Showing papers by "Oswaldo Cruz Foundation published in 2023"

Chronic Mercury Exposure and GSTP1 Polymorphism in Munduruku Indigenous from Brazilian Amazon

Abstract: Genetic polymorphisms may be involved with mercury levels and signs and symptoms of intoxication from this exposure. Therefore, the aims were to describe the frequency of the GSTP1 polymorphism and to evaluate its effects on mercury levels and neurological signs in three Munduruku indigenous villages in the Brazilian Amazon. One-hundred-and-seven indigenous (over 12 years old) were included and genotyped (rs1695) using a TaqMan validated assay. Then, associations were evaluated by binary logistic regression, using odds ratios (OR) and 95% confidence intervals (CI). Mean age was 27.4 ± 13.9 years old, 52.3% were male, mean hair mercury concentration was 8.5 ± 4.3, exceeding the reference limit (≥6.0 µg/g), and were different among the three villages: 13.5 ± 4.6 µg/g in Sawré Aboy, 7.4 ± 2.3 µg/g in Poxo Muybu and 6.9 ± 3.5 µg/g in Sawré Muybu. The minor allele frequency of GSTP1 G was significantly different among the villages: 57% Sawré Muybu, 21% Poxo Muybu and 15% Sawré Aboy. Finally, after adjustment, GSTP1 GG and GA genotypes were associated with lower levels of Hg (OR = 0.13; CI95% = 0.03–0.49) and abnormal somatosensory signs (OR = 3.7; 95%IC = 1.5–9.3), respectively. In conclusion, monitoring this population is imperative to identify individuals at higher risk of developing signs of chronic mercury exposure based on the genetic profile.

When she looked at the pandemic: one photograph, countless movies

Abstract: Abstract The sensory and affective impact of a photograph in a news story may mitigate the message of the headline and even the content of the story as a whole, with the potential to lead the reader to countless parallel stories. These stories can be imbricated to the theme explored by the media, but can also expand to several other issues, memories, events, authors unveiling to the reader other situational layers that transcend the sphere of simple reading leading to reflexivity. This text has, therefore, the intention to reflect at length about the inequities in health, especially in relation to the distribution of vaccines against Covid-19 on the planet, from an image published by France Presse in a story about a day with record application of doses of the vaccine against Covid-19 in India.

Excessive Consumption of Alcoholic Beverages and Extremely High Levels of High-Density Lipoprotein Cholesterol (HALP) in the ELSA-Brasil Cohort Baseline

Abstract: Background: It has already been established that the consumption of alcoholic beverages increases high-density lipoprotein cholesterol (HDL-C) levels in dose–response. Methods and Results:A cross-sectional analysis was carried out with 6132 participants of both sexes aged between 35 and 74 years, who were active and retired workers from six Brazilian states. Heavy drinkers were categorized by sex: men > 210 g/week and women > 140 g/week; moderate drinkers: men ≤ 209 g/week and women ≤ 139 g/week. The HDL-C level was dichotomized into normal (40 mg/dL–82.9 mg/dL) and extremely high (≥83 mg/dL). We used binary logistic regression to assess associations between baseline alcohol intake and HDL-C, which were adjusted for sex, age, income, physical activity, kilocalories and body mass index (BMI), and we found an positive association between extremely high HDL-C and the excessive consumption of alcoholic beverages. These participants were mostly women with a high income, lower waist circumference, kilocalorie consumption and also a higher consumption in all categories of alcoholic beverages. Conclusion: Excessive alcohol consumption was associated with a higher probability of extremely high HDL-C.

Multiplexed snRNA-seq from frozen human brain samples v2

Abstract: This is a protocol to save your experiment and successfully demultiplex nuclei using CellPlex, the cholesterol-modified-oligo (CMO) barcodes from 10x Genomics on your nuclei from snap-frozen tissue. Generally speaking, nuclei from snap-frozen tissue perform really poorly with CMOs. This crucial information was not in their initial documentation when the product was launched, leading to costly losses. However, if you are multiplexing samples from different human donors, there is a way of saving your data and still demultiplexing your samples quickly and reliably. By taking advantage of known SNPs in the human genome, it is possible to demultiplex samples originating from different donors using freemuxlet.

