Morphometric analysis of the thymic epithelial cell network using integrated and orthogonal digital pathology approaches

Maria K Lagou; Maria P Markaki; Stepan Vodopyanov; Dimitrios G Argyris; George K Karagiorgis; Leslie Gunther-Cummins; Alexandros Hardas; Christos Panorias; Sophia DesMarais; Conner Entenberg; Randall S Carpenter; Panagiota S Filippou; Hillary Guzik; Xheni Nishku; Joseph Churaman; Vera DesMarais; Frank P Macaluso; Maria Maryanovich; Katerina Angelopoulou; Theofilos Poutahidis; George S Karagiannis

doi:10.1038/s42003-026-10038-w

. 2026 Apr 14;9:811. doi: 10.1038/s42003-026-10038-w

Morphometric analysis of the thymic epithelial cell network using integrated and orthogonal digital pathology approaches

Maria K Lagou ^1,^2,¹⁶, Maria P Markaki ³, Stepan Vodopyanov ^1,², Dimitrios G Argyris ^1,^2,⁴, George K Karagiorgis ³, Leslie Gunther-Cummins ^5,^6,⁷, Alexandros Hardas ⁸, Christos Panorias ⁹, Sophia DesMarais ¹, Conner Entenberg ¹, Randall S Carpenter ^6,¹⁰, Panagiota S Filippou ¹¹, Hillary Guzik ^5,^7,¹², Xheni Nishku ^5,⁷, Joseph Churaman ^5,⁷, Vera DesMarais ^5,^6,^7,¹², Frank P Macaluso ^5,^6,^7,¹², Maria Maryanovich ^6,^10,¹³, Katerina Angelopoulou ³, Theofilos Poutahidis ¹⁴, George S Karagiannis ^1,^2,^4,^12,^13,^15,^✉

PMCID: PMC13265849 PMID: 41981116

Abstract

The thymus, a central organ for T cell development, lacks standardized pathology methods to assess its structure and function. Most studies rely on flow cytometry, which profiles thymic cell populations but offers limited spatial insight into tissue organization. This gap restricts our understanding of how physiological or pathological conditions remodel thymic architecture and impact immune competence. Here, we introduce a digital pathology pipeline that integrates analytical algorithms with rationalized morphometric assessments at tissue-wide, microanatomical, and ultrastructural levels. We develop and validate it using a mouse model of age-dependent thymic involution, characterized by cortical atrophy, medullary expansion, and epithelial disorganization, providing a benchmark for quantifying known lesions and testing reproducibility. As a proof-of-concept, we apply the pipeline to a DMBA-induced carcinogenesis model combined with early-life cholera toxin (CT) exposure, to evaluate whether environmental stimulation (“hygiene hypothesis”) mitigates late-stage thymic pathology. CT normalizes select parameters, particularly in the medulla, including reduced perivascular space and mTEC density, suggesting enhanced thymocyte occupancy. Overall, this versatile and scalable pipeline provides a standardized framework for quantitative evaluation of thymic structure, supporting both basic and translational studies of thymic biology and disease.

Subject terms: Imaging the immune system, Ageing, 3-D reconstruction

A new digital pathology pipeline enables multiscale, quantitative analysis of thymic structure, revealing age-dependent epithelial remodeling and showing improved medullary regeneration after early-life immune modulation in a carcinogenesis model.

Introduction

The thymus is a primary lymphoid organ, in which T cell development occurs¹. The thymic environment subjects developing T cells, also known as thymocytes, to a series of crucial and intricate steps, necessary for their continued growth and competence. To support an assortment of specialized and unique developmental functions^2–4, the thymus is compartmentalized into an outer cortical and an inner medullary zone. In general, the cortex oversees the homing, survival, and commitment of bone-marrow-derived early thymic progenitors (ETPs) to the T cell lineage. It also provides the requisite signals for the development of functional T cell receptors (TCR) through the process of positive selection, and promotes the proliferation/expansion of positively-selected thymocytes, promoting them into single-positive (SP) thymocytes expressing either CD4⁺ (SP4) or CD8⁺ (SP8)^2,4,5. The medulla serves equally fundamental roles, such as the establishment of central tolerance and the development of regulatory T cells (T_regs)^2,4,5. The above processes are orchestrated via a complex, joint effort of an intrathymic stromal cell network, consisting of mesenchymal, endothelial, neuroendocrine, dendritic, and other immune cells, although the main protagonists are the highly specialized thymic epithelial cells (TECs)^4,6–9.

Like other lymphoid organs, the thymus relies heavily on its distinct zonation to function properly^10,11. Recent evidence has revealed that the traditional view of corticomedullary compartmentalization oversimplifies the complex nature of the thymic environments, which includes additional, highly-specialized sub-niches. A prime example is a unique medullary sub-niche, known as the “perivascular space” (PVS), which was originally overlooked for its functional significance due to its appearance as an “epithelium-free area” in histological representations^12,13. The entry of positively-selected single-positive (SP) thymocytes into the medulla triggers an interaction with the AutoImmune Regulator⁺ (AIRE⁺) mTEC^hi (CD80^hi, MHCII^hi) and the post-AIRE mimetic cells, crucial for establishing self-tolerant T cells^14,15. After passing through this test, self-tolerant CD4⁺ and CD8⁺ T cells emigrate from the medullary parenchyma into the peripheral blood, a process known as “thymocyte egress”¹⁶. This process is tightly regulated by a complex network of thymic stromal cells, including endothelial cells, pericytes, and fibroblasts, which together orchestrate chemokine networks, shaping the microanatomy of the perivascular space^17–23. As our understanding on thymic function deepens, it therefore becomes increasingly clear that this primary lymphoid organ is much more intricate and contextually organized than previously believed.

Given its pivotal role in adaptive immunity, the thymus has been predominantly explored via cell population analyses, utilizing flow cytometry^7,24,25. This approach, however, has largely neglected the organ’s spatiotemporal dynamics, which are crucial for comprehensive, quantitative studies. In response, the Societies of Toxicological Pathology^1,26–28 have provided the International Harmonization of Nomenclature and Diagnostic Criteria (INHAND) histopathology guidelines that, while valuable, are subjectively interpreted and lean towards a less objective framework. This gap in precise and universally accepted assessments hampers our understanding, especially as various conditions and pathologies can significantly alter thymic histological structure^27,28.

In recent years, substantial progress has been made in thymus imaging aimed at providing contextual and functional insights into thymic organization. High-resolution imaging, including ultrastructural analyses, has illuminated the remarkable TEC heterogeneity^29–31. In addition, real-time intravital imaging (IVI) has achieved moderate success in visualizing thymocyte trafficking dynamics³², while functional imaging approaches, particularly MALDI mass spectrometry imaging (MSI), have been applied to delineate the biochemical remodeling of the thymus across various pathophysiological contexts^33–35. Undoubtedly, the most transformative advances are expected from spatial transcriptomics, as recent studies have achieved an unprecedented spatially resolved atlas of the human thymus, offering powerful new perspectives on the spatiotemporal regulation of thymopoiesis in health and disease³⁶. Despite these advances, to our knowledge, no digital pathology frameworks currently exist that enable rapid, accurate, and comprehensive morphometric characterization of the TEC network. To address these issues, we have developed an orthogonal digital pathology pipeline that spans from macro- to microanatomical, and ultimately to ultrastructural analyses. Our innovative approach is anchored in a thoroughly vetted mouse model, chosen for its well-documented portrayal of age-related thymic involution. Through this proof-of-concept, we demonstrate the pipeline’s potential in meticulously charting thymic alterations across a spectrum of pathological conditions, offering a new lens through which to examine intrathymic remodeling.

Results

Mouse cohort necropsy – general findings and progressive age-associated lesions

To develop a digital pathology framework for assessing thymic lesions (Fig. 1A), we employed a mouse model of age-dependent thymic involution, in which mice were sacrificed at diverse age groups, and experimental endpoints were assessed (Fig. 1B). As expected, we found that the total body weight of the mice significantly increased over time (Fig. 1C). Because it is typical for aged mice to confer spontaneous age-associated lesions, we performed whole body necropsy, to determine the possibility of unreported pathologies that may bias our study. In general, we found no severe pathologies in the Young-Adult and Adult mouse groups. However, certain Old mice developed spontaneous lesions, diagnosed by veterinary pathologists (Figs. S1A–F; Supplementary Data 1): Multifocal mild-to-moderate neutrophilic, histiocytic, lymphoplasmacytic, necrotizing steatitis with mineralization and hemosiderin-laden macrophages in the omentum (Figs. S1A-A’); Adrenal gland adenoma type B and pheochromocytoma (Figs. S1B-B’); Multifocal solid, bronchiolo-alveolar lung carcinoma (Fig S1C-C’); Multifocal lymphocytic proliferation of the renal pelvis and associated vasculitis of the perirenal fat vessels (Fig. S1D-D’); Bone marrow hyperplasia (Fig. S1E-E’); Splenic hyperplasia (Fig. S1F-F’). These lesions are typical and expected to spontaneously rise in mice during aging³⁷. Because there is no literature linking any of these lesions to specific thymic defects, except that they may represent general age-associated stressors, we reasoned that our cohort was suitable for further analysis.

Of all the aforementioned lesions, and consistent with prior literature³⁸, splenic enlargement based on absolute spleen weight was noted in a subgroup of aged mice (Fig. S2A). However, the splenic weight index (SWI) did not reveal a significant increase (Fig. S2B), indicating that splenic enlargement either accompanies or occurs at a lower rate compared to body weight gain with aging. Moreover, spleens from Old-aged mice presented with extramedullary hematopoiesis, characterized by red pulp megakaryocytes with hemosiderosis (Fig. S1F-F’). To determine the relative contribution of individual splenic compartments to the overall enlargement, we developed two independent morphometric methods, an HE-based and an IF-based, both calculating the White-to-Red-Pulp Ratio (WRPR) (Fig. S2C). Both methods correlated well with each other, and revealed a trend for WRPR decline with aging, but neither showed statistical significance across the age groups (Fig. S2D-F). In addition, we assessed the white and red pulp fractions independently, but again, neither method revealed any changes with aging (Fig. S2G–L). In conclusion, splenic enlargement is an incidental finding in only a subset of Old mice, and seems to occur due to the concerted increase of both white and red pulp.

We then macroscopically assessed the extracted thymi and found a significant decline of thymic weight with aging (Fig. 1D). In consistency, the thymic weight index (TWI) continuously declined from a mean of ~3.46 at the Young-Adult to a mean of <0.7 at the Old group (Fig. 1E). This rapid decline in thymic size (Fig. 1C, D) is well-characterized in mice³⁹, and is consistent with thymic decline in humans after reaching puberty^40–42. In conclusion, a macroscopic analysis of the thymus in female mice reveals evidence of age-dependent thymic involution^39–42, and as such, it represents an efficient in vivo model for the development of microanatomical and ultrastructural morphometries, using our digital pathology framework (Fig. 1A, B).

Microscopic assessment of cortex-medulla compartmentalization

Our subsequent studies focused on establishing morphometric assessments of the thymic structure, to draw functional correlates of thymic function and thymopoietic capacity. Based on International Harmonization of Nomenclature and Diagnostic Criteria (INHAND) histopathology guidelines by multiple Societies of Toxicological Pathology^27,28,43, the volumetric ratio of thymic cortex to medulla (i.e., corticomedullary ratio, CMR) is the strongest indicator of functional status, and may be altered in pathology and/or pharmacological interventions, such as in endogenous cortisol production in response to chronic stress, or following exogenous glucocorticoid or cyclosporin administration^{1,27,28,43,44}. The guidelines provided by INHAND, however, are based on evaluating the ratio subjectively, thus leading to operator-dependent biases²⁷. Based on the guidelines²⁷, here, we sought to standardize the “corticomedullary ratio” as a reproducible and measurable index, to inform on the structural and functional status of the thymic parenchyma.

Our pipeline involves three independent methods that calculate corticomedullary ratio in a digital pathology framework (Fig. 2A-A”’). Thymus sections (Fig. 2A) are first stained either with a Keratin-5/Keratin-8 (KRT5/8) dual immunofluorescence staining protocol (Slide 1), or with HE (Slide 2), and digital whole slide-scanning is used to generate whole tissue sectioned images. The cortical and medullary compartments are then digitally demarcated as individual regions of interest (ROIs) in either the KRT5/8 or the HE modality (Fig. 2A’). In the HE modality, the cortex appears darker, because cortical thymocytes are smaller in size and tightly arranged in their physical space, when compared to their medullary counterparts, thus generating higher density of hematoxylin⁺ nuclei (Fig. S3A). In the IF modality contrariwise, the specific keratin filament immunolabeling signifies topology of specific thymic epithelial cell (TEC) compartments, with KRT5 exclusively labeling medullary TEC (mTEC), whereas KRT8 labeling all cortical TEC (cTEC), along with a prominent subset of mTEC (Fig. S3B-B”)^45–48. The cortical parenchyma is thus defined as KRT5^-KRT8⁺, while the medullary parenchyma as KRT5⁺KRT8⁺ (Figs. S3B’–B”). Besides the positivity/negativity in the keratin signal, the nature of the cTEC/mTEC networks is also distinctly diverse, with cTEC appearing with “meshwork” architecture while mTEC densely distributed as isolated cellular entities (Fig. S3B–B”). In either modality (HE or KRT5/8 IF), a clear distinction of the corticomedullary junction (CMJ) is demonstrated (Fig. S3A, B), and is used for the digital demarcation of the non-overlapping cortical and medullary ROIs (Fig. 2A”). CMR is then defined as the ratio of surface areas of the two thymic compartments, and abbreviated as CMR_WT-HE (for the HE modality), and CMR_WT-IF (for the IF modality) (Fig. 2A”’ and equation #1).

The whole-thymus approaches described above may sample a large area for analysis, but they are more prone to errors regarding the orientation of the organ’s axis in the paraffin block, as well as to the depth of block “trimming” prior to sectioning (e.g., superficial trimming would result in biased cortex domination due to the tangential cutting of the lobules). To circumvent these issues, we developed another independent method, by standardizing the quantified ROI across different samples. The method exclusively relies on images of individual thymic lobules, whose corticomedullary junction orientation is arranged in parallel fashion across the two sides of the medullary parenchyma. In these lobules, a one-dimensional line is then drawn perpendicularly to the orientation of the corticomedullary junction (Fig. S3C). The cortical and medullary segment widths are then measured on this vector (Fig. S3C), and the ratio of absolute widths is calculated for each mouse (Fig. 2A”’ and equation 2). Due to the measurement of corticomedullary ratios using standardized microanatomical features, i.e., thymic lobules with a predetermined orientation, this method (CMR_TL) is more consistent among replicate experiments. It could be limited, however, by pathologies resulting in loss of thymic lobulation, etc.

An in-depth analysis of functional adaptations of thymic compartmentalization during aging has previously demonstrated the numeric decline of cTEC, leading to a reduced thymopoietic potential of the aged thymus⁴⁹. Interestingly these data have further documented that mTEC do not seem to decline as fast with aging, although many mTEC subsets tend to lose their terminal differentiation and present with a progenitor-like phenotype⁴⁹. These observations suggest an expected reduction of the corticomedullary ratio in aged mice²⁸. To confirm this hypothesis, we assessed CMR using all three methods described, and successfully captured its expected decrease with aging (Fig. 2B–G). Although the tendency of decline was clear after adolescence, none of the three methods captured any difference (p > 0.05) between Young-Adult and Adult groups (Fig. 2E–G). However, there was significant reduction between Young-Adult and Old groups for all three methods (Fig. 2E–G), suggesting that our morphometries can efficiently capture the expected disturbances in thymic compartmentalization with older age.

To determine a potential deviation of our proposed morphometries from the accepted and established methods of corticomedullary ratio calculation, we performed a subjective analysis of corticomedullary ratio, based on the INHAND guidelines^26–28. The measurements were made by a board-certified veterinary pathologist, who was “blinded” to the mouse age-groups, using an arbitrary scale of categorical CMR values. This method also successfully captured the expected decrease of the corticomedullary ratio with aging (Fig. 2H). Importantly, the pathologist’s scores correlated best with the CMR_WT-HE, when compared to the CMR_WT-IF and CMR_TL approaches (Fig. 2I, J and S3D–F), which was expected, given that the blinded pathologist also utilized whole-thymus HE slides, to perform quantifications. When the three newly-developed methods were compared to one another independently, we found that the best correlation was between whole-thymus approaches (i.e., CMR_WT-HE and CMR_WT-IF) with a Spearman rho of 0.79 (p < 0.001). As expected, the worst correlations were observed between the CMR_TL and other approaches (Fig. 2I and S3G–I). Despite that the CMR_TL deviated numerically from all other methods, it was the method that presented with the least variability (i.e., standard deviation) for each mouse age-group (Fig. 2G), an expected finding, given that it is exclusively based on a pre-defined microanatomical unit. These observations demonstrate that the proposed methods exhibit high orthogonality and could be used complementarily to evaluate thymic compartmentalization.

Topological assessment of the cortical thymic epithelial cell (cTEC) network

In post-embryonic thymus, cTEC develop with meshwork architecture, and their functional maturation requires thymocyte signals beyond the DN1 stage^2,50. Reciprocally, the cTEC are incremental for properly educating thymocytes and for their progressive maturation through most developmental stages^2,4,50,51. The meshwork architecture of the cTEC population is dismantled in certain thymic pathologies, including the natural age-dependent thymic decline, as well as acute thymic involution induced by cytotoxic stressors^{7,24,26–28,52}. As the next step of our digital pathology pipeline, we therefore propose a rationalized approach to establish quantifiable features of cTEC meshwork architecture, to characterize its functional thymopoietic capacity.

The first feature, herewith termed “cTEC Volume Index” (VI_cTEC), reports the cTEC volume as a fraction of the parenchymal mass of the cortex. The analysis involves an initial binarization of the KRT8⁺ signal based on intensity thresholding to generate a “cTEC mask”, then reporting the KRT8⁺ surface area as a proportion of the total surface area of the cortical parenchyma (Fig. 3A–C”, 3F and equation 3). Because there are no other cells known to express KRT8 in the thymic cortex apart from the cTEC stroma⁴⁸, an increased ratio would thus indicate cTEC hyperplasia or hypertrophy (i.e., increased cTEC volume per unit of thymic parenchyma), while a decreased ratio would indicate cTEC atrophy or hypoplasia (i.e., decreased cTEC volume supporting thymopoiesis) (Fig. 3G). It should be noted though that the dual KRT5/KRT8 immunofluorescence is critical for the assessment of this feature, because KRT5⁺KRT8⁺ regions correspond to the medulla, and thus should be excluded from the analysis (Fig. 3A–A’, S3B-B”).

Fig. 3 — A–F Digital pathology pipeline for the topological assessment of cTEC meshwork architecture. A–A’ Whole thymi are stained for KRT5/8 immunofluorescence, and cortical regions are identified as KRT5^-KRT8⁺. B–B’ Thymic cortex field-of-view (FOV), depicting meshwork architecture of the cTEC network. C–C’ The binarization of the image shown in (B) via KRT8 signal intensity thresholding, creates a “mask” of the cTEC network. Accompanying formula indicated with the arrow. D–D’ Development of outline from the binary mask shown in (C), revealing the interface area of the cTEC network. Accompanying formula indicated with the arrow. E, E’ Skeletonization of the binary mask shown in (C), revealing the branching characteristics of the cTEC meshwork. Accompanying formulas indicated with the arrow. F Illustration of a cTEC network part that can be used as “key” for the formulas shown in (C–E). Abbreviations: VI_cTEC cTEC volume index, II_cTEC cTEC interface index, BI_cTEC cTEC branching index, SA_cTEC cTEC surface area, SA_Pc surface area of cortical parenchyma, P_cTEC cTEC perimeter, jn junction, br branch. G Theoretical adaptations of the cTEC network topology, along with expected changes in cTEC network morphometries following these adaptations. Representative IF images before (H) and after skeletonization (H’) from Young Adult, Adult, and Old mice, showcasing the spatiotemporal changes in VIcTEC, IIcTEC and BIcTEC I–M Morphometric assessment of cTEC network topology, using cTEC volume index (VI_cTEC) (I), cTEC interface index (II_cTEC) (J); cTEC branching indices, i.e., BI1_cTEC (K), BI2_cTEC (L), and BI3_cTEC (M). One-Way ANOVA; *p $\leq$ 0.05; **p $\leq$ 0.01; ***p $\leq$ 0.001; ****p $\leq$ 0.0001; ns, not-significant. Error bars; Standard Error of Mean. Total N = 26 mice; mice may have been excluded from individual variables, based on exclusion criteria.

