Periodic Reporting for period 4 - EMcapsulins (Electron Microscopy gene reporters based on bioengineered encapsulin nanocompartments)
Berichtszeitraum: 2025-03-01 bis 2025-08-31
Despite substantial advances in genomics, transcriptomics, and proteomics, the molecular basis of most cellular dysfunctions remains poorly understood, limiting development of targeted molecular therapies. Genetically encoded fluorescent proteins have transformed molecular imaging through super-resolution techniques; however, mapping the nanoscale distribution of key molecular components—proteins, mRNA, and signaling processes—remains fundamentally challenging when cellular ultrastructure must be preserved and individual cells must be visualized within multicellular networks across millimeter-scale tissue volumes.
Electron microscopy is the most established technique for imaging such volumes with nanometer resolution when operated in "volume EM" mode with sequential mechanical or ion beam sectioning. EM provides dense labeling of cellular membranes, organelles, DNA, and ribosomes through heavy-metal stains, whereas fluorescence microscopy requires active inclusion of these anatomical references, severely limiting available spectral channels for molecular multiplexing.
Current connectomics initiatives mapping neuronal wiring diagrams capture only static structural snapshots that cannot resolve signaling molecules, proteomes, or transcriptomes—the molecular machinery governing synaptic plasticity that changes long before morphological correlates become visible. This missing information must be obtained through labor-intensive correlative measures involving fluorescence imaging, electrophysiology, and transcriptomics, presenting severe technical challenges in co-registering data across vastly different spatial scales.
Existing semi-genetic EM labeling methods—including immunogold labeling, enzymatic DAB polymerization via HRP or APEX2, and photooxidation via miniSOG—have critical limitations: epitope masking, limited antibody diffusion, spatial blurring of contrast that precludes geometric multiplexing, and dependence on exogenous compounds impeding high-throughput serial EM. Ferritin, while fully genetic, produces particles (~12 nm) too small for reliable detection at resolutions used for high-throughput volume EM.
Understanding the relationship between cellular ultrastructure and function contributes directly to mechanistic comprehension of cellular processes with implications for therapeutic strategies. In the brain, connectomics aims to map neuronal wiring diagrams to identify architectural motifs supporting specific computations. However, structural data alone lacks molecular identity information essential for interpreting circuit function. Beyond neuroscience, this methodology enables deciphering structure-function relationships in other cellular networks including the immune system, with applications in cellular therapies, organoid generation, tissue engineering, and synthetic biology.
The overarching goal of EMcapsulins is to accelerate morphometric analysis in biomedicine—quantitative analysis of cellular ultrastructure in relation to tissue function—by creating genetically controlled EM markers and reporters that geometrically encode cellular identity and activation history readable by electron microscopy.
Electron microscopy is the most established technique for imaging such volumes with nanometer resolution when operated in "volume EM" mode with sequential mechanical or ion beam sectioning. EM provides dense labeling of cellular membranes, organelles, DNA, and ribosomes through heavy-metal stains, whereas fluorescence microscopy requires active inclusion of these anatomical references, severely limiting available spectral channels for molecular multiplexing.
Current connectomics initiatives mapping neuronal wiring diagrams capture only static structural snapshots that cannot resolve signaling molecules, proteomes, or transcriptomes—the molecular machinery governing synaptic plasticity that changes long before morphological correlates become visible. This missing information must be obtained through labor-intensive correlative measures involving fluorescence imaging, electrophysiology, and transcriptomics, presenting severe technical challenges in co-registering data across vastly different spatial scales.
Existing semi-genetic EM labeling methods—including immunogold labeling, enzymatic DAB polymerization via HRP or APEX2, and photooxidation via miniSOG—have critical limitations: epitope masking, limited antibody diffusion, spatial blurring of contrast that precludes geometric multiplexing, and dependence on exogenous compounds impeding high-throughput serial EM. Ferritin, while fully genetic, produces particles (~12 nm) too small for reliable detection at resolutions used for high-throughput volume EM.
Understanding the relationship between cellular ultrastructure and function contributes directly to mechanistic comprehension of cellular processes with implications for therapeutic strategies. In the brain, connectomics aims to map neuronal wiring diagrams to identify architectural motifs supporting specific computations. However, structural data alone lacks molecular identity information essential for interpreting circuit function. Beyond neuroscience, this methodology enables deciphering structure-function relationships in other cellular networks including the immune system, with applications in cellular therapies, organoid generation, tissue engineering, and synthetic biology.
The overarching goal of EMcapsulins is to accelerate morphometric analysis in biomedicine—quantitative analysis of cellular ultrastructure in relation to tissue function—by creating genetically controlled EM markers and reporters that geometrically encode cellular identity and activation history readable by electron microscopy.
The EMcapsulins project successfully advanced multiplexed genetic reporters based on prokaryotic encapsulin nanocompartments, achieving substantial progress across all objectives. The foundational Nature Biotechnology publication (DOI: 10.1038/s41587-023-01713-y) demonstrated that six concentric barcodes can be automatically segmented and classified, with subcellular localization serving as an additional differentiator. Multiple barcodes were assembled into distinct patterns using rigid cross-linkers, establishing combinatorial space for geometric multiplexing comparable to fluorescent proteins for spectral multiplexing. In vivo validation demonstrated robust detection in Drosophila and mouse brains without toxicity, confirming translational feasibility.
