Neuropathology 2020; 40, 358–366 doi:10.1111/neup.12660 Original Article The autophagic marker p62 highlights Alzheimer type II astrocytes in metabolic/hepatic encephalopathy Ellen Gelpi,1,2 Jasmin Rahimi,3 Sigrid Klotz,1 Susanne Schmid,1 Gerda Ricken,1 Sara Forcen-Vega,4 Herbert Budka1 and Gabor G. Kovacs1,5,6,7 1 Division of Neuropathology and Neurochemistry, Department of Neurology, Medical University of Vienna, 3 Department of Neurology and Karl Landsteiner Institute for Neuroimmunological and Neurodegenerative Disorders, Danube Hospital, Vienna, Austria, 2Neurological Tissue Bank of the Biobank of Hospital Clinic, August Pi i Sunyer Biomedical Research Institute (IDIBAPS), 4Neurology Department, Germans Trias i Pujol Hospital, Autonomous University of Barcelona, Barcelona, Spain, 5Center for Neurodegenerative Disease Research (CNDR), Institute on Aging and Department of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA, 6Tanz Centre for Research in Neurodegenerative Disease (CRND) and Department of Laboratory Medicine and Pathobiology, University of Toronto and 7Laboratory Medicine Program, University Health Network, Toronto, Ontario, Canada Metabolic/hepatic encephalopathy is neuropathologically regions, usually involving the deep cortical layers, basal gang- characterized by the presence of Alzheimer type II astro- lia, and pontine nuclei.1 It is, therefore, considered to reflect a cytes (AA II) with large and clear nuclear morphology. gliopathy. To date, there is no good immunohistochemical marker to AA II are characterized by a larger nucleus than in resting better identify these cells. Here, we assessed cases of or reactive fibrillary or gemistocytic astrocytes, a clear chroma- hepatic encephalopathy of different etiologies by immuno- tin (Fig. 1G), and scarce cellular processes. They are frequently histochemistry using an anti-p62 antibody. We observed forming pairs: doublets, or even triplets1 (Fig. 1H). However, peripheral or diffuse nuclear staining of variable intensity AA II may be difficult to visualize on sections stained with in AA II in all cases but not in normal controls or reactive hematoxylin and eosin (HE), showing a spectrum of nuclear astrocytes. We conclude that p62 is a useful immunohisto- changes, from slight enlargement and chromatin loosening to a chemical marker for the identification of AA II and may completely clear or empty appearance of the nucleus with a be helpful for the neuropathological diagnosis of meta- well-defined membrane rim and peripheral dot-like condensa- bolic/hepatic encephalopathy in difficult or equivocal tion (Fig. 1I, arrow). There is currently no good marker to spe- cases. cifically identify AA II: they are characteristically not or are poorly stained by glial fibrillary acidic protein (GFAP) immu- Key words: Alzheimer type II astrocytes, astrogliopathy, nohistochemistry, while they can be depicted using anti-S-100 hepatic encephalopathy, metabolic encephalopathy, p62. protein antibodies.1,2 However, S-100 protein is not a specific marker of astrocytes and labels most glioneuronal elements. INTRODUCTION Because the neuropathological diagnosis of metabolic enceph- The presence of so-called Alzheimer type II astrocytes alopathy may be difficult in its early or less severe disease (AA II) in the human brain usually reflects a metabolic distur- stages, the application of a reliable immunohistochemical bance caused by renal or, more frequently, hepatic dysfunc- marker would be helpful to objectively support this diagnosis tion. Indeed, hepatic encephalopathy has been associated with in the routine diagnostic or experimental