The second feature, herewith termed “cTEC Interface Index” (II_cTEC), informs on the surface area of the cTEC meshwork, readily exposed for direct cTEC-thymocyte interactions. As the developing thymocytes travel through the cortex, they are juxtaposed and continuously come in contact with MHC I/II-self-peptide complexes that are presented in the surface of cTEC^4,53–55. To draw functional correlates of this process, we implemented an extension of the previous analysis that generates the cTEC mask based on KRT8⁺ signal intensity thresholding. Besides calculating the VI_cTEC from this binary mask, an “outline” of the mask is generated (Fig. 3D, D”), and its perimeter is measured and normalized to a given surface area of cortical parenchyma bFig. 3F and equation 4). In case of VI_cTEC increase for instance, a corresponding II_cTEC would inform on whether the observed cTEC volume increase is due to increase in cTEC cell body thickness, or due to increase of cTEC network tortuosity, branching morphogenesis, etc. (Fig. 3G).

The third feature, herewith termed “cTEC Branching Index” (BI_cTEC), informs on more detailed topological and structural descriptors of the cTEC network. Because cTEC exhibit dendritic morphology, they create a vast meshwork, characterized by establishment of cytoplasmic junctions (nodes) and branches Fig. S3B, B’). As above, this feature also considers a cTEC mask, as generated to calculate VI_cTEC and II_cTEC, but at higher magnification. The cTEC mask is “skeletonized” using an ImageJ plugin⁵⁶, an algorithm that reduces the binary cTEC mask to one-pixel thickness but retaining the original object representation (Fig. 3E, E”). This transformation allows for straightforward quantification of network descriptors, such as number of branches and junctions, both normalized to a given cortical surface area. With this method, the “average branch length” of all branches in the cTEC skeleton is also automatically calculated, when the network skeleton is analyzed (Fig. 3E, E”, 3F and equations 5–7). With regards to interpretation, the three BI_cTEC features, i.e., BI1_cTEC, BI2_cTEC, and BI3_cTEC, collectively illustrate the nature of the cTEC meshwork complexity, specifically its tortuosity and branching capacity, and offer complementary information when assessed along with other descriptors, such as VI_cTEC and II_cTEC (Fig. 3G).

A previous study has performed quantitative assessments of age-dependent lesions in cTEC and mTEC, and reported structural and functional deficits of the former compartment in the murine thymus⁵⁷. In agreement, our results indicate a significant increase (p < 0.001) of VI_cTEC in older mice, compared to younger mice (Fig. 3H, I). However, this cTEC volume “increase” should be cautiously interpreted in the context of overall numerical cTEC and thymocyte decline, known to occur with aging⁵⁸. The cTEC network is generated through long, looping structures forming an intracellular labyrinth of multiple “voids” in each TEC, inside which groups of 2–10 cortical thymocytes reside⁵⁷. As such, a reduced number of thymocytes due to aging could lead to the “narrowing” of these voids, and the artificial resemblance of the cTEC network as more “massive” in nature. Along these lines, the remaining network descriptors, e.g., II_cTEC, BI1_cTEC, and BI2_cTEC also revealed a significant increase in older mice (Fig. 3J–L), demonstrating the dynamic shift of cTEC/thymocyte mass ratio. In consistency with these data for example, the relative area occupied by CXCL12⁺ cells, a homing chemokine predominantly produced by cTEC, was found to be larger in older thymi in humans, again due to the reduction of thymus size with aging⁵⁹.

When we specifically evaluated the branching capacity of the cTEC network in older mice, we found significant increase (p < 0.05) in the relative number of branches and junctions, when compared to the younger mice (Fig. 3H’, 3K–L). This unorthodox observation could be explained by the hypertrophic growth of the few surviving TEC in older thymi, and/or the simultaneous cTEC network collapse, due to the sparse distribution of resident thymocyte populations within the cTEC labyrinthine voids. One study has recently reported inconspicuous development of cTEC projections in old mice⁵⁷, thus challenging our observations. In consistency with our results however, other studies have reported cTEC hypertrophy as a result of Lymphotoxin-β receptor (LTβR) reduced expression or deficiency⁶⁰, which is normally observed during aging in lymphoid organs^61,62. In accordance with our finding showing higher number of branches and junctions, we also found significant decrease in our final network descriptor, the average branch length (BI3_cTEC), in the Old mouse group (Fig. 3M). This finding is highly consistent with the hypertrophic nature of the surviving cTEC at old age, because newly formed branches typically have less time to grow, and are thus expected to be shorter (Fig. 3F–H’).

In summary, a detailed analysis of cTEC network topology has demonstrated a substantial remodeling towards hypertrophic growth with aging. cTEC hypertrophy in this case, however, is not characterized by an abnormal thickening of individual cTEC branches, but by the de novo branching morphogenesis and increased tortuosity of the network. These findings collectively suggest that our newly proposed morphometries must be used non-redundantly, to describe the precise topology of the cTEC network across thymus health and disease.

Ultrastructural topography of the cTEC network

To validate our findings and gain further resolution on cTEC network remodeling with aging, as visualized through the morphometric assessments using immunofluorescence (Fig. 3), we combined scanning electron microscopy of OTO-sections (OTO-SEM) with transmission electron microscopy (TEM) in mice undergoing age-dependent thymic involution. The cellular component of the cTEC meshwork, also known as “cytoreticulum”, is identified in both OTO-SEM and TEM by the characteristic presence of cytoplasmic tonofilaments, vacuoles, and the occasional presence of desmosomes between interconnecting cTEC projections (Fig. S4A–D), as reported^63–67. These distinguishing epithelial features provide validation of cTEC identity, especially when they are projected in proximity to cTEC nuclei, or when the cytoplasmic projection is shown to bud off the cTEC on the same z-plane (Fig. S4E; asterisk). As opposed to thymocyte nuclei, which often contain conspicuous heterochromatin patches giving a “rosette- or flower-shaped” impression, the cTEC nuclei are more euchromatic and easily distinguished (Fig. S4E, F). However, the identification of such epithelial features becomes more crucial when tracking trajectories of cytoreticulum segments among developing thymocytes, either because the cTEC nucleus is further away from the respective field-of-view or exists in a different z plane (Fig. S4F; asterisk).

Because the cTEC network develops meshwork architecture, single section analyses would provide valuable, albeit limited information. We thus obtained 3-dimensional reconstructions at high resolution using 3D OTO-SEM Array Tomography⁶⁸. Backscattered OTO-SEM revealed the ultrastructural details of cTEC cytoreticulum, while it partially or completely enveloped the thymocytes under various stages of their development in the outer cortex (Fig. 4A, B, Videos S1–2). The cytoreticulum (with/without inclusion of the cTEC nucleus or nuclei) was then outlined in each z-stack independently, and reconstructed (Fig. 4A, B, Videos S1–2). 3D reconstruction views of the cTEC meshwork confirmed that the cytoreticulum formed well-circumscribed labyrinthine “voids”⁵⁷, each homing groups of ~1–10 developing thymocytes (Fig. 4A, B, Videos S1, 2). However, viewing the cytoreticulum in single OTO-SEM sections fails to demonstrate the continuity of this unique architectural pattern (Fig. 4C–E), as evidenced by the 3D reconstruction (Fig. 4A, B). Thymocytes within these voids are in partial or complete contact with the adjacent cytoreticulum (Fig. 4A–E), suggesting that the cytoreticulum reflects to regions rich in MHC I/II-peptide complexes directly presented to enveloped thymocytes. At first glance, one could erroneously assume that cTEC-thymocyte interactions improve with older age, given the numerical decline of thymocytes and the concurrent cTEC hypertrophy (Figs. 3 and 4A–E); in other words, fewer thymocytes now correspond to larger cTEC volumes. However, single section analyses using TEM reveals that cortical thymocytes tend to lose their strong adherence to the nearby cTEC meshwork, thus creating more profound “empty spaces” around thymocytes with aging (Fig. 4F–H; arrows). These data collectively suggest that the ability of thymocytes to adhere to the cTEC network within the labyrinthine voids, is corrupted in aged mice.

Evidence of cTEC cytoreticulum surrounding small groups of thymocytes was confirmed in both Young-Adult (Fig. 4A) and Old mice (Fig. 4B). In Young-Adult mice, however, cTEC were significantly smaller in size, each creating 1–2 labyrinthine voids surrounding a substantial number (~10) of lymphocytes each (Fig. 4A). On the contrary, cTEC were more massive in Old mice, each creating a large number of labyrinthine voids but surrounding a significantly smaller number of thymocytes each, i.e., 1-2 thymocytes (Fig. 4B). These data are consistent with our novel premise that age-dependent hypertrophy of the cytoreticulum represents a compensatory mechanism to the numerical decline of its population, in an effort to preserve the thymopoietic capacity of the cortical parenchyma.

An ultrastructural feature that could further support the establishment of cTEC hypertrophy during aging as a means to compensate for the numerical decline of the cTEC population, is the emergence of multinucleated cTEC in the Old group (Fig. 4I–K, O). In general, multinucleation in thymic epithelial cells has been observed in involuted thymi using electron microscopy⁶⁹, confirming the legitimacy of our observations. It is uncertain whether such multinucleated cTECs develop via cTEC fusion or via a pathological mechanism that includes disturbed cell cycle control. However, these cTEC appear non-apoptotic (Fig. 4I–K), implying they might imitate other cells in the body adopting syncytial configuration, in response to increased metabolic demands⁷⁰. Indeed, ~24-month-old mice exhibited cTEC with up to 5–6 nuclei arranged in a “caterpillar” pattern (Fig. 4K). Binucleated cells, however, were present in all age groups, with an increasing trend during aging (Fig. 4O). These observations could formulate the working hypothesis that the rate of cTEC multinucleation in the cortex is directly proportional with age, although a statistical conclusion may not be easily drawn from qualitative TEM data, due to the low yield of observable events. As a newly proposed feature in our pipeline, however, cTEC multinucleation could represent a key remodeling event observed in various thymic pathologies, especially those associated with underlying involution.

Cytoplasmic vacuoles in TEC had been observed in electron microscopy studies, but their precise functional roles were not appreciated at that time^64,71. It is now established that cTEC employ both thymoproteasome-dependent and macroautophagy-dependent mechanisms to digest, produce, load, and present self-peptides on their surface via the major histocompatibility complex (MHC)^72,73. These degradation pathways are tightly controlled in the normal thymus, to prevent an unrestrained disturbance of TEC function and homeostasis. Ultrastructural analysis of cTEC vacuoles defined their identity as auto(phago)lysosomes (Fig. 4L–N, P–R, S4A–D), implicating them in the macroautophagy machinery⁷⁴. We found that cTEC autolysosomes are distributed throughout the cTEC body, including the perinuclear cytoplasm, as well as more distally within the cytoplasmic projections (Fig. 4L–N and S4E, F). It is very possible that these vacuoles are strategically positioned near the sites of self-peptide loading into MHC molecules. Apart from lysosomes, we distinguished 5 types of vacuoles, along with their relative proportions (Fig. 4P, Q), based on their ultrastructural morphology, as described in previous studies^74–80: Autophagosomes (APGs); Early and Late Autolysosomes (early AUTs and late AUTs, respectively); Multilamellar Bodies (MLBs); Multivesicular Bodies (MVBs) (see Materials and Methods). It has been previously reported the total number of TEC vacuoles decrease with age⁸¹. Even though we noticed changes amongst most of the vacuole types’ percentages, we further studied the early and late AUTs, as they represented the majority of the vacuoles (Fig. 4Q), by calculating their Surface Area (SA_vac) (Fig. 4R) and their Electron-Dense content Ratio (EDR_vac) (Fig. S5A–J). Interestingly, we noticed statistically significant decrease in the SA_vac between the Young-Adult and Old groups, in all AUT types. For early AUTs, the decreased surface area in the Old group might decipher the concurrent increased EDR_vac, possibly indicating no significant changes in absolute content numbers (Fig. 4P and S5J, K). Yet, the fact that the EDR_vac remains unchanged in the Old late AUTs, while the SA_vac is decreased, might hint at probable proportional cargo absolute number decrease, as well (Figs. 4P and S5J, L).

Our final ultrastructural focus regarded cortical macrophages. In general, they exert multiple functions, with the most exemplified ones being the comprising phagocytosis of thymocytes that fail positive selection, and maintenance of blood-thymus barrier, i.e., scavenging of foreign antigens that leak into the thymic parenchyma⁸². Due to their increased phagocytic activity, macrophages are often deemed the “tingible-body” or “starry sky” pattern description in routine histology²⁷. Varying degrees of macrophage activity was observed in mice across the different age groups, with the most prominent feature the presence of secondary lysosomes in their cytoplasm (Fig. S6A–C). With aging, however, lipid-laden macrophages (LLM) were clearly evident (Fig. S6C), and associated with lipid droplets of various sizes and densities in their cytoplasm, as reported⁸³. These data confirm previous findings that cortical macrophages undergo age-dependent attrition, as opposed to medullary counterparts that persist through life⁸². In general, LLMs have been associated with disease progression in inflammatory conditions⁸⁴, and LLM indexes have been successfully used as diagnostic tools⁸⁵. LLM presence in the cortex, as assessed by our TEM analysis, could thus be explored for its potential to inform on the functional characterization and disease progression in the thymus.

Together, the combined OTO-SEM and TEM studies detailed here, complement the digital pathology pipeline discussed in the previous sections. By offering ultrastructural insights into the cTEC meshwork remodeling with aging, these studies not only validate established knowledge in the field, but also enhance our understanding of thymic structure and function under various thymic pathologies, illustrating the intricate changes at the ultrastructural level.

Morphometric analysis of the medullary perivascular space (PVS)

After successful negative selection, the mature SP4/8 thymocytes egress from the medullary parenchyma via tightly controlled chemokine signals, contextualized in microanatomical regions, called perivascular spaces (PVS). Medullary vessels that are specialized for thymocyte egress have been demonstrated to contain double membranes, an inner boundary consisting of pericytes wrapped around blood vessel endothelial cells and an outer boundary consisting of adventitial cells of mesenchymal/fibroblast origin^17,86–91. The two boundaries simultaneously circumnavigate the medullary blood vessel and encompass a well-demarcated area that excludes mTEC, thus appearing as an epithelium-free region⁸⁸ (Fig. 5A). SP4/8 thymocyte egress is dependent upon chemokine signals, most notably sphingosine-1-phosphate (S1P), whose controlled biogenesis/degradation, orchestrates their stepwise emigration, first from the medulla into the PVS and then, from the PVS into the blood vessel^92–94. Because many natural or pathological conditions affect the PVS and thymic output^95–98, the composition and size of the PVS constitutes an important topological feature reflecting to the ability of the medulla to pump out mature thymocytes to the periphery.

Fig. 5 — A Representative image of thymic medulla at high magnification, highlighting the mTEC (KRT5⁺) network and the epithelium-free areas, corresponding to the perivascular spaces (PVS). A special blood vessel stain has not been performed, but blood vessel lumens are apparent in these sections, based on absence of cell nuclei (DRAQ5). Yellow-dashed line, outer boundary of the PVS; red-dotted line, inner boundary of the PVS, corresponding to the blood vessel wall. B, B” Digital pathology pipeline for assessing the fraction of the PVS within the thymus. KRT5/8 IF staining reveals the cortical (KRT5^-KRT8⁺) parenchyma, the epithelium-containing medullary (KRT5⁺KRT8⁺) parenchyma, and the medullary PVS (KRT5^-KRT8^-), here depicted in blue mask (B’). Morphometric assessment of the PVS as a fraction of only the medulla (B’), or the entire thymus (B”). The regions-of-interest used for the normalization, i.e., medulla only, and whole thymus, are demarcated with yellow color in (B’) and (B”), respectively. The formulas used to calculate the two versions of PVS fraction are shown in the upper right corner. Abbreviations: SA_PVS, surface area of the PVS; SA_Pm, surface area of medullary parenchyma; SA_WT, surface area of whole thymus. Representative images of thymi along with magnified inserts (C’–E’) across the three age groups, for morphometric assessment of the PVS. Shown from left-to-right: Young-Adult (C, C’), Adult (D, D’), and Old (E, E’) groups. F, G Quantification of PVS fraction, using digital pathology approaches. Shown in sequence: PVS as fraction of the medulla (PVS_Pm) (F), and PVS as fraction of the whole thymus (PVS_WT) (G). One-Way ANOVA; *p $\leq$ 0.05; **p $\leq$ 0.01; ***p $\leq$ 0.001; ****p $\leq$ 0.0001; ns not-significant. H Analysis of linear correlation between the two established morphometric methods. The correlations are assessed using Spearman’s correlation coefficient analysis with Spearman rho and p-values, and visualized with a scatterplot with simple linear regression line between the two methods. Error bars; Standard Error of Mean. Total N = 26 mice; mice may have been excluded from individual variables, based on exclusion criteria.

Previous studies have reported the quantification of the PVS fraction in the parenchyma of the thymus, either in a non-compartmentalized (i.e., in relation to the surface area of the whole thymus), or a compartmentalized (i.e., as a fraction of medullary parenchyma only) fashion^12,96. The two strategies are informative when assessing either the medullary functions specifically, or in respect to the total thymopoietic capacity, respectively. Based on prior knowledge that the medullary PVS is an epithelium-free area¹⁷, we drew regions of interest (ROIs) devoid of the KRT5⁺KRT8⁺ epithelium around the medullary PVS (Fig. 5A, B), and calculated the proportion of the surface area covered by the demarcated PVS within the KRT5⁺KRT8⁺ surface area of the medullary region (PVS_Pm) (Fig. 5B’ and equation 8), or within the KRT8⁺ (whole-thymus) surface area (PVS_WT) (Fig. 5B” and equation 9). In consistency with prior observations on human thymi^96,99,100, we found that the PVS fraction is significantly increased over time, regardless of the morphometry used (Fig. 5C–G and S7A–C, A’C’). Impressively, approximately 50% of the medullary parenchyma and 25% of the entire thymic parenchyma correspond to the PVS zone in the Old group (Fig. 5F, G). In addition, there is a significant linear correlation (Spearman rho = 0.96) between the two morphometries (Fig. 5H), indicating that the methods are in fact orthogonal, and as such, equally suitable for studying factors affecting thymocyte emigration.

Topological assessment of the medullary thymic epithelial cell (mTEC) network

As opposed to cTEC, mTEC are scattered in freeform arrangement without developing any meshwork architecture in the medulla⁴. The interaction among different mTEC subsets and the single-positive (SP) thymocytes, controls the fate and success of negative selection, and eventually thymocyte egress⁴. Due to the absence of meshwork architecture, an identical pipeline to that of cTEC, e.g., including skeleton analysis, could not be implemented for mTEC topology. However, as in the case of cTEC, SP4/8 thymocytes also interact through direct physical contact with various mTEC subsets to complete their developmental journey⁴. As such, the two features, “mTEC Volume Index” (VI_mTEC) and “mTEC Interface Index” (II_mTEC), were employed in a similar fashion, because they are equally informative with the cTEC equivalents (Fig. 6A–D’ and equations 10-11). To this end, the KRT5/8 double-IF-stained sections used above for the assessment of cTEC network topology, were also utilized for mTEC morphometry development. To circumvent biases from potential inclusion of the epithelium-free regions corresponding to medullary PVS, we quantified the VI_mTEC and II_mTEC features, after excluding the PVS. Our analysis suggested that mTEC did not significantly decline with aging (Fig. 6E, F). At first glance, this observation implies a relative resilience of the mTEC compartment, especially when compared to the cTEC. This outcome aligns with recent literature suggesting that aging predominantly affects the cTEC population, resulting in numerical and functional impairments, in turn, compromising thymopoiesis. In contrast, the impact of aging on mTEC is less profound^49,57,100.

However, prior reports have demonstrated multiple functional mTEC impairments with aging, either by measuring the expression of key genes of mTEC functionality and differentiation, or by assessing mTEC progenitor activity⁴⁹. Given that certain medullary functions, such as negative selection, are uniquely dependent on highly-specialized mTEC subsets, such as AIRE⁺ mTEC^hi and the newly discovered thymic mimetic cells, we reasoned that our topological investigations using digital pathology should additionally focus on mTEC subsets, instead of only on the entire mTEC network. To this end, we developed two non-redundant morphometric features of mTEC subset distribution, the density and the homogeneity of distribution.

The methods rely on implementing lineage markers that will help discern mTEC subset(s) of interest from background mTEC populations (Fig. 6G–I). Here, we demonstrate the methods for two distinct mTEC subsets, the AIRE⁺ mTEC^hi subset, a master-regulator of negative selection^101–103, and the post-AIRE DCLK1⁺ tuft cells, which belong to the thymic mimetic cells and mediate neuroendocrine and immunological functions in the medulla^104–107. To distinguish these two mTEC subsets from other mTEC, we introduced antibodies, specifically targeting AIRE and DCLK1 (Fig. 6G’–I’), which exhibited patterns of expected expression in the nucleus and the cytoplasm, respectively. Individual mTEC subsets were identified automatically, using a pixel/particle classifier (Fig. 6I-I’). Next, the medullary parenchyma was demarcated following exclusion of the PVS, thus limiting all ensuing analyses to the epithelium-containing areas.

