Overview
Connectomics — the systematic mapping of neural connections — has progressed from a single-organism, single-lab endeavor to a global, multi-petabyte field in roughly four decades. Understanding this trajectory helps practitioners appreciate why certain methods exist, what lessons have been learned (sometimes painfully), and where the field is headed.
Instructor script: a chronological tour
The pre-connectome era (1880s-1980s)
Santiago Ramón y Cajal (1852-1934) established the neuron doctrine: the brain is composed of discrete cells (neurons) that communicate at specialized junctions. Using the Golgi staining method — which, by a still-mysterious mechanism, labels only ~1% of neurons in a tissue sample — Cajal produced hundreds of exquisite drawings of neural circuits. These drawings were, in effect, the first wiring diagrams, though they captured morphology rather than verified synaptic connections.
Cajal’s key insight: neurons have a directional flow of information — dendrites receive, axons transmit. This “law of dynamic polarization” anticipated the directed graphs that modern connectomics produces.
Key limitation of the Cajal era: Golgi staining shows individual cell morphology beautifully but cannot reliably identify synaptic connections between specific neurons. You can see the tree, but not which branches actually touch.
Electron microscopy enters neuroscience (1950s-1960s): The development of transmission EM revealed synaptic ultrastructure for the first time. Gray (1959) classified synapses into Type I (asymmetric, excitatory) and Type II (symmetric, inhibitory). Palay and colleagues established the ultrastructural vocabulary — vesicles, clefts, postsynaptic densities — that connectomics annotators still use today.
The C. elegans connectome (1970s-1986)
The landmark: White JG, Southgate E, Thomson JN, Brenner S (1986). “The structure of the nervous system of the nematode Caenorhabditis elegans.” Philosophical Transactions of the Royal Society of London B 314:1-340.
This 340-page monograph reported the first complete connectome of any organism: 302 neurons, approximately 7,000 chemical synapses, and approximately 900 gap junctions (electrical synapses).
Why C. elegans was feasible:
- Fixed cell lineage: every individual has exactly the same 302 neurons, with the same cell names and positions (Sulston et al. 1983)
- Small size: the entire nervous system fits within a ~1 mm body
- Transparent body: enables correlative light microscopy
- Genetic tractability: Sydney Brenner chose C. elegans specifically as a model for relating genes to behavior (Brenner 1974)
Method: Manual serial-section TEM. Physical ultrathin sections were cut, collected on grids, imaged on a TEM, and traced by hand. The project took approximately 15 years. Much of the tracing was performed by Nichol Thomson, working through stacks of photographic prints.
Key findings from the original connectome:
- Connectivity is non-random: certain connection patterns (motifs) occur far more often than expected by chance
- Bilateral symmetry with individual variation: left-right neuron pairs have similar but not identical connectivity
- The nerve ring (circumpharyngeal ring) is a central processing hub containing ~180 neurons
- Sexually dimorphic circuits: male and hermaphrodite wiring differs in specific circuits (later elaborated by Cook et al. 2019)
Teaching point: “Even with 302 neurons, the connectome took 15 years and required multiple subsequent revisions. This sets expectations for the challenge of larger organisms.”
The term “connectome” is coined (2005)
Two groups independently introduced the word:
-
Sporns O, Tononi G, Kötter R (2005) “The human connectome: a structural description of the human brain.” PLoS Computational Biology 1(4):e42. — Defined three scales: macro-connectome (brain region connections, from MRI), meso-connectome (neuron population projections, from tract tracing), and micro/nano-connectome (individual synapses, from EM).
-
Hagmann P (2005) PhD thesis, EPFL. — Independently coined the same term in the context of diffusion MRI tractography.
The term catalyzed the field by providing a unified vocabulary and framing brain mapping as a systematic, completable project rather than an open-ended exploration.
The macro-connectome era: Human Connectome Project (2010-present)
The NIH-funded Human Connectome Project (Van Essen et al. 2013) mapped brain-wide connectivity in ~1,200 healthy adults using diffusion MRI and resting-state fMRI. This is a different enterprise from EM connectomics — it maps tract-level connections between brain regions (macroscale), not synapse-level connections between individual neurons (nanoscale).
Relevance to this course: The HCP demonstrated that large-scale, systematic brain mapping is scientifically productive and fundable. It created infrastructure, data standards, and analysis tools that influenced the EM connectomics community. But it cannot resolve individual synapses — for that, we need electron microscopy.
The volume EM revolution (2004-2015)
Serial block-face SEM (SBEM): Denk & Horstmann (2004) demonstrated automated volume EM: a diamond knife inside the SEM chamber cuts a thin layer, the exposed surface is imaged, and the process repeats. No manual section handling, no lost sections, automatic z-alignment.
Focused ion beam SEM (FIB-SEM): Knott et al. (2008) showed that a gallium ion beam could mill even finer layers (4-8 nm), achieving isotropic resolution for the first time.
ATUM (Automated Tape-collecting Ultramicrotome): Hayworth et al. (2014) enabled automated collection of thousands of ultrathin sections on tape, dramatically scaling serial-section TEM.
These technologies transformed connectomics from an artisanal craft to a scalable data-generation pipeline.
Retinal connectomics proves functional relevance (2011-2013)
Briggman KL, Helmstaedter M, Denk W (2011) “Wiring specificity in the direction-selectivity circuit of the retina.” Nature 471:183-188. — Combined SBEM with calcium imaging to show that direction-selective ganglion cells receive wiring that specifically implements their directional preference. This was a landmark because it connected structure (which neurons are connected) to function (direction selectivity) at the synaptic level.
