Results of the European Society of Toxicologic Pathology Survey on the Use of Artificial Intelligence in Toxicologic Pathology.

The European Society of Toxicologic Pathology (ESTP) initiated a survey through its Pathology 2.0 workstream in partnership with sister professional societies in Europe and North America to generate a snapshot of artificial intelligence (AI) usage in the field of toxicologic pathology. In addition to demographic information, some general questions explored AI relative to (1) the current status of adoption across organizations; (2) technical and methodological aspects; (3) perceived business value and finally; and (4) roadblocks and perspectives. AI has become increasingly established in toxicologic pathology with most pathologists being supportive of its development despite some areas of uncertainty. A salient feature consisted of the variability of AI awareness and adoption among the responders, as the spectrum extended from pathologists having developed familiarity and technical skills in AI, to colleagues who had no interest in AI as a tool in toxicologic pathology. Despite a general enthusiasm for these techniques, the overall understanding and trust in AI algorithms as well as their added value in toxicologic pathology were generally low, suggesting room for the need for increased awareness and education. This survey will serve as a basis to evaluate the evolution of AI penetration and acceptance in this domain.

European Society of Toxicologic Pathology (Pathology 2.0 Molecular Pathology Special Interest Group): Review of In Situ Hybridization Techniques for Drug Research and Development.

In situ hybridization (ISH) is used for the localization of specific nucleic acid sequences in cells or tissues by complementary binding of a nucleotide probe to a specific target nucleic acid sequence. In the last years, the specificity and sensitivity of ISH assays were improved by innovative techniques like synthetic nucleic acids and tandem oligonucleotide probes combined with signal amplification methods like branched DNA, hybridization chain reaction and tyramide signal amplification. These improvements increased the application spectrum for ISH on formalin-fixed paraffin-embedded tissues. ISH is a powerful tool to investigate DNA, mRNA transcripts, regulatory noncoding RNA, and therapeutic oligonucleotides. ISH can be used to obtain spatial information of a cell type, subcellular localization, or expression levels of targets. Since immunohistochemistry and ISH share similar workflows, their combination can address simultaneous transcriptomics and proteomics questions. The goal of this review paper is to revisit the current state of the scientific approaches in ISH and its application in drug research and development.

HistoNet: A Deep Learning-Based Model of Normal Histology.

We introduce HistoNet, a deep neural network trained on normal tissue. On 1690 slides with rat tissue samples from 6 preclinical toxicology studies, tissue regions were outlined and annotated by pathologists into 46 different tissue classes. From these annotated regions, we sampled small 224 × 224 pixels images (patches) at 6 different levels of magnification. Using 4 studies as training set and 2 studies as test set, we trained VGG-16, ResNet-50, and Inception-v3 networks separately at each magnification level. Among these model architectures, Inception-v3 and ResNet-50 outperformed VGG-16. Inception-v3 identified the tissue from query images, with an accuracy up to 83.4%. Most misclassifications occurred between histologically similar tissues. Investigation of the features learned by the model (embedding layer) using Uniform Manifold Approximation and Projection revealed not only coherent clusters associated with the individual tissues but also subclusters corresponding to histologically meaningful structures that had not been annotated or trained for. This suggests that the histological representation learned by HistoNet could be useful as the basis of other machine learning algorithms and data mining. Finally, we found that models trained on rat tissues can be used on non-human primate and minipig tissues with minimal retraining.

Society of Toxicologic Pathology Digital Pathology and Image Analysis Special Interest Group Article*: Opinion on the Application of Artificial Intelligence and Machine Learning to Digital Toxicologic Pathology.

Toxicologic pathology is transitioning from analog to digital methods. This transition seems inevitable due to a host of ongoing social and medical technological forces. Of these, artificial intelligence (AI) and in particular machine learning (ML) are globally disruptive, rapidly growing sectors of technology whose impact on the long-established field of histopathology is quickly being realized. The development of increasing numbers of algorithms, peering ever deeper into the histopathological space, has demonstrated to the scientific community that AI pathology platforms are now poised to truly impact the future of precision and personalized medicine. However, as with all great technological advances, there are implementation and adoption challenges. This review aims to define common and relevant AI and ML terminology, describe data generation and interpretation, outline current and potential future business cases, discuss validation and regulatory hurdles, and most importantly, propose how overcoming the challenges of this burgeoning technology may shape toxicologic pathology for years to come, enabling pathologists to contribute even more effectively to answering scientific questions and solving global health issues. [Box: see text].

Correlation Between Nasal Epithelial Injury and In Vitro Cytotoxicity Using a Series of Small Molecule Protein Tyrosine Phosphatase 1B Inhibitors Investigated for Reversal of Leptin Resistance in Obesity.