The first morphometric feature, i.e., density of AIRE⁺ (d_aire) and DCLK1⁺ (d_dclk1) mTEC, is expressed as AIRE⁺ and DCLK1⁺ mTEC numbers per mm² of epithelium-containing medullary parenchyma (Fig. 6J, J’ and equations 12-13). For the second morphometry, i.e., homogeneity of AIRE⁺ (h_aire) and DCLK1⁺ (h_dclk1), each IF image is first segmented into multiple, equal, rectangular sectors and then, auto-fluorescent erythrocytes and other imaging artifacts are manually removed, before subjecting the images to homogeneity assessments (Fig. 6K, K’, S8A–C, B’, C’). In each individual, artifact-corrected sector, the densities of AIRE⁺ and DCLK1⁺ mTEC are first calculated (equations 14-15), and the homogeneity among sectors is then assessed (equations 16-19). If the mTEC subset of interest was distributed in a homogenous arrangement within the medullary parenchyma, each sector would exhibit similar density and as such the homogeneity index for the respective mTEC subset would be high (Fig. 6L). If, however, the mTEC subset of interest was heterogeneously arranged, i.e., it was arranged in small cellular clusters within the medullary parenchyma, then such cell clusters would stochastically fall within only a small portion of these sectors, thus leaving other sectors devoid of cells. The latter arrangement would result in dissimilar density profiles among the different sectors, and as such, a lower homogeneity index for the respective mTEC subset (Fig. 6L). Based on this principle, the h_aire and h_dclk1 calculate the fraction of the medullary sectors in each thymus, whose individual sector d_aire and d_dclk1 fall within a single referent standard deviation from their mean density. Based on these assumptions (equations 16-19), the h index gets values between “0” and “1”, with a “1” implying a perfectly homogeneous mTEC cell arrangement where all sectors present with similar cell density, while lower h values express varying degrees of clustered arrangement.

We assessed the morphometric features of mTEC subset distribution in our mouse model of age-dependent thymic involution. It has been previously demonstrated that the absolute number of AIRE⁺ mTEC^hi declines significantly during aging¹⁰⁸. Using the d_aire index, however, we found that their density remains unaltered during aging (Fig. 6M, N, S8D). This observation implies that the numeric decline of AIRE⁺ mTEC^hi cells follows proportionally the reduction of the medullary volume, thus keeping a consistent density throughout aging. Moreover, there are no changes in the h_aire index (Fig. 6P), suggesting that aging does not impact their spatial arrangement, either. Together, these observations indicate that despite the documented mTEC numerical decline¹⁰⁸, the interactions between SP4/8 thymocytes and AIRE⁺ mTEC^hi cells are not essentially compromised from age-dependent intrathymic remodeling.

Finally, we performed similar investigations on DCLK1⁺ tuft cells, a thymic mimetic cell subset that has undergone further terminal differentiation from the AIRE⁺ mTEC^hi cells, thus, overall belonging to the post-AIRE mTEC subsets^4,105,109. Thymic tuft cells, similar to other mimetic cells in the thymus, present developing SP thymocytes with a unique subset of lineage-restricting tissue-restricted antigens, while AIRE⁺ exert broader functions, indicating complementarity of the two populations^{15,104,105,109}. Interestingly, we uncovered a significant decline of thymic tuft cells with aging, as assessed through d_dclk1 measurement (Fig. 6M, O, S8E). This decline began early at the Adult group, but did not reveal any signs of further decline in the Old mouse group (Fig. 6M, O, S8E). Accompanying the density decrease, we strikingly found an increase in h_dclk1, thus suggesting their sparser arrangement in the aged thymus (Fig. 6Q). Being fewer in number than AIRE⁺ mTEC^hi, thymic tuft cells attempt to achieve more strategic positioning via spatial remodeling during aging, to increase the probability of interacting with SP developing thymocytes. Together, these observations indicate that combining density and homogeneity may provide detailed resolution of dynamic shifts of mTEC subsets in thymus disease.

Early-life cholera-toxin (CT) exposure improves medullary architecture in a mouse model of DMBA-driven thymic injury and carcinogenesis

Next, we sought to establish that our morphometric pipeline (Fig. 1A) is applicable beyond age-dependent thymic involution, serving as a generalized tool for assessing intrathymic remodeling. The carcinogen 7,12-dimethylbenz[a]anthracene (DMBA) induces apoptosis in lymphoid organs such as the thymus, spleen, and Peyer’s patches¹¹⁰, yet its effects on the cTEC/mTEC network and thymic regeneration remain unclear. Elsewhere, early-life exposure to cholera toxin (CT), known to mitigate cancer and inflammatory conditions^111–114, has been shown to reduce the incidence and multiplicity of DMBA-induced tumors¹¹⁵. In the same study, CT-exposed, DMBA-treated mice exhibited lower neutrophil infiltration and higher expression of CD3, CD8, and FOXP3 in the thymic medulla, thus suggesting reduced inflammation and improved thymocyte occupancy during regeneration¹¹⁵. Consistent with evidence that microbial stimuli promote immune regulation and protect from autoimmunity¹¹⁶, we thus hypothesized that early-life CT exposure primarily enhances thymic medullary infrastructure. To test this, we reanalyzed thymic tissues from a small cohort (DMBA-ONLY n = 3; DMBA + CT n = 3) from our original study¹¹⁵, using our morphometric pipeline to evaluate thymopoiesis and to visualize CT-mediated changes.

Animals received DMBA from weeks 8 to 13 after birth, while only half were exposed to CT early in life (weeks 2–5) (Fig. 7A). Thymi collected two weeks after finalizing DMBA treatments were analyzed using dual KRT5/KRT8 immunofluorescence, revealing distinct cortex-medulla compartmentalization (Fig. 7B), similar to age-dependent involution (Figs. 2–6). Notably though, CT-exposed mice showed an expanded medullary area and a proportionally reduced cortical area (Fig. 7B–D), resulting in a significantly lower CMR_WT-IF (Fig. 7E). This ratio, averaging ~2 in DMBA + CT mice, is consistent with a morphologically “normal” thymus²⁷, and interestingly, it was similar to the CMR observed in the Adult mice (4–8-month-old) from the mouse model of age-dependent thymic involution (compare Fig. 2E with Fig. 7E), which could be considered as a “matched-age” control for the CT-exposed DMBA-treated mice. On the other side, DMBA-only-treated mice exhibited CMR of ~5–6, indicating medullary atrophy and poor cortical support, consistent with prior studies on acute thymic involution by cytotoxic drugs¹¹⁷. Together, these findings support our hypothesis that early-life CT exposure enhances medullary regeneration following DMBA-induced thymic injury.

An in-depth analysis of cTEC network topologies, i.e., VI_cTEC, II_cTEC, BI1_cTEC, BI2_cTEC, and BI3_cTEC, revealed no major differences in TEC distribution between DMBA-treated animals with or without early-life CT exposure (Fig. 8A–I). Morphometric quantifications followed the same procedures as in the age-dependent involution model. To further illustrate the pipeline’s versatility, we performed regional analyses within the cortex, separately assessing the outer (subcapsular) and inner (deep) cortical zones (Fig. 8E–I and 8E–I’, respectively). These refined measurements could theoretically uncover region-specific remodeling processes, given that developing thymocytes migrate along a U-shaped path from the corticomedullary junction to the outer cortex and back^23,118–120. While some local variations were noted, such as a significant increase in VI_cTEC within the outer cortex of CT-exposed animals (Fig. 8E), the deep cortex remained unaffected (Fig. 8E’). As the thymic mesenchyme adjacent to the capsule interacts with subcapsular (outer cortex) cTECs and supports post-injury regeneration^3,19,121, these findings indicate that CT exposure may enhance this interface after DMBA-induced damage. Overall, however, the cortex appeared largely unaffected by early-life CT exposure.

Fig. 8 — Representative regions of murine thymi (from Fig. 7B, B’) treated with DMBA-only (A–D) or DMBA + CT (A’–D’), with a focus on cortical parenchyma. In sequence: Tissue overview from low-power fields (LPFs) stained with dual KRT5/KRT8 immunofluorescence (IF), to demarcate cortex-medulla distinction (white-dotted line) in DMBA-only (A) and DMBA + CT (A’) mice; magnified insets of the corresponding white-dotted boxes shown in (A–A’), revealing only the KRT8 channel along with DAPI in DMBA-only (B) and DMBA + CT (B’) mice; the same images as (B, B’) are shown after thresholding that creates a binary KRT8/cTEC mask (C, C’); the same images as (B, B’) are shown after skeletonization (D, D’). Quantification of cTEC network topologies in the outer (E–I) and deep (E’–I’) cortex independently, in mice treated with DMBA-only or DMBA + CT. In sequence: morphometric assessment of outer/deep cTEC network topology, using cTEC volume index (VI_cTEC) (E, E’); cTEC interface index (II_cTEC) (F, F’); cTEC branching indices, i.e., BI1_cTEC (G, G’), BI2_cTEC (H, H’), and BI3_cTEC (I, I’). Unpaired Student’s t test; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001; ns not significant. Error bars; Standard Error of Mean. Total N = 6 mice.

We next examined the medullary morphometries, which included analyses of both PVS (Fig. 9A–D) and mTEC network topography (Fig. 9E–I). Interestingly, there was no significant difference in PVS when expressed as a fraction of the total thymic parenchyma (PVS_WT) (Fig. 9C). However, when normalized to the medullary compartment alone (PVS_Pm), a marked reduction was observed in mice that received DMBA + CT (Fig. 9D). This is not a discrepancy but rather a normalization effect, since PVS_WT incorporates the cortical parenchyma, which was more extensive in CT-exposed animals (Fig. 7B, C), thereby diluting medulla-specific differences when assessed across the whole organ. The pronounced decrease in PVS_Pm therefore reflects a localized alteration in medullary organization, potentially linked to changes in thymocyte dynamics. Specifically, a smaller PVS fraction could signify either reduced retention of thymocytes within the medulla prior to negative selection or enhanced perivascular turnover associated with increased thymocyte egress. Considering that our previous study using this model¹¹⁵ demonstrated improved immune functionality following early-life CT exposure, it is more plausible that the observed reduction in PVS_Pm corresponds to enhanced thymocyte egress efficiency. Consistent with these findings, the VI_mTEC and II_mTEC morphometries (Fig. 9E–I) were calculated within the medullary parenchyma after excluding the epithelium-free PVS regions to better assess thymocyte-mTEC interactions. While both parameters exhibited modest reductions (Fig. 9H, I), the significantly decreased II_mTEC value suggests increased thymocyte occupancy within the medulla, where the presence of more abundant SP4⁺/SP8⁺ thymocytes may physically displace KRT5⁺ mTECs, effectively reducing their surface area per defined medullary unit.

Together, these analyses indicate that early-life CT exposure promotes improved medullary regeneration and thymocyte egress following DMBA-induced thymic injury. More broadly, these findings demonstrate that our morphometric pipeline sensitively captures context-specific alterations in TEC architecture, even in models mechanistically unrelated to aging, highlighting its versatility as a generalized framework for assessing thymic remodeling.

Data integration using principal component analysis (PCA)

Finally, we integrated all morphometric parameters derived from our pipeline for both animal models into a principal component analysis (PCA), to uncover patterns of structural variation and identify the dominant morphometric trajectories associated with thymic aging (Fig. 10B) and acute remodeling (Fig. 10B).

In the mouse model of age-dependent thymic involution (Fig. 10A), PCA revealed that thymic architecture segregates strongly by age along PC1, which accounted for 44% of the total variance. This axis represents the dominant structural trajectory of thymic involution, which mostly encompasses the adaptive remodeling of the cTEC network (Fig. 3I–M), enlargement of the perivascular spaces (Fig. 5F, G), and reorganization of mTEC subset topology (Fig. 6N–Q). The clear left-to-right progression of Young-Adult, Adult, and Old thymi underscores a continuous morphometric transition consistent with age-related remodeling (Fig. 10A). Notably, within the intermediate “Adult” group, the 4- and 8-month-old animals formed two discernible subclusters along the PC1, only partly overlapping, thus illustrating the sensitivity of the pipeline in detecting even subtler morphometric shifts that precede an overt structural involution (Fig. 10A). On the other hand, PC2 (14.7%) captured subtler, intra-group variability likely reflecting more localized remodeling rather than age-dependent changes (Fig. 10A).

Interestingly, in the DMBA-induced thymic injury model, PCA revealed a strong segregation of samples along the first principal component (PC1), which accounted for 66% of the total variance, indicating a more distinct separation between experimental groups compared to the age-dependent model (Fig. 10B). The DMBA-ONLY and DMBA + CT groups formed two clearly defined clusters along PC1, demonstrating that the analysis captured robust, condition-dependent architectural differences in thymic organization (Fig. 10Β), as also confirmed by the individual topological measurements (Figs. 7–9). In contrast, PC2 (~14.1%) again accounted for localized or within-group variability without contributing significantly to the overall separation (Fig. 10B). Together, PCA analyses from both models demonstrate the robustness and sensitivity of our morphometric pipeline, confirming its ability to capture both gradual, age-dependent transitions and discrete, condition-driven architectural alterations in thymic structure.

Discussion

In this study, we developed a comprehensive digital pathology pipeline to analyze thymus microanatomy, combining multiple orthogonal approaches. Traditionally, research in the field has used population-based methods, such as flow cytometry, to study intrathymic dynamic changes, thus focusing on numerical ratios among cell subsets of interest¹²¹. However, these methods lack contextual information, such as the spatial distribution of stromal cells and thymocytes, within these complex thymic environments. Despite significant advancements in imaging methods, the microanatomy of the thymus remains underexplored, compared to other, particularly lymphoid organs. Our pipeline circumvents this knowledge gap, offering insights into various pathologies, including acute or chronic thymic involution, and specific diseases affecting the thymus. Overall, our pipeline serves as a versatile tool for immunology labs specializing in thymus biology, homeostasis, and regeneration.

From development until preadolescence, the thymus generates a critical lifelong T cell repertoire to support systemic immunity, immune surveillance, and central tolerance^39,122,123. However, it undergoes an age-dependent decline shortly after reaching the peak of functional maturity in adolescence. In late adulthood, systemic immunity and peripheral immunosurveillance becomes less dependent on the thymus, and more dependent on expansion of peripheral memory T cells^124–126. Nevertheless, a recent epidemiologic study has challenged this premise¹²⁷, suggesting that the thymus significantly contributes to systemic immunity even at older ages. Age-dependent thymic decline correlates with higher cancer incidence^5,128–130, suggesting that targeting thymic involution could help prevent cancer and promote longevity¹³¹. Among thymic disorders, age-dependent involution is the most well-characterized thymic pathology¹³², and as such, served as the primary focus for demonstrating our pipeline’s utility. Additionally, the thymus is vulnerable to acute involution^9,133,134, from factors like chemo/radiotherapy¹³⁵, infections¹³³, sexual steroids^136,137, pregnancy^138,139, and obesity^140,141. Therefore, our pipeline may be utilized to offer insights in studies aimed at dissecting the molecular mechanisms and reversing thymic involution.

In addition, our digital pathology pipeline is a vital tool for studying thymic structure in various immunodeficiency and autoimmune disorders. For instance, MHC-II deficiency (MHCII-D), also known as Bare Lymphocyte Syndrome (BLS), a combined immunodeficiency syndrome due to mutations in genes regulating expression of MHC molecules¹⁴², is marked by increased corticomedullary ratio due to hypoplastic medulla, and a decline of AIRE⁺ mTEC^hi cells¹⁴³. Similarly, AIRE deficiency leading to the autoimmune-polyendocrinopathy-candidiasis (APECED) syndrome, is characterized by loss of self-tolerance affecting multiple organs¹⁰². The syndrome is associated with medullary atrophy and diminished AIRE⁺ mTEC^hi cells¹⁴⁴. Interestingly, it has been surmised that underlying the widespread loss of self-tolerance in peripheral tissues may be the dysregulation of terminally-differentiated post-AIRE subsets¹⁴⁵. Of note, thymic tuft and myoid cells, two well-characterized thymic “mimetic” cell subsets, present unique chemosensory cell- and muscle cell-specific Tissue-Restricted Antigens (TRAs), respectively, to developing SP4/8 thymocytes, hinting on the APECED symptoms occurring in various endocrine organs and skeletal muscle^{104,105,107,109,146,147}. A deeper understanding of the developmental pathways and spatial distribution of specific mTEC subsets, as demonstrated in our morphometric analyses, could thus offer insights on the onset and progression of primary immunodeficiencies, given that relevant mouse models are currently available^148–150. In summary, our pipeline may be instrumental in exploring cTEC/mTEC adaptations in preclinical mouse models for rare immunological disorders.

A key feature of our morphometric approaches is their “orthogonality” as they offer alternate strategies for evaluating morphological characteristics of thymic environments and their cells. The most conspicuous example of orthogonality is the measurement of corticomedullary ratios, using either a “whole-thymus” or a “thymic-lobule” approaches. Interestingly, different imaging modalities within the same approach, i.e., IF or HE for the “whole-thymus” approach, revealed high correlation, suggesting they provide overlapping information. However, the same imaging modality, i.e., HE across different approaches: “whole-thymus” versus “thymic-lobule”, have been able to both capture the decreased corticomedullary ratio with aging, albeit via offering non-overlapping information, such as the status of thymic lobulation in the aged thymus. Although pathologists evaluate whole thymus slides using HE staining, according to the INHAND guidelines^1,26,28, and the method is widely adopted in laboratories, whole-thymus approaches may suffer from significant limitations, such as sample orientation in the paraffin block, depth of sectioning, and tissue preservation, raising issues of reproducibility. Orthogonal approaches in morphometric measurements will thus provide high credibility of the reported data.

In imaging, “skeletonization” transforms binary objects to one-pixel-thick representations, facilitating topological analysis and reducing dimensionality. This technique, applicable to both 2D and 3D models of any branched network^151,152, is notably used in studying vasculature development and angiogenesis¹⁵³. Here, we adapted an ImageJ plugin⁵⁶, to assess topological descriptors, such as number of branches/junctions, leveraging the meshwork architecture of the cTEC network, which resembles angiogenic vasculature^4,154. To our knowledge, this is the first instance of using “skeletonization” to study the topology of the cTEC network in the murine thymus. Previous research documented the decline of the thymic cortex versus the medulla with aging, leading to decreased cTEC number and thymopoietic potential^28,39,155. Our findings in Old mice confirmed this decline via an “artificial” increase in cTEC network density, due to fewer thymocytes percolating through the network^39,155. Interestingly, we additionally observed that the cTEC network undergoes hypertrophic growth in response to this decline, adjusting to the increased functional demands of thymopoiesis. The unique set of morphometric features we developed extracts key details of cTEC meshwork topology, revealing that cTEC adaptations relate to increased branching capacity without alterations in cell body thickness or tortuosity. Together, our pipeline not only corroborates earlier findings on age-dependent thymic changes, but also provides detailed insights into the nature of these changes. For instance, it unravels new possibilities for investigating cytoskeletal dynamics, as a putative strategy to tackle cortical adaptations in the aged thymus towards achieving better regenerative potential.

Multinucleation in TEC in both mice and humans has been sporadically reported. An older study using the C3H mouse strain found multinucleated cTEC and mTEC, with strong prevalence in the latter, with incidence increasing with age, and linked to thymic involution¹⁵⁶. Similarly, binucleated and multinucleated cTEC were demonstrated in an acute thymic involution model, with mouse exposure to high altitudes, and confirmed in human patients⁶⁹. These studies suggest that cTEC become multinucleated under certain (patho)physiological conditions, although the molecular prerequisites remain undetermined. In general, multinucleation arises either from cell fusion resulting in “syncytia”, or from nuclear division without cytokinesis resulting in “coenocytes”¹⁵⁷. It has been speculated that multinucleation outside normal tissue development may enhance functional capacity, and facilitate rapid responses to high metabolic and regeneration demands^158–160. Notable examples of this process are monocytes fusing into osteoclasts to facilitate bone remodeling⁷⁰, and fused hepatocytes that promote liver regeneration¹⁶¹. Similarly, in the thymus, multinucleation might reflect a response to increased thymopoietic demands, which promote cTEC hypertrophy and syncytial configuration during involution. Electron microscopy has accurately displayed the potential of syncytial adaptation in thymic involution⁶⁹. However, these studies are unable to probe the molecular basis of this adaptation, highlighting the need for future research to understand the functional decline of the aging thymus.