Toward Real-World Computational Nephropathology

Abstract: In pathology, defining consistent criteria for disease diagnosis and prognostication requires ever increasing effort. Nephropathology, in particular, demands greater attention to the correspondence between distinctive patterns of kidney lesions and their associated clinical diagnoses. Some examples include changes in the glomeruar basement membrane and nephrotic syndrome, endocapillary hypercellularity and nephritic syndrome, and extensive glomerular crescents and renal dysfunction. Although consensus has been achieved for diagnostic criteria and assessing the activity and chronicity of prevalent nephropathies, there remains a lack of substantial agreement in the classification of certain kidney lesions.1 Some diagnoses require complex procedures to perform lesion scoring, which are labor intensive, time-consuming, and error prone. Automating the steps involved in diagnosis will represent an important achievement in the future of nephropathology. To attain this goal, digital pathology combined with machine intelligence evolves toward automating diagnostic processes in nephropathology. Recent methods exploit the fundamentals of computational analysis to classify, detect, and segment histological structures in digital whole slide images (WSIs) using visual pattern recognition. The availability of artificial intelligence (AI) systems capable of counting histological structures and scoring disease activity and chronicity in WSI would affect nephropathology practice. Although many machine learning (ML)–based methods have achieved promising results in academic research, evidence needs to be provided regarding applying these methods in the real world.2 It is due to differences between the settings of experiments conducted under controlled academic conditions versus the wilderness of clinical scenarios.3 Such differences may become exacerbated when considering the limitations and methodological restraints of controlled studies in computational pathology. These shortcomings have motivated recent proposals detailing general requirements for ML applications in medical imaging.4 Nevertheless, significant progress must be made regarding improving systems capable of accurate performance in a clinical environment. Despite the importance of incremental advances, most experimental validation continues to be performed using data with limited representativeness, and optimistic validation procedures result in poor generalization assessments and unreliable outcomes. Although these issues restrict progress, improvements in ML methods and computational infrastructure have enabled conducting of more robust experimental research aimed at performing in a real-world context. Because these works are prone to multiple biases (e.g., observer, recall, and sampling), efforts are needed toward building unbiased histological datasets representative of populations. Recently, National Institutes of Health (NIH) has established several consortia projects for hosting well-curated renal tissue images and omics data to be used by the community. Notable ones are the Human Biomolecular Atlas Program and the Kidney Precision Medicine Project. Other NIH consortiums with data for computational pathology studies are NEPTUNE, CureGN, and GUDMAP. This context prompts reflection on the status quo of computational nephropathology (CNP) to shape future research designed to produce intelligent tools suitable for real-world tasks. Although some works5 broadly overview the current state of the CNP domain, this article addresses the practical experimentation challenges and directions for this field. Figure 1 presents a perspective on managing critical aspects to better conduct academic research in CNP.Figure 1: The perspective of a real-world computational nephropathology (CNP) includes key aspects that must be improved to better conduct academic research. Specifically, future work must consider the following: (1) the gathering and curation of representative labeled and unlabeled data from multiple laboratories (with QC/QA) as well as the training of generalizable ML models even in cases with few data samples; (2) the adoption of validation methodologies on the basis of rigorous cross-validation protocols and adequate metrics, enriched with cross-laboratory, cross-equipment, and cross-staining experiments and the exploitation of interactions between nephropathologists and the computational tools (human-in-the-loop); (3) the use of complementary qualitative and quantitative analysis, thorough statistical significance assessments, and powerful explainability resources; (4) be grounded on robust experimental designs to enable more reliable and context-aware conclusions; and (5) adhering to technical and scientific reporting guidelines as well as the sharing of data, source code, and pretrained models following privacy, safety, and ethical standards. Notably, avoiding biases is a transversal concern in CNP studies and must be considered throughout the workflow. All that should be empowered by the next-generation workforce, which must better understand the clinical needs. ML, machine learning; QA, quality assessment; QC, quality control.Data WSI constitutes the core data source for ML in CNP. Although CNP experimentation must consider a plethora of lesion presentations in tissue samples and