settings. the presence of AA II in the gray matter of different brain Through staining of postmortem brains for other diagnos- tic purposes, we observed an intense staining of AA II nuclei using p62 immunohistochemistry in patients with metabolic Correspondence: Ellen Gelpi, MD, PhD, EFN, Division of Neuropa- thology and Neurochemistry, Department of Neurology, Medical Uni- encephalopathy. This observation prompted us to systemati- versity of Vienna, Austria, AKH Waehringer Guertel 18-20, Leitstelle cally assess glial p62 immunoreactivity in hepatic encephalop- 4J, 1090 Vienna, Austria. Email: ellen.gelpi@meduniwien.ac.at athy of different etiologies and to compare the staining Received 25 February 2020; revised 20 March 2020; accepted 21 pattern with other conditions characterized by prominent March 2020; published online 02 June 2020. reactive gliosis. © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of Neuropathology. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. p62 in metabolic encephalopathy 359 Fig. 1 Microphotographs of brain sections of hepatic encephalopathy (A-L) and mito- chondrial encephalopathy/Leigh syndrome (M-O). Laminar microvacuolation of the neuropil in cortical layers of the cerebrum is observed in a case of severe hepatic encephalopathy at a low magnification (arrows in A, B) and at a higher magnification (D, E). Abundant p62-positive nuclei are identified in deep layers at a low magnification (arrows in C) and at a higher magnification (F). AA II have the characteris- tic enlarged cell nuclei and clear chromatin (G). AA II forming a triplet of nuclei (H). Small punc- tate condensations along the nuclear membrane in an AA II (I). p62 immunohistochemistry strongly labels the nuclei of AA II, including the peripheral nuclear membrane condensa- tions (arrows in L). (M-O) In a case of Leigh syndrome, glial nuclei in the basal ganglia are also enlarged but show more prominent cytoplasm on HE sta- ined sections than typical AA II (M). p62 shows intense and dif- fuse labeling of enlarged glial nuclei (N, O). HE (A, D, E, G, H, M), LFB-HE (B), p62 immu- nohistochemistry (C, F, J-L, N, O). Scale bars: 200 μm (A–C), 50 μm (N), 20 μm (D–H, J, K–M), 10 μm (I). MATERIALS AND METHODS Formalin-fixed, paraffin-embedded tissue blocks from the Postmortem brains were selected from the archives of the frontal, temporal and occipital cortices, anterior and posterior Institute of Neurology of the Medical University of basal ganglia, thalamus, midbrain, pons and cerebellum were Vienna and the Neurological Tissue Bank of the selected, and 5-μm-thick sections were stained with hematox- IDIBAPS Biobank in Barcelona. The use of brain tissue ylin and eosin (HE) and Luxol fast blue (LFB) and HE (LFB- for research was approved by the respective institutional HE), where the presence of AA II was assessed as present/ ethics committees and conforms to the provisions of the absent (Table 1). One region per case with obvious AA II was Declaration of Helsinki. first stained by immunohistochemistry using a commercial © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of Neuropathology. Table 1 Main features of cases with hepatic encephalopathy of different etiologies and anatomical distribution of suspected AA II 360 Order Age Gender Cause of Metabolic Neuropathological Suspected Alzheimer II astrocytes on HE stained sections hepatic/renal encephalopathy findings Frontal Cingulum Parietal Temporal Occipital Hippocampus BBGG Amygdala Thalamus Midbrain Pons Medulla cbl damage severity HE obl + dentate 1 79 m Hepatitis C and liver Prominent 1) Metabolic + + + + + + + + + + + + + cirrhosis