Advances in toponomics drug discovery: Imaging cycler microscopy correctly predicts a therapy method of amyotrophic lateral sclerosis

Abstract

An imaging cycler microscope (ICM) is a fully automated (epi)fluorescence microscope which overcomes the spectral resolution limit resulting in parameter- and dimension-unlimited fluorescence imaging. This enables the spatial resolution of large molecular systems with their emergent topological properties (toponome) in morphologically intact cells and tissues displaying thousands of multi protein assemblies at a time. The resulting combinatorial geometry of these systems has been shown to be key for in-vivo/in-situ detection of lead proteins controlling protein network topology and (dys)function: If lead proteins are blocked or downregulated the corresponding disease protein network disassembles. Here, correct therapeutic predictions are exemplified for ALS. ICM drug target studies have discovered an 18-dimensional cell surface molecular system in ALS-PBMC with a lead drug target protein, whose therapeutic downregulation is now reported to show statistically significant effect with stop of disease progression in one third of the ALS patients. Together, this clinical and the earlier experimental validations of the ICM approach indicate that ICM readily discovers in vivo robustness nodes of disease with lead proteins controlling them. Breaking in vivo robustness nodes using drugs against their lead proteins is likely to overcome current high drug attrition rates.

Keywords: ALS; MELC; drug discovery; functional superresolution; high content analysis; imaging cycler microscopy; topology; toponome.

Figure 1

Functional super resolution of large molecular networks. (a-d) Dermoepithelial junction in human tissue. Imaging cycler microscopy based discovery of molecular networks in situ . (a) Direct realtime protein profiling in 100-dimensional ICM data set using an algorithm based on the similarity mapping approach ,,. Each data point has a PCMD of 256¹⁰⁰. Note: sharp images at the junctional area discriminating between Lamina Fibroreticularis (LF), Lamina Densa (LD, green profile in d), Lamina Lucida (LL) and the basal ceratinocyte layer (BC, red profile in d), as known from transmission electron microscopy (c). (b) Same area as in (a), displaying traditional triple fluorescence imaging. (e) 3-dimensinal ICM imaging of distinct 32-component multi protein complexes on the cell surface of a blood T-lymphocyte . Multi protein complexes are composed of differential combination of 32 proteins/glycotopes listed in (f). (g) Examples are marked with asterisks (number 1 to 3) and detailed as CMPs with proteins present (1) or absent (0) together characterised as individual CMPs. Bars: 10 µm (a, b), 50 nm (c), 1 µm (e). A similar figure is featured in http://www.toposnomos.com/huto/tis.html. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 2

Discovery of disease specific 100-dimansional protein profiles simultaneously and in real time in morphology intact tissue sections at a PCMD of 256¹⁰⁰ per pixel, exemplified in human skin. (a) List of 100 co-mapped biomolecules and selected protein profiles (0, 1, 3, 32 – 35) specific for diseased (d, f) and normal skin (e, g). (b,c) Diseased (b) and normal skin (c) are highlighted by pseudo colouring as histological stain for morphological orientation. Note that, by moving the cursor over the pixels, the software directly recognises in realtime which protein profiles are specific for the diseased (d, f) or normal skin (e, g). For example, pixel protein profiles (PPP) with numbers 0 and 1 are specific for the normal skin (e), and PPP 3, as well as PPPs 32 – 35 (d, f, respectively) are specific for the diseased skin. Note: For many similar applications at real time see webpage of the human toponome (HUTO) project (www.huto.toposnomos.com). (h) Power law (Zipf’s law) substantiates highly organised protein systems, as seen in (d-g). If 49 molecules are co-mapped Zipf’s law applies (blue line), but does not apply, if <15 molecules are co-mapped (red and green lines). This is revealed by plotting the log-log-relationship of thousands of distinct protein assemblies in toponome data sets ,. Bar: 100 µm (b–g). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 3

ALS specific motif as predictive toponome biomarker visualized simultaneously in ALS PBMC (a) as directly compared with healthy control (b). The motif was revealed by fully automated ICM based co-mapping of 18 cell surface proteins (c) on isolated PBMC. The motif as a whole is composed of 200 distinct CMPs. The visual field of (a) shows cells, each of which displays one ALS-specific CMP out of the whole motif (different colours). The motif contains CD16 and CD45RA as lead proteins denoted (1 = lead protein, present in all CMPs) (d), while other proteins are either not associated (0 = anticolocated), or variably associated with the lead proteins (* = wild cards) (d, motif M1) –. Courtesy of HUTO Project. Bar: 100 μm. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]