This research provides a cautionary example when evaluating changes in behavioral end points with respect to postulated pharmacologic activity. Various small molecule substrate mimetic protein tyrosine phosphatase 1B (PTP1B) inhibitors were investigated as pharmacologic agents for decreasing food consumption using intranasal (IN) dosing as a means for direct nose-to-brain delivery along the olfactory/trigeminal nerve pathways. Although food consumption was decreased in diet-induced obese (DIO) mice, nasal discharge was observed. Studies were conducted to investigate local effects on the nasal airway and to develop structure-activity relationships. Intranasal administration of PTP1B inhibitors at ≥0.03 mg/d to DIO mice produced dose-dependent injury to various cell types of the nasal epithelia. Protein tyrosine phosphatase 1B inhibitors with calculated log octanol >3.0 were the most toxic. Whereas a pharmacologically inactive analog of a PTP1B inhibitor produced nasal injury, along with decreased food consumption, the marketed IN drug ketorolac produced no lesions at the same dose of 0.3 mg/d and only minor changes at 3 mg/d. Rat skin fibroblast cells were exposed in vitro to PTP1B inhibitors, ketorolac, paraquat, and the detergent sodium dodecylbenzene sulfonate (NDS) followed by measures of cytotoxicity. The most potent PTP1B inhibitors were similar to NDS, whereas ketorolac was the least toxic compound. Cytotoxic potency in vitro was similar to in vivo. In conclusion, PTP1B inhibitors injured nasal epithelium through a mechanism independent of PTP1B inhibition and likely due to nonspecific cytotoxicity such as disruption of the cell membrane. Decreased food consumption in DIO mice was due to toxicity rather than a pharmacologic mode of action.

Examination of the role of galectins in cell migration and re-epithelialization of wounds.

Re-epithelialization is a crucial step for wound healing. As galectins play important roles in re-epithelialization, we describe here protocols for in vivo, ex vivo and in vitro examination of the role of galectins in cell migration and in re-epithelialization of wounds. For in vivo models, mouse corneas are wounded by a variety of techniques and the rate of re-epithelialization is quantified. For ex vivo organ culture models, mouse corneas are wounded in situ, the eyes are enucleated, the eyeballs are cultured in the presence or absence of galectins and the rate of re-epithelialization is quantified. For cell cultured-based in vitro assays, we examine formation of lamellipodia and activation of focal adhesion kinase in various epithelial cells.

Analysis of differential expression of glycosyltransferases in healing corneas by glycogene microarrays.

It is generally accepted that the glycans on the cell surface and extracellular matrix proteins play a pivotal role in the events that mediate re-epithelialization of wounds. Yet, the global alteration in the structure and composition of glycans, specifically occurring during corneal wound closure remains unknown. In this study, GLYCOv2 glycogene microarray technology was used for the first time to identify the differentially expressed glycosylation-related genes in healing mouse corneas. Of approximately 2000 glycogenes on the array, the expression of 11 glycosytransferase and glycosidase enzymes was upregulated and that of 19 was downregulated more than 1.5-fold in healing corneas compared with the normal, uninjured corneas. Among them, notably, glycosyltransferases, beta3GalT5, T-synthase, and GnTIVb, were all found to be induced in the corneas in response to injury, whereas, GnTIII and many sialyltransferases were downregulated. Interestingly, it appears that the glycan structures on glycoproteins and glycolipids, expressed in healing corneas as a result of differential regulation of these glycosyltransferases, may serve as specific counter-receptors for galectin-3, a carbohydrate-binding protein, known to play a key role in re-epithelialization of corneal wounds. Additionally, many glycogenes including a proteoglycan, glypican-3, cell adhesion proteins dectin-1 and -2, and mincle, and mucin 1 were identified for the first time to be differentially regulated during corneal wound healing. Results of glycogene microarray data were confirmed by qRT-PCR and lectin blot analyses. The differentially expressed glycogenes identified in the present study have not previously been investigated in the context of wound healing and represent novel factors for investigating the role of carbohydrate-mediated recognition in corneal wound healing.

Galectin-3 promotes lamellipodia formation in epithelial cells by interacting with complex N-glycans on alpha3beta1 integrin.

Recent studies have shown that galectin-3 (Gal-3; also known as LGALS3), a beta-galactoside-binding lectin, promotes cell migration during re-epithelialization of corneal wounds. The goal of this study was to characterize the molecular mechanism by which Gal-3 stimulates cell migration. We demonstrate here that exogenous Gal-3, but not Gal-1 or Gal-8, promotes cell scattering and formation of lamellipodia in human corneal epithelial cells in a beta-lactose-inhibitable manner. alpha3beta1 integrin was identified as the major Gal-3-binding protein in corneal epithelial cells by affinity chromatography of cell lysates on a Gal-3-Sepharose column. Preincubation of cells with anti-alpha3 integrin function-blocking antibody significantly inhibited the induction of lamellipodia by Gal-3. Furthermore, exogenous Gal-3 activated both focal adhesion kinase, a key regulator of integrin-dependent intracellular signaling, and Rac1 GTPase, a member of the family of Rho GTPases, well known for its role in the reorganization of the actin cytoskeleton and formation of lamellipodial extensions. Experiments involving knockdown of beta-1,6-N-acetylglucosaminytransferase V, an enzyme that synthesizes high-affinity glycan ligands for Gal-3, revealed that carbohydrate-mediated interaction between Gal-3 and complex N-glycans on alpha3beta1 integrin plays a key role in Gal-3-induced lamellipodia formation. We propose that Gal-3 promotes epithelial cell migration by cross-linking MGAT5-modified complex N-glycans on alpha3beta1 integrin and subsequently activating alpha3beta1-integrin-Rac1 signaling to promote lamellipodia formation.

