Brain induces the expression of an early cell surface marker for blood-brain barrier-specific endothelium.

Capillaries derived from the perineural vascular plexus invade brain tissue early in embryonic development. Considerably later they differentiate into blood-brain barrier (BBB)-forming blood vessels. In the chick, the BBB as defined by impermeability for the protein horseradish peroxidase develops around embryonic day 13. We have previously found that brain endothelial cells start to express a number of proteins at around the same time, suggesting that these proteins play a role in BBB function. Here we describe a 74 kd protein defined by the monoclonal antibody HT7 that is expressed on the surface of chick embryonic blood cells and brain endothelial but on no other endothelial cells. This protein is not detectable on early embryonic brain endothelium, but is expressed by these cells on embryonic day 10. It is absent in choroid plexus endothelial cells which represent permeable fenestrated endothelial cells. The antigen is expressed on choroid plexus epithelium which is the site of the blood-cerebrospinal fluid barrier. Since it is also found in basolateral membranes of kidney tubules, it may be involved in specific carrier mechanisms. Embryonic mouse brain tissue transplanted on the chick chorio-allantoic membrane induces the expression of this antigen on endothelial cells derived from the chorio-allantois. Brain tissue can therefore induce in endothelial cells in vivo the expression of a molecule characteristic of brain endothelium.     

Developing nervous tissue induces formation of blood-brain barrier characteristics in invading endothelial cells: A study using quail-chick transplantation chimeras

Brain capillaries have structural and functional characteristics that constitute a regulatory interface, or “barrier,” between the blood and the brain. We have investigated the role of the neural tissue environment in the differentiation of the endothelial barrier, by transplanting embryonic brain fragments to the coelomic cavity, where they were vascularized by nonneural vessels, and fragments of embryonic mesoderm to the brain, where they were vascularized by neural vessels. A major problem in this approach is that when embryonic tissues are transplanted to an ectopic site, their own blood vessels survive and form a part of the new vascular system. This has made the results of previous experiments difficult to interpret. We overcame this problem by transplanting fragments of tissue that had not yet been vascularized from very young quail embryos to host chick embryos. These grafts did not contain vascular channels that could form part of a new vascular system. Furthermore, the distinctive quail nuclear morphology allowed us to demonstrate that the grafted tissue was, in fact, vascularized by the host vessels. Abdominal vessels vascularizing grafted neural tissue formed structural, functional, and histochemical features of the blood-brain barrier. In contrast, brain vessels vascularizing grafted mesodermal tissue were devoid of barrier characteristics. These results indicate that endothelial blood-brain barrier characteristics develop in response to some aspect of the neural environment.     

Neurothelin: an inducible cell surface glycoprotein of blood-brain barrier-specific endothelial cells and distinct neurons

The blood-brain barrier is characterized by still poorly understood barrier and transport functions performed by specialized endothelial cells. Hybridoma technology has been used to identify a protein termed neurothelin that is specific for these endothelial cells. Neurothelin is defined by the species-specific mouse mAb 1W5 raised against lentil-lectin-binding proteins of neural tissue from embryonic chick. In the posthatch chick, neurothelin expression is found on endothelial cells within the brain but not on those of the systemic vascular system. Injection of the monoclonal antibody in vivo leads to labeling of brain capillaries, indicating that the corresponding antigen is expressed on the luminal surface of brain endothelial cells. Transplantation of embryonic mouse brain onto the chick chorioallantoic membrane results in rodent brain vascularization by the avian vascular system. Subsequently, normally mAb 1W5-negative endothelial cells, originating from blood vessels of the chick chorioallantoic membrane, are induced to express neurothelin when they are in contact with mouse neural tissue. In contrast to differentiated brain neurons that do not express neurothelin, neurons of the nonvascularized chick retina synthesize neurothelin. However, neurothelin is not found on retinal ganglion cell axons terminating on 1W5-negative brain cells. 1W5 immunoreactivity was also found in the pigment epithelium that forms the blood-eye barrier. Putting epithelial cells into culture results in concentration of neurothelin at cell-cell contact sites, leaving other cell surface areas devoid of antigen. Therefore, the distribution of neurothelin appears to be regulated by cell-cell interactions. In Western blot analysis, neurothelin was identified as a protein with a molecular mass of approximately 43 kD. The protein bears at least one intramolecular disulfide bridge and sulfated glucuronic acid as well as alpha-D-substituted mannose/glucose moieties. The exclusive neurothelin expression in the posthatch chick on endothelial cells of the central nervous system but not on systemic endothelial cells makes neurothelin a marker specific for blood-brain barrier-forming endothelial cells. The spatiotemporally regulated neurothelin expression in neurons suggests an interaction between vascularization and neuronal differentiation.     

