Activated macrophages in HIV encephalitis and a macaque model show increased [H](R)-PK11195 binding in a PI3-kinase-dependent manner

HIV encephalitis (HIVE) is a neurodegenerative disease seen in approximately one in four terminally infected patients. Macaques infected with the simian immunodeficiency virus develop encephalitis (SIVE) very similar to the human disease. Neurodegeneration in both these conditions occurs from the effects of toxic viral proteins and neurotoxins derived from activated brain macrophages. Activated macrophages in the brain of macaques with SIVE can be labeled in vivo using positron emission tomography (PET) using PK11195, a ligand that binds the peripheral benzodiazepine receptor (PBR). However, the functional significance and mechanisms mediating increased PK11195 binding in activated brain macrophages are not known. Using post mortem tissues from macaques with SIVE and macrophages cell cultures activated with lipopolysaccharide (LPS), we show that [³H](R)-PK11195 binding is increased in activated macrophages. Increased [³H](R)-PK11195 binding in LPS-activated macrophages was reversed by pharmacologically inhibiting class III phosphatidylinositol-3 kinase (PI3-kinase), but was not altered by inhibiting the mitogen-activated protein kinase (MAP-kinase) pathway. Our results suggest that activated macrophages in lentiviral encephalitis show increased [³H](R)-PK11195 binding in a PI3-kinase-dependent fashion which may help elucidate the function of PBR in activated brain macrophages in HIVE and other neuroinflammatory diseases.     

PK11195 labels activated microglia in Alzheimer's disease and in vivo in a mouse model using PET

Activated microglia may promote neurodegeneration in Alzheimer's disease (AD) and may also help in amyloid clearance in immunization therapies. In vivo imaging of activated microglia using positron emission tomography (PET) could assist in defining the role of activated microglia during AD progression and therapeutics. We hypothesized that PK11195, a ligand that binds activated microglia, could label these cells in postmortem AD tissues and in vivo in an animal model of AD using PET. [(3)H](R)-PK11195 binding was significantly higher in AD frontal cortex compared to controls and correlated mainly with the abundance of immunohistochemically labeled activated microglia. With age, the brains of APP/PS1 transgenic mice showed progressive increase in [(3)H](R)-PK11195 binding and [(11)C](R)-PK11195 retention in vivo assessed using microPET, which correlated with the histopathological abundance of activated microglia. These results suggest that PK11195 binding in AD postmortem tissue and transgenic mice in vivo correlates with the extent of microglial activation and may help define the role of activated microglia in the pathogenesis and treatment of AD.     

Longitudinal analysis of monocyte/macrophage infection in simian immunodeficiency virus-infected, CD8+ T-cell-depleted macaques that develop lentiviral encephalitis

The histopathological hallmark of lentiviral-associated encephalitis is an abundance of infected and activated macrophages. Why a subset of infected hosts develops lentiviral encephalitis and others do not is unknown. Using a CD8(+) T-cell depletion model of simian immunodeficiency virus (SIV)-infected rhesus macaques, we examined the relationship between peripheral SIV infection of monocytes/macrophages and the development of encephalitis. At the same time that cerebral spinal fluid viral load increased in macaques that developed encephalitis, we observed that monocyte-derived macrophages from these macaques produced more virus than those from macaques that did not develop encephalitis. However, during the course of infection, the number of blood monocyte-associated SIV DNA copies did not distinguish macaques that developed simian immunodeficiency virus encephalitis from macaques that did not develop encephalitis. Paradoxically, in this model, macaques that developed encephalitis had fewer SIV-infected macrophages in lungs and thymus at postmortem than macaques that did not develop encephalitis. These findings suggest that inherent differences in host monocyte viral production are related to development of encephalitis.     

Macrophages relate presynaptic and postsynaptic damage in simian immunodeficiency virus encephalitis

Neurodegeneration observed in lentiviral-associated encephalitis has been linked to viral-infected and -activated central nervous system macrophages. We hypothesized that lentivirus, macrophages, or both lentivirus and macrophages within distinct microenvironments mediate synaptic damage. Using the simian immunodeficiency virus (SIV)-infected macaque model, we assessed the relationship between virus, macrophages, and neurological damage in multiple brain regions using laser confocal microscopy. In SIV-infected macaques with SIV encephalitis (SIVE), brain tissue concentrations of SIV RNA were 5 orders of magnitude greater than that observed in nonencephalitic animals. In SIVE, staining for postsynaptic protein microtubule-associated protein-2 was significantly decreased in the caudate, hippocampus, and frontal cortical gray matter compared to nonencephalitic controls, whereas staining for presynaptic protein synaptophysin was decreased in SIV-infected macaques with and without encephalitis. These data suggest that presynaptic damage occurs independent of pathological changes associated with SIVE, whereas postsynaptic damage is more tightly linked to regional presence of both activated and infected macrophages.     