encephalopathy; 2) Incidental LB pathology (Braak 2); 3) PART (Braak II) + mild CAA 2 77 m Hepatitis C Moderate 1) Metabolic + + + + − + + − + − + + + encephalopathy; 2) Alzheimer’s disease neuropathologic changes A3B3C2; 3) Incidental LB pathology olfactory bulb only 3 63 f Hepatitis B, liver Moderate 1) Metabolic + + + + + + + + + + + + + cirrhosis encephalopathy; 2) Acute ischemic stroke 4 85 f Alcohol abuse, chronic Moderate 1) Metabolic 3/+ 3/+ 2/+ 3/+ 1/+ 0/s 2/+ 0/s 1/+ 2/+ 0/+ 0/− 0/+ renal insufficiency encephalopathy, 2) PART (Braak II) + mild CAA; 3) LATE 5 71 f Alcohol abuse, liver Moderate 1) Metabolic + − − − − + + + + + + + + cirrhosis, chronic renal encephalopathy; 2) Mild insufficiency cerebellar atrophy; 3) PART (Braak II) 6 59 f Alcohol abuse, liver Moderate 1) Metabolic + + + + + + + + + + + + + cirrhosis encephalopathy; 2) Acute Wernicke encephalopathy; 3) SVD 7 59 m Alcohol abuse, liver Prominent 1) Metabolic + + + + + + + + + + + + + cirrhosis, encephalopathy hepatocarcinoma 8 65 m Liver cirrhosis Prominent 1) Metabolic + + + + + + + − + + + + + encephalopathy; 2) Morel cortical laminar sclerosis; 3) Focal subarachnoid bleeding 9 70 f Primary biliary Prominent 1) Metabolic + + + + + + + + + + + + + cirrhosis encephalopathy with focal spongy polio- encephalopathy 10 64 f Autoimmune hepatitis, Prominent 1) Metabolic + + + + + + + + + + + + + liver cirrhosis encephalopathy; 2) Posthypoxic, postictal encephalopathy and bilateral hippocampal sclerosis (Continues) Neuropathology. © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of E Gelpi et al. Table 1 (Continued) Order Age Gender Cause of Metabolic Neuropathological Suspected Alzheimer II astrocytes on HE stained sections hepatic/renal encephalopathy findings Frontal Cingulum Parietal Temporal Occipital Hippocampus BBGG Amygdala Thalamus Midbrain Pons Medulla cbl damage severity HE obl + dentate 11 84 f Hepatocellular Mild 1) Metabolic + + + + + + + + + + + + + carcinoma encephalopathy; 2) PART Neuropathology. (Braak IV); 3) Incidental LB pathology (Braak 2) 12 58 m Hepatocellular Prominent 1) Metabolic + + + + + + + + + + + + + carcinoma encephalopathy; 2) AgD Saito I 13 51 m Metastatic unknown Prominent 1) Metabolic + + + + + + + + + + + + + primary tumor with encephalopathy; 2) Pontine p62 in metabolic encephalopathy subtotal liver micro-metastasis carcinoma destruction 14 81 f Metastatic pancreas Moderate 1) Metabolic + + + + + + + + + + + + + carcinoma, liver encephalopathy; 2) Mild necrosis ARP (Braak II, CERAD B) 15 85 m Metastatic prostate Mild 1) Metabolic + − − − + + + + + + + − + carcinoma including encephalopathy; 2) liver Subdural hemorrhage; 3) Acute hypoxic-ischemic neuronal damage in pons and cerebellum 16 78 f Metastatic colon Mild 1) Metabolic + + + − + + + − − + + + + carcinoma including encephalopathy; 2) Mild liver ARP (Braak I, CERAD A) + mild CAA 17 58 f B-cell lymphoma Moderate 1) Metabolic + + + + + + + + + + + + + diffuse, acute renal encephalopathy; 2) and hepatic failure Lymphomatosis meningea 18 68 m Sepsis Moderate 1) Metabolic + − − − + + + + + + + + + encephalopathy; 2) PART (Braak II) 19 52 m Sepsis, Moderate 1) Metabolic + − − − + + + − + + + + + cardiopulmonary encephalopathy; 2) reanimation posthypoxic-postischemic encephalopathy 20 15 f Sepsis, vasculitis Moderate 1) Metabolic + + + + + + + + + + + + + c-ANCA encephalopathy; 2) Multiple microbleeds 21 84 m Sepsis, renal Mild 1) Metabolic + − + + + + + − + − + − + insufficiency encephalopathy; 2) SVD; 3) PART (Braak I) 22 0.5 f Sepsis; pulmonary Prominent 1) Metabolic + − − − − + + − − − + + + transplantation, fungal