Surfactant protein A promoter directs the expression of Cre recombinase in brain microvascular endothelial cells of transgenic mice.

Brain microvascular endothelial cells (ECs) have unique characteristics distinguished from peripheral ECs and play important roles in blood-brain barrier (BBB). To investigate the physiological control of the brain ECs, we generated a transgenic mouse line in which the expression of Cre recombinase was driven by the promoter of the mouse surfactant protein A (SP-A) gene. The Cre activity was detected in blood vessels of brain, alveolar type II cells of lung and epithelium of gland stomach. In brain ECs, the Cre activity started at embryonic day 11.5, indicating that the subpopulation of ECs in brain could be molecularly defined at early embryonic stages. The use of SP-A-Cre mice should facilitate analysis of gene function in the brain ECs.     

Avian coronary endothelium is a mosaic of sinus venosus‐ and ventricle‐derived endothelial cells in a region‐specific manner

The origin of coronary endothelial cells (ECs) has been investigated in avian species, and the results showed that the coronary ECs originate from the proepicardial organ (PEO) and developing epicardium. Genetic approaches in mouse models showed that the major source of coronary ECs is the sinus venosus endothelium or ventricular endocardium. To clarify and reconcile the differences between avian and mouse species, we examined the source of coronary ECs in avian embryonic hearts. Using an enhanced green fluorescent protein‐Tol2 system and fluorescent dye labeling, four types of quail‐chick chimeras were made and quail‐specific endothelial marker (QH1) immunohistochemistry was performed. The developing PEO consisted of at least two cellular populations in origin, one was sinus venosus endothelium‐derived inner cells and the other was surface mesothelium‐derived cells. The majority of ECs in the coronary stems, ventricular free wall, and dorsal ventricular septum originated from the sinus venosus endothelium. The ventricular endocardium contributed mainly to the septal artery and a few cells to the coronary stems. Surface mesothelial cells of the PEO differentiated mainly into a smooth muscle phenotype, but a few differentiated into ECs. In avian species, the coronary endothelium had a heterogeneous origin in a region‐specific manner, and the sources of ECs were basically the same as those observed in mice.     

Plasticity of endothelial cells during arterial-venous differentiation in the avian embryo

Remodeling of the primary vascular system of the embryo into arteries and veins has long been thought to depend largely on the influence of hemodynamic forces. This view was recently challenged by the discovery of several molecules specifically expressed by arterial or venous endothelial cells. We here analysed the expression of neuropilin-1 and TIE2, two transmembrane receptors known to play a role in vascular development. In birds, neuropilin-1 was expressed by arterial endothelium and wall cells, but absent from veins. TIE2 was strongly expressed in embryonic veins, but only weakly transcribed in most arteries. To examine whether endothelial cells are committed to an arterial or venous fate once they express these specific receptors, we constructed quail-chick chimeras. The dorsal aorta, carotid artery and the cardinal and jugular veins were isolated together with the vessel wall from quail embryos between embryonic day 2 to 15 and grafted into the coelom of chick hosts. Until embryonic day 7, all grafts yielded endothelial cells that colonized both host arteries and veins. After embryonic day 7, endothelial plasticity was progressively lost and from embryonic day 11 grafts of arteries yielded endothelial cells that colonized only chick arteries and rarely reached the host veins, while grafts of jugular veins colonized mainly host veins. When isolated from the vessel wall, quail aortic endothelial cells from embryonic day 11 embryos were able to colonize both host arteries and veins. Our results show that despite the expression of arterial or venous markers the endothelium remains plastic with regard to arterial-venous differentiation until late in embryonic development and point to a role for the vessel wall in endothelial plasticity and vessel identity.     