YKL-40, a marker of simian immunodeficiency virus encephalitis, modulates the biological activity of basic fibroblast growth factor

Human immunodeficiency virus encephalitis causes dementia in acquired immune deficiency syndrome patients. Using proteomic analysis of postmortem cerebrospinal fluid (CSF) and brain tissue from the simian immunodeficiency virus primate model, we demonstrate here a specific increase in YKL-40 that was tightly associated with lentiviral encephalitis. Longitudinal analysis of CSF from simian immunodeficiency virus-infected pigtailed macaques showed an increase in YKL-40 concentration 2 to 8 weeks before death from encephalitis. This increase in YKL-40 correlated with an increase in CSF viral load; it may therefore represent a biomarker for the development of encephalitis. Analysis of banked human CSF from human immunodeficiency virus-infected patients also demonstrated a correlation between YKL-40 concentration and CSF viral load. In vitro studies demonstrated increased YKL-40 expression and secretion by macrophages and microglia but not by neurons or astrocytes. We found that YKL40 displaced extracellular matrix-bound basic fibroblast growth factor (bFGF) as well as inhibited the mitogenic activity of both fibroblast growth factor receptor 1-expressing BaF3 cells and bFGF-induced axonal branching in hippocampal cultures. Taken together, these findings demonstrate that during lentiviral encephalitis, YKL-40 may interfere with the biological activity of bFGF and potentially of other heparin-binding growth factors and chemokines that can affect neuronal function or survival.     

Induction of pathogenic sets of genes in macrophages and neurons in NeuroAIDS

The etiology of the central nervous system (CNS) alterations after human immunodeficiency virus (HIV) infection, such as dementia and encephalitis, remains unknown. We have used microarray analysis in a monkey model of neuroAIDS to identify 98 genes, many previously unrecognized in lentiviral CNS pathogenesis, whose expression is significantly up-regulated in the frontal lobe of simian immunodeficiency virus-infected brains. Further, through immunohistochemical illumination, distinct classes of genes were found whose protein products localized to infiltrating macrophages, endothelial cells and resident glia, such as CD163, Glut5, and ISG15. In addition we found proteins induced in cortical neurons (ie, cyclin D3, tissue transglutaminase, alpha1-antichymotrypsin, and STAT1), which have not previously been described as participating in simian immunodeficiency virus or HIV-related CNS pathology. This molecular phenotyping in the infected brains revealed pathways promoting entry of macrophages into the brain and their subsequent detrimental effects on neurons. These data support the hypothesis that in HIV-induced CNS disease products of activated macrophages and astrocytes lead to CNS dysfunction by directly damaging neurons, as well as by induction of altered gene and protein expression profiles in neurons themselves which are deleterious to their function.     

CD163, a marker of perivascular macrophages, is up-regulated by microglia in simian immunodeficiency virus encephalitis after haptoglobin-hemoglobin complex stimulation and is suggestive of breakdown of the blood-brain barrier

Macrophages and microglia are the major cell types infected by human immunodeficiency virus and simian immunodeficiency virus (SIV) in the central nervous system. Microglia are likely infected in vivo, but evidence of widespread productive infection (ie, presence of viral RNA and protein) is lacking. This conclusion is controversial because, unlike lymphocytes, macrophages and microglia cannot be discreetly immunophenotyped. Of particular interest in the search for additional monocyte/macrophage-lineage cell markers is CD163; this receptor for haptoglobin-hemoglobin (Hp-Hb) complex, which forms in plasma following erythrolysis, is expressed exclusively on cells of monocyte/macrophage lineage. We examined CD163 expression in vitro and in vivo by multiple techniques and at varying times after SIV infection in macaques with or without encephalitis. In normal and acutely SIV-infected animals, and in SIV-infected animals without encephalitis, CD163 expression was detected in cells of monocyte/macrophage lineage, including perivascular macrophages, but not in parenchymal microglia. However, in chronically infected animals with encephalitis, CD163 expression was detected in activated microglia surrounding SIV encephalitis lesions in the presence of Hp-Hb complex, suggesting leakage of the blood-brain barrier. CD163 expression was also induced on microglia in vitro after stimulation with Hp-Hb complex. We conclude that CD163 is a selective marker of perivascular macrophages in normal macaques and during the early phases of SIV infection; however, later in infection in animals with encephalitis, CD163 is also expressed by microglia, which are probably activated as a result of vascular compromise.     

