from albino rat eyes and pancreas, bovine and sheep pineals and human placenta containing TRH-like immunoreactivity were chromatographed on silica-gel plates. Comparison of elution profiles for TRH-like immunoreactivity with that of TRH revealed the presence of substances other than TRH in these samples. Chromatography of TRH-like immunoreactivity obtained fromRana pipiens skin eluted in two solvent systems produced elution profiles identical with that of synthetic Pyroglu-His-Pro-NH₂ consistent with reports that frog skin contains large quantities of TRH. Implications of these findings are discussed.     

TRH-like immunoreactivity in rat pancreas and eye, bovine and sheep pineals, and human placenta: non-identify with synthetic Pyroglu-His-Pro-NH2 (TRH).

Extracts from albino rat eyes and pancreas, bovine and sheep pineals and human placenta containing TRH-like immunoreactivity were chromatographed on silica-gel plates. Comparison of elution profiles for TRH-like immunoreactivity with that of TRH revealed the presence of substances other than TRH in these samples. Chromatography of TRH-like immunoreactivity obtained from Rana pipiens skin eluted in two solvent systems produced elution profiles identical with that of synthetic Pyroglu-His-Pro-NH2 consistent with reports that frog skin contains large quantities of TRH. Implications of these findings are discussed.     

Serial DST, TRH test, and TRH-like immunoreactivity measurements in major affective disorders

Neuroendocrine functions in depressed patients with major affective disorders were serially investigated by the Dexamethasone Suppression Test (DST), the thyrotropin-releasing hormone (TRH) stimulation test, and the plasma TRH-like immunoreactivity (TRH-LI) measurement. Prior to antidepressant therapy, the sensitivity for nonsuppression to the DST was 36.0%, whereas that for blunted thyroid-stimulating hormone (TSH) response to TRH was 28.0%. Both DST nonsuppression and TSH blunting appeared to be state-related markers for depressed patients. Specifically, a significant increase of maximum TSH response to TRH after 4 weeks of antidepressant therapy was associated with clinical improvement. Plasma TRH-LI in depressed patients was significantly lower than that of healthy controls. It is possible that the lower plasma TRH-LI level is related to the pathophysiology of some depressed patients with major affective disorders.     

Lipopolysaccharide modulation of thyrotropin-releasing hormone (TRH) and TRH-like peptide levels in rat brain and endocrine organs

Lipopolysaccharide (LPS) is a proinflammatory and depressogenic agent whereas thyrotropin-releasing hormone (TRH; pGlu-His-Pro-NH₂) is an endogenous antidepressant and neuroprotective peptide. LPS and TRH also have opposing effects on K⁺ channel conductivity. We hypothesized that LPS can modulate the expression and release of not only TRH but also TRH-like peptides with the general structure pGlu-X-Pro-NH₂, where “X” can be any amino acid residue. The response might be “homeostatic,” that is, LPS might increase TRH and TRH-like peptide release, thereby moderating the cell damaging effects of this bacterial cell wall constituent. On the other hand, LPS might impair the synthesis and release of these neuropeptides, thus facilitating the induction of early response genes, cytokines, and other downstream biochemical changes that contribute to the “sickness syndrome.” Sprague-Dawley rats (300 g) received a single intraperitoneal injection of 100 μg/kg LPS. Animals were then decapitated 0, 2, 4, 8, and 24 h later. Serum cytokines and corticosterone peaked 2 h after intraperitoneal LPS along with a transient decrease in serum T3. TRH and TRH-like peptides were measured by a combination of high-performance liquid chromatography and radioimmunoassay. TRH declined in the nucleus accumbens and amygdala in a manner consistent with LPS-accelerated release and degradation. Various TRH-like peptide levels increased at 2 h in the anterior cingulate, hippocampus, striatum, entorhinal cortex, posterior cingulate, and cerebellum, indicating decreased release and clearance of these peptides. These brain regions are part of aneuroimmunomodulatory system that coordinates the behavioral, endocrine, and immune responses to the stresses of sickness, injury, and danger. A sustained rise in TRH levels in pancreatic β-cells accompanied LPS-impaired insulin secretion. TRH and Leu-TRH in prostate and TRH in epididymis remained elevated 2-24h after intraperitoneal LPS. We conclude that these endogenous neuroprotective and antidep resant-like peptides both mediate and moderate some of the behavioral and toxic effects of LPS.     