encephalopathy; 2) Diffuse sepsis, pulmonary gliosis fibrosis, hepato-splenomegalia and hepatic steatosis (Continues) © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of 361 Table 1 (Continued) 362 Order Age Gender Cause of Metabolic Neuropathological Suspected Alzheimer II astrocytes on HE stained sections hepatic/renal encephalopathy findings Frontal Cingulum Parietal Temporal Occipital Hippocampus BBGG Amygdala Thalamus Midbrain Pons Medulla cbl damage severity HE obl + dentate 23 69 m Sepsis, cor pulmonale, Mild 1) Metabolic + + + + + + + + + + + + + hepatic steatosis encephalopathy - Wernicke encephalopathy; 2) Incidental LB pathology (Braak 3); 3) PART (Braak II). 24 7 m Sepsis, cardial Prominent 1) Metabolic + + + + + + + + + + + + + transplantation, fungal encephalopathy; 2) pneumonia and sepsis, Hypoxic-ischemic damage acute liver necrosis with cortical necrosis; 3) Fungal microabscesses 25 64 m Pericardial Mild 1) Metabolic + + + + + + + + + + + + + tamponade, congestive encephalopathy; 2) Mild liver acute hypoxic-ischemic neuronal damage; 3) PART (Braak I) 26 69 f Cardiac and hepatic Mild 1) Metabolic + + + + + + + + + + + + + fibrosis, anemia encephalopathy; 2) perniciosa, colitis Arteriolosclerosis and ulcerosa status cribrosus; 3) PART (Braak II) 27 81 f Cardial insufficiency Moderate 1) Metabolic + + + + + + + + + + + + + encephalopathy; 2) PART (Braak III) 28 1 m Cardial malformation Prominent 1) Metabolic or hypoxic? + − + − + + + − + + + + + with insufficiency and encephalopathy; 2) hepatosplenomegalia malformation: Dandy-Walker like and agenesis of olfactorius 29 74 f Suprarenal Mild 1) Metabolic + + + + + + + − + + + − + insufficiency, encephalopathy with insulinoma, multiple Wernicke-like changes and complications Morel cortical laminar sclerosis; 2) Mild ARP (Braak I, CERAD B) 30 85 f Hepatic insufficiency, Moderate 1) Metabolic + − + − + + + + + + + + + unknown origin encephalopathy; 2) PART (Braak II); 3) Old infarct 31 2 m Mitochondrial Moderate 1) Metabolic-mitochondrial − + + − − + + − + + − + + encephalopathy - encephalopathy consistent Leigh syndrome; with Leigh-syndrome hepatomegaly In case 4, a detailed anatomical mapping of p62 immunoreactivity in relation to the presence of AA II was performed (0, negative; 1, sparse stained glial nuclei; 2, moderate density of sta- ined glial nuclei; 3, high density of stained nuclei. s, single; +, present; −, absent; n.a, not available. ARP, Alzheimer’s disease-related pathology; CAA, amyloid angiopathy; LATE, limbic age-related TDP43 encephalopathy; LB, Lewy body; PART, primary age-related tauopathy; SVD, small vessel disease. Neuropathology. © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of E Gelpi et al. p62 in metabolic encephalopathy 363 monoclonal anti-p62 antibody (clone 3/p62 ligand, dilution protein 43 kDa (TDP-43), and Creutzfeldt–Jakob disease 1:500; BD-Transduction Laboratories, Franklin Lakes, NJ, (one case each), as well as one normal brain. USA). Then, selected cases of hepatic diseases (Table 2) were Details of cases with hepatic/metabolic encephalopathy immunostained for p62 in the frontal cortex, basal ganglia, and are shown in Table 1. pons, and in one case, a detailed mapping of p62 distribution was performed (Table 1). Antigen retrieval was performed by RESULTS boiling the sections in citrate buffer at pH 6.0 for 20 min. The Nuclear p62 staining was detected in enlarged glial cells of immunoreaction was visualized by the polymer-immuno- the gray matter that were consistent with AA II on HE- complex method using an Envision System kit (Dako, Glo- stained sections (Fig. 1J–L). This immunopositivity