Induction of cell-cycle regulators in simian immunodeficiency virus encephalitis

Neuronal degeneration associated with human immunodeficiency virus encephalitis has been attributed to neurotoxicity of signaling molecules secreted by activated, infected macrophages. We hypothesized that the barrage of signals present in the extracellular milieu of human immunodeficiency virus-infiltrated brain causes inappropriate activation of neuronal cell-cycle machinery. We examined the presence of three members of the cell-cycle control machinery: pRb, E2F1, and p53 in the simian immunodeficiency virus encephalitis (SIVE) model. Compared to noninfected and simian immunodeficiency virus-infected, nonencephalitic controls, we observed increased protein expression of E2F1 and p53 and aberrant cellular localization of E2F1 and pRb. In SIVE, E2F1 was abundant in the cytoplasm of neurons in both neurons and astrocytes proximal to SIVE pathology in the basal ganglia. pRb staining was nuclear and cytoplasmic in cortical neurons of SIVE cases. Antibodies to phosphorylated pRb also labeled the cytoplasm of cortical neurons. These data suggest that in SIVE, cell signaling results in phosphorylation of pRb which may result in subsequent alteration in E2F1 activity. As increased E2F1 and p53 activities have been linked to cell death, these data suggest that the neurodegeneration in SIVE could in part be because of changes in expression and activity of cell-cycle machinery.     

Correlation of acute humoral response with brain virus burden and survival time in pig-tailed macaques infected with the neurovirulent simian immunodeficiency virus SIVsmmFGb

Infection of pig-tailed macaques with the simian immunodeficiency virus (SIV) isolate SIVsmmFGb frequently results in SIV encephalitis (SIVE) in addition to immunodeficiency and acquired immune deficiency syndrome. We used in situ hybridization to quantitate the number of SIV-infected cells in brain parenchyma, choroid plexus, and meninges from 17 macaques that developed acquired immune deficiency syndrome after infection with SIVsmmFGb. SIV-infected cells and histopathological lesions of SIVE were identified in 15 of 17 animals (88.2%), including 12 of 12 rapid progressors (RP) and 3 of 5 slow progressors (SP). The parenchymal virus burden was much greater in RP macaques than in the three SP macaques with SIVE (median values of 24.3 versus 0.3 infected cells/mm(2), respectively; P < 0.05). Viral load differences between RP and SP with SIVE were less marked in choroid plexus (29.6 versus 12.8 infected cells/mm(2), respectively) and meninges (133.0 versus 34.2 infected cells/mm(2), respectively). A significant negative correlation was observed between the magnitude of the anti-SIV antibody titer at 1 month after inoculation and brain virus burden at necropsy (r = -0.614; P < 0.01). The close association between immune response and SIVE in this model should prove useful for identifying correlates of immune protection against primate lentiviral encephalitis.     

CD163 identifies perivascular macrophages in normal and viral encephalitic brains and potential precursors to perivascular macrophages in blood

Perivascular macrophages are uniquely situated at the intersection between the nervous and immune systems. Although combined myeloid marker detection differentiates perivascular from resident brain macrophages (parenchymal microglia), no single marker distinguishes perivascular macrophages in humans and mice. Here, we present the macrophage scavenger receptor CD163 as a marker for perivascular macrophages in humans, monkeys, and mice. CD163 was primarily confined to perivascular macrophages and populations of meningeal and choroid plexus macrophages in normal brains and in brains of humans and monkeys with human immunodeficiency virus or simian immunodeficiency virus (SIV) encephalitis. Scattered microglia in SIV encephalitis lesions and multinucleated giant cells were also CD163 positive. Consistent with prior findings that perivascular macrophages are primary targets of human immunodeficiency virus and SIV, all SIV-infected cells in the brain were CD163 positive. Using fluorescent dyes that definitively and selectively label perivascular macrophages in vivo, we confirmed that dye-labeled simian perivascular macrophages were CD163 positive and able to repopulate the central nervous system within 24 hours. Flow cytometric studies demonstrated a subset of monocytes (CD163(+)CD14(+)CD16(+)) that were immunophenotypically similar to brain perivascular macrophages. These findings recognize CD163(+) blood monocytes/macrophages as a source of brain perivascular macrophages and underscore the utility of this molecule in studying the biology of perivascular macrophages and their precursors in humans, monkeys, and mice.     