Thymocyte development relies on the presentation of MHC-self-peptide complexes by the cTEC/mTEC¹⁶². The wide repertoire of cortical self-peptides is orchestrated by unique proteolytic systems in cTEC, such as thymoproteosome β5t⁷², cathepsin-L¹⁶³, thymus-specific protease PRSS16¹⁶⁴, and the macroautophagy pathway¹⁶⁵. Macroautophagy, a process capturing part of the cTEC/mTEC cytoplasm including organelles and vesicles for digestion in autophagolysosomes, is crucial for generating self-peptides, especially for MHC-II presentation^73,166,167. Prior electron microscopy studies have identified autophagolysosomes in TECs, but the complexity of TEC heterogeneity and the specific identification of these vacuoles in different TEC subsets was not appreciated at the time^64,168. Although we did not specifically explore the functions of this proteolytic system, we found that autophagolysosomes are strategically positioned both near the perinuclear cytoplasm and along cytoplasmic extensions surrounding thymocytes, possibly optimizing the sites for self-peptide loading. Age-dependent alterations in the number and content of these autophagolysosomes were evident, and possibly useful for extrapolating functional correlates of the cTEC network. Although not developed fully in our study, incorporating ultrastructural morphometry into our digital pathology pipeline has greatly enhanced our ability to identify the structural and functional deficits of the cTEC network with aging.

Upon successful negative selection, mature SP4/8 thymocytes leave the thymus trough a microanatomical compartment called the “perivascular space” (PVS)^{17,86–88,90,96,98,169}. The PVS, which encircles medullary blood vessels, provides a distinct egress path, consisting of two concentric mesenchymal sheaths, an outer boundary made of adventitial fibroblasts and an inner boundary corresponding to the pericyte coating of the blood vessel³. To exit the thymus, SP4/8 thymocytes first migrate into the PVS passing through the first boundary, and then into the blood passing through the second boundary, following chemokine gradients, such as sphingosine-1 phosphate (S1P)^{21,93,170–174}. The size and distribution of the PVS is therefore a direct functional correlate of thymic output¹⁷. Although it is barely visible in infants, the PVS significantly expands with age in humans, indicating egress defects in older individuals⁹⁶. In our study, the age-related PVS expansion was observed using our novel morphometric analyses, demonstrating again, the orthogonality across different methods. Interestingly, the PVS serves as a niche for B cells and plasma cells, contributing to immune defense in healthy individuals⁸⁹, but also to autoimmune responses under pathological conditions (e.g., Myasthenia gravis)¹⁷⁵. Many of these conditions may critically affect PVS microanatomy without necessarily altering the overall corticomedullary architecture¹⁴⁵, and as such, our morphometric approaches are suitable for providing insights on thymocyte emigration in thymic involution and disease.

Assessing both the density and homogeneity of various mTEC subsets within the medullary parenchyma is crucial for understanding the biological processes underlying intrathymic remodeling across different thymic pathologies, such as in the case of age-dependent decline. By analyzing these two metrics concurrently, we gained valuable insights into the intricate balance of mTEC subsets maintained within the aged thymus. Our findings revealed that while the density and homogeneity of AIRE⁺ mTEC^hi cells are relatively preserved, indicating a stable potential for interactions between AIRE⁺ mTEC^hi and SP4/8 thymocytes–and thus ensuring exposure to a broad spectrum of tissue-restricted antigens, there was notable decline in the density of DCLK1⁺ tuft cells, a post-AIRE thymic mimetic cell derivative. Interestingly, despite their reduced density, tuft cells demonstrate a more homogenous distribution within the aged thymus, possibly reflecting their strategic repositioning to optimize their interactions with SP4/8 thymocytes. This observation suggests a deliberate biological adaptation, favoring the preservation of AIRE⁺ mTEC^hi cells, which present a wider array of TRAs compared to that of tuft cells^{15,104,105,109}. This strategic preservation of AIRE⁺ over DCLK1⁺ mTEC cells may be an evolutionary compromise to minimize the impact on central tolerance mechanisms and prevent the escape of self-reactive T cells¹⁰⁸. Our study underscores the importance of considering both cell density and distribution patterns in understanding adaptive responses of the thymus to aging or other pathologies, and its implications for maintaining immune competence and preventing autoimmunity.

While our digital pathology pipeline provides powerful insights into thymic remodeling and spatial organization, its results must be interpreted with caution. The assays presented here are morphological rather than functional, and extrapolating functional conclusions from structural metrics alone may be misleading. For instance, proximity-based measurements such as the TEC-thymocyte interface (documented here as II_ctec and II_mtec) do not necessarily confirm the presence of MHC-peptide/TCR-based immunological contacts at this resolution. The true strength of this pipeline emerges when integrated with functional assays. In our DMBA-induced thymic injury model for example, early-life cholera toxin exposure enhanced medullary occupancy by CD8⁺ and FOXP3⁺ T cells¹¹⁵, coinciding with reduced PVS area and altered mTEC network configuration (Fig. 9). When considered alongside these functional data, the morphometric findings suggest improved thymocyte trafficking and egress. However, without such functional context, these same structural changes could be misinterpreted as thymocyte retention or stalling, underscoring the need for cautious and integrative interpretation.

Multivariate approaches such as principal component analysis (PCA) are widely used to provide an integrated overview of complex, multidimensional datasets¹⁷⁶, including in morphometric imaging¹⁷⁷. In our study, PCA served as a powerful exploratory tool to visualize global patterns of thymic remodeling and to identify combinations of morphometric variables that best distinguished experimental groups. However, PCA alone cannot specify which individual parameters drive these separations or in what direction; rather, it reveals overarching variance structures that must be interpreted in the context of these underlying morphometries. Thus, detailed analysis of each parameter remains paramount to assign biological meaning, while functional correlates are required to draw safe mechanistic conclusions. For instance, our two models, age-dependent thymic involution and DMBA-induced injury with/ without CT exposure, demonstrated that similar PCA segregation patterns (i.e., clear separation along PC1) can emerge from fundamentally distinct remodeling processes. Whereas aging was characterized mostly by cortical attrition and relative medullary resilience (Figs. 3–6), the DMBA with/without CT model, showed another trend with predominant medullary adaptation and cortical preservation (Figs. 8–9). This comparison underscores both the strengths and the limitations of PCA: it effectively summarizes structural variance across models but must be anchored to the specific morphometric signatures driving these changes to avoid overgeneralized interpretations.

In conclusion, our comprehensive digital pathology pipeline not only offers a holistic view of the thymic environments with a broad spectrum of morphometric analyses, but also serves as a practical toolset for other laboratories interested in thymus pathology (Supplementary Information). This integrated approach enables the detailed exploration of thymic architecture and its adaptations to aging and disease, providing a platform for the study of cellular distributions, density, and structural dynamics. By offering a versatile and accessible resource, our work facilitates unified investigations into thymic biology, paving the way for novel therapeutic strategies aimed at enhancing thymic regeneration and bolstering immune resilience across the lifespan.

Methods

Animal subjects

Ethics statement

All studies involving mice were conducted in accordance with NIH regulations concerning the care and use of experimental animals with the approval of IACUC of Molecular Imaging, Inc. (Ann Harbor, MI), a facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC), or with the approval of Albert Einstein College of Medicine Animal Care and Use Committee. In this study, we have complied with all relevant ethical regulations for animal use.

Mouse model of age-dependent thymic involution

Wild-type female FVB/NJ mice were acquired from NCI at 5–6 weeks of age, housed, and maintained in groups of 5 animals. After a post-arrival adaptation of 1-week, the mice were introduced to an “aging protocol”, during which they underwent weekly clinical and behavioral assessments, including body weight and pain assessment, based on an established grimace scale^178,179. None of the mice fulfilled any endpoint criteria that could be regarded as confounding conditions in our study. Mice were humanely sacrificed as soon as they reached pre-specified age endpoints, and classified as “Young-Adult” (1.5-months, n = 5), “Adult” (4–8 months, n = 10), and “Old” (>18-months, n = 11), into which they were allocated randomly. Euthanasia was performed after isoflurane anesthesia followed by cervical dislocation. Mice housed under this aging protocol were treated identically until reaching the endpoint criteria. Due to established gender-related differences in the development of T cells^{141,155,180,181}, it is generally not recommended to pool together male and female mice in studies related to thymic functions. When compared to males for example, aged female mice tend to have higher numbers of TEC, including the AIRE⁺ mTEC subpopulation, which is critical for the process of negative selection^155,180. For purposes of method development, the current study has included female mice only. However, our proposed readouts can be performed in mice of any sex, age, strain, and body condition, without limitations. Finally, for evaluating strain-dependent alterations during aging, three 24-month-old C57Bl/6 J mice, provided by the Maria-Cuervo lab, were included in the “Old” group, in addition to the FVB/NJ mice.

Mouse model of DMBA-driven chemical carcinogenesis

A mouse model of chemical carcinogenesis induced by 7,12-dimethylbenz[a]anthracene (DMBA) was employed, incorporating two cohorts: one pre-treated with cholera toxin (CT) and one serving as a control. Mouse housing conditions, experimental design, and tissue collection/processing procedures have been described in detail elsewhere¹¹⁵. Briefly, female FVB/N mice were used. CT was administered by gastric gavage once weekly for four consecutive weeks, starting at 2 weeks of age, at doses of 5, 7.5, 10, and 10 µg per mouse at weeks 2, 3, 4, and 5, respectively. DMBA was subsequently administered by gastric gavage at 1 mg/mouse weekly for six consecutive weeks, beginning at 8 weeks of age. Mice were sacrificed two weeks after completion of DMBA treatment (at 15 weeks of age). In the original study, experimental group sizes were as follows: DMBA-ONLY (n = 11) and DMBA + CT (n = 12). For the present proof-of-principle analysis, three representative cases (n = 3) from each group were selected, in which the thymic tissue remained abundant and intact. Histological verification of thymic integrity was performed by H&E staining.

Necropsy procedures

Complete necropsies were conducted on sacrificed mice, according to standardized protocols^182,183, and age-associated lesions were recorded. Before sacrificing, whole body weight measurements were documented, and gross examinations were conducted on the extracted thymus and spleen after trimming peri-thymic and peri-splenic adipose tissue, respectively. The organs were weighted, fixed into 10% neutral buffered formalin (Fisher Scientific) for 24–48 h, and paraffin embedded (FFPE). The weights were reported both as absolute and as “relative” values, normalized to body weights [Thymus Weight Index (TWI); Spleen Weight Index (SWI)].

Histology and immunofluorescence staining, and image acquisition

Tissue preparation

Unless otherwise stated, all slides, 5 μm thick, were first dewaxed twice in xylene (Fisher Scientific) for 10 min, then serially dehydrated twice in 100^o ethanol (Fisher Scientific) for 10 min, twice in 95^o ethanol (Fisher Scientific) for 10 min, and once in 70^o ethanol (Fisher Scientific) for 10 min, and finally rehydrated in distilled water for 5 min.

Hematoxylin & Eosin (HE) staining

All slides were stained using modified Meyer’s hematoxylin (Fisher Scientific) for 5 min, then differentiated in 1% HCl, 70% ethanol (Fisher Scientific) for 5 s, re-stained using Eosin Y (Fisher Scientific) for 1 min 45 s, cleared (95^o ethanol, 100^o ethanol, and xylene), and finally mounted with Permount (Fisher Scientific) medium.

Immunofluorescence staining

Heat-Induced Antigen Retrieval (HIER) was performed with Citrate Buffer, pH 6.0 (Novus Biologicals) or EDTA, pH 8.0 (Epredia) in a steamer for 21 min, followed by incubation with blocking solution (Goat serum, dilution 1:20 in Phosphate Buffered Saline Tween-20, PBST, 0.05% (v/v), Fisher Scientific) for 1 h at room temperature. Primary antibody staining was performed overnight at 4°C, using: anti-Keratin-5 (Chicken anti-mouse; polyclonal; dilution 1:500 in blocking solution; Biolegend); anti-Keratin-8 (Guinea Pig anti-mouse; polyclonal; dilution 1:100 in blocking solution; Origene); anti-AIRE (Rat anti-mouse; monoclonal: 5H12 clone; dilution 1:500 in blocking solution; Invitrogen); anti-DCLK1 (Rabbit anti-mouse; monoclonal: JA11-03 clone; dilution 1:200 in blocking solution; Invitrogen); anti-CD3 (Rabbit anti-mouse; monoclonal: SP7 clone; dilution 1:100 in blocking solution; Abcam); and anti-B220 (Rat anti-mouse; monoclonal: RA3-6B2 clone; dilution 1:100 in blocking solution; R&D systems). Secondary antibody staining was performed for 50 min at room temperature with: Goat anti-Rabbit Alexa488 (Invitrogen); Goat anti-Chicken Alexa488 (Invitrogen); Goat anti-Guinea Pig Alexa555 (Invitrogen); Goat anti-Rabbit Alexa555 (Invitrogen); Goat anti-Rat Alexa555 (Invitrogen); and Goat anti-Rat Alexa647 (Invitrogen), all diluted at 1:200 in blocking solution. Nuclear counterstaining was performed for 5 min with 4′,6-diamidino-2-phenylindole (DAPI, dilution 1:1000 in PBST 0,05%; Novus Biologicals), or with Deep Red Anthraquinone 5 (DRAQ5, dilution 1:500 in PBST 0,05%; Invitrogen), before mounting with ProLong Gold anti-fade reagent (Invitrogen).

Microscopy and slide scanning

After completion of the HE or IF staining as described above, the slides were kept in the dark at room temperature, until imaged on a 3D HISTECH P250 Flash III automated slide scanner, using a 20 × 0.75NA objective lens. The instrument automatically focuses and stitches the images into one single image file. For the IF in particular, each channel was acquired separately, with a set exposure time, and then combined into a multichannel overlay image. Further image analyses and morphometric development were performed using CaseViewer, ImageJ, and QuPath, as detailed in Supplementary Information. A protocol-style guide describing image analysis and the extraction of raw data for calculating the morphometries using the provided equations is presented in Supplementary Information.

Imaging manipulation

All imaging data were processed under standardized conditions. No non-linear manipulations (e.g., gamma correction) were applied to artificially alter image properties. Adjustments to brightness and contrast for presentation and visualization purposes were uniformly applied. All images used for quantification were processed using identical acquisition and analysis parameters.

Electron microscopy

Tissue preparation

Fresh thymic tissues were processed for Transmission Electron Microscopy (TEM) (n = 7), or with an Osmium-Thiocarbohydrazide-Osmium (OTO) method (n = 2), for 3D-SEM Array Tomography, as will be discussed below.

Transmission electron microscopy (TEM)

Samples were fixed with 2.5% glutaraldehyde, 2% paraformaldehyde in 0.1 M sodium cacodylate buffer, postfixed with 1% osmium tetroxide followed by 2% uranyl acetate, dehydrated through a graded series of ethanol and embedded in LX112 resin (LADD Research Industries, Burlington VT). Ultrathin sections were cut on a Leica Ultracut UC7, stained with uranyl acetate followed by lead citrate and viewed on a JEOL 1400 Plus transmission electron microscope at 120 kV.

OTO processing and scanning electron microscopy (OTO-SEM)

Samples were immersion fixed in 2.0% paraformaldehyde, 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer, then processed using a modified National Center for Microscopy and Imaging method of OTO¹⁸⁴. In brief, samples were post fixed with reduced osmium, treated with thiocarbohydrazide, further stained with osmium, en bloc stained with uranyl acetate, further stained with lead aspartate, dehydrated in a graded series of ethanol, and embedded into LX112 resin. 55 nm-thick sections were cut on a Leica Artos microtome using a Diatome AT 35° knife, and picked up on freshly glowed silicon wafers. Sections were observed on Zeiss Supra 40 Field Emission Scanning Electron Microscope in backscatter mode, using an acceleration voltage of 8.0 kV.

Three-dimensional reconstruction of the cTEC meshwork

Regions of interest were collected with ATLAS 5.0, using a pixel size of 6.0 × 6.0 nm and a dwell time of 6.0 µs. Images were aligned and segmentation was performed using the IMOD suite of programs, as described^185,186. Briefly, stacks of images were aligned in IMOD using Midas and cells of interest were manually segmented out using the drawing tools in 3DMOD. cTEC cells were identified by their large, euchromatic nucleus, nucleolus shape, and the presence of tonofilaments throughout the cell.

cTEC vacuole analysis

All TEM images of Young Adult and Old thymi samples, as mentioned in the TEM section above, were used to identify cTEC vacuole types and to compare their relative proportions between age groups. Apart from lysosomes, there were 5 distinguishable types of vacuoles, based on ultrastructural morphology, as described in previous studies^{74–76,78–80}: (i) Autophagosomes (APGs), characterized by a double membrane and various content; (ii),(ii) Early and Late Autolysosomes (early AUTs and late AUTs, respectively), with usually a single membrane and decreasing amounts of electron-dense content, respectively; (iv) Multilamellar Bodies (MLBs), having numerous internal concentric membranes; (v) Multivesicular Bodies (MVBs), with variable micro-vesicles.

A total of 35 TEM images (×600°C2500 magnification, µm scale bars embedded) containing cTEC segments were used. In each image, the identification of vacuoles and focusing on their center was performed using DigitalMicrograph (Gatan) software. Each vacuole was exported as tif in ImageJ. After classifying all vacuoles in the cTEC network based on the 5-type scheme shown above, the relative type proportions were calculated. In total, 735 vacuoles were analyzed (Young-Adult group, N = 372; Old group N = 363). Given that early and late AUTs were the majority of the vacuoles, additional morphometries were performed for these two types in further detail, as described (Supplementary Information).

Statistics and reproducibility

Statistical package

GraphPad Prism 10 software was used for generation of all graphs and for statistical hypothesis testing.

Statistical hypothesis testing

Results are presented as hybrid scatterplot/bar graphs, with means and Standard Error of the Mean (SEM). Linear correlations are graphed using correlation matrices or scatterplots with simple linear regression lines and are analyzed using Spearman’s correlation and coefficient of determination (r). Comparison of variables among age groups is performed with One-Way ANOVA and post-hoc analysis using Dunnett’s multiple comparisons test. Comparison of variables between DMBA-treated animals receiving or not receiving early-life exposure to CT is performed with unpaired Student’s t test. For all graphs, p values of $\leq$ 0.05 are stated as statistically significant, while p $>$ 0.05 are as non-significant (ns). For all graphs, (*) corresponds to p $\leq$ 0.05, (**) corresponds to p $<$ 0.01, (***) corresponds to p $<$ 0.001, and (****) corresponds to p $<$ 0.0001.

Exclusion criteria for individual cases

All datasets underwent outlier calculator testing using an alpha of 0.05. Any values falling outside the 95% confidence interval in specific variables were excluded. Individual cases were also excluded if they lacked sufficient thymic or splenic tissue, to generate enough or representative regions of interest (ROIs) for analysis.

Blindness in image analysis

All pathologists involved in histopathological diagnoses and quantifications were blinded to the age group allocations. Scientists performing AIRE⁺ mTEC homogeneity analysis, AIRE⁺ and DCLK1⁺ mTEC density analyses, and electron microscopy 3D reconstructions, were also completely blinded. All other variables were calculated by scientists that were not blinded to age group allocation. However, to eliminate all the operator-dependent biases, all the samples were treated and processed in the same way, i.e., snapshots from each case were captured and analyzed together in one batch, using the same settings (e.g., intensity thresholding, etc.) in the associated digital pathology software.

Reproducibility

All morphometric analyses were performed using standardized, version-controlled pipelines with fixed parameter settings across all samples. Biological replicates were incorporated by dividing mice into at least three independent cohorts, which were analyzed and subsequently pooled to account for both biological and technical variability. Key morphometric outputs were validated across orthogonal methods (e.g., HE vs IF-based measurements), demonstrating consistent trends. All statistical tests, inclusion/exclusion criteria, and preprocessing steps were pre-defined, and analyses were performed under blinded conditions where applicable to ensure robustness.

Principal component analysis (PCA)

PCA combining all individual morphometric variables was performed, to reduce multi-dimensionality, and 2D PCA plots were generated for values that were subjected to ln(x) transformation for the experimental groups by importing data matrix into ClustVis webtool¹⁸⁷. R packages for plot generation and analysis have been described¹⁸⁷. Data pre-processing was performed using the unit variance scaling method. The default method “SVD with imputation”, which performs imputation and Singular Value Decomposition (SVD) iteratively until estimates of missing values converge, was applied to calculate principal components as described¹⁸⁷. PCA plots showing scatterplots with axes, each corresponding to the two different principal components (PC1 and PC2) were generated for both mouse models.