Distribution of thyrotropin-releasing hormone (TRH) immunoreactivity in the brain of urodele amphibians

To improve knowledge of the peptidergic systems in the brain of amphibians we have conducted a comparative analysis of the distribution of TRH immunoreactive cell bodies and fibers in three species of urodeles. Fiber labeling was observed in all main brain subdivisions suggesting different control functions for TRH in extrahypothalamic systems. However, as in other vertebrates, TRH neurons were abundant in the hypothalamic nuclei that presumably project to the median eminence and the neural lobe of the hypophysis. Considerable interspecies differences were noted mainly related to innervation of the olfactory and visual centers (thalamus and mesencephalic tectum) and the precise localization of immunoreactive cell bodies, which was assessed by double labeling with tyrosine hydroxylase. The comparison of the distribution of TRH immunoreactive neurons and fibers found in urodeles with those reported for other vertebrates, in particular with anamniotes, reveals a strong resemblance but also notable variations not only across vertebrate classes but also within the same class. In this respect, the virtual lack in urodeles of TRH innervation of the intermediate lobe of the hypophysis clearly contrasts with the innervation found in anurans. Therefore, the important role of skin color adaptation proposed for TRH in anurans on the basis of the direct innervation of the intermediate lobe is not applicable for urodeles.     

Distribution of thyrotropin-releasing hormone (TRH) immunoreactivity in the brain of the zebrafish (Danio rerio)

The distribution of thyrotropin-releasing hormone (TRH) in the brain of the adult zebrafish was studied with immunohistochemical techniques. In the telencephalon, abundant TRH-immunoreactive (TRHir) neurons were observed in the central, ventral, and supra- and postcommissural regions of the ventral telencephalic area. In the diencephalon, TRHir neurons were observed in the anterior parvocellular preoptic nucleus, the suprachiasmatic nucleus, the lateral hypothalamic nucleus, the rostral parts of the anterior tuberal nucleus and torus lateralis, and the posterior tuberal nucleus. Some TRHir neurons were also observed in the central posterior thalamic nucleus and in the habenula. The mesencephalon contained TRHir cells in the rostrodorsal tegmentum, the Edinger-Westphal nucleus, the torus semicircularis, and the nucleus of the lateral lemniscus. Further TRHir neurons were observed in the interpeduncular nucleus. In the rhombencephalon, TRHir cells were observed in the nucleus isthmi and the locus coeruleus, rostrally, and in the vagal lobe and vagal motor nucleus, caudally. In the forebrain, TRHir fibers were abundant in several regions, including the medial and caudodorsal parts of the dorsal telencephalic area, the ventral and commissural parts of the ventral telencephalic area, the preoptic area, the posterior tubercle, the anterior tuberal nucleus, and the posterior hypothalamic lobe. The dorsal thalamus exhibited moderate TRHir innervation. In the mesencephalon, the optic tectum received a rich TRHir innervation between the periventricular gray zone and the stratum griseum centrale. A conspicuous TRHir longitudinal tract traversed the tegmentum and extended to the rhombencephalon. The medial and lateral mesencephalic reticular areas and the interpeduncular nucleus were richly innervated by TRHir fibers. In the rhombencephalon, the secondary gustatory nucleus received abundant TRHir fibers. TRHir fibers moderately innervated the ventrolateral and ventromedial reticular area and richly innervated the vagal lobe and Cajal's commissural nucleus. Some TRHir fibers coursed in the lateral funiculus of the spinal cord. Some TRHir amacrine cells were observed in the retina. The wide distribution of TRHir neurons and fibers observed in the zebrafish brain suggests that TRH plays different roles. These results in the adult zebrafish reveal a number of differences with respect to the TRHir systems reported in other adult teleosts but were similar to those found during late developmental stages of trout (Díaz et al., 2001).     

