No description available.

Lactating dairy cows heavily rely on mammary gland functionality to maximize milk production. The number and activity of secretory mammary epithelial cells (MEC) plays a pivotal role in defining the synthesis potential of the gland. This dissertation aimed to investigate the effects of increased milking frequency (IMF), heat stress (HS), and cell heterogeneity as key contributors to the regulation of mammary gland milk synthesis capacity in lactating Holstein cows. The first study evaluated the implementation of IMF with 2x and 4x udder halves at early and mid-lactation for 21 and 20 d on milk yield (MY) and its association with changes in cistern and alveolar capacity. Results showed that udder halves milked 4x produced 2.27 kg more MY. Additionally, cows milked during early and mid-lactation had increased cistern capacity, while alveolar capacity remained unaffected. This suggests that increased cistern capacity may support MY enhancement through possible systemic responses caused by IMF. The second study examined the effects of 4 days of HS on mammary gland tissue structure, MEC number, and activity using a pair feeding model. Heat stress reduced MY of 4.3 kg/d. At the tissue level, HS decreased alveolar area and increased alveoli number and nucleated MEC per area. Gene expression analysis revealed unaffected activity-related targets but showed reduced phosphorylation of protein synthesis (pSTAT5) and cell survival (pS6K1) markers, as well as upregulation of an autophagosome-related protein (LC3 II). These findings indicate impaired pathways that could explain the reduction in MY after acute HS. The final study utilized single-cell RNA sequencing (scRNA-seq) to characterize the heterogeneity of epithelial and immune cell subpopulations in milk. Analysis revealed multiple subpopulations with distinct gene expression profiles, including different subtypes of mammary epithelial cells expressing representative marker genes (CSN3, CSN2, CSN1S1, CSN1S2, and LALBA) and immune cell types such as T cells, granulocytes (including neutrophils), macrophages, and B cells. Understanding the populations of hematopoietic cells in milk provides valuable insights into mammary gland function during lactation. The investigation of factors influencing cell number and activity in MEC is crucial for optimizing milk production and maintaining udder health. By identifying and addressing these factors, dairy farmers and researchers can implement strategies to enhance mammary gland function, improve milk production efficiency, and ensure the overall well-being of dairy cows. / Doctor of Philosophy / Milk production capacity in dairy cows relies on specialized cells in the mammary gland called secretory mammary epithelial cells (MEC). This study investigated how management practices, environmental factors, and individual cow factors affect the regulation of milk synthesis in Holstein cows. In the first study, we compared milking frequency in udder halves milked two times or four times per day during early and mid-lactation. The cows that were milked four times produced 2.27 kg/d of additional milk. This perhaps happened because the mammary gland's storage capacity increased with more frequent milking. Next, we studied the effects of short-term heat stress on the structure of the mammary gland tissue and the number and activity of MEC. Heat stress lowered milk production by 4.3 kg/d. We observed changes in the size and number of certain cells in the mammary gland, which likely affected the observed milk production findings. We also noticed differences in the activity of proteins related to protein production, cell survival, and the recycling of cell materials. In the final part of the study, we used single-cell characterization techniques to examine the different types of MEC and immune cells in milk. We found that there are various subgroups of MEC, as well as different types of immune cells such as T cells, neutrophils, macrophages, and B cells. Understanding the variety and abundance of these cell populations helps us learn more about how the mammary gland works during milk production. Studying the factors that influence the number and activity of MEC is essential for optimizing milk production in dairy cows. By identifying and addressing these factors, dairy farmers and researchers can develop strategies to enhance mammary gland function, improve milk production efficiency, and ensure the overall well-being of dairy cows.