Supplementary information

Transparent Peer Review file^{(171.8KB, pdf)}

Supplementary Information^{(21.7MB, pdf)}

42003_2026_10038_MOESM3_ESM.pdf^{(29.9KB, pdf)}

Description of Additional Supplementary Files

Supplementary Data 1^{(90.2KB, pdf)}

Supplementary Data 2^{(137.7KB, pdf)}

Supplementary Video 1^{(295.8MB, 7z)}

Supplementary Video 2^{(253.4MB, 7z)}

Acknowledgements

We thank Dr. Ana-Maria Cuervo for providing the 2-year-old C57BL/6 J mice for our studies, as well as Dr. Susmita Kaushik of the Proteostasis of Aging Core for the TEM vacuole identification and valuable analysis consultations. We also thank the Histopathology Core for the processing and paraffin-embedding of thymic and splenic tissues. Several investigators would like to acknowledge the following for funding support: the new investigator start-up funds, provided to G.S. Karagiannis from the Montefiore-Einstein Comprehensive Cancer Center (P30 CA013330); the Careers in Immunology Fellowship (2022), awarded to M.K. Lagou and G.S. Karagiannis from the American Association of Immunologists; the NIH R21AG091008 provided to G.S. Karagiannis and M. Maryanovich, the American Society of Hematology Junior Faculty Scholar From thymus to tissues and tumorAward (2023), provided to M. Maryanovich; the NIH T32 HL144456 (NHLBI), the NIH F32 HL158084 (NHLBI), the Gottesman Stem Cell Institute Paul S. Frenette Scholars Award, and the Cancer Research Institute Irvington Postdoctoral Fellowship (CRI4994), provided to R.S. Carpenter. We would also like to acknowledge the use of the Zeiss Supra 40 Field Emission Scanning Electron Microscope (1S10RR025554-01A1) and the 3DHISTECH P250 High-Capacity Slide Scanner (1S10OD026852-01A1) in the Analytical Imaging Facility (AIF), both acquired through Shared Instrumentation Grants. The AIF was also partially funded by the NCI Cancer Center Support grant (P30CA013330).

Author contributions

M.K. Lagou, M.P. Markaki, K. Angelopoulou, T. Poutahidis, and G.S. Karagiannis designed the study; M.K. Lagou, M.P. Markaki, S. Vodopyanov, D.G. Argyris, L. Gunther-Cummins and C. Entenberg performed experiments; L. Gunther-Cummins, H. Guzik, X. Nishku, J. Churaman performed sample preparation for all the imaging studies; M.K. Lagou, M.P. Markaki, S. Vodopyanov, D.G. Argyris, L. Gunther-Cummins, S. DesMarais, C. Entenberg, V. DesMarais, and F.P. Macaluso optimized imaging parameters and performed image analysis; A. Hardas, and T. Poutahidis performed reviews of all thymic and splenic slides and diagnostic pathology, as well as reviews on the spontaneous lesions of aged mice; M.K. Lagou, M.P. Markaki, D.G. Argyris, C. Panorias, P.S. Filippou, and G.S. Karagiannis performed data analyses and statistical analyses. R.S. Carpenter, M. Maryanovich, M.K. Lagou, P.S. Filipoou, K. Angelopoulou, T. Poutahidis, and G.S. Karagiannis engaged in critical interpretation of the experimental results and conclusions. M.K. Lagou, M. Markaki, D.G. Argyris, S. Vodopyanov, and G.S. Karagiannis wrote the final report. All authors contributed to the editing and approval of the final report. All figures were created with Abode InDesign and Illustrator.

Peer review

Peer review information

Communications Biology thanks Yideng Cai and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Shan E Ahmed Raza and Laura Rodríguez Pérez. A peer review file is available.

Data availability

All data supporting the findings of this study are available from the corresponding author upon reasonable request. The numerical source data for the graphs in all main and Supplementary Figs. (i.e., the values of the individual data points), are provided as Supplementary Data 2. No other large-scale datasets were generated or analyzed that would necessitate deposition into public repositories.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-026-10038-w.