Specific bindings of prostaglandin D2, E2 and F2α in postmortem human brain

Binding activities specific for each of [³H]prostaglandin (PG) D₂, E₂ and F_2α were detected in the P₂ fraction of the human brain homogenates. The bindings were time-dependent, saturable and of high affinity;K_dvalues were 30 nM for all the PG bindings. Regional distribution of these binding activities was determined by measuring specific bindings with 10 nM [³H]PG-D₂, [³H]PG-E₂ and [³H]PG-F_2α in the P₂ fractions from 17 brain regions. The PG-D₂ binding activity was high in the hypothalamus, amygdala and hippocampus followed by cerebellar nuclei, thalamus, nucleus accumbens and cerebral cortex. The PG-E₂ binding sites were similarly concentrated in the hypothalamus and the limbic system, but, unlike the PG-D₂ binding, no significant binding of [³H]PG-3₂ was observed in cerebellar nuclei, cerebellar cortex and putamen. Compared with these two PG bindings, PG-F_2α binding activity was low in many areas, but significant binding was detected in the amygdala, cingulate cortex, cerebellar medulla, hippocampus, nucleus accumbens, midbrain and hypothalamus. These results suggest the presence and specific distribution of three distinct types of PG binding activities, i.e. specific binding of PG-D₂, PG-E₂ and PG-F_2α, in the human brain.     

Hypothalamic and extrahypothalamic sauvagine-like immunoreactivity in the bullfrog (Rana catesbeiana) central nervous system

In the present study, immunocytochemistry and radioimmunoassay were used to investigate the presence of sauvagine in both hypothalamic and extrahypothalamic areas of the central nervous system (CNS) of the bullfrog (Rana catesbeiana) using a specific antiserum raised against synthetic non-conjugated sauvagine (SVG), a frog (Phyllomedusa sauvagei) skin peptide of the corticotropin-releasing factor (CRF) family. Sauvagine-immunoreactive (SVG-ir) bipolar neurons were found in the nucleus of the fasciculus longitudinalis medialis located in the rostral mesencephalic tegmentum. In the tectal mesencephalon, beaded SVG-ir fibres were present in the optic tectum, and in the torus semicircularis. Abundant SVG-ir varicose fibres were seen in the granulosa layer of the cerebellum, the nucleus isthmi, and the obex of the spinal cord. SVG-ir fibres were also seen by the alar plate of the rombencephalon. In the diencephalon, the antiserum stained parvocellular neurons of the preoptic nucleus (PON) which extended their dendrites into the cerebro-spinal fluid (CSF) of the third ventricle and projected their ependymofugal fibres to the zona externa (ZE) of the median eminence. Immunopositive fibres were also present in the medial forebrain bundle at the chiasmatic field, the posterior thalamus, the pretectal gray, and the ventrocaudal hypothalamus. In the telencephalon (forebrain), SVG-ir fibres were seen in the medial septum, the lateral septum, and the amygdala. The SVG immunoreactivity could not be detected after using the SVG antiserum previously immunoabsorbed with synthetic SVG (0.1 μM), but immunoblock of the antiserum with sucker (Catostomus commersoni) urotensin I (sUI), sole (Hippoglossoides elassodon) urotensin I, sucker CRF, rat/human CRF, or ovine CRF (0.1–10 μM) did not eliminate visualization of the immunoreactivity. In radioimmunoassay, the SVG antiserum did not crossreact with sUI, or the SVG fragments SVG_1–16, SVG_16–27, and SVG_26–34, but it recognized the C-terminal fragment SVG_35–40. Crossreaction with mammalian ovine CRF and rat/human CRF was negligible. Both hypothalamic and mesencephalic extracts gave parallel displacement curves to SVG. The results suggest the presence in the bullfrog brain of a SVG-like neuropeptide, i.e., a peptide of the CRF family, that either is SVG or shares high homology with the C-terminus of that peptide. The function of this neuropeptide in amphibians is not known at this time, but based on its anatomical distribution to the ZE it could affect the release of adrenocorticotropin (ACTH) or other substances from the amphibian pars distalis. Involvement of the SVG-like peptide in behavioural (forebrain), visual (thalamus-tegmentum mesencephali-pretectal gray-optic tectum), motor coordination (cerebellum), and autonomic (spinal) functions, as well as an undefined interaction with the CSF in the bullfrog seems likely. © 1996 Wiley-Liss, Inc.     