Cell-to-cell variations and intracellular processes such as cytoskeletal organization and organelle dynamics exhibit massive heterogeneity. Advances in imaging and optics have enabled researchers to access spatiotemporal information in living cells efficiently. Even though current imaging technologies allow us to acquire an unprecedented amount of cell images, it is challenging to extract valuable information from the massive and complex dataset to interpret heterogeneous biological processes. Machine learning (ML), referring to a set of computational tools to acquire knowledge from data, provides promising solutions to meet this challenge. In this dissertation, we developed ML pipelines for deconvolution of subcellular protrusion heterogeneity from live cell imaging and molecular diagnostic from lens-free digital in-line holography (LDIH) imaging. Cell protrusion is driven by spatiotemporally fluctuating actin assembly processes and is morphodynamically heterogeneous at the subcellular level. Elucidating the underlying molecular dynamics associated with subcellular protrusion heterogeneity is crucial to understanding the biology of cellular movement. Traditional ensemble averaging methods without characterizing the heterogeneity could mask important activities. Therefore, we established an ACF (auto-correlation function) based time series clustering pipeline called HACKS (deconvolution of heterogeneous activities in coordination of cytoskeleton at the subcellular level) to identify distinct subcellular lamellipodial protrusion phenotypes with their underlying actin regulator dynamics from live cell imaging. Using our method, we discover “accelerating protrusion”, which is driven by the temporally ordered coordination of Arp2/3 and VASP activities. Furthermore, deriving the merits of ML, especially Deep Learning (DL) to learn features automatically, we advanced our pipeline to learn fine-grained temporal features by integrating the prior ML analysis results with bi-LSTM (bi-direction long-short term memory) autoencoders to dissect variable-length time series protrusion heterogeneity. By applying it to subcellular protrusion dynamics in pharmacologically and metabolically perturbed epithelial cells, we discovered fine differential response of protrusion dynamics specific to each perturbation. This provides an analytical framework for detailed and quantitative understanding of molecular mechanisms hidden in their heterogeneity. Lens-free digital in-line holography (LDIH) is a promising microscopic tool that overcomes several drawbacks (e.g., limited field of view) of traditional lens-based microscopy. Numerical reconstruction for hologram images from large-field-of-view LDIH is extremely time-consuming. Until now, there are no effective manual-design features to interpret the lateral and depth information from complex diffraction patterns in hologram images directly, which limits LDIH utility for point-of-care applications. Inherited from advantages of DL to learn generalized features automatically, we proposed a deep transfer learning (DTL)-based approach to process LDIH images without reconstruction in the context of cellular analysis. Specifically, using the raw holograms as input, the features extracted from a well-trained network were able to classify cell categories according to the number of cell-bounded microbeads, which performance was comparable with that of object images as input. Combined with the developed DTL approach, LDIH could be realized as a low-cost, portable tool for point-of-care diagnostics. In summary, this dissertation demonstrate that ML applied to cell imaging can successfully dissect subcellular heterogeneity and perform cell-based diagnosis. We expect that our study will be able to make significant contributions to data-driven cell biological research.

Cell-to-cell variations and intracellular processes such as cytoskeletal organization and organelle dynamics exhibit massive heterogeneity. Advances in imaging and optics have enabled researchers to access spatiotemporal information in living cells efficiently. Even though current imaging technologies allow us to acquire an unprecedented amount of cell images, it is challenging to extract valuable information from the massive and complex dataset to interpret heterogeneous biological processes. Machine learning (ML), referring to a set of computational tools to acquire knowledge from data, provides promising solutions to meet this challenge. In this dissertation, we developed ML pipelines for deconvolution of subcellular protrusion heterogeneity from live cell imaging and molecular diagnostic from lens-free digital in-line holography (LDIH) imaging. Cell protrusion is driven by spatiotemporally fluctuating actin assembly processes and is morphodynamically heterogeneous at the subcellular level. Elucidating the underlying molecular dynamics associated with subcellular protrusion heterogeneity is crucial to understanding the biology of cellular movement. Traditional ensemble averaging methods without characterizing the heterogeneity could mask important activities. Therefore, we established an ACF (auto-correlation function) based time series clustering pipeline called HACKS (deconvolution of heterogeneous activities in coordination of cytoskeleton at the subcellular level) to identify distinct subcellular lamellipodial protrusion phenotypes with their underlying actin regulator dynamics from live cell imaging. Using our method, we discover “accelerating protrusion”, which is driven by the temporally ordered coordination of Arp2/3 and VASP activities. Furthermore, deriving the merits of ML, especially Deep Learning (DL) to learn features automatically, we advanced our pipeline to learn fine-grained temporal features by integrating the prior ML analysis results with bi-LSTM (bi-direction long-short term memory) autoencoders to dissect variable-length time series protrusion heterogeneity. By applying it to subcellular protrusion dynamics in pharmacologically and metabolically perturbed epithelial cells, we discovered fine differential response of protrusion dynamics specific to each perturbation. This provides an analytical framework for detailed and quantitative understanding of molecular mechanisms hidden in their heterogeneity. Lens-free digital in-line holography (LDIH) is a promising microscopic tool that overcomes several drawbacks (e.g., limited field of view) of traditional lens-based microscopy. Numerical reconstruction for hologram images from large-field-of-view LDIH is extremely time-consuming. Until now, there are no effective manual-design features to interpret the lateral and depth information from complex diffraction patterns in hologram images directly, which limits LDIH utility for point-of-care applications. Inherited from advantages of DL to learn generalized features automatically, we proposed a deep transfer learning (DTL)-based approach to process LDIH images without reconstruction in the context of cellular analysis. Specifically, using the raw holograms as input, the features extracted from a well-trained network were able to classify cell categories according to the number of cell-bounded microbeads, which performance was comparable with that of object images as input. Combined with the developed DTL approach, LDIH could be realized as a low-cost, portable tool for point-of-care diagnostics. In summary, this dissertation demonstrate that ML applied to cell imaging can successfully dissect subcellular heterogeneity and perform cell-based diagnosis. We expect that our study will be able to make significant contributions to data-driven cell biological research.