References

1.Pearse, G. Normal structure, function and histology of the thymus. Toxicol. Pathol.34, 504–514 (2006). [DOI] [PubMed] [Google Scholar]
2.Han, J. & Zuniga-Pflucker, J. C. A 2020 view of thymus stromal cells in T cell development. J. Immunol.206, 249–256 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.James, K. D., Jenkinson, W. E. & Anderson, G. Non-epithelial stromal cells in thymus development and function. Front. Immunol.12, 634367 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Kadouri, N., Nevo, S., Goldfarb, Y. & Abramson, J. Thymic epithelial cell heterogeneity: TEC by TEC. Nat. Rev. Immunol.20, 239–253 (2020). [DOI] [PubMed] [Google Scholar]
5.Granadier, D., Iovino, L., Kinsella, S. & Dudakov, J. A. Dynamics of thymus function and T cell receptor repertoire breadth in health and disease. Semin. Immunopathol.43, 119–134 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Abramson, J. & Anderson, G. Thymic epithelial cells. Annu. Rev. Immunol.35, 85–118 (2017). [DOI] [PubMed] [Google Scholar]
7.Alawam, A. S., Anderson, G. & Lucas, B. Generation and regeneration of thymic epithelial cells. Front. Immunol.11, 858 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Alexandropoulos, K. & Danzl, N. M. Thymic epithelial cells: antigen presenting cells that regulate T cell repertoire and tolerance development. Immunol. Res.54, 177–190 (2012). [DOI] [PubMed] [Google Scholar]
9.Chaudhry, M. S., Velardi, E., Dudakov, J. A. & van den Brink, M. R. Thymus: the next (re)generation. Immunol. Rev.271, 56–71 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Swann, J. B. & Boehm, T. Everything in its place. Immunity31, 856–858 (2009). [DOI] [PubMed] [Google Scholar]
11.Griffith, A. V. et al. Spatial mapping of thymic stromal microenvironments reveals unique features influencing T lymphoid differentiation. Immunity31, 999–1009 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Garcia-Ceca, J., Montero-Herradon, S., Alfaro, D. & Zapata, A. G. Increased epithelial-free areas in thymuses with altered EphB-mediated thymocyte-thymic epithelial cell interactions. Histochem. Cell Biol.148, 381–394 (2017). [DOI] [PubMed] [Google Scholar]
13.Clarke, A. G. & Kendall, M. D. Histological changes in the thymus during mouse pregnancy. Thymus14, 65–78 (1989). [PubMed] [Google Scholar]
14.Kondo, K., Ohigashi, I. & Takahama, Y. Thymus machinery for T-cell selection. Int. Immunol.31, 119–125 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Michelson, D. A., Hase, K., Kaisho, T., Benoist, C. & Mathis, D. Thymic epithelial cells co-opt lineage-defining transcription factors to eliminate autoreactive T cells. Cell185, 2542–2558 e2518 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.James, K. D., Jenkinson, W. E. & Anderson, G. T-cell egress from the thymus: should I stay or should I go? J. Leukoc. Biol.104, 275–284 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Mori, K., Itoi, M., Tsukamoto, N., Kubo, H. & Amagai, T. The perivascular space as a path of hematopoietic progenitor cells and mature T cells between the blood circulation and the thymic parenchyma. Int. Immunol.19, 745–753 (2007). [DOI] [PubMed] [Google Scholar]
18.James, K. D. et al. Endothelial cells act as gatekeepers for LTbetaR-dependent thymocyte emigration. J. Exp. Med.215, 2984–2993 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Nitta, T. Mesenchymal stromal cells in the thymus. Inflamm. Regen.42, 33 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Nitta, T., Ota, A., Iguchi, T., Muro, R. & Takayanagi, H. The fibroblast: an emerging key player in thymic T cell selection. Immunol. Rev.302, 68–85 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Zachariah, M. A. & Cyster, J. G. Neural crest-derived pericytes promote egress of mature thymocytes at the corticomedullary junction. Science328, 1129–1135 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Bautista, J. L. et al. Single-cell transcriptional profiling of human thymic stroma uncovers novel cellular heterogeneity in the thymic medulla. Nat. Commun.12, 1096 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Bleul, C. C. & Boehm, T. Chemokines define distinct microenvironments in the developing thymus. Eur. J. Immunol.30, 3371–3379 (2000). [DOI] [PubMed] [Google Scholar]
24.Dudakov, J. A. et al. Interleukin-22 drives endogenous thymic regeneration in mice. Science336, 91–95 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Lucas, B. et al. Progressive changes in CXCR4 expression that define thymocyte positive selection are dispensable for both innate and conventional alphabetat-cell development. Sci. Rep.7, 5068 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Pearse, G. Histopathology of the thymus. Toxicol. Pathol.34, 515–547 (2006). [DOI] [PubMed] [Google Scholar]
27.Elmore, S. A. Enhanced histopathology of the thymus. Toxicol. Pathol.34, 656–665 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Willard-Mack, C. L. et al. Nonproliferative and proliferative lesions of the rat and mouse hematolymphoid system. Toxicol. Pathol.47, 665–783 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Vodopyanov, S. et al. Mapping murine thymic epithelial cells: functional ultrastructure and implications for thymopoiesis. J. Leukoc. Biol.117. 10.1093/jleuko/qiaf115 (2025). [DOI] [PubMed]
30.Vodopyanov, S. et al. Beyond cysts: organization of epithelial networks in the murine thymus. J. Cell Sci. 10.1242/jcs.264079 (2025). [DOI] [PMC free article] [PubMed]
31.Nakagawa, Y. et al. Thymic nurse cells provide microenvironment for secondary T cell receptor alpha rearrangement in cortical thymocytes. Proc. Natl. Acad. Sci. USA109, 20572–20577 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Halin, C., Mora, J. R., Sumen, C. & von Andrian, U. H. In vivo imaging of lymphocyte trafficking. Annu. Rev. Cell Dev. Biol.21, 581–603 (2005). [DOI] [PubMed] [Google Scholar]
33.Aguilan, J. T. et al. Identifying space-resolved proteins of the murine thymus, by combining MALDI-MSI and proteomics. Life Sci Alliance9, e202503205 (2025). [DOI] [PMC free article] [PubMed]
34.Meyer, M. B., Lee, S. M., Cichanski, S. R., Cobice, D. F. & Pike, J. W. Spatial detection and consequences of nonrenal calcitriol production as assessed by targeted mass spectrometry imaging. JCI Insight9. 10.1172/jci.insight.181763 (2024). [DOI] [PMC free article] [PubMed]
35.Tsuji, Y., Yamaguchi, S., Nakamura, T. & Ikegawa, M. Mass Spectrometry Imaging (MSI) delineates thymus-centric metabolism in vivo as an effect of systemic administration of dexamethasone. Appl. Sci.11, 11038 (2021).
36.Yayon, N. et al. A spatial human thymus cell atlas mapped to a continuous tissue axis. Nature635, 708–718 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Brayton, C. F., Treuting, P. M. & Ward, J. M. Pathobiology of aging mice and GEM:background strains and experimental design. Vet. Pathol.49, 85–105 (2012). [DOI] [PubMed] [Google Scholar]
38.Loukov, D., Naidoo, A., Puchta, A., Marin, J. L. & Bowdish, D. M. Tumor necrosis factor drives increased splenic monopoiesis in old mice. J. Leukoc. Biol.100, 121–129 (2016). [DOI] [PubMed] [Google Scholar]
39.Baran-Gale, J. et al. Ageing compromises mouse thymus function and remodels epithelial cell differentiation. eLife9, e56221 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Kumar, B. V., Connors, T. J. & Farber, D. L. Human T cell development, localization, and function throughout life. Immunity48, 202–213 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Linton, P. J. & Dorshkind, K. Age-related changes in lymphocyte development and function. Nat. Immunol.5, 133–139 (2004). [DOI] [PubMed] [Google Scholar]
42.Palmer, D. The effect of age on thymic function. Front. Immunol.4. 10.3389/fimmu.2013.00316 (2013). [DOI] [PMC free article] [PubMed]
43.Elmore, S. A. Enhanced Histopathology Evaluation of Lymphoid Organs. Methods Mol. Biol.1803, 147–168 (2018). [DOI] [PubMed] [Google Scholar]
44.Kavanagh, M., Pulido, O., Caldwell, D. & Gill, S. 229-243 (2005).
45.Farr, A. G. & Braddy, S. C. Patterns of keratin expression in the murine thymus. Anat. Rec.224, 374–378 (1989). [DOI] [PubMed] [Google Scholar]
46.Gupta, R. et al. Cytokeratin (CK5, CK8, CK14) expression and presence of progenitor stem cells in human fetal thymuses. Clin. Anat.29, 711–717 (2016). [DOI] [PubMed] [Google Scholar]
47.Lee, E. N. et al. Characterization of the expression of cytokeratins 5, 8, and 14 in mouse thymic epithelial cells during thymus regeneration following acute thymic involution. Anat. Cell Biol.44, 14–24 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Odaka, C. et al. Keratin 8 is required for the maintenance of architectural structure in thymus epithelium. PLoS ONE8, e75101 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Nusser, A. et al. Developmental dynamics of two bipotent thymic epithelial progenitor types. Nature606, 165–171 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Grevellec, A. & Tucker, A. S. The pharyngeal pouches and clefts: development, evolution, structure and derivatives. Semin. Cell Dev. Biol.21, 325–332 (2010). [DOI] [PubMed] [Google Scholar]
51.Vaidya, H. J., Briones Leon, A. & Blackburn, C. C. FOXN1 in thymus organogenesis and development. Eur. J. Immunol.46, 1826–1837 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Wertheimer, T. et al. Production of BMP4 by endothelial cells is crucial for endogenous thymic regeneration. Sci Immunol3. 10.1126/sciimmunol.aal2736 (2018). [DOI] [PMC free article] [PubMed]
53.Wekerle, H. & Ketelsen, U.-P. Thymic nurse cells—Ia-bearing epithelium involved in T-lymphocyte differentiation? Nature283, 402–404 (1980). [DOI] [PubMed] [Google Scholar]
54.Nakagawa, Y. et al. Thymic nurse cells provide microenvironment for secondary T cell receptor α rearrangement in cortical thymocytes. Proc. Natl. Acad. Sci. USA109, 20572–20577 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Guyden, J. C. & Pezzano, M. Thymic nurse cells: a microenvironment for thymocyte development and selection. Int. Rev. Cytol.223, 1–37 (2003). [DOI] [PubMed] [Google Scholar]
56.Arganda-Carreras, I., Fernandez-Gonzalez, R., Munoz-Barrutia, A. & Ortiz-De-Solorzano, C. 3D reconstruction of histological sections: application to mammary gland tissue. Microsc. Res. Tech.73, 1019–1029 (2010). [DOI] [PubMed] [Google Scholar]
57.Venables, T., Griffith, A. V., DeAraujo, A. & Petrie, H. T. Dynamic changes in epithelial cell morphology control thymic organ size during atrophy and regeneration. Nat. Commun.10, 4402 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Griffith, A. V., Fallahi, M., Venables, T. & Petrie, H. T. Persistent degenerative changes in thymic organ function revealed by an inducible model of organ regrowth. Aging Cell11, 169–177 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Hernandez-Lopez, C. et al. The CXCL12/CXCR4 pair in aged human thymus. Neuroimmunomodulation17, 217–220 (2010). [DOI] [PubMed] [Google Scholar]
60.Milicevic, N. M., Nohroudi, K., Milicevic, Z. & Westermann, J. Activation of cortical and inhibited differentiation of medullary epithelial cells in the thymus of lymphotoxin-beta receptor-deficient mice: an ultrastructural study. J. Anat.212, 114–124 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Borelli, A. & Irla, M. Lymphotoxin: from the physiology to the regeneration of the thymic function. Cell Death Differ.28, 2305–2314 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Kwok, T. et al. Age-associated changes to lymph node fibroblastic reticular cells. Front. Aging3, 838943 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Cavallotti, C., D’Andrea, V., Tonnarini, G., Cavallotti, C. & Bruzzone, P. Age-related changes in the human thymus studied with scanning electron microscopy. Microsc. Res. Tech.71, 573–578 (2008). [DOI] [PubMed] [Google Scholar]
64.De Waal, E. J. & Rademakers, L. H. Heterogeneity of epithelial cells in the rat thymus. Microsc. Res. Tech.38, 227–236 (1997). [DOI] [PubMed] [Google Scholar]
65.De Waal, E. J., Rademakers, L. H., Schuurman, H. J., Van Loveren, H. & Vos, J. G. Ultrastructure of the cortical epithelium of the rat thymus after in vivo exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Arch. Toxicol.67, 558–564 (1993). [DOI] [PubMed] [Google Scholar]
66.Erokhina, V. & Avilova, O. Ultramicroscopic changes of rats parathyroid glands and thymus after single administration of cyclophosphamide at the different periods of observation. Wiad. Lek.72, 362–367 (2019). [PubMed] [Google Scholar]
67.Mandel, T. Ultrastructure of epithelial cells in the cortex of guinea pig thymus. Z. Zellforsch. Mikros. Anat.92, 159–168 (1968). [DOI] [PubMed] [Google Scholar]
68.Radulovic, S., Sunkara, S., Rachel, R. & Leitinger, G. Three-dimensional SEM, TEM, and STEM for analysis of large-scale biological systems. Histochem. Cell Biol.158, 203–211 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Kendall, M. D. The syncytial nature of epithelial cells in the thymic cortex. J. Anat.147, 95–106 (1986). [PMC free article] [PubMed] [Google Scholar]
70.Vignery, A. Osteoclasts and giant cells: macrophage-macrophage fusion mechanism. Int. J. Exp. Pathol.81, 291–304 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Mandel, T. Epithelial cells and lymphopoiesis in the cortex of guinea-pig thymus. Aust. J. Exp. Biol. Med. Sci.47, 153–155 (1969). [DOI] [PubMed] [Google Scholar]
72.Nitta, T. et al. Thymoproteasome shapes immunocompetent repertoire of CD8+ T cells. Immunity32, 29–40 (2010). [DOI] [PubMed] [Google Scholar]
73.Rodrigues, P. M. et al. LAMP2 regulates autophagy in the thymic epithelium and thymic stroma-dependent CD4 T cell development. Autophagy19, 426–439 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Jung, M., Choi, H. & Mun, J. Y. The autophagy research in electron microscopy. Appl. Microsc.49, 11 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Hariri, M. et al. Biogenesis of multilamellar bodies via autophagy. Mol. Biol. Cell11, 255–268 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Klionsky, D. J. et al. Guidelines for the use and interpretation of assays for monitoring autophagy in higher eukaryotes. Autophagy4, 151–175 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
77.Nagy, P., Varga, Á, Kovács, A. L., Takáts, S. & Juhász, G. How and why to study autophagy in Drosophila: It’s more than just a garbage chute. Methods75, 151–161 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
78.Clague, M. J. & Urbé, S. Multivesicular bodies. Curr. Biol.18, R402–R404 (2008). [DOI] [PubMed] [Google Scholar]
79.Eskelinen, E.-L., Reggiori, F., Baba, M., Kovács, A. L. & Seglen, P. O. Seeing is believing: the impact of electron microscopy on autophagy research. Autophagy7, 935–956 (2011). [DOI] [PubMed] [Google Scholar]
80.Ylä-Anttila, P., Vihinen, H., Jokitalo, E. & Eskelinen, E. L. Monitoring autophagy by electron microscopy in Mammalian cells. Methods Enzymol.452, 143–164 (2009). [DOI] [PubMed] [Google Scholar]
81.Uddin, M. N., Nishio, N., Ito, S., Suzuki, H. & Isobe, K. Autophagic activity in thymus and liver during aging. Age34, 75–85 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Zhou, T. A. et al. Thymic macrophages consist of two populations with distinct localization and origin. Elife11. 10.7554/eLife.75148 (2022). [DOI] [PMC free article] [PubMed]
83.Dung, H. C. Electron microscopic study of involuting thymus of “lethargic” mutant mice. Anat. Rec.177, 585–601 (1973). [DOI] [PubMed] [Google Scholar]
84.Lee-Rueckert, M., Lappalainen, J., Kovanen, P. T. & Escola-Gil, J. C. Lipid-Laden macrophages and inflammation in atherosclerosis and cancer: an integrative view. Front. Cardiovasc. Med.9, 777822 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Kazachkov, M. Y., Muhlebach, M. S., Livasy, C. A. & Noah, T. L. Lipid-laden macrophage index and inflammation in bronchoalveolar lavage fluids in children. Eur. Respir. J.18, 790–795 (2001). [DOI] [PubMed] [Google Scholar]
86.Bearman, R. M., Bensch, K. G. & Levine, G. D. The normal human thymic vasculature: an ultrastructural study. Anat. Rec.183, 485–497 (1975). [DOI] [PubMed] [Google Scholar]
87.Kato, S. Thymic microvascular system. Microsc. Res. Tech.38, 287–299 (1997). [DOI] [PubMed] [Google Scholar]
88.Kendall, M. D. The morphology of perivascular spaces in the thymus. Thymus13, 157–164 (1989). [PubMed] [Google Scholar]
89.Nunez, S. et al. The human thymus perivascular space is a functional niche for viral-specific plasma cells. Sci. Immunol.1. 10.1126/sciimmunol.aah4447 (2016). [DOI] [PMC free article] [PubMed]
90.Ushiki, T. & Takeda, M. Three-dimensional ultrastructure of the perivascular space in the rat thymus. Arch. Histol. Cytol.60, 89–99 (1997). [DOI] [PubMed] [Google Scholar]
91.Odaka, C., Morisada, T., Oike, Y. & Suda, T. Distribution of lymphatic vessels in mouse thymus: immunofluorescence analysis. Cell Tissue Res.325, 13–22 (2006). [DOI] [PubMed] [Google Scholar]
92.Baeyens, A. A. L. & Schwab, S. R. Finding a way out: S1P signaling and immune cell migration. Annu. Rev. Immunol.38, 759–784 (2020). [DOI] [PubMed] [Google Scholar]
93.Kumar, A. & Saba, J. D. Regulation of immune cell migration by Sphingosine-1-Phosphate. Cell Mol. Biol.61, 121 (2015). [PMC free article] [PubMed]
94.Weinreich, M. A. & Hogquist, K. A. Thymic emigration: when and how T cells leave home. J. Immunol.181, 2265–2270 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
95.Colomb, E. et al. Genetic control of giant perivascular space formation in the thymus of NOD mice. Diabetes45, 1535–1540 (1996). [DOI] [PubMed] [Google Scholar]
96.Flores, K. G., Li, J., Sempowski, G. D., Haynes, B. F. & Hale, L. P. Analysis of the human thymic perivascular space during aging. J. Clin. Investig.104, 1031–1039 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
97.Funk, J., Ruehl-Fehlert, C., Leonard, C., Kellner, R. & Rittinghausen, S. Immunohistochemical characterization of proliferative lesions in the thymus of aging CD-1 mice from two studies on the RITA database, with special reference to the perivascular space. Toxicol. Pathol.50, 308–328 (2022). [DOI] [PubMed] [Google Scholar]
98.Kato, S. & Schoefl, G. I. Microvasculature of normal and involuted mouse thymus. Light- Electron-Microsc. Study Acta Anat.135, 1–11 (1989). [DOI] [PubMed] [Google Scholar]
99.Steinmann, G. G., Klaus, B. & Muller-Hermelink, H. K. The involution of the ageing human thymic epithelium is independent of puberty. A morphometric study. Scand. J. Immunol.22, 563–575 (1985). [DOI] [PubMed] [Google Scholar]
100.Liang, Z., Dong, X., Zhang, Z., Zhang, Q. & Zhao, Y. Age-related thymic involution: mechanisms and functional impact. Aging Cell21, e13671 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
101.Mitre, T. M., Pietropaolo, M. & Khadra, A. The dual role of autoimmune regulator in maintaining normal expression level of tissue-restricted autoantigen in the thymus: a modeling investigation. Math. Biosci.287, 12–23 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
102.Zuklys, S. et al. Normal thymic architecture and negative selection are associated with Aire expression, the gene defective in the autoimmune-polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED). J. Immunol.165, 1976–1983 (2000). [DOI] [PubMed] [Google Scholar]
103.Chen, J., Yang, W., Yu, C. & Li, Y. Autoimmune regulator initiates the expression of promiscuous genes in thymic epithelial cells. Immunol. Invest37, 203–214 (2008). [DOI] [PubMed] [Google Scholar]
104.Cabric, V. & Brown, C. C. Thymic mimicry: the art of imitation. J. Exp. Med.220. 10.1084/jem.20231010 (2023). [DOI] [PMC free article] [PubMed]
105.Michelson, D. A. & Mathis, D. Thymic mimetic cells: tolerogenic masqueraders. Trends Immunol.43, 782–791 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
106.Miller, C. N. et al. Thymic tuft cells promote an IL-4-enriched medulla and shape thymocyte development. Nature559, 627–631 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
107.Bornstein, C. et al. Single-cell mapping of the thymic stroma identifies IL-25-producing tuft epithelial cells. Nature559, 622–626 (2018). [DOI] [PubMed] [Google Scholar]
108.Srinivasan, J. et al. Age-related changes in thymic central tolerance. Front. Immunol.12, 676236 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
109.Givony, T. et al. Thymic mimetic cells function beyond self-tolerance. Nature622, 164–172 (2023). [DOI] [PubMed] [Google Scholar]
110.Burchiel, S. W. et al. DMBA-induced cytotoxicity in lymphoid and nonlymphoid organs of B6C3F1 mice: relation of cell death to target cell intracellular calcium and DNA damage. Toxicol. Appl. Pharm.113, 126–132 (1992). [DOI] [PubMed] [Google Scholar]
111.Argyris, D. G. et al. Suppression of chemically induced mammary cancer by early-life oral administration of cholera toxin in mice is associated with aberrant regulation of Bmp and Notch signaling pathways. Mol. Biol. Rep.52, 150 (2025). [DOI] [PubMed] [Google Scholar]
112.Doulberis, M. et al. Cholera-toxin suppresses carcinogenesis in a mouse model of inflammation-driven sporadic colon cancer. Carcinogenesis36, 280–290 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
113.Poutahidis, T., Angelopoulou, K. & Erdman, S. E. Old enemies meet new friends for colon cancer prevention. Oncoimmunology4, e1027474 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
114.Varian, B. J. et al. Consuming cholera toxin counteracts age-associated obesity. Oncotarget10, 5497–5509 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
115.Afaloniati, H. et al. Peri-weaning cholera toxin consumption suppresses chemically-induced carcinogenesis in mice. Int. J. Cancer154, 1097–1110 (2024). [DOI] [PubMed] [Google Scholar]
116.Bach, J. F. The hygiene hypothesis in autoimmunity: the role of pathogens and commensals. Nat. Rev. Immunol.18, 105–120 (2018). [DOI] [PubMed] [Google Scholar]
117.Fletcher, A. L. et al. Ablation and regeneration of tolerance-inducing medullary thymic epithelial cells after cyclosporine, cyclophosphamide, and dexamethasone treatment. J. Immunol.183, 823–831 (2009). [DOI] [PubMed] [Google Scholar]
118.Schulz, O., Hammerschmidt, S. I., Moschovakis, G. L. & Forster, R. Chemokines and chemokine receptors in lymphoid tissue dynamics. Annu. Rev. Immunol.34, 203–242 (2016). [DOI] [PubMed] [Google Scholar]
119.Halkias, J., Melichar, H. J., Taylor, K. T. & Robey, E. A. Tracking migration during human T cell development. Cell Mol. Life Sci.71, 3101–3117 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
120.Itoi, M., Tsukamoto, N., Yoshida, H. & Amagai, T. Mesenchymal cells are required for functional development of thymic epithelial cells. Int. Immunol.19, 953–964 (2007). [DOI] [PubMed] [Google Scholar]
121.Lee, J. Y. M. & Love, P. E. A simplified approach to evaluating T Cell development by flow cytometry. Methods Mol. Biol.2580, 131–149 (2023). [DOI] [PubMed] [Google Scholar]
122.Cahill, R. N., Kimpton, W. G., Cunningham, C. P. & Washington, E. A. An immune system switch at birth triggers a change in the lifespan of peripheral T cells. Semin. Immunol.9, 355–363 (1997). [DOI] [PubMed] [Google Scholar]
123.Webb, S. & Haniffa, M. Large-scale single-cell RNA sequencing atlases of human immune cells across lifespan: possibilities and challenges. Eur. J. Immunol. e2250222. 10.1002/eji.202250222 (2023). [DOI] [PubMed]
124.Minato, N., Hattori, M. & Hamazaki, Y. Physiology and pathology of T-cell aging. Int Immunol.32, 223–231 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
125.Ostan, R. et al. Immunosenescence and immunogenetics of human longevity. Neuroimmunomodulation15, 224–240 (2008). [DOI] [PubMed] [Google Scholar]
126.Palmer, S., Albergante, L., Blackburn, C. C. & Newman, T. J. Thymic involution and rising disease incidence with age. Proc. Natl. Acad. Sci. USA115, 1883–1888 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
127.Kooshesh, K. A., Foy, B. H., Sykes, D. B., Gustafsson, K. & Scadden, D. T. Health Consequences of Thymus Removal in Adults. N. Engl. J. Med389, 406–417 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
128.Gulla, S. et al. Role of thymus in health and disease. Int. Rev. Immunol. 42, 347–363 (2022). [DOI] [PubMed]
129.Chopp, L., Redmond, C., O’Shea, J. J. & Schwartz, D. M. From thymus to tissues and tumors: a review of T cell biology. J Allergy Clin. Immunol. 10.1016/j.jaci.2022.10.011 (2022). [DOI] [PMC free article] [PubMed]
130.Wang, W., Thomas, R., Sizova, O. & Su, D. M. Thymic function associated with cancer development, relapse, and antitumor immunity - a mini-review. Front. Immunol.11, 773 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
131.Sansoni, P. et al. The immune system in extreme longevity. Exp. Gerontol.43, 61–65 (2008). [DOI] [PubMed] [Google Scholar]
132.Marinova, T. T. Epithelial framework reorganization during human thymus involution. Gerontology51, 14–18 (2005). [DOI] [PubMed] [Google Scholar]
133.Ansari, A. R. & Liu, H. Acute thymic involution and mechanisms for recovery. Arch. Immunol. Ther. Exp.65, 401–420 (2017). [DOI] [PubMed] [Google Scholar]
134.Kinsella, S. & Dudakov, J. A. When the damage is done: injury and repair in thymus function. Front. Immunol.11, 1745 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
135.Lagou, M. K., Anastasiadou, D. P. & Karagiannis, G. S. A proposed link between acute thymic involution and late adverse effects of chemotherapy. Front. Immunol.13, 933547 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
136.Rijhsinghani, A. G., Thompson, K., Bhatia, S. K. & Waldschmidt, T. J. Estrogen blocks early T cell development in the thymus. Am. J. Reprod. Immunol.36, 269–277 (1996). [DOI] [PubMed] [Google Scholar]
137.Taves, M. D. & Ashwell, J. D. Effects of sex steroids on thymic epithelium and thymocyte development. Front. Immunol.13, 975858 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
138.Laan, M., Haljasorg, U., Kisand, K., Salumets, A. & Peterson, P. Pregnancy-induced thymic involution is associated with suppression of chemokines essential for T-lymphoid progenitor homing. Eur. J. Immunol.46, 2008–2017 (2016). [DOI] [PubMed] [Google Scholar]
139.Swami, S., Tong, I., Bilodeau, C. C. & Bourjeily, G. Thymic involution in pregnancy: a universal finding? Obstet. Med. 5, 130–132 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
140.Yang, H. et al. Obesity accelerates thymic aging. Blood114, 3803–3812 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
141.Lagou, M. K. & Karagiannis, G. S. Obesity-induced thymic involution and cancer risk. Semin Cancer Biol.93, 3–19 (2023). [DOI] [PubMed] [Google Scholar]
142.Reith, W. & Mach, B. The bare lymphocyte syndrome and the regulation of MHC expression. Annu Rev. Immunol.19, 331–373 (2001). [DOI] [PubMed] [Google Scholar]
143.Ferrua, F. et al. Thymic epithelial cell alterations and defective thymopoiesis lead to central and peripheral tolerance perturbation in MHCII Deficiency. Front. Immunol.12, 669943 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
144.Costa-Carvalho, B. T. et al. A remarkable depletion of both naive CD4+ and CD8+ with high proportion of memory T cells in an IPEX infant with a FOXP3 mutation in the forkhead domain. Scand. J. Immunol.68, 85–91 (2008). [DOI] [PubMed] [Google Scholar]
145.Marx, A. et al. Thymus and autoimmunity. Semin Immunopathol.43, 45–64 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
146.White, A. J. et al. Lymphotoxin signals from positively selected thymocytes regulate the terminal differentiation of medullary thymic epithelial cells. J. Immunol.185, 4769–4776 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
147.Van de Velde, R. L. & Friedman, N. B. Thymic myoid cells and myasthenia gravis. Am. J. Pathol.59, 347–368 (1970). [PMC free article] [PubMed] [Google Scholar]
148.Kekalainen, E. et al. Defective central tolerance in Aire-deficient mice is not sufficient to induce symptomatic autoimmunity during lymphopenia-induced T cell proliferation. Scand. J. Immunol.74, 71–79 (2011). [DOI] [PubMed] [Google Scholar]
149.Kekalainen, E., Miettinen, A. & Arstila, T. P. Does the deficiency of Aire in mice really resemble human APECED? Nat. Rev. Immunol.7, 1 (2007). [DOI] [PubMed] [Google Scholar]
150.Wildin, R. S. & Freitas, A. IPEX and FOXP3: clinical and research perspectives. J. Autoimmun.25, 56–62 (2005). [DOI] [PubMed] [Google Scholar]
151.Saha, P. K., Strand, R. & Borgefors, G. Digital topology and geometry in medical imaging: a survey. IEEE Trans. Med. Imaging34, 1940–1964 (2015). [DOI] [PubMed] [Google Scholar]
152.Saha, P. K., Borgefors, G. & di Baja, G. S. A survey on skeletonization algorithms and their applications. Pattern Recognit. Lett.76, 3–12 (2016). [Google Scholar]
153.Ramakrishnan, V. et al. 3D visualization, skeletonization and branching analysis of blood vessels in angiogenesis. Int. J. Mol. Sci.24. 10.3390/ijms24097714 (2023). [DOI] [PMC free article] [PubMed]
154.Anderson, G. & Takahama, Y. Thymic epithelial cells: working class heroes for T cell development and repertoire selection. Trends Immunol.33, 256–263 (2012). [DOI] [PubMed] [Google Scholar]
155.Hun, M. L. et al. Gender disparity impacts on thymus aging and lhrh receptor antagonist-induced thymic reconstitution following chemotherapeutic damage. Front. Immunol.11, 302 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
156.Scheiff, J. M. Multinucleate cells in thymus of C3H mice. Cell Tissue Res.170, 305–314 (1976). [DOI] [PubMed] [Google Scholar]
157.Daubenmire, R. F. The use of the terms coenocyte and syncytium in biology. Science84, 533 (1936). [DOI] [PubMed] [Google Scholar]
158.Olona, A. et al. Adipoclast: a multinucleated fat-eating macrophage. BMC Biol.19, 246 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
159.Brunner, J. S. et al. Environmental arginine controls multinuclear giant cell metabolism and formation. Nat. Commun.11, 431 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
160.Kreutz, C. et al. Hepatocyte ploidy is a diversity factor for liver homeostasis. Front. Physiol.8, 862 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
161.Heinke, P. et al. Diploid hepatocytes drive physiological liver renewal in adult humans. Cell Syst.13, 499–507 e412 (2022). [DOI] [PubMed] [Google Scholar]
162.Shah, D. K. & Zuniga-Pflucker, J. C. An overview of the intrathymic intricacies of T cell development. J. Immunol.192, 4017–4023 (2014). [DOI] [PubMed] [Google Scholar]
163.Nakagawa, T. et al. Cathepsin L: critical role in Ii degradation and CD4 T cell selection in the thymus. Science280, 450–453 (1998). [DOI] [PubMed] [Google Scholar]
164.Gommeaux, J. et al. Thymus-specific serine protease regulates positive selection of a subset of CD4+ thymocytes. Eur. J. Immunol.39, 956–964 (2009). [DOI] [PubMed] [Google Scholar]
165.Nedjic, J., Aichinger, M., Emmerich, J., Mizushima, N. & Klein, L. Autophagy in thymic epithelium shapes the T-cell repertoire and is essential for tolerance. Nature455, 396–400 (2008). [DOI] [PubMed] [Google Scholar]
166.Aichinger, M., Wu, C., Nedjic, J. & Klein, L. Macroautophagy substrates are loaded onto MHC class II of medullary thymic epithelial cells for central tolerance. J. Exp. Med.210, 287–300 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
167.Patterson, N. L. & Mintern, J. D. Intersection of autophagy with pathways of antigen presentation. Protein Cell3, 911–920 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
168.Mandel, T. Differentiation of epithelial cells in the mouse thymus. Z. Zellforsch. Mikros. Anat.106, 498–515 (1970). [DOI] [PubMed] [Google Scholar]
169.Kato, S. & Schoefl, G. I. The vasculature of the guinea-pig thymus: topographic studies by light and electron microscopy. Arch. Histol. Jpn50, 299–314 (1987). [DOI] [PubMed] [Google Scholar]
170.Maeda, Y. et al. S1P lyase in thymic perivascular spaces promotes egress of mature thymocytes via up-regulation of S1P receptor 1. Int. Immunol.26, 245–255 (2014). [DOI] [PubMed] [Google Scholar]
171.Yagi, H. et al. Immunosuppressant FTY720 inhibits thymocyte emigration. Eur. J. Immunol.30, 1435–1444 (2000). [DOI] [PubMed] [Google Scholar]
172.Cyster, J. G. & Schwab, S. R. Sphingosine-1-phosphate and lymphocyte egress from lymphoid organs. Annu. Rev. Immunol.30, 69–94 (2012). [DOI] [PubMed] [Google Scholar]
173.Kuchler, L. et al. Elevated intrathymic sphingosine-1-phosphate promotes thymus involution during sepsis. Mol. Immunol.90, 255–263 (2017). [DOI] [PubMed] [Google Scholar]
174.Matloubian, M. et al. Lymphocyte egress from thymus and peripheral lymphoid organs is dependent on S1P receptor 1. Nature427, 355–360 (2004). [DOI] [PubMed] [Google Scholar]
175.Vincent, A., Scadding, G. K., Thomas, H. C. & Newsom-Davis, J. In-vitro synthesis of anti-acetylcholine-receptor antibody by thymic lymphocytes in myasthenia gravis. Lancet1, 305–307 (1978). [DOI] [PubMed] [Google Scholar]
176.Ivosev, G., Burton, L. & Bonner, R. Dimensionality reduction and visualization in principal component analysis. Anal. Chem.80, 4933–4944 (2008). [DOI] [PubMed] [Google Scholar]
177.Suarez-Najera, L. E. et al. Morphometric study of adipocytes on breast cancer by means of photonic microscopy and image analysis. Microsc. Res. Tech.81, 240–249 (2018). [DOI] [PubMed] [Google Scholar]
178.Langford, D. J. et al. Coding of facial expressions of pain in the laboratory mouse. Nat. Methods7, 447–449 (2010). [DOI] [PubMed] [Google Scholar]
179.Keating, S. C., Thomas, A. A., Flecknell, P. A. & Leach, M. C. Evaluation of EMLA cream for preventing pain during tattooing of rabbits: changes in physiological, behavioural and facial expression responses. PLoS ONE7, e44437 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
180.Dumont-Lagacé, M., St-Pierre, C. & Perreault, C. Sex hormones have pervasive effects on thymic epithelial cells. Sci. Rep.5, 12895 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
181.Velardi, E. et al. Sex steroid blockade enhances thymopoiesis by modulating Notch signaling. J. Exp. Med.211, 2341–2349 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
182.Treuting, P. M. & Snyder, J. M. Mouse necropsy. Curr. Protoc. Mouse Biol.5, 223–233 (2015). [DOI] [PubMed] [Google Scholar]
183.Rasmussen, L. & McInnes, E. Chapter 1: Necropsy of the mouse. in A Practical Guide to the Histology of the Mouse. (ed. Scudamore, C. L.) 1-24 (Chichester: Wiley-Blackwell, 2014).
184.Deerinck, T.J., Bushong, E.A., Ellisman, M.H. &Thor, A. Preparation of biological tissues for serial block face Scanning Electron Microscopy (SBEM) V2. protocils.io (2022). 10.17504/protocols.io.36wgq7je5vk5/v2
185.Mastronarde, D. N. Dual-axis tomography: an approach with alignment methods that preserve resolution. J. Struct. Biol.120, 343–352 (1997). [DOI] [PubMed] [Google Scholar]
186.Kremer, J. R., Mastronarde, D. N. & McIntosh, J. R. Computer visualization of three-dimensional image data using IMOD. J. Struct. Biol.116, 71–76 (1996). [DOI] [PubMed] [Google Scholar]
187.Metsalu, T. & Vilo, J. ClustVis: a web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap. Nucleic Acids Res.43, W566–W570 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Transparent Peer Review file^{(171.8KB, pdf)}

Supplementary Information^{(21.7MB, pdf)}

42003_2026_10038_MOESM3_ESM.pdf^{(29.9KB, pdf)}

Description of Additional Supplementary Files

Supplementary Data 1^{(90.2KB, pdf)}

Supplementary Data 2^{(137.7KB, pdf)}

Supplementary Video 1^{(295.8MB, 7z)}

Supplementary Video 2^{(253.4MB, 7z)}

Data Availability Statement

[CR1] 1.Pearse, G. Normal structure, function and histology of the thymus. Toxicol. Pathol.34, 504–514 (2006). [DOI] [PubMed] [Google Scholar]

[CR2] 2.Han, J. & Zuniga-Pflucker, J. C. A 2020 view of thymus stromal cells in T cell development. J. Immunol.206, 249–256 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.James, K. D., Jenkinson, W. E. & Anderson, G. Non-epithelial stromal cells in thymus development and function. Front. Immunol.12, 634367 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Kadouri, N., Nevo, S., Goldfarb, Y. & Abramson, J. Thymic epithelial cell heterogeneity: TEC by TEC. Nat. Rev. Immunol.20, 239–253 (2020). [DOI] [PubMed] [Google Scholar]