Microglial activation in healthy aging

Healthy brain aging is characterized by neuronal loss and decline of cognitive function. Neuronal loss is closely associated with microglial activation and postmortem studies have indeed suggested that activated microglia may be present in the aging brain. Microglial activation can be quantified in vivo using (R)-[(11)C]PK11195 and positron emission tomography. The purpose of this study was to measure specific binding of (R)-[(11)C]PK11195 in healthy subjects over a wide age range. Thirty-five healthy subjects (age range 19-79 years) were included. In all subjects 60-minute dynamic (R)-[(11)C]PK11195 scans were acquired. Specific binding of (R)-[(11)C]PK11195 was calculated using receptor parametric mapping in combination with supervised cluster analysis to extract the reference tissue input function. Increased binding of (R)-[(11)C]PK11195 with aging was found in frontal lobe, anterior and posterior cingulate cortex, medial inferior temporal lobe, insula, hippocampus, entorhinal cortex, thalamus, parietal and occipital lobes, and cerebellum. This indicates that activated microglia appear in several cortical and subcortical areas during healthy aging, suggesting widespread neuronal loss.     

Progressive multifocal leukoencephalopathy and retroviral encephalitis in acquired immunodeficiency syndrome

Antigens of human polyomaviruses, the etiologic agents of progressive multifocal leukoencephalopathy (PML), and of human immunodeficiency virus were localized in paraffin sections from brains of six patients who died with the acquired immunodeficiency syndrome. Immunostaining revealed polyomaviral antigens in oligodendrocytes and in some astrocytes. Human immunodeficiency (retro) virus antigens were immunostained in mononuclear macrophages, glial cells, and vascular endothelial cells. Both viral types were found ultrastructurally. The lesions of PML were more destructive than is usually seen in cases without the acquired immunodeficiency syndrome. The retroviral encephalitis could have occurred before the onset of PML. However, a secondary retroviral encephalitis could have resulted if the monocytes responding to an initial polyomaviral lesion were already infected with human immunodeficiency virus before they differentiated into macrophages.     

Chemokine expression in simian immunodeficiency virus-induced AIDS encephalitis.

The pathogenesis of neurological dysfunction associated with human immunodeficiency (HIV)-1 infection is uncertain. However, the presence of macrophage infiltrates in the central nervous system is a key feature of HIV encephalitis and is correlated with HIV-associated dementia. Moreover, it has been demonstrated that HIV-infected monocyte/macrophages can produce toxic substances that may play a critical role in the development of HIV-associated dementia. However, the exact mechanisms responsible for HIV infection and leukocyte recruitment to the central nervous system remain speculative. Similar to HIV-infected patients, simian immunodeficiency virus (SIV)-infected macaque monkeys develop immunosuppression and acquired immune deficiency syndrome (AIDS)-related inflammatory disorders, including AIDS encephalitis. In this study, we demonstrate that encephalitic brain from SIV-infected animals has elevated immunohistochemical expression of the C-C chemokines, macrophage inflammatory protein-1 alpha and -beta, RANTES, and monocyte chemotactic protein-3, and the C-X-C chemokine interferon-inducible protein-10. These findings suggest that one or all of of these chemokines could be involved in leukocyte recruitment to the brain in SIV-infected macaque monkeys.     

Transfer of neuropathogenic simian immunodeficiency virus with naturally infected microglia.