Cell-cell junctions of dermal microvascular endothelial cells contain tight and adherens junction proteins in spatial proximity

Endothelial cell-cell contacts control the vascular permeability, thereby regulating the flow of solutes, macromolecules, and leukocytes between blood vessels and interstitial space. Because of specific needs, the endothelial permeability differs significantly between the tight blood-brain barrier endothelium and the more permeable endothelial lining of the non-brain microvasculature. Most likely, such differences are due to a differing architecture of the respective interendothelial cell contacts. However, while the molecules and junctional complexes of macrovascular endothelial cells and the blood-brain barrier endothelium are fairly well characterized, much less is known about the organization of intercellular contacts of microvascular endothelium. Toward this end, we developed a combined cross-linking and immunoprecipitation protocol which enabled us to map nearest neighbor interactions of junctional proteins in the human dermal microvascular endothelial cell line HMEC-1. We show that proteins typically located in tight or adherens junctions of epithelial cells are in the proximity in HMEC-1 cells. This contrasts with the separation of the different types of junctions observed in polarized epithelial cells and "tight" endothelial layers of the blood-brain barrier and argues for a need of the specific junctional contacts in microvascular endothelium possibly required to support an efficient transendothelial migration of leukocytes.     

Evidence for a cyclic renewal of lymphocyte precursor cells in the embryonic chick thymus

Experiments involving sequential transplantations of the chick embryonic thymus at E9 to E12 into a first 3-day host quail embryo and then into a second chick host allowed demonstration of the cyclic periodicity of hemopoietic cell seeding of the embryonic thymus. After a first wave of colonization occurring between E6.5 and E8, the thymus becomes refractory to hemopoietic cell entry for about 4 days. It resumes its capacity to be seeded by a second wave of blood-borne stem cells at E12. After a second period of non receptivity starting at E14, a third wave of incoming cells reaches the thymus around E18. Therefore, with a slightly different periodicity, the same cyclic mechanism regulates the renewal of lymphocytes in chick and quail embryos. Quail hemopoietic cells were immunostained in the chimeric thymuses, with a species specific monoclonal antibody (anti-MB1) which recognizes a common surface antigenic determinant on all endothelial and blood cells of the quail (except erythrocytes). Two steps could thus be distinguished in the seeding process. When the thymus becomes receptive for hemopoietic cells, the latter first accumulate in the intrathymic blood vessels before penetrating massively in the thymic parenchyma. The quail chick-chimera system combined with the use of a species- and cell-type-specific antibody provides a unique tool for studying thymic colonization by lymphocyte precursors.     

Synthesis and localization of plasma proteins in the developing human brain. Integrity of the fetal blood-brain barrier to endogenous proteins of hepatic origin

The distribution and possible origins of plasma proteins in the human embryonic and fetal brain at different stages of development have been investigated by a combination of isolation and translation of mRNAs and immunocytochemistry using specific antisera. As many as 23 plasma-like proteins have been identified using immunocytochemical methods at the light microscopical level. The presence of mRNAs for 13 of the immunocytochemically positive plasma proteins was demonstrated by in vitro and in ovo translation followed by crossed immunoelectrophoresis and autoradiography; this indicates in situ synthesis of these proteins (e.g., alpha-fetoprotein, alpha 1-antitrypsin, GC-globulin, alpha 2-macroglobulin, pseudocholinesterase, and transferrin) in some brain regions. The regional distribution of some proteins and the absence of some mRNAs suggest that the presence of certain plasma proteins in developing brain may be accounted for by uptake from csf or via nerve processes extending beyond the blood-brain barrier. In several cases, specific proteins appear to be associated with defined cell types, e.g., alpha-fetoprotein, GC-globulin, and ceruloplasmin with neurons, alpha 2-macroglobulin with endothelial cells, and ferritin with glial cells. Some proteins were associated with two or three cell types, e.g., alpha 1-antitrypsin with neurons and glia, and transferrin and alpha 2HS-glycoprotein with neurons, glia, and endothelial cells. Comparison of the expression of mRNAs from fetal brain and liver injected into Xenopus oocytes showed that a few proteins (transferrin and ceruloplasmin) were secreted when liver mRNA was injected, but not when brain mRNA was injected. This suggests that there may be an important difference in the structure and/or processing of these proteins in the brain which may reflect a function different from that associated with them when they originate from the liver. Staining was generally intracellular rather than extracellular; plasma proteins were not associated with the areas immediately around blood vessels although there was a strong immunoprecipitation for each protein within the lumen of cerebral blood vessels. These immunocytochemical findings together with the identification of mRNAs for a large number of plasma proteins in immature brain are discussed in relation to animal experimental work which suggests that the blood-brain barrier to protein is present even at very early stages of brain development.     