[CR5] 5.Granadier, D., Iovino, L., Kinsella, S. & Dudakov, J. A. Dynamics of thymus function and T cell receptor repertoire breadth in health and disease. Semin. Immunopathol.43, 119–134 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Abramson, J. & Anderson, G. Thymic epithelial cells. Annu. Rev. Immunol.35, 85–118 (2017). [DOI] [PubMed] [Google Scholar]

[CR7] 7.Alawam, A. S., Anderson, G. & Lucas, B. Generation and regeneration of thymic epithelial cells. Front. Immunol.11, 858 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Alexandropoulos, K. & Danzl, N. M. Thymic epithelial cells: antigen presenting cells that regulate T cell repertoire and tolerance development. Immunol. Res.54, 177–190 (2012). [DOI] [PubMed] [Google Scholar]

[CR9] 9.Chaudhry, M. S., Velardi, E., Dudakov, J. A. & van den Brink, M. R. Thymus: the next (re)generation. Immunol. Rev.271, 56–71 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Swann, J. B. & Boehm, T. Everything in its place. Immunity31, 856–858 (2009). [DOI] [PubMed] [Google Scholar]

[CR11] 11.Griffith, A. V. et al. Spatial mapping of thymic stromal microenvironments reveals unique features influencing T lymphoid differentiation. Immunity31, 999–1009 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Garcia-Ceca, J., Montero-Herradon, S., Alfaro, D. & Zapata, A. G. Increased epithelial-free areas in thymuses with altered EphB-mediated thymocyte-thymic epithelial cell interactions. Histochem. Cell Biol.148, 381–394 (2017). [DOI] [PubMed] [Google Scholar]

[CR13] 13.Clarke, A. G. & Kendall, M. D. Histological changes in the thymus during mouse pregnancy. Thymus14, 65–78 (1989). [PubMed] [Google Scholar]

[CR14] 14.Kondo, K., Ohigashi, I. & Takahama, Y. Thymus machinery for T-cell selection. Int. Immunol.31, 119–125 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Michelson, D. A., Hase, K., Kaisho, T., Benoist, C. & Mathis, D. Thymic epithelial cells co-opt lineage-defining transcription factors to eliminate autoreactive T cells. Cell185, 2542–2558 e2518 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.James, K. D., Jenkinson, W. E. & Anderson, G. T-cell egress from the thymus: should I stay or should I go? J. Leukoc. Biol.104, 275–284 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Mori, K., Itoi, M., Tsukamoto, N., Kubo, H. & Amagai, T. The perivascular space as a path of hematopoietic progenitor cells and mature T cells between the blood circulation and the thymic parenchyma. Int. Immunol.19, 745–753 (2007). [DOI] [PubMed] [Google Scholar]

[CR18] 18.James, K. D. et al. Endothelial cells act as gatekeepers for LTbetaR-dependent thymocyte emigration. J. Exp. Med.215, 2984–2993 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Nitta, T. Mesenchymal stromal cells in the thymus. Inflamm. Regen.42, 33 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Nitta, T., Ota, A., Iguchi, T., Muro, R. & Takayanagi, H. The fibroblast: an emerging key player in thymic T cell selection. Immunol. Rev.302, 68–85 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Zachariah, M. A. & Cyster, J. G. Neural crest-derived pericytes promote egress of mature thymocytes at the corticomedullary junction. Science328, 1129–1135 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Bautista, J. L. et al. Single-cell transcriptional profiling of human thymic stroma uncovers novel cellular heterogeneity in the thymic medulla. Nat. Commun.12, 1096 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Bleul, C. C. & Boehm, T. Chemokines define distinct microenvironments in the developing thymus. Eur. J. Immunol.30, 3371–3379 (2000). [DOI] [PubMed] [Google Scholar]

[CR24] 24.Dudakov, J. A. et al. Interleukin-22 drives endogenous thymic regeneration in mice. Science336, 91–95 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Lucas, B. et al. Progressive changes in CXCR4 expression that define thymocyte positive selection are dispensable for both innate and conventional alphabetat-cell development. Sci. Rep.7, 5068 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Pearse, G. Histopathology of the thymus. Toxicol. Pathol.34, 515–547 (2006). [DOI] [PubMed] [Google Scholar]

[CR27] 27.Elmore, S. A. Enhanced histopathology of the thymus. Toxicol. Pathol.34, 656–665 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Willard-Mack, C. L. et al. Nonproliferative and proliferative lesions of the rat and mouse hematolymphoid system. Toxicol. Pathol.47, 665–783 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Vodopyanov, S. et al. Mapping murine thymic epithelial cells: functional ultrastructure and implications for thymopoiesis. J. Leukoc. Biol.117. 10.1093/jleuko/qiaf115 (2025). [DOI] [PubMed]

[CR30] 30.Vodopyanov, S. et al. Beyond cysts: organization of epithelial networks in the murine thymus. J. Cell Sci. 10.1242/jcs.264079 (2025). [DOI] [PMC free article] [PubMed]

[CR31] 31.Nakagawa, Y. et al. Thymic nurse cells provide microenvironment for secondary T cell receptor alpha rearrangement in cortical thymocytes. Proc. Natl. Acad. Sci. USA109, 20572–20577 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Halin, C., Mora, J. R., Sumen, C. & von Andrian, U. H. In vivo imaging of lymphocyte trafficking. Annu. Rev. Cell Dev. Biol.21, 581–603 (2005). [DOI] [PubMed] [Google Scholar]

[CR33] 33.Aguilan, J. T. et al. Identifying space-resolved proteins of the murine thymus, by combining MALDI-MSI and proteomics. Life Sci Alliance9, e202503205 (2025). [DOI] [PMC free article] [PubMed]

[CR34] 34.Meyer, M. B., Lee, S. M., Cichanski, S. R., Cobice, D. F. & Pike, J. W. Spatial detection and consequences of nonrenal calcitriol production as assessed by targeted mass spectrometry imaging. JCI Insight9. 10.1172/jci.insight.181763 (2024). [DOI] [PMC free article] [PubMed]

[CR35] 35.Tsuji, Y., Yamaguchi, S., Nakamura, T. & Ikegawa, M. Mass Spectrometry Imaging (MSI) delineates thymus-centric metabolism in vivo as an effect of systemic administration of dexamethasone. Appl. Sci.11, 11038 (2021).

[CR36] 36.Yayon, N. et al. A spatial human thymus cell atlas mapped to a continuous tissue axis. Nature635, 708–718 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Brayton, C. F., Treuting, P. M. & Ward, J. M. Pathobiology of aging mice and GEM:background strains and experimental design. Vet. Pathol.49, 85–105 (2012). [DOI] [PubMed] [Google Scholar]

[CR38] 38.Loukov, D., Naidoo, A., Puchta, A., Marin, J. L. & Bowdish, D. M. Tumor necrosis factor drives increased splenic monopoiesis in old mice. J. Leukoc. Biol.100, 121–129 (2016). [DOI] [PubMed] [Google Scholar]

[CR39] 39.Baran-Gale, J. et al. Ageing compromises mouse thymus function and remodels epithelial cell differentiation. eLife9, e56221 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Kumar, B. V., Connors, T. J. & Farber, D. L. Human T cell development, localization, and function throughout life. Immunity48, 202–213 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Linton, P. J. & Dorshkind, K. Age-related changes in lymphocyte development and function. Nat. Immunol.5, 133–139 (2004). [DOI] [PubMed] [Google Scholar]

[CR42] 42.Palmer, D. The effect of age on thymic function. Front. Immunol.4. 10.3389/fimmu.2013.00316 (2013). [DOI] [PMC free article] [PubMed]

[CR43] 43.Elmore, S. A. Enhanced Histopathology Evaluation of Lymphoid Organs. Methods Mol. Biol.1803, 147–168 (2018). [DOI] [PubMed] [Google Scholar]

[CR44] 44.Kavanagh, M., Pulido, O., Caldwell, D. & Gill, S. 229-243 (2005).

[CR45] 45.Farr, A. G. & Braddy, S. C. Patterns of keratin expression in the murine thymus. Anat. Rec.224, 374–378 (1989). [DOI] [PubMed] [Google Scholar]

[CR46] 46.Gupta, R. et al. Cytokeratin (CK5, CK8, CK14) expression and presence of progenitor stem cells in human fetal thymuses. Clin. Anat.29, 711–717 (2016). [DOI] [PubMed] [Google Scholar]

[CR47] 47.Lee, E. N. et al. Characterization of the expression of cytokeratins 5, 8, and 14 in mouse thymic epithelial cells during thymus regeneration following acute thymic involution. Anat. Cell Biol.44, 14–24 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Odaka, C. et al. Keratin 8 is required for the maintenance of architectural structure in thymus epithelium. PLoS ONE8, e75101 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Nusser, A. et al. Developmental dynamics of two bipotent thymic epithelial progenitor types. Nature606, 165–171 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Grevellec, A. & Tucker, A. S. The pharyngeal pouches and clefts: development, evolution, structure and derivatives. Semin. Cell Dev. Biol.21, 325–332 (2010). [DOI] [PubMed] [Google Scholar]

[CR51] 51.Vaidya, H. J., Briones Leon, A. & Blackburn, C. C. FOXN1 in thymus organogenesis and development. Eur. J. Immunol.46, 1826–1837 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Wertheimer, T. et al. Production of BMP4 by endothelial cells is crucial for endogenous thymic regeneration. Sci Immunol3. 10.1126/sciimmunol.aal2736 (2018). [DOI] [PMC free article] [PubMed]

[CR53] 53.Wekerle, H. & Ketelsen, U.-P. Thymic nurse cells—Ia-bearing epithelium involved in T-lymphocyte differentiation? Nature283, 402–404 (1980). [DOI] [PubMed] [Google Scholar]

[CR54] 54.Nakagawa, Y. et al. Thymic nurse cells provide microenvironment for secondary T cell receptor α rearrangement in cortical thymocytes. Proc. Natl. Acad. Sci. USA109, 20572–20577 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Guyden, J. C. & Pezzano, M. Thymic nurse cells: a microenvironment for thymocyte development and selection. Int. Rev. Cytol.223, 1–37 (2003). [DOI] [PubMed] [Google Scholar]

[CR56] 56.Arganda-Carreras, I., Fernandez-Gonzalez, R., Munoz-Barrutia, A. & Ortiz-De-Solorzano, C. 3D reconstruction of histological sections: application to mammary gland tissue. Microsc. Res. Tech.73, 1019–1029 (2010). [DOI] [PubMed] [Google Scholar]

[CR57] 57.Venables, T., Griffith, A. V., DeAraujo, A. & Petrie, H. T. Dynamic changes in epithelial cell morphology control thymic organ size during atrophy and regeneration. Nat. Commun.10, 4402 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Griffith, A. V., Fallahi, M., Venables, T. & Petrie, H. T. Persistent degenerative changes in thymic organ function revealed by an inducible model of organ regrowth. Aging Cell11, 169–177 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Hernandez-Lopez, C. et al. The CXCL12/CXCR4 pair in aged human thymus. Neuroimmunomodulation17, 217–220 (2010). [DOI] [PubMed] [Google Scholar]

[CR60] 60.Milicevic, N. M., Nohroudi, K., Milicevic, Z. & Westermann, J. Activation of cortical and inhibited differentiation of medullary epithelial cells in the thymus of lymphotoxin-beta receptor-deficient mice: an ultrastructural study. J. Anat.212, 114–124 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Borelli, A. & Irla, M. Lymphotoxin: from the physiology to the regeneration of the thymic function. Cell Death Differ.28, 2305–2314 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Kwok, T. et al. Age-associated changes to lymph node fibroblastic reticular cells. Front. Aging3, 838943 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Cavallotti, C., D’Andrea, V., Tonnarini, G., Cavallotti, C. & Bruzzone, P. Age-related changes in the human thymus studied with scanning electron microscopy. Microsc. Res. Tech.71, 573–578 (2008). [DOI] [PubMed] [Google Scholar]

[CR64] 64.De Waal, E. J. & Rademakers, L. H. Heterogeneity of epithelial cells in the rat thymus. Microsc. Res. Tech.38, 227–236 (1997). [DOI] [PubMed] [Google Scholar]

[CR65] 65.De Waal, E. J., Rademakers, L. H., Schuurman, H. J., Van Loveren, H. & Vos, J. G. Ultrastructure of the cortical epithelium of the rat thymus after in vivo exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Arch. Toxicol.67, 558–564 (1993). [DOI] [PubMed] [Google Scholar]

[CR66] 66.Erokhina, V. & Avilova, O. Ultramicroscopic changes of rats parathyroid glands and thymus after single administration of cyclophosphamide at the different periods of observation. Wiad. Lek.72, 362–367 (2019). [PubMed] [Google Scholar]

[CR67] 67.Mandel, T. Ultrastructure of epithelial cells in the cortex of guinea pig thymus. Z. Zellforsch. Mikros. Anat.92, 159–168 (1968). [DOI] [PubMed] [Google Scholar]

[CR68] 68.Radulovic, S., Sunkara, S., Rachel, R. & Leitinger, G. Three-dimensional SEM, TEM, and STEM for analysis of large-scale biological systems. Histochem. Cell Biol.158, 203–211 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR69] 69.Kendall, M. D. The syncytial nature of epithelial cells in the thymic cortex. J. Anat.147, 95–106 (1986). [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Vignery, A. Osteoclasts and giant cells: macrophage-macrophage fusion mechanism. Int. J. Exp. Pathol.81, 291–304 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Mandel, T. Epithelial cells and lymphopoiesis in the cortex of guinea-pig thymus. Aust. J. Exp. Biol. Med. Sci.47, 153–155 (1969). [DOI] [PubMed] [Google Scholar]

[CR72] 72.Nitta, T. et al. Thymoproteasome shapes immunocompetent repertoire of CD8+ T cells. Immunity32, 29–40 (2010). [DOI] [PubMed] [Google Scholar]

[CR73] 73.Rodrigues, P. M. et al. LAMP2 regulates autophagy in the thymic epithelium and thymic stroma-dependent CD4 T cell development. Autophagy19, 426–439 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR74] 74.Jung, M., Choi, H. & Mun, J. Y. The autophagy research in electron microscopy. Appl. Microsc.49, 11 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR75] 75.Hariri, M. et al. Biogenesis of multilamellar bodies via autophagy. Mol. Biol. Cell11, 255–268 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR76] 76.Klionsky, D. J. et al. Guidelines for the use and interpretation of assays for monitoring autophagy in higher eukaryotes. Autophagy4, 151–175 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR77] 77.Nagy, P., Varga, Á, Kovács, A. L., Takáts, S. & Juhász, G. How and why to study autophagy in Drosophila: It’s more than just a garbage chute. Methods75, 151–161 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR78] 78.Clague, M. J. & Urbé, S. Multivesicular bodies. Curr. Biol.18, R402–R404 (2008). [DOI] [PubMed] [Google Scholar]

[CR79] 79.Eskelinen, E.-L., Reggiori, F., Baba, M., Kovács, A. L. & Seglen, P. O. Seeing is believing: the impact of electron microscopy on autophagy research. Autophagy7, 935–956 (2011). [DOI] [PubMed] [Google Scholar]

[CR80] 80.Ylä-Anttila, P., Vihinen, H., Jokitalo, E. & Eskelinen, E. L. Monitoring autophagy by electron microscopy in Mammalian cells. Methods Enzymol.452, 143–164 (2009). [DOI] [PubMed] [Google Scholar]

[CR81] 81.Uddin, M. N., Nishio, N., Ito, S., Suzuki, H. & Isobe, K. Autophagic activity in thymus and liver during aging. Age34, 75–85 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR82] 82.Zhou, T. A. et al. Thymic macrophages consist of two populations with distinct localization and origin. Elife11. 10.7554/eLife.75148 (2022). [DOI] [PMC free article] [PubMed]

[CR83] 83.Dung, H. C. Electron microscopic study of involuting thymus of “lethargic” mutant mice. Anat. Rec.177, 585–601 (1973). [DOI] [PubMed] [Google Scholar]

[CR84] 84.Lee-Rueckert, M., Lappalainen, J., Kovanen, P. T. & Escola-Gil, J. C. Lipid-Laden macrophages and inflammation in atherosclerosis and cancer: an integrative view. Front. Cardiovasc. Med.9, 777822 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR85] 85.Kazachkov, M. Y., Muhlebach, M. S., Livasy, C. A. & Noah, T. L. Lipid-laden macrophage index and inflammation in bronchoalveolar lavage fluids in children. Eur. Respir. J.18, 790–795 (2001). [DOI] [PubMed] [Google Scholar]

[CR86] 86.Bearman, R. M., Bensch, K. G. & Levine, G. D. The normal human thymic vasculature: an ultrastructural study. Anat. Rec.183, 485–497 (1975). [DOI] [PubMed] [Google Scholar]

[CR87] 87.Kato, S. Thymic microvascular system. Microsc. Res. Tech.38, 287–299 (1997). [DOI] [PubMed] [Google Scholar]

[CR88] 88.Kendall, M. D. The morphology of perivascular spaces in the thymus. Thymus13, 157–164 (1989). [PubMed] [Google Scholar]

[CR89] 89.Nunez, S. et al. The human thymus perivascular space is a functional niche for viral-specific plasma cells. Sci. Immunol.1. 10.1126/sciimmunol.aah4447 (2016). [DOI] [PMC free article] [PubMed]

[CR90] 90.Ushiki, T. & Takeda, M. Three-dimensional ultrastructure of the perivascular space in the rat thymus. Arch. Histol. Cytol.60, 89–99 (1997). [DOI] [PubMed] [Google Scholar]

[CR91] 91.Odaka, C., Morisada, T., Oike, Y. & Suda, T. Distribution of lymphatic vessels in mouse thymus: immunofluorescence analysis. Cell Tissue Res.325, 13–22 (2006). [DOI] [PubMed] [Google Scholar]

[CR92] 92.Baeyens, A. A. L. & Schwab, S. R. Finding a way out: S1P signaling and immune cell migration. Annu. Rev. Immunol.38, 759–784 (2020). [DOI] [PubMed] [Google Scholar]

[CR93] 93.Kumar, A. & Saba, J. D. Regulation of immune cell migration by Sphingosine-1-Phosphate. Cell Mol. Biol.61, 121 (2015). [PMC free article] [PubMed]

[CR94] 94.Weinreich, M. A. & Hogquist, K. A. Thymic emigration: when and how T cells leave home. J. Immunol.181, 2265–2270 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR95] 95.Colomb, E. et al. Genetic control of giant perivascular space formation in the thymus of NOD mice. Diabetes45, 1535–1540 (1996). [DOI] [PubMed] [Google Scholar]

[CR96] 96.Flores, K. G., Li, J., Sempowski, G. D., Haynes, B. F. & Hale, L. P. Analysis of the human thymic perivascular space during aging. J. Clin. Investig.104, 1031–1039 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR97] 97.Funk, J., Ruehl-Fehlert, C., Leonard, C., Kellner, R. & Rittinghausen, S. Immunohistochemical characterization of proliferative lesions in the thymus of aging CD-1 mice from two studies on the RITA database, with special reference to the perivascular space. Toxicol. Pathol.50, 308–328 (2022). [DOI] [PubMed] [Google Scholar]

[CR98] 98.Kato, S. & Schoefl, G. I. Microvasculature of normal and involuted mouse thymus. Light- Electron-Microsc. Study Acta Anat.135, 1–11 (1989). [DOI] [PubMed] [Google Scholar]

[CR99] 99.Steinmann, G. G., Klaus, B. & Muller-Hermelink, H. K. The involution of the ageing human thymic epithelium is independent of puberty. A morphometric study. Scand. J. Immunol.22, 563–575 (1985). [DOI] [PubMed] [Google Scholar]

[CR100] 100.Liang, Z., Dong, X., Zhang, Z., Zhang, Q. & Zhao, Y. Age-related thymic involution: mechanisms and functional impact. Aging Cell21, e13671 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR101] 101.Mitre, T. M., Pietropaolo, M. & Khadra, A. The dual role of autoimmune regulator in maintaining normal expression level of tissue-restricted autoantigen in the thymus: a modeling investigation. Math. Biosci.287, 12–23 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR102] 102.Zuklys, S. et al. Normal thymic architecture and negative selection are associated with Aire expression, the gene defective in the autoimmune-polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED). J. Immunol.165, 1976–1983 (2000). [DOI] [PubMed] [Google Scholar]

[CR103] 103.Chen, J., Yang, W., Yu, C. & Li, Y. Autoimmune regulator initiates the expression of promiscuous genes in thymic epithelial cells. Immunol. Invest37, 203–214 (2008). [DOI] [PubMed] [Google Scholar]