variations in the slide preparation and digitalization (fixing, embedding, slicing, staining, and scanning), most assessments of CNP methods are performed in narrowly controlled scenarios, involving small datasets that do not reflect real-world data variability and heterogeneity. These limitations hinder the development of accurate and practical computational models. Therefore, future CNP research should consider data from multiple laboratories6 and seek to represent wide-ranging kidney lesion presentations. To facilitate such data generation, modern auxiliary computing tools can aid in gathering, curation, labeling, augmentation, and performing quality control and quality assessment in larger and varied data collections. In addition, it is necessary to research generalizable ML methods capable of learning from relatively few labeled samples and/or larger unlabeled sets that are easier to obtain. Validation and Analysis Using large volumes of data with robust representativeness, complex ML models can be trained, thanks to modern high-performance computing infrastructure. The association between big data and computational power can leverage quantitative evaluations on the basis of rigorous cross-validation protocols, adequate metrics, and thorough statistical analysis to yield reliable generalization assessments.7 The entire validation process can also be enriched with cross-laboratory, cross-equipment, and cross-staining studies. In addition, visual error analysis is expected to improve discussions on CNP experimentation by providing qualitative information. Along with clear descriptions of experimental limitations, meticulous qualitative and quantitative analysis will allow for deeper assessments of proposed methods and a clearer understanding of their capabilities, shortcomings, and potential to evolve. Moreover, interactions between nephropathologists and computational tools can guide the evolution of ML models through human-in-the-loop supervision (annotation and feedback) and model training while at the same time calling attention to improving explainability—a key step to achieving clinical translatability. Reporting and Sharing The evolution toward real-world–oriented CNP encompasses standardized practices to ensure proper peer review, reproducibility, replicability, and repeatability. Adopting technical and scientific reporting guidelines can bring strategic advances to the field. Many ongoing initiatives are developing reporting guidelines for AI in health care, such as STARD-AI8 and TRIPOD-AI,9 which focus on the preclinical and offline validation of predictive models for diagnosis and prognosis. With respect to sharing, although gaps remain in the standardization of digital nephrology image data,10 advances in CNP methodology can be powerfully leveraged if researchers make data, source code, and pretrained models available. Using transfer learning through pretrained deep-learning models allows for knowledge sharing and accelerates system development while avoiding sensitive data reveal. Sharing these elements must follow privacy, safety, and ethical standards to protect patient anonymity, guarantee controlled access, and adhere to fair research principles. Reliable Conclusions Considering the effects of medical imaging analysis on the practice of nephropathology, it is important for preclinical studies to be grounded by robust experimental design. Otherwise, it might bias conclusions because of the lack of representative supporting data, proper validation methodology, and complementary quantitative and qualitative analysis.3 In other words, real-world applications demand substantial datasets properly representing visual phenomena. It is also necessary for a well-defined experimental methodology that allows for comprehensive performance assessments, strict statistical validations that confirm the performance of the methods, and clear reporting of all research performed. These elements enable studies to draw more reliable and clear context-aware conclusions. Concluding Remarks Parallel advancements in photonics have enabled the identification of molecular markers in digital histology images, vertebrate codex using the gene breaking protein trap library, and deep genomics information using spatial transcriptomics. Another milestone on the future roadmap of using AI as an assistive tool for clinical diagnosis will be the capability of inferring subcellular diversity abundance using specific molecular motifs on the basis of digital histology image data alone. Integrating the workforce from the engineering and clinical domains is of critical importance so that these groups can exchange ideas more effectively and not work in isolation. This process will empower the next-generation workforce to understand clinical needs better while also using high-performance computational tools to address specific requirements. Finally, advances in ML-based CNP prompt a debate on the plausibility of digitally assisted diagnosis in clinical practice. The reliability of AI systems remains a concern,11 and to gain trust, we suggest deep diving into the dimensions discussed here regarding representative real-world data curation, rigorous validation protocols, critical and enriched analysis, meticulous reporting, and broader sharing to ultimately ground experimental designs and support reliable conclusions.