was strup, Denmark), and 3,3’-diaminobenzidine was used as observed in all cases with hepatic encephalopathy of dif- chromogen. For double immunofluorescence labeling, the ferent etiologies, except for some of septic origin. Nuclear anti-p62 antibody was combined with antibodies against S-100 staining for p62 in AA II was particularly intense in a case protein (rabbit polyclonal, dilution 1:2000; Dako), GFAP (rab- of mitochondrial encephalopathy (Leigh syndrome) bit polyclonal, dilution 1:5000; Dako), and tubulin polymeriza- (Fig. 1M–O). Double immunofluorescence revealed tion-promoting protein (TPPP/p25 (rabbit polyclonal, dilution p62-positive nuclei in some delicate GFAP-positive 1:2000; non-commercial). It has been shown that TPPP is (Fig. 2A) and diffuse S-100 protein-positive cells (Fig. 2B) mainly expressed in differentiated oligodendrocytes of the but not in TPPP/p25-positive oligodendrocytes (Fig. 2C), central nervous system (CNS),3 After blocking of thus supporting the astrocytic nature of the cells. autofluorescence with Sudan Black B, antibody binding immu- In cortical areas, AA II were best identified in deep layers. noreactivities was visualized with secondary antibodies such as anti-mouse IgG conjugated with Alexa Fluor488 (Thermo In severely affected cases, laminar microvacuolation of the Fisher Scientific, Waltham, MA, USA) at a dilution of 1:800 neuropil in deep layers could be observed at a low magnifica- and anti-rabbit IgG conjugated with Cy3 (Thermo Fisher Sci- tion (arrows in Fig. 1A, B) and at a higher magnification (Fig. entific) at a dilution of 1:1000. 1D, E). Here, abundant p62-positive nuclei were identified at We selected 31 cases with hepatic encephalopathy of viral a low magnification (arrows in Fig. 1C) and at a higher magni- (Hepatitis C), neoplastic (hepatocarcinoma and liver metas- fication (Fig. 1F). When AA II showed the characteristic tasis), alcoholic (liver cirrhosis), systemic (sepsis), and mito- enlarged nuclei with clear chromatin, immunoreactivity was chondrial (Leigh syndrome) origins (Table 1). For additional enhanced along the nuclear membrane and in the small punc- comparison of the immunostaining pattern, we assessed tate condensations (Fig. 1L). In cells with less obvious nuclear different pathologies, including subacute stage of cerebral change, immunoreactivity was more diffuse. infarction with prominent reactive gliosis, as well as different The distribution and intensity of p62 immunoreactivity neurodegenerative diseases with variable degrees of chronic in AA II nuclei was not homogeneous among different reactive gliosis, including Alzheimer’s disease, corticobasal brain areas of the same patient and between patients. The degeneration, progressive supranuclear palsy, Parkinson’s strongest signal was generally observed in cortical areas disease, frontotemporal lobar degeneration with inclusions and was lower in the basal ganglia and pontine nuclei, but immunoreactive for transactivation response DNA-binding this was not uniform (Table 2). There were cases Table 2 p62 Immunoreactivity in the frontal cortex, basal ganglia, and pons in selected patients with different ages, etiologies, and for- malin fixation times Order Age Weeks in Hepatic pathology Glial p62 nuclear immunoreactivity formalin Frontal cortex Basal ganglia Pons 1 79 1 Hepatitis C/cirrhosis 0/+ 1/+ 0/s 4 85 1 Alcohol abuse/cirrhosis 3/+ 3/+ 0/s 22 0,5 2 Hepatic steatosis 1/+ 2/+ 1/+ 20 15 2 Sepsis, vasculitis 2/+ 0/+ 2/+ 19 52 3 Sepsis 0/+ 0/+ 0/+ 13 51 4 Metastasis 2/+ 2/+ 0/+ 31 2 4 Mitochondrial disorder/Leigh syndrome 0/− 3/+ 2/+ 14 81 >4 Metastasis 3/+ 3/+ 0/+ 7 59 6 Alcohol abuse/cirrhosis/hepatocellular carcinoma 3/+ 3/+ 2/+ 11 84 6 Hepatocellular carcinoma 1/+ 0/+ 0/+ 17 58 14 Acute hepatic and renal failure, B cell lymphoma 0/+ 0/+ 0/+ 28 1 > 25 Cardiac malformation, hepatosplenomegaly 3/+ 0/+ 0/+ 21 84 n.a. Sepsis 0/+ 0/+ 0/+ © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of Neuropathology. 