Extracellular expression of heat shock protein 90 (eHsp90) by tumor cells is correlated with malignancy. Development of small molecule probes that can detect eHsp90 in vivo may therefore have utility in the early detection of malignancy. We synthesized a cell impermeable far-red fluorophore-tagged Hsp90 inhibitor to target eHsp90 in vivo. High resolution confocal and lattice light sheet microscopy show that probe-bound eHsp90 accumulates in punctate structures on the plasma membrane of breast tumor cells and is actively internalized. The extent of internalization correlates with tumor cell aggressiveness, and this process can be induced in benign cells by over-expressing p110HER2. Whole body cryoslicing, imaging and histology of flank and spontaneous tumor-bearing mice strongly suggests that eHsp90 expression and internalization is a phenomenon unique to tumor cells in vivo and may provide an 'Achilles heel' for the early diagnosis of metastatic disease and targeted drug delivery.

Fluorescent proteins (FPs) that can be optically highlighted enable PALM super-resolution microscopy and pulse-chase experiments of cellular molecules. Most FPs evolved in cytoplasmic environments either in the original source organism or in the cytoplasm of bacteria during the course of optimization for research applications. Consequently, many FPs may fold incorrectly in the chemically distinct environments in subcellular organelles. Here, we describe the first monomeric photoswitchable (from green to bright red) FP adapted for oxidizing environments.

Electrical recordings from a large array of electrodes give us access to neural population activity with single-cell, single-spike resolution. These recordings contain extracellular spikes which must be correctly detected and assigned to individual neurons. Despite numerous spike-sorting techniques developed in the past, a lack of high-quality ground-truth datasets hinders the validation of spike-sorting approaches. Furthermore, existing approaches requiring manual corrections are not scalable for hours of recordings exceeding 100 channels. To address these issues, we built a comprehensive spike-sorting pipeline that performs reliably under noise and probe drift by incorporating a channel-covariance feature and a clustering based on fast density-peak finding. We validated performance of our workflow using multiple ground-truth datasets that recently became available. Our software scales linearly and processes a 1000-channel recording in real-time using a single workstation. Accurate, real-time spike sorting from large recording arrays will enable more precise control of closed-loop feedback experiments and brain-computer interfaces.