The central nervous system (CNS) is a target for human immunodeficiency virus infection, and, in individuals with acquired immune deficiency syndrome, this can lead to a devastating dementia. Only certain viral variants appear capable of invading the CNS and infecting microglia and brain macrophages. To determine whether the virus entering the brain may be particularly pathogenic to the CNS, we isolated microglia from the brains of simian immunodeficiency virus-infected rhesus monkeys. Serial transfer of these cells into naive animals indicated that productive simian immunodeficiency virus infection could indeed be transferred. Furthermore, CNS infection occurred within a relatively short time span and was associated with viral gene expression in the brain and pathology characteristic of human immunodeficiency virus encephalitis. While demonstrating that neuropathogenic variants partition into the CNS, our approach will allow the dissection of functional neuropathogenic elements present in these viruses.     

Monocyte adhesion to endothelium in simian immunodeficiency virus-induced AIDS encephalitis is mediated by vascular cell adhesion molecule-1/alpha 4 beta 1 integrin interactions.

Because the mechanisms associated with recruitment of monocytes to brain in AIDS encephalitis are unknown, we used tissues from rhesus monkeys infected with simian immunodeficiency virus (SIV) to examine the relative contributions of various adhesion pathways in mediating monocyte adhesion to endothelium from encephalitic brain. Using a modified Stamper and Woodruff tissue adhesion assay, we found that the human monocytic cell lines, THP-1 and U937, and the B cell line, Ramos, preferentially bound to brain vessels from monkeys with AIDS encephalitis. Using a combined tissue adhesion/immunohistochemistry approach, these cells only bound to vessels expressing vascular cell adhesion molecule-1 (VCAM-1). Furthermore, pretreatment of tissues with antibodies to VCAM-1 or cell lines with antibodies to VLA-4 (CD49d) inhibited adhesion by more than 70%. Intercellular adhesion molecule-1 (ICAM-1)/beta 2 integrin interactions were not significant in mediating cell adhesion to the vasculature in encephalitic simian brain using a cell line (JY) capable of binding rhesus monkey ICAM-1. In addition, selectin-mediated interactions did not significantly contribute to cell binding to encephalitic brain as there was no immunohistochemical expression of E-selectin and P-selectin in either normal or encephalitic brain, nor was there a demonstrable adhesive effect from L-selectin using L-selectin-transfected 300.19 cells on simian encephalitic brain. These results demonstrate that using the tissue adhesion assay, THP-1, U937, and Ramos cells bind to vessels in brain from animals with AIDS encephalitis using VCAM-1/alpha 4 beta 1 integrin interactions and suggest that VCAM-1 and VLA-4 may be integral for monocyte recruitment to the central nervous system during the development of AIDS encephalitis.     

Imaging microglial activation and glucose consumption in a mouse model of Alzheimer's disease

In Alzheimer's disease (AD), persistent microglial activation as sign of chronic neuroinflammation contributes to disease progression. Our study aimed to in vivo visualize and quantify microglial activation in 13- to 15-month-old AD mice using [¹¹C]-(R)-PK11195 and positron emission tomography (PET). We attempted to modulate neuroinflammation by subjecting the animals to an anti-inflammatory treatment with pioglitazone (5-weeks' treatment, 5-week wash-out period). [¹¹C]-(R)-PK11195 distribution volume values in AD mice were significantly higher compared with control mice after the wash-out period at 15 months, which was supported by immunohistochemistry data. However, [¹¹C]-(R)-PK11195 μPET could not demonstrate genotype- or treatment-dependent differences in the 13- to 14-month-old animals, suggesting that microglial activation in AD mice at this age and disease stage is too mild to be detected by this imaging method.     

Proliferating cellular nuclear antigen expression as a marker of perivascular macrophages in simian immunodeficiency virus encephalitis