Vasculogenesis and angiogenesis in the mouse embryo studied using quail/mouse chimeras

Using quail/chick chimeras, we have previously shown that different embryonic territories are vascularized through two distinct mecanisms, angiogenesis and vasculogenesis. Angiogenesis occurs in tissues of somatopleural origin, vasculogenesis occurs in territories of splanchnopleural origin. The aim of this work was to establish if these modes of vascularization were conserved in the mammalian embryo. Since in vivo manipulations with mammalian embryos are difficult to perform, we used a quail/mouse chimera approach. Mouse limb buds of somatopleural origin, and visceral organ rudiments of splanchnopleural origin, were grafted into the coelomic cavity of 2.5 day-old quail embryos. After four to seven days, the hosts were killed and the origin of the endothelial cells in the mouse tissues was determined by double staining with the quail endothelial and hematopoietic cell-specific marker, QH1 and mouse-specific VEGFR2 and VEGFR3 probes. Our findings show that the great majority of vessels which developed in the mouse limbs was QH1+, indicating that these tissues were vascularized by angiogenesis. Conversely, visceral organs were vascularized through the vasculogenesis process by mouse endothelial cells which differentiated in situ. These results demonstrate for the first time that in the mouse embryo, as previously shown in avian species, the tissues from somatopleural origin are vascularized by angiogenesis, while rudiments of a splanchnopleural origin are vascularized by vasculogenesis, both at vascular and lymphatic levels.     

Production of a heparin-binding angiogenesis factor by the embryonic kidney

Embryonic mouse kidneys induce angiogenesis when transplanted on the quail chorioallantoic membrane (Ekblom, P., H. Sariola, M. Karkinen, and L. Saxén, 1982, Cell Differ., 11:35-39). In these experiments all blood vessels were derived from the quail host, suggesting that kidney endothelium is derived from outside blood vessels. We have now analyzed whether kidney angiogenesis is regulated by kidney-derived soluble factors that stimulate the growth of new blood vessels. In the rabbit cornea, 11-d embryonic kidneys induced angiogenesis, whereas uninduced 11-d kidney mesenchymes did not. To characterize and purify this activity from an embryonic organ, we dissected between 600 and 1,000 14-17-d-old embryonic mouse kidneys for each purification experiment. Growth factor activity for capillary endothelial cells was found to bind to heparin-Sepharose and eluted at 0.9-1.1 M sodium chloride. Gel filtration revealed a molecular weight of 16,000-20,000 of this factor. A major 18,000-mol-wt band was seen after gel electrophoresis and silver staining of partially purified growth factor material. The chromatographed factor is mitogenic for endothelial cells but not for smooth muscle cells and stimulates angiogenesis in vivo in the rabbit cornea. Adult kidneys contained two heparin-binding endothelial cell growth factors. The differentiation-dependent production of an angiogenesis factor by the embryonic kidney suggests an important role of angiogenesis in organogenesis.     

Detection of endogenous epidermal growth factor-like activity in the developing chick embryo

Epidermal growth factor-like activity has been detected by radioreceptor assay and radioimmunoassay in the developing chick embryo. Very little activity could be detected prior to Day 8 of embryonic life (hatching is at Day 21). A peak of EGF activity was detectable by both assays over Days 10 to 12. The EGF activity then fell to virtually undetectable levels during Days 14 to 17. A later rise in RRA detectable EGF like activity was then observed over Days 18-20. The EGF activity from a Day 11 embryo chromatographed on high-performance liquid chromatography as a single peak, with very high recovery of activity, at a later elution position than mouse EGF or human EGF.     

