It has been proved feasible to generate attenuation maps (mu-maps) from cardiac SPECT using deep learning. However, this assumed that the training and testing datasets were acquired using the same scanner, tracer, and protocol. We investigated a robust generation of CT-derived mu-maps from cardiac SPECT acquired by different scanners, tracers, and protocols from the training data. We first pre-trained a network using 120 studies injected with (99m)Tc-tetrofosmin acquired from a GE 850 SPECT/CT with 360-degree gantry rotation, which was then fine-tuned and tested using 80 studies injected with (99m)Tc-sestamibi acquired from a Philips BrightView SPECT/CT with 180-degree gantry rotation. The error between ground-truth and predicted mu-maps by transfer learning was 5.13 +/- 7.02%, as compared to 8.24 +/- 5.01% by direct transition without fine-tuning and 6.45 +/- 5.75% by limited-sample training. The error between ground-truth and reconstructed images with predicted mu-maps by transfer learning was 1.11 +/- 1.57%, as compared to 1.72 +/- 1.63% by direct transition and 1.68 +/- 1.21% by limited-sample training. It is feasible to apply a network pre-trained by a large amount of data from one scanner to data acquired by another scanner using different tracers and protocols, with proper transfer learning.

Upon binding to intestinal epithelial cells, enterohemorrhagic Escherichia coli (EHEC), enteropathogenic E. coli (EPEC), and Citrobacter rodentium trigger formation of actin pedestals beneath bound bacteria. Pedestal formation has been associated with enhanced colonization, and requires intimin, an adhesin that binds to the bacterial effector translocated intimin receptor (Tir), which is translocated to the host cell membrane and promotes bacterial adherence and pedestal formation. Intimin has been suggested to also promote cell adhesion by binding one or more host receptors, and allelic differences in intimin have been associated with differences in tissue and host specificity. We assessed the function of EHEC, EPEC, or C. rodentium intimin, or a set of intimin derivatives with varying Tir-binding abilities in animal models of infection. We found that EPEC and EHEC intimin were functionally indistinguishable during infection of gnotobiotic piglets by EHEC, and that EPEC, EHEC, and C. rodentium intimin were functionally indistinguishable during infection of C57BL/6 mice by C. rodentium. A derivative of EHEC intimin that bound Tir but did not promote robust pedestal formation on cultured cells was unable to promote C. rodentium colonization of conventional mice, indicating that the ability to trigger actin assembly, not simply to bind Tir, is required for intimin-mediated intestinal colonization. Interestingly, streptomycin pre-treatment of mice eliminated the requirement for Tir but not intimin during colonization, and intimin derivatives that were defective in Tir-binding still promoted colonization of these mice. These results indicate that EPEC, EHEC, and C. rodentium intimin are functionally interchangeable during infection of gnotobiotic piglets or conventional C57BL/6 mice, and that whereas the ability to trigger Tir-mediated pedestal formation is essential for colonization of conventional mice, intimin provides a Tir-independent activity during colonization of streptomycin pre-treated mice.

Attachment to host cells by enterohaemorrhagic Escherichia coli (EHEC) is associated with the formation of a highly organized cytoskeletal structure containing filamentous actin, termed an attaching and effacing (AE) lesion. Intimin, an outer membrane protein of EHEC, is required for the formation of AE lesions, as is Tir, a bacterial protein that is translocated into the host cell to function as a receptor for intimin. We established a yeast two-hybrid assay for intimin-Tir interaction and, after random mutagenesis, isolated 24 point mutants in intimin, which disrupted Tir recognition in this system. Analysis of 11 point mutants revealed a correlation between recognition of recombinant Tir and the ability to trigger AE lesions. Many of the mutations fell within a 50-residue region near the C-terminus of intimin. Alanine-scanning mutagenesis of this region revealed four residues (Ser890, Thr909, Asn916 and Asn927) that are critical for Tir recognition. Mapping the sequences of EHEC intimin and Tir onto the crystal structure of the intimin-Tir complex of enteropathogenic E. coli predicts that each of these four intimin residues lies at the intimin-Tir interface and contributes to a pocket that interacts with Ile298 of EHEC Tir. Thus, this genetic approach to intimin function both identified residues critical for Tir binding and demonstrated a correlation between the ability to bind Tir and the ability to trigger actin focusing.