Regional changes in β1 thyroid hormone receptor immunoreactivity in rat brain after thyroidectomy

Quantitative [¹²⁵I]protein G-based immunohistochemistry was used to map the distribution of β₁ thyroid hormone receptor (TRβ₁) in normal and thyroidectomized adult rat brain, using a previously characterized polyclonal antibody. The distribution of TRβ₁-like immunoreactivity in normal brain was largely but not perfectly concordant with previous accounts of TRβ₁ mRNA distribution in rat brain. Thyroidectomy resulted in increased immunolabeling in most brain regions (mean increase: 14%, range: −4% to +25%), with statistically significant effects being observed in 9 of the 36 brain regions examined. Brain regions showing the most pronounced effects included the habenular nucleus (+22%), the oriens layer of the hippocampal CA3 region (+24%), and the lateral geniculate nucleus of the thalamus (+23%). These results demonstrate that the TRβ₁ protein in brain is capable of plastic changes in response to adult-onset alterations in TH levels. The observed pattern of brain regional receptor changes following thyroidectomy may provide clues for functional effects of thyroid function alterations in adults.     

Selective antagonism of prostaglandin (PG) E1, PGD2 and prostacyclin (PGI2) on human and rabbit platelets by DI-4-phloretin phosphate (DPP)

The inhibition of human and rabbit platelet aggregation by prostaglandin (PG) E₁, PGD₂ and prostacyclin (PGI₂) was examined to determine if the three PGs inhibit platelets via a common receptor. Di-4-phloretin phosphate (DPP) was found to antagonise PGD₂, but not PGE₁ or PGI₂ on human platelets. In contrast, on rabbit platelets, DPP antagonised PGI₂ and PGE₁ but not PGD₂. The results suggest that there are two distinct types of PG receptors on human and rabbit platelets activated either by prostacyclin or PGD₂.     

胞浆型磷脂酶A2基因多态性与精神分裂症的关系

目的：分析印度人群钙依赖性胞浆型磷脂酶A2（cPLA2)BanⅠ限制性内切酶基因多态性与精神分裂症的相互关系。方法：应用聚合酶链式反应（PCR）限制性片段长度多态性（RFLP）方法,在89例精神分裂症患者和78例健康人群中观察比较cPLA2等位基因和基因型频数分布。结果：PCR产物的BanⅠ限制性酶切片段于cPLA2基因第一非编码区显示多态性位点,命名为位点A;患者组和健康对照组cPLA2等位基因频数呈显著差异（P＜0．02）;精神分裂症患者显示A2／A2纯合基因型显著增加（P＜0．02）。结论：cPLA2基因多态性与印度人群精神分裂症相关联;cPLA2基因可能为精神分裂症候选基因之一,或与其他致病基因呈连锁不平衡。     

Typical and atypical antipsychotic occupancy of D2 and S2 receptors: An autoradiographic analysis in rat brain

In thin sections of rat brain, [3H]spiperone binds to D2 sites in the basal ganglia (caudate-putamen, nucleus accumbens, olfactory tubercle) and S2 sites in the claustrum and motor cortex. The in vitro displacement of [3H]spiperone from these regions was quantified autoradiographically with the "atypical" neuroleptics clozapine and thioridazine, which ameliorate psychosis, a "typical" neuroleptic, haloperidol, which also induces extrapyramidal side effects, or with metoclopramide, which induces extrapyramidal side effects but is an ineffective antipsychotic. Whereas metoclopramide was equipotent at D2 sites, haloperidol was less potent and clozapine and thioridazine more potent by 2- to 3-fold at competing for D2 sites in the nucleus accumbens or olfactory tubercle than in the caudate-putamen. As measured autoradiographically or with tissue homogenates, clozapine, thioridazine, and five other atypical neuroleptics were 4- to 800-times more potent at competing for S2 sites in the frontal cortex than for D2 sites in the basal ganglia. A preference of atypical antipsychotics for D2 receptors in the nucleus accumbens and olfactory tubercle and for the S2 receptor may explain the relative lack of extrapyramidal side effects produced by these compounds.     