364 E Gelpi et al. Fig. 2 Microscopic findings of double immunofluorescence staining for p62 (green signal in A-C) with GFAP (red signal in A), S-100 protein (red signal in B)m and TPPP (red signal in C). p62 immunoreactivity is localized in some cells with delicate GFAP-positive branching pro- cesses (A) and in diffusely sta- ined S-100 protein-positive branching processes (B). In con- trast, they do not coincide with TPPP-positive oligodendro- cytes (C). (e.g. case 4) showing a patchy distribution of immunoreac- plaques and large reactive astrocytes, respectively, were tivity in the basal ganglia and cerebral cortex. In rare moderately labeled. cases, cells that were considered to be AA II on This observation suggests p62 as a very useful neuro- HE-stained sections were not or were only faintly immu- pathological marker of metabolic gliosis, particularly in noreactive for p62. Inversely, some cases with p62-positive hepatic encephalopathy. nuclei were not always clearly identifiable as AA II: for Immunohistochemistry using anti-p62 antibodies has example, in the case of Leigh syndrome (Fig.1M–O) proved very useful in the study of neurodegenerative dis- where immunopositivity filled the whole nucleus, in con- eases, as it is commonly found in neuronal cytoplasmic or trast to other cases with peripheral nuclear immuno- nuclear inclusions (e.g. Alzheimer’s disease, frontotemporal staining. Bergmann’s glia also showed relatively lobar degenerations, Lewy body diseases, or trinucleotide prominent p62 nuclear staining in one case of metabolic repeat disorders such as Huntington’s disease). The presence encephalopathy that had no prominent Purkinje cell loss of p62 is also a useful predictor of C9orf72 expansion muta- (case 30). In contrast, other pathologies associated with tion when accumulated in granular neurons of the cerebellar Bergmann gliosis remained negative. Aquaporin 4 immu- cortex or hippocampal neurons.4,5 noreactivity was undetectable in AA II. p62 or sequestosome-1 is a protein encoded by No immunostaining in glial nuclei was observed in the SQSTM1 and is thought to target protein aggregates for control and most neurodegenerative conditions with reac- lysosomal degradation, by binding to ubiquitinated pro- tive astrogliosis, except for one case of corticobasal degener- teins, among other functions.6–8 It is, therefore, considered ation. In that case, p62-positive nuclei corresponded to to be an indicator of autophagic degradative activity. p62 those of tau-positive astrocytic plaques on adjacent tissue itself is also degraded by autophagy. When autophagy is sections. Moreover, one case of sucacute stage cerebral induced, it remains at low levels in the cell, while it accu- infarction showed moderate nuclear immunoreactivity of mulates when autophagy is deficient. It is also involved in large reactive astrocytes and also of some “eosinophilic neu- protein aggregation, as shown for several proteinopathies rons.” There were no differences in staining intensity that associated with neurodegenerative conditions.9 could be related to fixation time (Table 2) or postmortem Hepatic/metabolic encephalopathy has been reported to delay (data not shown). No data on ammonia levels were underlie several complex metabolic alterations,10 