We presenta robust, long‐range optical autofocus system for microscopy utilizing machine learning. This can be useful for experiments with long image data acquisition times that may be impacted by defocusing resulting from drift of components, for example due to changes in temperature or mechanical drift. It is also useful for automated slide scanning or multiwell plate imaging where the sample(s) to be imaged may not be in the same horizontal plane throughout the image data acquisition. To address the impact of (thermal or mechanical) fluctuations over time in the optical autofocus system itself, we utilize a convolutional neural network (CNN) that is trained over multiple days to account for such fluctuations. To address the trade‐off between axial precision and range of the autofocus, we implement orthogonal optical readouts with separate CNN training data, thereby achieving an accuracy well within the 600 nm depth of field of our 1.3 numerical aperture objective lens over a defocus range of up to approximately +/–100 μm. We characterize the performance of this autofocus system and demonstrate its application to automated multiwell plate single molecule localization microscopy.

In tuberculosis (TB), the innate inflammatory immune response drives tissue destruction, morbidity, and mortality. Monocytes secrete matrix metalloproteinases (MMPs), which have key roles in local tissue destruction and cavitation. We hypothesized that integrin signaling might regulate monocyte MMP secretion in pulmonary TB during cell adhesion to the extracellular matrix (ECM). Adhesion to type I collagen and fibronectin by Mycobacterium tuberculosis–stimulated monocytes increased MMP-1 gene expression by 2.6-fold and 4.3-fold respectively, and secretion by 60% (from 1208.1 ± 186 to 1934.4 ± 135 pg/ml; p < 0.0001) and 63% (1970.3 ± 95 pg/ml; p < 0.001). MMP-10 secretion increased by 90% with binding to type I collagen and 55% with fibronectin, whereas MMP-7 increased 57% with collagen. The ECM did not affect the secretion of tissue inhibitors of metalloproteinases-1 or -2. Integrin αVβ3 surface expression was specifically upregulated in stimulated monocytes and was further increased after adhesion to type I collagen. Binding of either β3 or αV integrin subunits increased MMP-1/10 secretion in M. tuberculosis–stimulated monocytes. In a cohort of TB patients, significantly increased integrin β3 mRNA accumulation in induced sputum was detected, to our knowledge, for the first time, compared with control subjects (p < 0.05). Integrin αVβ3 colocalized with areas of increased and functionally active MMP-1 on infected monocytes, and αVβ3 blockade markedly decreased type I collagen breakdown, and impaired both monocyte adhesion and leukocyte migration in a transwell system (p < 0.0001). In summary, our data demonstrate that M. tuberculosis stimulation upregulates integrin αVβ3 expression on monocytes, which upregulates secretion of MMP-1 and -10 on adhesion to the ECM. This leads to increased monocyte recruitment and collagenase activity, which will drive inflammatory tissue damage.

The cause of chronic inflammation in ulcerative colitis (UC) is incompletely understood. Here we tested the hypothesis that an excessive acute inflammatory response to bacteria contributes to the pathogenesis. Acute inflammatory responses were provoked in vivo in UC patients and healthy controls by intradermal inoculation with bacteria. Vascular responses were quantified by laser Doppler. Inflammatory exudates were recovered in superimposed suction blisters and cells measured by polychromatic flow cytometry, cytokines by multiplex array, and inflammatory lipids by mass spectrometry. Vascular responses in UC patients were heightened at 24h after bacterial injection (p=0·03), and remained abnormally high at 48h (p=0·0005) and this amplified response was seen in UC with Gram-positive as well as Gram-negative organisms (p=0·01). The cellular infiltrate over the injection site, composed largely of neutrophils at 4 hours a was greater in UC (p=0·002). At 48h, the increased numbers of cells in UC were composed of neutrophils (p=0·001) and CD4 lymphocytes (p=0·001). The exaggerated inflammation in UC was not a cytokine-driven phenomenon. Exaggerated onset was normalised in patients taking 5aminosalicylates, accompanied by increased concentrations of hydroxy fatty acids 9-oxooctadecadienoic acid (OxoODE; p=0·05) and 13-OxoODE (p=0·01) in resolving exudates. In vitro, these compounds suppressed macrophage inflammatory cytokine secretion through PPARγ (p<0·0001). Conversely, 5-aminosalicylates did not inhibit early inflammatory reactions in control participants. Acute inflammatory responses to bacteria in UC are both overly exuberant and slow to resolve. Neutrophils accumulate in excess and persist, in keeping with the pathological appearances of disease flares. These studies also provide new insight into the mechanism of 5-aminosalicylate (5ASA) drugs, which act as pro-resolution rather than indiscriminate anti-inflammatory agents by promoting formation of immunomodulatory hydroxy 3 lipids. While production of these lipids is not defective as part of the underlying disease process, this identifies a novel mechanism of drug action harnessing pro-resolution pathways. SummaryWysoczanski and colleagues demonstrate that the inflammatory response to injected bacteria is exaggerated and prolonged in ulcerative colitis. This disordered inflammation appears to be associated with increased secretion of PGE2. 5-aminosalicylate drugs, which are used to treat this condition, normalize inflammation and PGE2 secretion, and appear to work through PPARγ 4

No abstract