Enterohemorrhagic E. coli (EHEC) has emerged as an important agent of diarrheal disease in the developed countries. Attachment to host cells, an essential step during intestinal colonization by EHEC, is associated with the formation of a highly organized cytoskeletal structure containing filamentous actin, termed attaching and effacing (A/E) lesion, directly beneath bound bacteria. The outer membrane protein, intimin, is required for the formation of this structure, as is Tir, a bacterial protein that is translocated into the host cell and thought to function as a receptor for intimin. In this thesis, we characterized A/E lesion formation by in vivo and in vitro-grown EHEC, aimed at testing whether bacterial adaptation to the mammalian host included up regulation of A/E lesion formation. Our results showed that actin signaling by EHEC was induced upon bacterial growth in vivo, and this induction was likely due to the up regulation of multiple activities by in vivo-grown EHEC. We also focused on the interaction between intimin and the host cell, an interaction that triggers actin condensation of A/E lesion formation. We evaluated the role of β1 integrins, one of the proposed receptors of intimin, in A/E lesion formation, and demonstrated that β1 integrins are not essential for intimin-mediated cell binding and actin condensation. To better understand intimin function, we mapped the functional domains of intimin, showed that the minimal cell binding domain of intimin correlates with the minimal Tir-binding domain. This minimal Tir-binding domain, when purified and coated on latex beads, was sufficient to trigger actin condensation on preinfected mammalian cells, suggesting that Tir-binding by intimin is critical in the final step of A/E lesion formation. To further demonstrate the significance of the interaction between intimin and Tir in A/E lesion formation, we developed a yeast two-hybrid system to identify intimin mutants diminished in Tir-binding, and then characterized those mutants for the ability to trigger actin condensation, the final step of A/E lesion formation. Finally, as a first step to study the downstream actin signaling pathway after Tir-binding, we mapped the domain of Tir involved in intimin-binding, and showed that the N-terminus and C-terminus of Tir are likely to be localized in the host cell cytoplasm, available to interact with downstream effectors in actin signaling.

Enterohaemorrhagic Escherichia coli (EHEC) has emerged as an important agent of diarrhoeal disease. Attachment to host cells, an essential step during intestinal colonization by EHEC, is associated with the formation of a highly organized cytoskeletal structure containing filamentous actin, termed an attaching and effacing (A/E) lesion, directly beneath bound bacteria. The outer membrane protein intimin is required for the formation of this structure, as is Tir, a bacterial protein that is translocated into the host cell and is thought to function as a receptor for intimin. To understand intimin function better, we fused EHEC intimin to a homologous protein, Yersinia pseudotuberculosis invasin, or to maltose-binding protein. The N-terminal 539 amino acids of intimin were sufficient to promote outer membrane localization of the C-terminus of invasin and, conversely, the N-terminal 489 amino acids of invasin were sufficient to promote the localization of the C-terminus of intimin. The C-terminal 181 residues of intimin were sufficient to bind mammalian cells that had been preinfected with an enteropathogenic E. coli strain that expresses Tir but not intimin. Binding of intimin derivatives to preinfected cells correlated with binding to recombinant Tir protein. Finally, the 181-residue minimal Tir-binding region of intimin, when purified and immobilized on latex beads, was sufficient to trigger A/E lesions on preinfected mammalian cells.

Intimin is a bacterial outer membrane protein required for intimate attachment of enterohemorrhagic and enteropathogenic Escherichia coli (EHEC and EPEC) to mammalian cells. beta1-chain integrins have been proposed as candidate receptors for intimin. We found that binding of mammalian cells to immobilized intimin was not detectable unless mammalian cells were preinfected with EPEC or EHEC. beta1-chain integrin antagonists or inactivation of the gene encoding the beta1-chain did not affect binding of preinfected mammalian cells to intimin or the actin condensation associated with the attachment of EPEC. The results indicate that beta1-chain integrins are not essential for intimin-mediated cell attachment or EPEC-mediated actin polymerization.