Blood-Brain Barrier Breakdown after Embolic Stroke in Rats Occurs without Ultrastructural Evidence for Disrupting Tight Junctions

The term blood-brain barrier (BBB) relates to the ability of cerebral vessels to hold back hydrophilic and large molecules from entering the brain, thereby crucially contributing to brain homeostasis. In fact, experimental opening of endothelial tight junctions causes a breakdown of the BBB evidenced as for instance by albumin leakage. This and similar observations led to the conclusion that BBB breakdown is predominantly mediated by damage to tight junction complexes, but evidentiary ultrastructural data are rare. Since functional deficits of the BBB contribute to an increased risk of hemorrhagic transformation and brain edema after stroke, which both critically impact on the clinical outcome, we studied the mechanism of BBB breakdown using an embolic model of focal cerebral ischemia in Wistar rats to closely mimic the essential human pathophysiology. Ischemia-induced BBB breakdown was detected using intravenous injection of FITC-albumin and tight junctions in areas of FITC-albumin extravasation were subsequently studied using fluorescence and electron microscopy. Against our expectation, 25 hours after ischemia induction the morphology of tight junction complexes (identified ultrastructurally and using antibodies against the transcellular proteins occludin and claudin-5) appeared to be regularly maintained in regions where FITC-albumin massively leaked into the neuropil. Furthermore, occludin signals along pan-laminin-labeled vessels in the affected hemisphere equaled the non-affected contralateral side (ratio: 0.966 vs. 0.963; P = 0.500). Additional ultrastructural analyses at 5 and 25 h after ischemia induction clearly indicated FITC-albumin extravasation around vessels with intact tight junctions, while the endothelium exhibited enhanced transendothelial vesicle trafficking and signs of degeneration. Thus, BBB breakdown and leakage of FITC-albumin cannot be correlated with staining patterns for common tight junction proteins alone. Understanding the mechanisms causing functional endothelial alterations and endothelial damage is likely to provide novel protective targets in stroke.     

Impact of chronic smoking on traumatic brain microvascular injury: An in vitro study

Traumatic brain injury (TBI) is a major reason of cerebrovascular and neurological damage. Premorbid conditions such as tobacco smoking (TS) can worsen post-TBI injuries by promoting vascular endothelial impairments. Indeed, TS-induced oxidative stress (OS) and inflammation can hamper the blood-brain barrier (BBB) endothelium. This study evaluated the subsequence of chronic TS exposure on BBB endothelial cells in an established in vitro model of traumatic cell injury. Experiments were conducted on confluent TS-exposed mouse brain microvascular endothelial cells (mBMEC-P5) following scratch injury. The expression of BBB integrity–associated tight junction (TJ) proteins was assessed by immunofluorescence imaging (IF), Western blotting (WB) and quantitative RT-PCR. We evaluated reactive oxygen species (ROS) generation, the nuclear factor 2–related (Nrf2) with its downstream effectors and several inflammatory markers. Thrombomodulin expression was used to assess the endothelial haemostatic response to injury and TS exposure. Our results show that TS significantly decreased Nrf2, thrombomodulin and TJ expression in the BBB endothelium injury models while increased OS and inflammation compared to parallel TS-free cultures. These data suggest that chronic TS exposure exacerbates traumatic endothelial injury and abrogates the protective antioxidative cell responses. The downstream effect was a more significant decline of BBB endothelial viability, which could aggravate subsequent neurological impairments.     

Neurothelin: molecular characteristics and developmental regulation in the chick CNS.

Neurothelin has recently been identified as a cell surface protein specific for chick endothelial cells forming the blood-brain barrier. Neurons of the adult brain are essentially devoid of neurothelin. In contrast, neurons of the chick retina, which lack blood vessels and accessory astrocytes, express neurothelin. Here we demonstrate that during chick brain development initially neurothelin is expressed probably in all neuroblasts. With proceeding cytodifferentiation, such as vascularization and gliogenesis, brain neurons become neurothelin negative. Coincidentally the endothelial cells forming the blood-brain barrier start to synthesize neurothelin. In contrast to brain neurons, in retina neurons, neurothelin expression increases by one order of magnitude during the course of histogenesis. Coculturing of chick retinal cells with purified rat astrocytes in vitro results in reduction of neural neurothelin expression as quantified by ELISA. Conversely, disruption of the glia-neuron interactions by culturing brain neurons as individualized cells in vitro leads to a reexpression of neurothelin. This is consistent with the hypothesis that astrocytes inhibit neurothelin expression in neurons. Biochemical characterization classifies neurothelin as an integral membrane protein. Temperature-induced-detergent phase separation, phospholipase C digestion and sodium carbonate treatment were employed to distinguish between integral membrane proteins, lipid-anchored proteins and peripheral membrane proteins. Two-dimensional gel electrophoresis reveals an isoelectric point of about 6.4 for neurothelin. Polysaccharide analysis by glycosidase digestion and lectin binding indicates that neurothelin is highly glycosylated. The relative molecular mass of glycosylated neurothelin is 41 x 10(3), whereas the peptide backbone is only 25 x 10(3). The very strict spatiotemporal regulation of neurothelin expression in the central nervous system suggests that neurothelin fulfils possibly a crucial function such as transport of low relative molecular mass components that are essential for neuronal metabolism. The proposed biological activity of neurothelin might be specifically affected by some of its distinct biochemical features.     