Previous work from our laboratory in the mdx mouse model of Duchenne muscular dystrophy (DMD) demonstrated that a fraction of muscle stem cells (i.e., satellite cells) (MuSCs) progressively lose the expression of myogenic markers during the progression of the disease. In the process of characterizing this aberrant behaviour, we serendipitously discovered that MuSCs might be separated into two distinct subpopulations based on the expression of the GPI-anchored surface protein CD90. Crucially, this separation does not correlate with a divergence from the myogenic lineage; instead, it separates the pool of MuSCs into two subpopulations, both maintaining myogenic properties in healthy muscles. These two newly identified subpopulations do not overlap with any previously reported subpopulation and may be prospectively isolated; present a different response in terms of kinetics of activation and differentiation during the regenerative process induced by acute muscle damage; show a different propensity to enter in GAlert state upon distal injury; display dissimilar pAMPK activity and contain a different amount of mitochondria; are present in different proportions in distinct muscle groups; differentially express ECM encoding genes during quiescence. Moreover, one of the two subpopulations can give rise to the other and therefore appears to be upstream in the lineage hierarchy. Notably, loss of function experiments, in which CD90 was downregulated in MuSCs, diminish the difference in activation displayed by the two subpopulations. This demonstrates that CD90 is a molecular determinant of MuSCs functional diversification. Importantly, although the two subpopulations of MuSCs are numerically similar in healthy limb muscles, one of the two subpopulations is lost with time in dystrophic mdx mice. Based on these data, we are hypothesizing that an imbalance between the two newly identified subpopulations may impair regeneration in the dystrophic muscles. These observations not only increase our knowledge of the molecular and cellular dynamics that are controlling normal and pathological muscle homeostasis but also open the possibility that restoring the proper functional equilibrium between subpopulations of MuSCs may counteract the progression of the dystrophic disease.

Cells adhere to each other and to the extracellular matrix (ECM) through protein molecules on the surface of the cells. The breaking and forming of adhesive bonds, a process critical in cancer invasion and metastasis, can be influenced by the mutation of cancer cells. Several molecules have been reported to play a crucial role in cellular adhesion and proliferation, and eventually in cancer progression, with TGF-β being one of the most important. In this thesis, we propose a general framework to model cancer cells movement and aggregation, in response to nonlocal social interactions (that is, attraction towards neighbours that are far away, repulsion from those that are near by, and alignment with neighbours at intermediate distances), as well as other molecules' effect, e.g., TGF-β. We develop nonlocal mathematical models describing cancer invasion and metastasis as a result of integrin-controlled cell-cell adhesion and cell-matrix adhesion, for two cancer cell populations with different levels of mutation. The models consist of nonlinear partial differential equations, describing the dynamics of cancer cells and TGF-β dynamics, coupled with nonlinear ordinary differential equations describing the ECM and integrins dynamics. We study our models analytically and numerically, and we demostrate a wide range of spatiotemporal patterns. We investigate the effect of mutation and TGF-β concentration on the speed on cancer spread, as well as the effect of nonlocal interactions on cancer cells' speed and turning behaviour.