including available. mitochondrial dysfunction in astrocytes due to increased ammonia levels in blood and in the brain,11,12 among others. Moreover, experimental studies have shown an involvement DISCUSSION of mitophagy and autophagy13 in the pathogenesis of We assessed the immunohistochemical expression of p62 hepatic encephalopathy. In particular, treatment of cultured in glial cells in different brain diseases and observed prom- rat astrocytes with low concentrations of NH4Cl induced inent nuclear staining of AA II in metabolic/hepatic autophagy, while with higher concentrations from 2 mM encephalopathy, including astrocytes with no typical onwards, NH4Cl inhibited autophagy in astrocytes in a time- “clear” morphology. This was not observed in reactive and dose-dependent manner.12 These findings may provide astrocytes of most chronic neurodegenerative diseases, one explanation for why high ammonia levels can induce the except for one case of corticobasal degeneration and sub- accumulation of p62 through inhibition of autophagy. In acute stage cerebral infarction, where nuclei of astrocytic addition, exposure of astrocytes to ammonia also induces © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of Neuropathology. p62 in metabolic encephalopathy 365 astrocytic swelling, which can be exacerbated by cytokines/ REFERENCES inflammatory mediators.14 Some experimental studies have 1. Kril J, Chimelli L, Morris CM, Harris JB. Nutritional and also shown that increased plasma membrane aquaporin toxic diseases. In: Love S, Budka H, Ironside JW, 4 levels contribute to the astrocytic swelling/brain edema in Perry A, (eds). Greenfield’s Neuropathology, 9th edn, hepatic encephalopathy.10 We found no increased immuno- vol. 1. Chapter 9. Boca Raton, FL: CRC Press, 2015; 596. reactivity for aquaporin 4 in AA II. However, the detailed 2. Kimura T, Budka H. Glial fibrillary acidic protein mechanism of peripheral and diffuse nuclear staining for (GFAP) and S-100 protein in human hepatic encepha- p62 in hepatic/metabolic encephalopathy remains to be lopathy: immunocytochemical demonstration of disso- elucidated. ciation of two glia-associated proteins. Acta We observed a somewhat uneven distribution of Neuropathol 1986; 70: 17–21. p62-positive AA II within the same brain area and 3. Lehotzky A, Lau P, Tokési N, Muja N, Hudson LD, between different brain areas and cases. It could be postu- Ovádi J. Tubulin polymerization-promoting protein lated that this might be related to levels of ammonia (TPPP/p25) is critical for oligodendrocyte differentia- and/or duration or even to the cause of hepatic damage, tion. Glia 2010; 58: 157–168. reflecting an evolutive process of metabolic alterations of 4. Al-Sarraj S, King A, Troakes C et al. p62 positive, astrocytes. While it is generally considered that ammonia TDP-43 negative, neuronal cytoplasmic and intra- levels are positively related to the severity of hepatic nuclear inclusions in the cerebellum and hippocampus encephalopathy, they are not always determinant as other define the pathology of C9orf72-linked FTLD and factors may exacerbate it,10,15 and they do not necessarily MND/ALS. Acta Neuropathol 2011; 122: 691–702. influence patient management.16 Unfortunately, we do not  5. Ramos-Campoy O, Avila-Polo R, Grau-Rivera O have enough data on ammonia levels or details on the et al. Systematic screening of ubiquitin/p62 aggregates duration of hepatic disease. Moreover, there was no par- in cerebellar cortex expands the neuropathological ticular difference in staining intensity depending on the phenotype of the C9orf72 expansion mutation. etiology of liver