Detection of differentially expressed wound-healing-related glycogenes in galectin-3-deficient mice.

PURPOSE: A prior study showed that exogenous galectin-3 (Gal-3) stimulates re-epithelialization of corneal wounds in wild-type (Gal-3(+/+)) mice but, surprisingly, not in galectin-3-deficient (Gal-3(-/-)) mice. In an effort to understand why the injured corneas of Gal-3(-/-) mice are unresponsive to exogenous Gal-3, the present study was designed to determine whether genes encoding the enzymes that regulate the synthesis of glycan ligands of Gal-3 are differentially expressed in Gal-3(-/-) corneas compared with the Gal-3(+/+) corneas. METHODS: Glycogene microarray technology was used to identify differentially expressed glycosyltransferases in healing Gal-3(+/+) and Gal-3(-/-) corneas. RESULTS: Of approximately 2000 glycogenes on the array, the expression of 8 was upregulated and that of 14 was downregulated more than 1.3-fold in healing Gal-3(-/-) corneas. A galactosyltransferase, beta3GalT5, which has the ability to synthesize Gal-3 ligands was markedly downregulated in healing Gal-3(-/-) corneas. The genes for polypeptide galactosaminyltransferases (ppGalNAcT-3 and -7) that are known to initiate O-linked glycosylation and N-aspartyl-beta-glucosaminidase, which participates in the removal of N-glycans, were found to be upregulated in healing Gal-3(-/-) corneas. Microarray data were validated by qRT-PCR. CONCLUSIONS: Based on the known functions of the differentially expressed glycogenes, it appears that the glycan structures on glycoproteins and glycolipids, synthesized as a result of the differential glycogene expression pattern in healing Gal-3(-/-) corneas may lead to the downregulation of specific counterreceptors for Gal-3. This may explain, at least in part, why, unlike healing Gal-3(+/+) corneas, the healing Gal-3(-/-) corneas are unresponsive to the stimulatory effect of exogenous Gal-3 on re-epithelialization of corneal wounds.

Milk components inhibit Acanthamoeba-induced cytopathic effect.

PURPOSE: Acanthamoebae provoke a vision-threatening corneal infection known as Acanthamoeba keratitis (AK). It is thought that Acanthamoeba-specific IgA antibodies present in mucosal secretions such as human tears, milk, and saliva provide protection against infection by inhibiting the adhesion of parasites to host cells. The goal of the present study was to determine whether human mucosal secretions have the potential to provide protection against the Acanthamoeba-induced cytopathic effect (CPE) by an additional mechanism that is independent of IgA. METHODS: Breast milk was used as a model of human mucosal secretions. In vitro CPE assays were used to examine the CPE inhibitory effect of IgA-depleted milk and various milk fractions obtained by gel filtration. The activity of amebic proteinases was examined by zymography. RESULTS: IgA-depleted milk inhibited the Acanthamoeba-induced CPE in a concentration-dependent manner. Milk proteins were separated into four major fractions (F1-F4) by gel filtration. Of these four fractions, CPE inhibitory activity was detected largely in fraction F3. In contrast, fractions F1, F2, and F4 lacked CPE inhibitory activity. Moreover, fraction F3, but not F1, F2, or F4, inhibited amebic proteinases. CONCLUSIONS: These data, in conjunction with published findings showing that amebic proteinases are responsible for the induction of Acanthamoeba CPE, led us to propose that human mucosal secretions have the potential to provide protection against Acanthamoeba-induced CPE by an additional mechanism that is independent of IgA and that involves the inhibition of cytotoxic proteinases of amebae.

Effect of human tears on acanthamoeba-induced cytopathic effect.

OBJECTIVE: To determine whether tears of healthy individuals provide protection against Acanthamoeba-induced cytopathic effect (CPE) in vitro. METHODS: Acanthamoebae were added to confluent cultures of corneal epithelium in 24-well plates, and co-cultures were incubated overnight in a serum-free medium containing varying amounts of tears or immunoglobulin A (IgA)-depleted tears. At the end of the incubation period, the cells were stained with Giemsa, and the extent of target cell damage (ie, CPE) was quantified. RESULTS: Acanthamoebae produced extensive CPE. The presence of even a low concentration of tears (10 microL of undiluted tears per milliliter of media) almost completely inhibited Acanthamoeba-induced CPE. The CPE was inhibited by pretreatment of the parasites with tears. In contrast, the pretreatment of host cells with tears was not protective. This finding suggests that the target of the inhibitory factor is the parasite. IgA-depleted tears also inhibited Acanthamoeba-induced CPE, albeit with a lower potency than total tears. CONCLUSION: In addition to known IgA-dependent protective factors, human tears contain factors that inhibit Acanthamoeba-induced CPE independently of IgA. Clinical Relevance Identification and characterization of factors that protect against Acanthamoeba-induced CPE should help in the development of novel, rationally designed strategies to manage and protect against keratitis.