Expression of the α1, α2 and α3 isoforms of the GABAA receptor in human alcoholic brain

The expression of the α₁, α₂ and α₃ isoforms of the GABA_A receptor was studied in the superior frontal and motor cortices of 10 control, 10 uncomplicated alcoholic and 7 cirrhotic alcoholic cases matched for age and post-mortem delay. The assay was based on competitive RT/PCR using a single set of primers specific to the α class of isoform mRNA species, and was normalized against a synthetic cRNA internal standard. The assay was shown to be quantitative for all three isoform mRNA species. Neither the patient's age nor the post-mortem interval significantly affected the expression of any isoform in either cortical area. The profile of expression was shown to be significantly different between the case groups, particularly because α₁ expression was raised in both groups of alcoholics cf controls. The two groups of alcoholics could be differentiated on the basis of regional variations in α₁ expression. In frontal cortex, α₁ mRNA expression was significantly increased when uncomplicated alcoholics were compared with control cases whereas alcoholic-cirrhotic cases were not significantly different from either controls or uncomplicated alcoholic cases. In the motor cortex, α₁ expression was elevated only when alcoholic-cirrhotic cases were compared with control cases. There was no significant difference between case groups or areas for any other isoform. © 1997 Elsevier Science B.V. All rights reserved.     

Dopamine D2 receptor agents, but not dopamine D1, modify brain glucose metabolism

The effects of recently described selective dopamine D₁ and D₂ agonists and antagonists on brain glucose metabolism were studied using the 2-[¹⁴C]deoxyglucose autoradiographic technique. The administration of LY-141865 or YM-09151-2, which behave as a specific D₂ agonist and antagonist respectively, modified brain glucose metabolism in a manner similar to that previously described for more classical dopaminergic agents, such as apomorphine and haloperidol. In contrast, the administration of SKF 38393 or SCH 23390, a specific D₁ agonist and antagonist respectively, was not followed by significant modifications of brain glucose metabolism in any of the brain regions studied. These results indicate that D₂ but not D₁ dopamine receptors are involved in the regulation of local brain glucose metabolism.     

Localization of ATP-gated P2X2 receptor immunoreactivity in the rat hypothalamus

Previous pharmacological studies have indicated that ATP receptors may be involved in the regulation of physiological functions in hypothalamus. In the present study, the distribution of P2X₂ receptor in the rat hypothalamus was studied with immunohistochemistry. It was shown that P2X₂ immunoreactivity-positive neurons and nerve fibres were localized in many hypothalamic nuclei. Intense labelling of both neuronal cell bodies and nerve fibres was observed in the paraventricular nucleus, arcuate nucleus, retrochiasmatic area, periventricular nucleus, and the ventral part of tuber cinereum area. In supraoptic, circular, and ventral tuberomammillary nuclei the neuronal cell bodies were strongly positive, but few nerve fibres were positive. Axons with strong P2X₂ immunoreactivity were found in the organum vasculosum of the lamina terminalis and median eminence. Some scattered positive neurons and nerve fibres were found in many hypothalamic nuclei including preoptic nucleus. The results of the present study demonstrated the existence of P2X receptors in hypothalamus, as a basis for detailed studies of the roles of P2X receptors in the regulation of hypothalamic functions.     

Effects of intracerebroventricularly infused histamine and selective H1, H2 and H3 agonists on food and water intake and urine flow in Wistar rats

The actions of intracerebroventricularly infused histamine and selective histamine H₁, H₂ and H₃ receptor agonists on food and water intake and urine flow were studied in rats. It was found that 100–800 nmoles of histamine significantly suppressed feeding. The H₁ agonist 2-(3-trifluoromethylphenyl)histamine (FMPH) decreased food intake, whereas the H₂ agonist dimaprit was without effect. Histamine- and FMPH-induced suppressions of feeding were attenuated by blockade of H₁ but not by H₂ receptors. The results clearly demonstrate that activation of brain H₁ receptors decreases food intake. In subsequent studies, we found that both metoprine and thioperamide, which increase histaminergic activity through different mechanisms, also reduced food intake. This finding indicates that the brain histaminergic system is associated with feeding behavior. The same is true with body water homeostasis. Histamine caused a long-lasting diuresis. Also dimaprit and metoprine increased urine flow and the blockade of H₂ receptors abolished the diuretic responses to histamine and dimaprit. On the other hand, the H₃ agonist (R)-α-methylhistamine elicited drinking and this effect could be prevented by thioperamide pretreatment. The results imply that activation of H₃ receptors predominantly provokes drinking, whereas central H₂ receptors mediate the diuretic effect of histamine.     