Brain perivascular macrophages are a major target of simian immunodeficiency virus (SIV) infection in rhesus macaques and HIV infection in humans. Perivascular macrophages are distinct from parenchymal microglia in their location, morphology, expression of myeloid markers, and turnover in the CNS. In contrast to parenchymal microglia, perivascular macrophages are continuously repopulated by blood monocytes, which undergo maturation to macrophages on entering the central nervous system (CNS). We studied differences in monocyte/macrophages in vivo that might account for preferential infection of perivascular macrophages by SIV. In situ hybridization for SIV and proliferating cellular nuclear antigen (PCNA) immunohistochemistry demonstrated that SIV-infected and PCNA-positive cells were predominantly found in perivascular cuffs of viremic animals and in histopathological lesions that characterize SIV encephalitis (SIVE) in animals with AIDS. Multilabel techniques including double-label immunohistochemistry and combined in situ hybridization and immunofluorescence confocal microscopy revealed numerous infected perivascular macrophages that were PCNA-positive. Outside the CNS, SIV-infected, PCNA-expressing macrophage subpopulations were found in the small intestine and lung of animals with AIDS. While PCNA is used as a marker of cell proliferation it is also strongly expressed in non-dividing cells undergoing DNA synthesis and repair. Therefore, more specific markers for cell proliferation including Ki-67, topoisomerase IIalpha, and bromodeoxyuridine (BrdU) incorporation were used which indicated that PCNA-positive cells within SIVE lesions were not proliferating. These observations are consistent with perivascular macrophages as terminally differentiated, non-dividing cells and underscores biological differences that could potentially define mechanisms of preferential, productive infection of perivascular macrophages in the rhesus macaque model of neuroAIDS. These studies suggest that within CNS and non-CNS tissues there exist subpopulations of macrophages that are SIV-infected and express PCNA.     

Elevated vascular cell adhesion molecule-1 in AIDS encephalitis induced by simian immunodeficiency virus.

AIDS encephalitis is a common sequela to HIV-1 infection in humans and simian immunodeficiency virus (SIVmac) infection in macaques. Although lentiviral-infected macrophages comprise parenchymal inflammatory infiltrates in affected brain tissue, the mechanisms responsible for leukocyte trafficking to the central nervous system in AIDS are unknown. In this study, we investigated the expression of various endothelial-derived leukocyte adhesion proteins in SIVmac-induced AIDS encephalitis. Encephalitic brains from SIVmac-infected macaques, but not uninflamed brains from other SIVmac-infected animals, were found to express abundant vascular cell adhesion molecule-1 (VCAM-1) protein on the majority of arteriolar, venular, and capillary endothelial cells. Soluble VCAM-1 concentrations in cerebrospinal fluid (CSF) from encephalitic animals were increased approximately 20-fold above those from animals without AIDS encephalitis. Expression of other endothelial-related adhesion molecules, including E-selectin, P-selectin, and intercellular adhesion molecule-1 (ICAM-1), was not uniformly associated with AIDS encephalitis. Thus, the presence of VCAM-1 in both brain and CSF was uniformly associated with SIVmac-induced disease of the central nervous system, and this expression may, at least in part, influence monocyte and lymphocyte recruitment to the central nervous system during the development of AIDS encephalitis. Moreover, measurement of soluble VCAM-1 in CSF may assist in the clinical assessment of animals or people with AIDS.     

Simian immunodeficiency virus infection of macaque bone marrow macrophages correlates with disease progression in vivo.

The pathogenesis of hematopoietic abnormalities associated with infection of susceptible hosts with either simian immunodeficiency virus (SIV) or human immunodeficiency virus (HIV) is not fully understood. To determine if bone marrow cells are infected with SIV and if the pattern of viral infection is correlated with the severity of disease and abnormalities in hematopoiesis, 23 SIV-infected rhesus monkeys were examined by immunohistochemistry and in situ hybridization. By immunohistochemistry, only four monkeys were positive for SIV core protein p27, while in situ hybridization revealed viral RNA in the bone marrow of 15 monkeys. Simian immunodeficiency virus RNA was consistently expressed in the bone marrow from monkeys with severe lymphoid depletion (11 of 11), but less so in monkeys with follicular hyperplasia (0 of 2) or mild lymphoid depletion (4 of 10). In animals with mild lymphoid depletion, bone marrow cells infected with SIV were mainly mononuclear cells that appeared to be of myelomonocytic lineage. In contrast, monkeys with severe lymphoid depletion had SIV RNA localized to larger mononuclear cells with abundant cytoplasm often located in small lucent areas of the stroma. These SIV RNA-positive mononuclear cells were positive for the macrophage determinant CD68 as demonstrated by immunohistochemistry. Furthermore the stage of simian acquired immune deficiency syndrome, as indicated by lymphoid morphology, and SIV localization in the bone marrow were correlated with the incidence of anemia, bone marrow hyperplasia, and abnormal distribution of macrophages in the bone marrow. These results indicate that, in common with other animal lentiviral infections, the macrophage is a major target of SIV infections in the bone marrow.