[CR104] 104.Cabric, V. & Brown, C. C. Thymic mimicry: the art of imitation. J. Exp. Med.220. 10.1084/jem.20231010 (2023). [DOI] [PMC free article] [PubMed]

[CR105] 105.Michelson, D. A. & Mathis, D. Thymic mimetic cells: tolerogenic masqueraders. Trends Immunol.43, 782–791 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR106] 106.Miller, C. N. et al. Thymic tuft cells promote an IL-4-enriched medulla and shape thymocyte development. Nature559, 627–631 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR107] 107.Bornstein, C. et al. Single-cell mapping of the thymic stroma identifies IL-25-producing tuft epithelial cells. Nature559, 622–626 (2018). [DOI] [PubMed] [Google Scholar]

[CR108] 108.Srinivasan, J. et al. Age-related changes in thymic central tolerance. Front. Immunol.12, 676236 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR109] 109.Givony, T. et al. Thymic mimetic cells function beyond self-tolerance. Nature622, 164–172 (2023). [DOI] [PubMed] [Google Scholar]

[CR110] 110.Burchiel, S. W. et al. DMBA-induced cytotoxicity in lymphoid and nonlymphoid organs of B6C3F1 mice: relation of cell death to target cell intracellular calcium and DNA damage. Toxicol. Appl. Pharm.113, 126–132 (1992). [DOI] [PubMed] [Google Scholar]

[CR111] 111.Argyris, D. G. et al. Suppression of chemically induced mammary cancer by early-life oral administration of cholera toxin in mice is associated with aberrant regulation of Bmp and Notch signaling pathways. Mol. Biol. Rep.52, 150 (2025). [DOI] [PubMed] [Google Scholar]

[CR112] 112.Doulberis, M. et al. Cholera-toxin suppresses carcinogenesis in a mouse model of inflammation-driven sporadic colon cancer. Carcinogenesis36, 280–290 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR113] 113.Poutahidis, T., Angelopoulou, K. & Erdman, S. E. Old enemies meet new friends for colon cancer prevention. Oncoimmunology4, e1027474 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR114] 114.Varian, B. J. et al. Consuming cholera toxin counteracts age-associated obesity. Oncotarget10, 5497–5509 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR115] 115.Afaloniati, H. et al. Peri-weaning cholera toxin consumption suppresses chemically-induced carcinogenesis in mice. Int. J. Cancer154, 1097–1110 (2024). [DOI] [PubMed] [Google Scholar]

[CR116] 116.Bach, J. F. The hygiene hypothesis in autoimmunity: the role of pathogens and commensals. Nat. Rev. Immunol.18, 105–120 (2018). [DOI] [PubMed] [Google Scholar]

[CR117] 117.Fletcher, A. L. et al. Ablation and regeneration of tolerance-inducing medullary thymic epithelial cells after cyclosporine, cyclophosphamide, and dexamethasone treatment. J. Immunol.183, 823–831 (2009). [DOI] [PubMed] [Google Scholar]

[CR118] 118.Schulz, O., Hammerschmidt, S. I., Moschovakis, G. L. & Forster, R. Chemokines and chemokine receptors in lymphoid tissue dynamics. Annu. Rev. Immunol.34, 203–242 (2016). [DOI] [PubMed] [Google Scholar]

[CR119] 119.Halkias, J., Melichar, H. J., Taylor, K. T. & Robey, E. A. Tracking migration during human T cell development. Cell Mol. Life Sci.71, 3101–3117 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR120] 120.Itoi, M., Tsukamoto, N., Yoshida, H. & Amagai, T. Mesenchymal cells are required for functional development of thymic epithelial cells. Int. Immunol.19, 953–964 (2007). [DOI] [PubMed] [Google Scholar]

[CR121] 121.Lee, J. Y. M. & Love, P. E. A simplified approach to evaluating T Cell development by flow cytometry. Methods Mol. Biol.2580, 131–149 (2023). [DOI] [PubMed] [Google Scholar]

[CR122] 122.Cahill, R. N., Kimpton, W. G., Cunningham, C. P. & Washington, E. A. An immune system switch at birth triggers a change in the lifespan of peripheral T cells. Semin. Immunol.9, 355–363 (1997). [DOI] [PubMed] [Google Scholar]

[CR123] 123.Webb, S. & Haniffa, M. Large-scale single-cell RNA sequencing atlases of human immune cells across lifespan: possibilities and challenges. Eur. J. Immunol. e2250222. 10.1002/eji.202250222 (2023). [DOI] [PubMed]

[CR124] 124.Minato, N., Hattori, M. & Hamazaki, Y. Physiology and pathology of T-cell aging. Int Immunol.32, 223–231 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR125] 125.Ostan, R. et al. Immunosenescence and immunogenetics of human longevity. Neuroimmunomodulation15, 224–240 (2008). [DOI] [PubMed] [Google Scholar]

[CR126] 126.Palmer, S., Albergante, L., Blackburn, C. C. & Newman, T. J. Thymic involution and rising disease incidence with age. Proc. Natl. Acad. Sci. USA115, 1883–1888 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR127] 127.Kooshesh, K. A., Foy, B. H., Sykes, D. B., Gustafsson, K. & Scadden, D. T. Health Consequences of Thymus Removal in Adults. N. Engl. J. Med389, 406–417 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR128] 128.Gulla, S. et al. Role of thymus in health and disease. Int. Rev. Immunol. 42, 347–363 (2022). [DOI] [PubMed]

[CR129] 129.Chopp, L., Redmond, C., O’Shea, J. J. & Schwartz, D. M. From thymus to tissues and tumors: a review of T cell biology. J Allergy Clin. Immunol. 10.1016/j.jaci.2022.10.011 (2022). [DOI] [PMC free article] [PubMed]

[CR130] 130.Wang, W., Thomas, R., Sizova, O. & Su, D. M. Thymic function associated with cancer development, relapse, and antitumor immunity - a mini-review. Front. Immunol.11, 773 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR131] 131.Sansoni, P. et al. The immune system in extreme longevity. Exp. Gerontol.43, 61–65 (2008). [DOI] [PubMed] [Google Scholar]

[CR132] 132.Marinova, T. T. Epithelial framework reorganization during human thymus involution. Gerontology51, 14–18 (2005). [DOI] [PubMed] [Google Scholar]

[CR133] 133.Ansari, A. R. & Liu, H. Acute thymic involution and mechanisms for recovery. Arch. Immunol. Ther. Exp.65, 401–420 (2017). [DOI] [PubMed] [Google Scholar]

[CR134] 134.Kinsella, S. & Dudakov, J. A. When the damage is done: injury and repair in thymus function. Front. Immunol.11, 1745 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR135] 135.Lagou, M. K., Anastasiadou, D. P. & Karagiannis, G. S. A proposed link between acute thymic involution and late adverse effects of chemotherapy. Front. Immunol.13, 933547 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR136] 136.Rijhsinghani, A. G., Thompson, K., Bhatia, S. K. & Waldschmidt, T. J. Estrogen blocks early T cell development in the thymus. Am. J. Reprod. Immunol.36, 269–277 (1996). [DOI] [PubMed] [Google Scholar]

[CR137] 137.Taves, M. D. & Ashwell, J. D. Effects of sex steroids on thymic epithelium and thymocyte development. Front. Immunol.13, 975858 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR138] 138.Laan, M., Haljasorg, U., Kisand, K., Salumets, A. & Peterson, P. Pregnancy-induced thymic involution is associated with suppression of chemokines essential for T-lymphoid progenitor homing. Eur. J. Immunol.46, 2008–2017 (2016). [DOI] [PubMed] [Google Scholar]

[CR139] 139.Swami, S., Tong, I., Bilodeau, C. C. & Bourjeily, G. Thymic involution in pregnancy: a universal finding? Obstet. Med. 5, 130–132 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR140] 140.Yang, H. et al. Obesity accelerates thymic aging. Blood114, 3803–3812 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR141] 141.Lagou, M. K. & Karagiannis, G. S. Obesity-induced thymic involution and cancer risk. Semin Cancer Biol.93, 3–19 (2023). [DOI] [PubMed] [Google Scholar]

[CR142] 142.Reith, W. & Mach, B. The bare lymphocyte syndrome and the regulation of MHC expression. Annu Rev. Immunol.19, 331–373 (2001). [DOI] [PubMed] [Google Scholar]

[CR143] 143.Ferrua, F. et al. Thymic epithelial cell alterations and defective thymopoiesis lead to central and peripheral tolerance perturbation in MHCII Deficiency. Front. Immunol.12, 669943 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR144] 144.Costa-Carvalho, B. T. et al. A remarkable depletion of both naive CD4+ and CD8+ with high proportion of memory T cells in an IPEX infant with a FOXP3 mutation in the forkhead domain. Scand. J. Immunol.68, 85–91 (2008). [DOI] [PubMed] [Google Scholar]

[CR145] 145.Marx, A. et al. Thymus and autoimmunity. Semin Immunopathol.43, 45–64 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR146] 146.White, A. J. et al. Lymphotoxin signals from positively selected thymocytes regulate the terminal differentiation of medullary thymic epithelial cells. J. Immunol.185, 4769–4776 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR147] 147.Van de Velde, R. L. & Friedman, N. B. Thymic myoid cells and myasthenia gravis. Am. J. Pathol.59, 347–368 (1970). [PMC free article] [PubMed] [Google Scholar]

[CR148] 148.Kekalainen, E. et al. Defective central tolerance in Aire-deficient mice is not sufficient to induce symptomatic autoimmunity during lymphopenia-induced T cell proliferation. Scand. J. Immunol.74, 71–79 (2011). [DOI] [PubMed] [Google Scholar]

[CR149] 149.Kekalainen, E., Miettinen, A. & Arstila, T. P. Does the deficiency of Aire in mice really resemble human APECED? Nat. Rev. Immunol.7, 1 (2007). [DOI] [PubMed] [Google Scholar]

[CR150] 150.Wildin, R. S. & Freitas, A. IPEX and FOXP3: clinical and research perspectives. J. Autoimmun.25, 56–62 (2005). [DOI] [PubMed] [Google Scholar]

[CR151] 151.Saha, P. K., Strand, R. & Borgefors, G. Digital topology and geometry in medical imaging: a survey. IEEE Trans. Med. Imaging34, 1940–1964 (2015). [DOI] [PubMed] [Google Scholar]

[CR152] 152.Saha, P. K., Borgefors, G. & di Baja, G. S. A survey on skeletonization algorithms and their applications. Pattern Recognit. Lett.76, 3–12 (2016). [Google Scholar]

[CR153] 153.Ramakrishnan, V. et al. 3D visualization, skeletonization and branching analysis of blood vessels in angiogenesis. Int. J. Mol. Sci.24. 10.3390/ijms24097714 (2023). [DOI] [PMC free article] [PubMed]

[CR154] 154.Anderson, G. & Takahama, Y. Thymic epithelial cells: working class heroes for T cell development and repertoire selection. Trends Immunol.33, 256–263 (2012). [DOI] [PubMed] [Google Scholar]

[CR155] 155.Hun, M. L. et al. Gender disparity impacts on thymus aging and lhrh receptor antagonist-induced thymic reconstitution following chemotherapeutic damage. Front. Immunol.11, 302 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR156] 156.Scheiff, J. M. Multinucleate cells in thymus of C3H mice. Cell Tissue Res.170, 305–314 (1976). [DOI] [PubMed] [Google Scholar]

[CR157] 157.Daubenmire, R. F. The use of the terms coenocyte and syncytium in biology. Science84, 533 (1936). [DOI] [PubMed] [Google Scholar]

[CR158] 158.Olona, A. et al. Adipoclast: a multinucleated fat-eating macrophage. BMC Biol.19, 246 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR159] 159.Brunner, J. S. et al. Environmental arginine controls multinuclear giant cell metabolism and formation. Nat. Commun.11, 431 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR160] 160.Kreutz, C. et al. Hepatocyte ploidy is a diversity factor for liver homeostasis. Front. Physiol.8, 862 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR161] 161.Heinke, P. et al. Diploid hepatocytes drive physiological liver renewal in adult humans. Cell Syst.13, 499–507 e412 (2022). [DOI] [PubMed] [Google Scholar]

[CR162] 162.Shah, D. K. & Zuniga-Pflucker, J. C. An overview of the intrathymic intricacies of T cell development. J. Immunol.192, 4017–4023 (2014). [DOI] [PubMed] [Google Scholar]

[CR163] 163.Nakagawa, T. et al. Cathepsin L: critical role in Ii degradation and CD4 T cell selection in the thymus. Science280, 450–453 (1998). [DOI] [PubMed] [Google Scholar]

[CR164] 164.Gommeaux, J. et al. Thymus-specific serine protease regulates positive selection of a subset of CD4+ thymocytes. Eur. J. Immunol.39, 956–964 (2009). [DOI] [PubMed] [Google Scholar]

[CR165] 165.Nedjic, J., Aichinger, M., Emmerich, J., Mizushima, N. & Klein, L. Autophagy in thymic epithelium shapes the T-cell repertoire and is essential for tolerance. Nature455, 396–400 (2008). [DOI] [PubMed] [Google Scholar]

[CR166] 166.Aichinger, M., Wu, C., Nedjic, J. & Klein, L. Macroautophagy substrates are loaded onto MHC class II of medullary thymic epithelial cells for central tolerance. J. Exp. Med.210, 287–300 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR167] 167.Patterson, N. L. & Mintern, J. D. Intersection of autophagy with pathways of antigen presentation. Protein Cell3, 911–920 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR168] 168.Mandel, T. Differentiation of epithelial cells in the mouse thymus. Z. Zellforsch. Mikros. Anat.106, 498–515 (1970). [DOI] [PubMed] [Google Scholar]

[CR169] 169.Kato, S. & Schoefl, G. I. The vasculature of the guinea-pig thymus: topographic studies by light and electron microscopy. Arch. Histol. Jpn50, 299–314 (1987). [DOI] [PubMed] [Google Scholar]

[CR170] 170.Maeda, Y. et al. S1P lyase in thymic perivascular spaces promotes egress of mature thymocytes via up-regulation of S1P receptor 1. Int. Immunol.26, 245–255 (2014). [DOI] [PubMed] [Google Scholar]

[CR171] 171.Yagi, H. et al. Immunosuppressant FTY720 inhibits thymocyte emigration. Eur. J. Immunol.30, 1435–1444 (2000). [DOI] [PubMed] [Google Scholar]

[CR172] 172.Cyster, J. G. & Schwab, S. R. Sphingosine-1-phosphate and lymphocyte egress from lymphoid organs. Annu. Rev. Immunol.30, 69–94 (2012). [DOI] [PubMed] [Google Scholar]

[CR173] 173.Kuchler, L. et al. Elevated intrathymic sphingosine-1-phosphate promotes thymus involution during sepsis. Mol. Immunol.90, 255–263 (2017). [DOI] [PubMed] [Google Scholar]

[CR174] 174.Matloubian, M. et al. Lymphocyte egress from thymus and peripheral lymphoid organs is dependent on S1P receptor 1. Nature427, 355–360 (2004). [DOI] [PubMed] [Google Scholar]

[CR175] 175.Vincent, A., Scadding, G. K., Thomas, H. C. & Newsom-Davis, J. In-vitro synthesis of anti-acetylcholine-receptor antibody by thymic lymphocytes in myasthenia gravis. Lancet1, 305–307 (1978). [DOI] [PubMed] [Google Scholar]

[CR176] 176.Ivosev, G., Burton, L. & Bonner, R. Dimensionality reduction and visualization in principal component analysis. Anal. Chem.80, 4933–4944 (2008). [DOI] [PubMed] [Google Scholar]

[CR177] 177.Suarez-Najera, L. E. et al. Morphometric study of adipocytes on breast cancer by means of photonic microscopy and image analysis. Microsc. Res. Tech.81, 240–249 (2018). [DOI] [PubMed] [Google Scholar]

[CR178] 178.Langford, D. J. et al. Coding of facial expressions of pain in the laboratory mouse. Nat. Methods7, 447–449 (2010). [DOI] [PubMed] [Google Scholar]

[CR179] 179.Keating, S. C., Thomas, A. A., Flecknell, P. A. & Leach, M. C. Evaluation of EMLA cream for preventing pain during tattooing of rabbits: changes in physiological, behavioural and facial expression responses. PLoS ONE7, e44437 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR180] 180.Dumont-Lagacé, M., St-Pierre, C. & Perreault, C. Sex hormones have pervasive effects on thymic epithelial cells. Sci. Rep.5, 12895 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR181] 181.Velardi, E. et al. Sex steroid blockade enhances thymopoiesis by modulating Notch signaling. J. Exp. Med.211, 2341–2349 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR182] 182.Treuting, P. M. & Snyder, J. M. Mouse necropsy. Curr. Protoc. Mouse Biol.5, 223–233 (2015). [DOI] [PubMed] [Google Scholar]

[CR183] 183.Rasmussen, L. & McInnes, E. Chapter 1: Necropsy of the mouse. in A Practical Guide to the Histology of the Mouse. (ed. Scudamore, C. L.) 1-24 (Chichester: Wiley-Blackwell, 2014).

[CR184] 184.Deerinck, T.J., Bushong, E.A., Ellisman, M.H. &Thor, A. Preparation of biological tissues for serial block face Scanning Electron Microscopy (SBEM) V2. protocils.io (2022). 10.17504/protocols.io.36wgq7je5vk5/v2

[CR185] 185.Mastronarde, D. N. Dual-axis tomography: an approach with alignment methods that preserve resolution. J. Struct. Biol.120, 343–352 (1997). [DOI] [PubMed] [Google Scholar]

[CR186] 186.Kremer, J. R., Mastronarde, D. N. & McIntosh, J. R. Computer visualization of three-dimensional image data using IMOD. J. Struct. Biol.116, 71–76 (1996). [DOI] [PubMed] [Google Scholar]

[CR187] 187.Metsalu, T. & Vilo, J. ClustVis: a web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap. Nucleic Acids Res.43, W566–W570 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Morphometric analysis of the thymic epithelial cell network using integrated and orthogonal digital pathology approaches

Maria K Lagou

Maria P Markaki

Stepan Vodopyanov

Dimitrios G Argyris

George K Karagiorgis

Leslie Gunther-Cummins

Alexandros Hardas

Christos Panorias

Sophia DesMarais

Conner Entenberg

Randall S Carpenter

Panagiota S Filippou

Hillary Guzik

Xheni Nishku

Joseph Churaman

Vera DesMarais

Frank P Macaluso

Maria Maryanovich

Katerina Angelopoulou

Theofilos Poutahidis

George S Karagiannis

Abstract

Introduction

Results

Mouse cohort necropsy – general findings and progressive age-associated lesions

Fig. 1. Mouse model of age-dependent thymic involution.

Microscopic assessment of cortex-medulla compartmentalization

Fig. 2. Morphometric assessment of corticomedullary ratio.

Topological assessment of the cortical thymic epithelial cell (cTEC) network

Fig. 3. Topological assessment of the cTEC network.

Ultrastructural topography of the cTEC network

Fig. 4. Ultrastructural topography of the cortical parenchyma and the cTEC meshwork.

Morphometric analysis of the medullary perivascular space (PVS)

Fig. 5. Morphometric assessment of the medullary perivascular space.

Topological assessment of the medullary thymic epithelial cell (mTEC) network

Fig. 6. Topological assessment of the mTEC network.

Early-life cholera-toxin (CT) exposure improves medullary architecture in a mouse model of DMBA-driven thymic injury and carcinogenesis

Fig. 7. Mouse model of DMBA-induced thymic injury with/without early-life exposure to cholera toxin (CT).

Fig. 8. Topological assessments of the cTEC network in a mouse model of DMBA-induced thymic injury with/without early-life exposure to CT.

Fig. 9. Topological assessments of the PVS and mTEC network in a mouse model of DMBA-induced thymic injury with/without early-life exposure to CT.

Data integration using principal component analysis (PCA)

Fig. 10. Principal Component Analysis (PCA) of thymic morphometric parameters in different mouse models of thymic pathology.

Discussion

Methods

Animal subjects

Ethics statement

Mouse model of age-dependent thymic involution

Mouse model of DMBA-driven chemical carcinogenesis

Necropsy procedures

Histology and immunofluorescence staining, and image acquisition

Tissue preparation

Hematoxylin & Eosin (HE) staining

Immunofluorescence staining

Microscopy and slide scanning

Imaging manipulation

Electron microscopy

Tissue preparation

Transmission electron microscopy (TEM)

OTO processing and scanning electron microscopy (OTO-SEM)

Three-dimensional reconstruction of the cTEC meshwork

cTEC vacuole analysis

Statistics and reproducibility

Statistical package

Statistical hypothesis testing

Exclusion criteria for individual cases

Blindness in image analysis

Reproducibility

Principal component analysis (PCA)

Supplementary information

Acknowledgements

Author contributions

Peer review

Peer review information

Data availability

Competing interests

Footnotes

Supplementary information

References