Evaluating the Performance of the DPP ZDC IgM/IgG Rapid Test for Chikungunya Virus Diagnosis in Febrile Outpatients: A Prospective, Diagnostic Accuracy Study

Abstract: Objective: Evaluate the performance of a novel antibody-based rapid diagnostic test (RDT) for detecting Chikungunya virus (CHIKV) infection in febrile patients in Rio de Janeiro, Brazil. Methods: We prospectively enrolled non-severe febrile patients aged 2-65 presenting as outpatients between October 2018 and July 2019. Serum samples were collected during acute and convalescent phases, tested for CHIKV antibodies using the DPP ZDC IgM/IgG rapid test, and compared against the reference test, CHIKV RT-PCR. We determined the seropositivity using ELISA IgM/IgG and evaluated the diagnostic performance of the WHO-endorsed CHIKV clinical definition against the reference test. Results: Of 500 participants, 226/261 (86.5%) tested ELISA IgM positive, 45/271 (16.6%) tested ELISA IgG positive, 100/294 (34%) CHIKV RT-PCR positive, and 117/495 (23.6%) RDT-antibody positive. During the acute phase [median 3 (2-4) days post illness onset], the sensitivity of IgM, IgG, and combined IgM/IgG ranged from 14.71-34.85%, while specificity ranged from 63.32-65.61%. During the convalescent phase [mean 16.5 (+5.5) days post-illness onset], sensitivity increased from 65.75% to 77.78%, and specificity ranged from 93.33 to 98.11%. The WHO's CHIKV clinical definition had a sensitivity, specificity, positive predictive value, and negative predictive value of 88 (79.9-93.6)%, 74 (68-80)%, 64.2 (58.2-69.8)%, and 92.3 (87.6-95.3)%, respectively. Conclusions: The DPP ZDC IgM/IgG accurately diagnosed CHIKV on samples collected during the convalescent phase. Field applications include investigating CHIKV in patients with sub-acute to chronic osteoarticular symptoms and conducting serosurveys to inform priority areas for CHIKV vaccine implementation. The WHO's clinical definition of CHIKV was accurate and could be deployed, especially in regions with limited diagnostic capacity.

Targeting temperature-sensitive transient receptor potential channels in hypertension: far beyond the perception of hot and cold

Abstract: Transient receptor potential (TRP) channels are nonselective cation channels and participate in various physiological roles. Thus, changes in TRP channel function or expression have been linked to several disorders. Among the many TRP channel subtypes, the TRP ankyrin type 1 (TRPA1), TRP melastatin type 8 (TRPM8), and TRP vanilloid type 1 (TRPV1) channels are temperature-sensitive and recognized as thermo-TRPs, which are expressed in the primary afferent nerve. Thermal stimuli are converted into neuronal activity. Several studies have described the expression of TRPA1, TRPM8, and TRPV1 in the cardiovascular system, where these channels can modulate physiological and pathological conditions, including hypertension. This review provides a complete understanding of the functional role of the opposing thermo-receptors TRPA1/TRPM8/TRPV1 in hypertension and a more comprehensive appreciation of TRPA1/TRPM8/TRPV1-dependent mechanisms involved in hypertension. These channels varied activation and inactivation have revealed a signaling pathway that may lead to innovative future treatment options for hypertension and correlated vascular diseases.

<i>In vitro</i> and <i>in silico</i> analysis of imatinib analogues as anti-<i>Trypanosoma cruzi</i> drug candidates

Abstract: Abstract Chagas disease (CD) is a neglected tropical disease caused by the intracellular protozoan Trypanosoma cruzi that remains a serious public health issue affecting more than 6 million people worldwide. The available treatment includes 2 nitro derivatives, benznidazole (BZ) and nifurtimox, that lack in efficacy in the later chronic phase and when administered against the several naturally resistant parasite strains and present several side-effects, demanding new therapeutic options. One strategy is based on repurposing by testing drugs already used for other illness that may share similar targets. In this context, our previous data on imatinib (IMB) and derivatives motivated the screening of 8 new IMB analogues. Our findings showed that all except 1 were active against bloodstream trypomastigotes reaching drug concentration capable of inducing a 50% of parasite lysis (EC 50 ) values < 12 μ m after 2 h while BZ was inactive. After 24 h, all derivatives were more potent than BZ, exhibiting EC 50 values 1.5–5.5 times lower. Against intracellular forms, 7 out of 8 derivatives presented high activity, with EC 50 values ≤ BZ. LS2/89 stood out as one of the most promising, reaching EC 90 values of 1.68 and 4.9 μ m on intracellular and trypomastigote forms, respectively, with the best selectivity index (>60) towards the proliferative forms. Physicochemical parameters as well as the absorption, distribution, metabolism, excretion and toxicity properties were predicted to be acceptable and with good chance of a favourable oral bioavailability. The promising results motivate further studies such as in vivo and combinatory assays aiming to contribute for a novel safer and effective therapy for CD.