The development of geometric multiplexing—encoding molecular information in nanoscopically discernible geometric features rather than spectral properties—represents a paradigm shift in reporter design. EMcapsulins produce robust, precisely confined, multiplexable EM contrast independently of synthetic compounds, representing a key advance over methods dependent on enzymatic or photoinduced polymerization resulting in unreliable epitope detection or spatially blurred precipitates.
Self-assembling protein nanocompartments functioning as nanoscopically readable barcode systems provide the first fully genetic multiplexing solution compatible with high-throughput serial EM, addressing a critical technological gap. The principle of multiplexing by nanoscopic shape has been adapted to correlative cryo-electron tomography, expansion microscopy, and X-ray nanotomography, demonstrating generalizability across imaging platforms.
A Nature Protocols manuscript (DOI: 10.1038/s41596-025-01260-7) established standardized methods for correlative fluorescence and EM workflows. This work introduced fluorescent variants capable of labeling target proteins, with successful labeling of endogenous connexins forming electrical gap junctions—demonstrating targeting to native structures without genetic modification of targets and fulfilling objectives for modular shell functionalization.
Results were disseminated through Nature Biotechnology and Nature Protocols publications, with open-source constructs deposited at Addgene. Additional publications describing automated multiplexed detection in mouse brain via SEM and FIB-SEM, and responsive variants, are in preparation. Patent application WO2023156638A1 protects core constructs and design principles. The research program is embedded within extensive local, national, and international collaborations.
The development of geometric multiplexing—encoding molecular information in nanoscopically discernible geometric features rather than spectral properties—represents a paradigm shift in reporter design. EMcapsulins produce robust, precisely confined, multiplexable EM contrast independently of synthetic compounds, representing a key advance over methods dependent on enzymatic or photoinduced polymerization resulting in unreliable epitope detection or spatially blurred precipitates.
Self-assembling protein nanocompartments functioning as nanoscopically readable barcode systems provide the first fully genetic multiplexing solution compatible with high-throughput serial EM, addressing a critical technological gap. The principle of multiplexing by nanoscopic shape has been adapted to correlative cryo-electron tomography, expansion microscopy, and X-ray nanotomography, demonstrating generalizability across imaging platforms.
A Nature Protocols manuscript (DOI: 10.1038/s41596-025-01260-7) established standardized methods for correlative fluorescence and EM workflows. This work introduced fluorescent variants capable of labeling target proteins, with successful labeling of endogenous connexins forming electrical gap junctions—demonstrating targeting to native structures without genetic modification of targets and fulfilling objectives for modular shell functionalization.
Results were disseminated through Nature Biotechnology and Nature Protocols publications, with open-source constructs deposited at Addgene. Additional publications describing automated multiplexed detection in mouse brain via SEM and FIB-SEM, and responsive variants, are in preparation. Patent application WO2023156638A1 protects core constructs and design principles. The research program is embedded within extensive local, national, and international collaborations.
The suite of genetically encoded barcodes for electron microscopy and correlative light microscopy represents a substantial advance beyond current state of the art. EMcapsulins constitute the first fully genetic multiplexing solution providing robust, spatially confined EM contrast without dependence on exogenous compounds, antibody penetration, or enzymatic polymerization.
Robust detection in FIB-SEM with 4 nm isotropic resolution was demonstrated (publication in preparation), enabling automated serial EM analysis for high-throughput volumetric applications. Multiple variants were expressed in mouse brain for detection via ultrasectioning/SEM and FIB-SEM. EMcapsulins are reliably discriminated from synaptic vesicles based on distinct size, rigid shells producing concentric contrast edges, and spherical morphology.
More detailed data on mapping subcellular protein distribution using EMcapsulins with titratable avidity was obtained than originally planned, enabling precise control over labeling density.
Substantial progress was made on calcium-responsive constructs monitored via live fluorescence microscopy followed by ultrastructural analysis, with publication forthcoming. This demonstrates feasibility of engineering stimulus-dependent geometric changes, extending the platform to dynamic reporters. Activity-dependent EMcapsulins establish a new research direction: responsive nanostructures for geometric multiplexing. Unlike spectral reporters, geometric integrators encode temporal information in structural features persisting through fixation and readable at EM resolution.
Robust detection in FIB-SEM with 4 nm isotropic resolution was demonstrated (publication in preparation), enabling automated serial EM analysis for high-throughput volumetric applications. Multiple variants were expressed in mouse brain for detection via ultrasectioning/SEM and FIB-SEM. EMcapsulins are reliably discriminated from synaptic vesicles based on distinct size, rigid shells producing concentric contrast edges, and spherical morphology.
More detailed data on mapping subcellular protein distribution using EMcapsulins with titratable avidity was obtained than originally planned, enabling precise control over labeling density.
Substantial progress was made on calcium-responsive constructs monitored via live fluorescence microscopy followed by ultrastructural analysis, with publication forthcoming. This demonstrates feasibility of engineering stimulus-dependent geometric changes, extending the platform to dynamic reporters. Activity-dependent EMcapsulins establish a new research direction: responsive nanostructures for geometric multiplexing. Unlike spectral reporters, geometric integrators encode temporal information in structural features persisting through fixation and readable at EM resolution.