damage, and it was apparently also not J Neuropathol Exp Neurol 2018; 77: 703–709. influenced by postmortem delay or formalin fixation time. 6. Cuyvers E, van der Zee J, Bettens K et al. Genetic In summary, the postmortem neuropathological diagno- variability in SQSTM1 and risk of early-onset sis of metabolic/hepatic encephalopathy has been a some- Alzheimer dementia: a European early-onset demen- what subjective, not always unequivocal diagnosis and can tia consortium study. Neurobiol Aging 2015; 36: 2005. represent a challenge, particularly in less obvious stages. e15–2005.e22. Even if not absolutely specific, we consider p62 as a useful 7. Seibenhener ML, Babu JR, Geetha T, Wong HC, immunohistochemical marker to visualize AA II. It can Krishna NR, Wooten MW. Sequestosome 1/p62 is a improve and objectify the identification of metabolic polyubiquitin chain binding protein involved in encephalopathy/gliopathy in postmortem brain tissue, in ubiquitin proteasome degradation. Mol Cell Biol 2004; supplementation of classical HE staining features. Why 24: 8055–8068. p62 accumulates in the nucleus is, however, still unclear 8. Bjørkøy G, Lamark T, Johansen T. p62/SQSTM1: a and deserves further investigation, particularly to better missing link between protein aggregates and the understand metabolic disturbances of astrocytes and their autophagy machinery. Autophagy 2006; 2: 138–139. relationship with autophagy. 9. Rué L, López-Soop G, Gelpi E, Martínez-Vicente M, Alberch J, Pérez-Navarro E. Brain region- and age- ACKNOWLEDGMENTS dependent dysregulation of p62 and NBR1 in a mouse model of Huntington’s disease. Neurobiol Dis 2013; We thank the brain donors and their families for generous 52: 219–228. brain donation for diagnostic and research purposes. We 10. Jayakumar AR, Norenberg MD. Hyperammonemia in are also indebted to the Neurological Tissue Bank of the hepatic encephalopathy. J Clin Exp Hepatol 2018; 8: IDIBAPS Biobank in Barcelona for sample and data pro- 272–280. curement. We thank Sara Charif, Veronica Santiago, and 11. Bai Y, Wang Y, Yang Y. Hepatic encephalopathy Judith Ludwig for technical assistance. changes mitochondrial dynamics and autophagy in the substantia nigra. Metab Brain Dis 2018; 33: DISCLOSURE 1669–1678. The authors received no financial support for this work 12. Lu KH. Involvement of autophagy and mitophagy in the and have no relationships that may pose a conflict of pathogenesis of hepatic encephalopathy (Doctoral The- interest. sis). 2018. Available from URL: https://docserv.uni- © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of Neuropathology. 366 E Gelpi et al. duesseldorf.de/servlets/DerivateServlet/Derivate-49390/ cultured astrocytes. J Neuroinflammation 2010; Lu%20Kaihui%20Thesis%202018%20pdfa-b1.pdf 7: 66. (Accessed January 30, 2020) 15. Reuter B, Walter K, Bissonnette J et al. Assessment 13. Soria LR, Brunetti-Pierri N. Ammonia and autophagy: of the spectrum of hepatic encephalopathy: a multi- an emerging relationship with implications for disor- center study. Liver Transpl 2018; 24: 587–594. ders with hyperammonemia. J Inherit Metab Dis 2019; 16. Haj M, Rockey DC. Ammonia levels do not guide 42: 1097–1104. clinical management of patients with hepatic encepha- 14. Rama Rao KV, Jayakumar AR, Tong X, lopathy caused by cirrhosis. Am J Gastroenterol 2019 Alvarez VM, Norenberg MD. Marked potentiation (Oct 14). [Epub ahead of print]. https://doi.org/10. of cell swelling by cytokines in ammonia- sensitized 14309/ajg.0000000000000343 © 2020 The Authors. Neuropathology published by John Wiley & Sons Australia, Ltd on behalf of Japanese Society of Neuropathology.