Neuroimmune interactions: From bench to bedside

Abstract: Neuroimmune interactions comprise the intricate crosstalk that exists between the brain and the immune system, which is only possible because they share a common language. As a result, body homeostasis and proper physiological functions are maintained. Interestingly, when this dialog malfunctions, many diseases emerge, such as neuroinflammatory and neuropsychiatric disorders. However, the knowledge of neuroimmune interactions that take place throughout the body is key to the development of new therapies to treat the aforementioned diseases. Therefore, in this chapter, we discuss in detail how the brain communicates with the immune system and vice versa. Furthermore, we focus on the translational aspects of this crosstalk and how they can be useful in the design of new therapies. Harnessing these interactions is key to understand disease pathophysiology and treatment.

Hepatitis E Virus Research in Brazil: Looking Back and Forwards

Abstract: Hepatitis E virus (HEV) has emerged as a public health concern in Brazil. From the first identification and characterization of porcine and human HEV-3 strains in the 2000s, new HEV subtypes have been identified from animal, human, and environmental isolates. As new potential animal reservoirs have emerged, there is a need to compile evidence on the zoonotic dissemination of the virus in animal hosts and the environment. The increasing amount of seroprevalence data on sampled and randomly selected populations must be systematically retrieved, interpreted, and considered under the One Health concept. This review focused on HEV seroprevalence data in distinct animal reservoirs and human populations reported in the last two decades. Furthermore, the expertise with experimental infection models using non-human primates may provide new insights into HEV pathogenesis, prevention, and environmental surveillance.

Quando ela olhou a pandemia: uma fotografia, incontáveis filmes

Abstract: Resumo O impacto sensorial e afetivo de uma fotografia em uma matéria jornalística pode mitigar a mensagem da lide e até mesmo o conteúdo da reportagem como um todo, com potencial para conduzir o leitor a um sem-número de histórias paralelas. Essas histórias podem se imbricar no tema explorado pela mídia, mas também podem se expandir para várias outras questões, memórias, eventos, autores, desvelando para o leitor outras camadas situacionais que transcendem a esfera da simples leitura conduzindo à reflexividade. Este texto tem, portanto, a pretensão de refletir detidamente acerca das iniquidades em saúde, em especial em relação à distribuição de vacinas contra a Covid-19 no planeta, por meio de uma imagem publicada pela France Presse em uma matéria sobre um dia com recorde de aplicação de doses dessa vacina na Índia.

Nuclei isolation from frozen human brain samples forsnRNA-seq v2

Abstract: Protocol to isolate nuclei from snap-frozen human brain samples for sn-RNAseq

“Social malaise... demonic agitation:” anarchism according to the criminology of the French physician Alexandre Lacassagne

Abstract: Abstract This article analyzes the way anarchism and its followers were understood in L’assassinat du président Carnot, by the French physician Alexandre Lacassagne. A few months before the book was published, in June 1894, the president of France, Sadi Carnot, had been killed by the Italian anarchist Sante Geronimo Caserio. Lacassagne was called upon to perform the autopsy of Carnot’s body and a psychiatric examination of Caserio. The results of these two analyses were published in the aforementioned book. He made his observations on the anarchist in the broader context of criminological debates pursued in the late nineteenth century, which were not restricted solely to the authors of Italian criminology.

Visiting Molecular Mimicry Once More: Pathogenicity, Virulence, and Autoimmunity

Abstract: The concept of molecular mimicry describes situations in which antigen sharing between parasites and hosts could benefit pathogen evasion from host immune responses. However, antigen sharing can generate host responses to parasite-derived self-like peptides, triggering autoimmunity. Since its conception, molecular mimicry and the consequent potential cross-reactivity following infections have been repeatedly described in humans, raising increasing interest among immunologists. Here, we reviewed this concept focusing on the challenge of maintaining host immune tolerance to self-components in parasitic diseases. We focused on the studies that used genomics and bioinformatics to estimate the extent of antigen sharing between proteomes of different organisms. In addition, we comparatively analyzed human and murine proteomes for peptide sharing with proteomes of pathogenic and non-pathogenic organisms. We conclude that, although the amount of antigenic sharing between hosts and both pathogenic and non-pathogenic parasites and bacteria is massive, the degree of this antigen sharing is not related to pathogenicity or virulence. In addition, because the development of autoimmunity in response to infections by microorganisms endowed with cross-reacting antigens is rare, we conclude that molecular mimicry by itself is not a sufficient factor to disrupt intact self-tolerance mechanisms.