Amphetamine-induced hypolocomotion in mice with more brain D2 dopamine receptors

The relationship between brain D₂ dopamine receptors and locomotor response to amphetamine was investigated in eight strains of mice. The D₂ receptor is defined as that dopaminergic site with high affinity (nanomolar) for neuroleptics and low affinity (micromolar) for agonists. D₂ receptors were measured in the striatum and olfactory tubercle using [³H]spiperone and 10 μM sulpiride to define specific binding. Four inbred strains of mice (CBA/J; C57BL/6J; DBA/2J; SEC/1ReJ) had low receptor densities of about 380 and 160 fmoles/mg protein in the striatum and olfactory tubercle, respectively; all these mice were essentially nonresponsive (i.e., locomotion) to low doses of amphetamine (0.5 and 1.0 mg/kg i.p.) or showed hyperlocomotion to high doses (5 mg/kg). Three other mouse strains (BALB/cJ; A/J; C3H/HeJ) had higher densities of about 600 and 230 fmoles/mg protein in the striatum and olfactory tubercle, respectively, and these mice all responded with hypolocomotion to the low doses and hyperlocomotion to the high dose of amphetamine. The two genetically different populations, one of which responded to amphetamine with hypolocomotion while the other did not, are analogous to hyperactive children, only 70% of whom respond to amphetamine-like drugs. Thus, the mice with high receptor density may serve as a model for studying the hyperactivity syndrome which may be associated with dopaminergic dysfunction.     

TRH immunoreactivity in rat brain regions, spinal cord and pancreas: validation by high-pressure liquid chromatography and thin-layer chromatography

Thyrotropin-releasing hormone (TRH) immunoreactivity in rat tissue was assayed using a new reversed-phase, ion-pair, high-pressure liquid chromatography (HPLC) method for separating TRH from its analogs. Virtually all of the TRH immunoreactivity in hypothalamus, brain stem, preoptic area, septum, striatum, frontal cortex, spinal cord and pancreas had chromatographic characteristics corresponding to those of synthetic TRH. When analyzed by two additional, previously described HPLC methods for neuropeptides, all TRH immunoreactivity in rat brain, spinal cord and pancreas also behaved identically to that of authentic TRH. Similarly, in two different thin-layer chromatography systems, all TRH immunoreactivity migrated with that of synthetic TRH. We conclude that TRH-immunoreactive material in rat brain, spinal cord and pancreas really is TRH. The perhaps surprising specificity of the TRH immunoassay may occur because antibodies directed against TRH (pyroglutamyl-histidyl-prolineamide) recognize two atypical amino acids in close proximity.     

A1 and A2A adenosine receptors and A1 mRNA in mouse brain: effect of long-term caffeine treatment

The effect of oral treatment with caffeine, in doses that are known to produce marked adaptive effects, was investigated on A₁ and A_2A receptors in the mouse brain. Caffeine (0.1, 0.3 or 1 g/l) was added to the drinking water and the animals were sacrificed after a 14-day treatment period. Ligand binding to A₁ receptors was studied, using quantitative autoradiography, with the agonist [³H]cyclohexyladenosine (CHA) and the antagonist [³H]1,3-dipropyl-8-cyclopentyl xanthine (DPCPX). Caffeine did not remain in the sections during the autoradiography experiments. Caffeine treatment (1 g/l, but not 0.1 or 0.3 g/l) tended to increase [³H]CHA binding to the CA3 subfield of the hippocampus, but in no other region studied. There was no change in the number of A₁ receptors since [³H]DPCPX binding to the CA3, cerebral and cerebellar cortex was not influenced by caffeine treatment. There was similarly no change in the ability of CHA to displace [³H]DPCPX binding, suggesting that there are no major changes in the proportion of A₁ receptors that are coupled to G-proteins. mRNA for the A₁ receptor, measured by in situ hybridization, did not differ significantly between caffeine-treated and control mice in the structures examined. Thus, higher doses of caffeine can cause an increase in A₁ agonist binding without a corresponding change in A₁ mRNA or in A₁ antagonist binding, suggesting that the adaptive changes seen upon prolonged caffeine treatment may be in sites different from A₁ receptors. Caffeine (1 g/l) increased A_2A receptors in the striatum measured as binding of the agonist [³H]CGS 21680 suggesting that up-regulation of A_2A receptors may be an adaptive effect of caffeine intake.