Saturable Retention of Vasopressin by Hippocampus Vessels In Vivo, Associated with Inhibition of Blood-Brain Transfer of Large Neutral Amino Acids

Vasopressin receptors have been reported in the endothelium of brain capillaries. The function of these receptors is not known. To test the prediction that vasopressin receptors in brain capillary endothelium affect amino acid transport across the blood-brain barrier and to assess the role of vasopressin transport across the cerebral vascular endothelium, we measured (a) the endothelial permeability to the large neutral amino acid leucine in the absence and presence of arginine vasopressin (AVP) and (b) the permeability of the blood-brain barrier to AVP relative to manitol. In brain regions protected by the blood-brain barrier, after circulation for 20 s, coinjection of leucine and AVP intravenously led to a decrease of leucine transport unrelated to changes of blood flow. The decrease was most pronounced in hippocampus (42%) and least pronounced in olfactory bulb and colliculi (17 and 19%, respectively). In the latter regions, the endothelial permeability to AVP did not significantly exceed that of mannitol. In hippocampus and in regions with no blood-brain barrier (pituitary and pineal glands), AVP retention in excess of mannitol retention was blocked by unlabeled AVP. The findings do not contradict the hypothesis of a role for AVP in the regulation of large neutral amino acid transfer into brain tissue.     

Astrocyte-endothelial cell relationships during the establishment of the blood-brain barrier in the chick embryo

Summary— The blood-brain barrier (BBB) preventing the passage of proteins is established at day 13 of development in the embryonic chick brain. We describe, as early as this stage, the existence of characteristic tight junctions between endothelial cells that is related to the time of appearance of the basal lamina. At earlier stages (E10, E12), when endothelial cells seem to be held back from the glio-neural neuropile by fibroblast-like cells identified by their appearance and position, the astrocyte plasma membranes already present a rare but characteristic molecular arrangement: the orthogonal arrays of particles (OAs). These OAs become progressively more abundant in astrocytic plasmalemmas contiguous to endothelial cells when these cells have been surrounded by the basal lamina since E15. The contact between astrocytes and basal lamina therefore seems to favor a high density of OAs, as has been shown in vertebrate astrocytes in contact with endothelial cells or leptomeninges. No correlation exists between the onset of the BBB and the time of appearance of OAs.     

Chronic excessive erythrocytosis induces endothelial activation and damage in mouse brain

Excessive erythrocytosis results in severely increased blood viscosity, which may have significant detrimental effects on endothelial cells and, ultimately, function of the vascular endothelium. Because blood-brain barrier stability is crucial for normal physiological function, we used our previously characterized erythropoietin-overexpressing transgenic (tg6) mouse line (which has a hematocrit of 0.8-0.9) to investigate the effect of excessive erythrocytosis on vessel number, structure, and integrity in vivo. These mice have abnormally high levels of nitric oxide (NO), a potent proinflammatory molecule, suggesting altered vascular permeability and function. In this study, we observed that brain vessel density of tg6 mice was significantly reduced (16%) and vessel diameter was significantly increased (15%) compared with wild-type mice. Although no significant increases in vascular permeability under normoxic or acute hypoxic conditions (8% O2 for 4 h) were detected, electron-microscopic analysis revealed altered morphological characteristics of the tg6 endothelium. Tg6 brain vascular endothelial cells appeared to be activated, with increased luminal protrusions reminiscent of ongoing inflammatory processes. Consistent with this observation, we detected increased levels of intercellular adhesion molecule-1 and von Willebrand factor, markers of endothelial activation and damage, in brain tissue. We propose that chronic excessive erythrocytosis and sustained high hematocrit cause endothelial damage, which may, ultimately, increase susceptibility to vascular disease.