Helmstaedter M et al. (2013) “Connectomic reconstruction of the inner plexiform layer in the mouse retina.” Nature 500:168-174. — Reconstructed ~1,000 neurons in the inner plexiform layer, revealing precise wiring rules between bipolar cell types and ganglion cell types.
Teaching point: “Retinal connectomics proved that EM wiring diagrams can answer functional questions, not just describe anatomy. This justified the enormous investment required for larger-scale projects.”
Drosophila connectomics timeline (2013-2024)
The fruit fly brain became the proving ground for scaling connectomics:
| Year | Milestone | Scale | Reference |
|---|---|---|---|
| 2013 | Medulla columns (7 columns) | ~800 neurons | Takemura et al. (2013) Nature |
| 2015 | Medulla expanded | ~1,500 neurons | Takemura et al. (2015) PNAS |
| 2018 | FAFB: complete adult brain EM volume | Imaging only (no full reconstruction) | Zheng et al. (2018) Cell |
| 2020 | Hemibrain connectome | ~25,000 neurons, ~20M synapses | Scheffer et al. (2020) eLife |
| 2023 | Larval brain connectome (complete) | ~3,016 neurons | Winding et al. (2023) Science |
| 2024 | FlyWire: whole adult brain connectome | ~139,255 neurons, ~54.5M synapses | Dorkenwald et al. (2024) Nature |
The progression: Column-level → hemisphere → complete adult brain. Each step required order-of-magnitude improvements in automated segmentation, proofreading infrastructure, and community coordination.
Mammalian cortex milestones (2015-2024)
| Year | Milestone | Scale | Reference |
|---|---|---|---|
| 2015 | Saturated reconstruction of mouse neocortex | ~1,500 μm³ (1,600 neurites) | Kasthuri et al. (2015) Cell |
| 2019 | Mouse cortex dense reconstruction | ~90,000 μm³ | Motta et al. (2019) Science |
| 2021 | MICrONS: mm³ mouse visual cortex + function | ~80,000 neurons, ~500M synapses | MICrONS Consortium (2021) bioRxiv |
| 2024 | H01: human cortex fragment | ~1 mm³, ~57,000 cells | Shapson-Coe et al. (2024) Science |
The BRAIN CONNECTS era (2023-present)
The NIH BRAIN Initiative CONNECTS program funds large-scale connectomics projects:
- MouseConnects/HI-MC: ~10 mm³ mouse hippocampus (Lichtman, Jain, and collaborators)
- Additional funded projects targeting other brain regions and species
This represents an institutional commitment to connectomics as infrastructure — not just individual lab projects but community resources.
Developmental connectomics (2021)
Witvliet et al. (2021) “Connectomes across development reveal principles of brain maturation.” Nature 596:257-261. — Mapped the C. elegans connectome at 8 developmental time points, from L1 larva to adult. Key finding: the overall architecture is established early, but significant rewiring occurs during development — some connections strengthen, others weaken, and new connections form.
Teaching point: “A connectome is a snapshot. Development reminds us that wiring is dynamic, even within an individual.”
Key lessons from the field’s history
- Completeness takes time: Even 302 neurons took 15 years (C. elegans). Whole fly brain took the community ~6 years from volume to connectome.
- Automation is necessary but insufficient: Every project required human proofreading after automated segmentation.
- Revisions are expected: The C. elegans connectome has been revised multiple times (Varshney 2011, Cook 2019). Published connectomes should be treated as living datasets.
- Structure informs but doesn’t determine function: Retinal connectomics showed structure-function links, but Bargmann & Marder (2013) caution that wiring alone doesn’t specify dynamics.
- Community infrastructure scales: FlyWire showed that 287 distributed proofreaders can collectively reconstruct an entire brain.
- The field is accelerating: The time from EM volume to published connectome is shrinking with each project.
Common misconceptions
| Misconception | Reality | Teaching note |
|---|---|---|
| “We’ve already mapped the brain” | We’ve mapped one worm, one fly, and small pieces of mouse and human cortex | Emphasize how much remains undone |
| “Cajal already drew the wiring diagram” | Cajal mapped morphology, not verified synaptic connections | Distinguish morphological inference from EM-verified connectivity |
| “The C. elegans connectome is finished” | It has been revised 3 times and still has gaps | Connectomes are living datasets |
| “Bigger volumes are always better” | The scientific question determines the required scale | Small, focused datasets can answer big questions |
References
- Bargmann CI, Marder E (2013) “From the connectome to brain function.” Nature Methods 10(6):483-490.
- Brenner S (1974) “The genetics of Caenorhabditis elegans.” Genetics 77(1):71-94.
- Briggman KL, Helmstaedter M, Denk W (2011) “Wiring specificity in the direction-selectivity circuit of the retina.” Nature 471:183-188.
- Cook SJ et al. (2019) “Whole-animal connectomes of both Caenorhabditis elegans sexes.” Nature 571:63-71.
- DeFelipe J (2010) “From the connectome to the synaptome: an epic love story.” Science 330(6008):1198-1201.
- Denk W, Horstmann H (2004) “Serial block-face scanning electron microscopy.” PLoS Biology 2(11):e329.
- Dorkenwald S et al. (2024) “Neuronal wiring diagram of an adult brain.” Nature 634:124-138.
- Hagmann P (2005) “From diffusion MRI to brain connectomics.” PhD thesis, EPFL.
- Hayworth KJ et al. (2014) “Ultrastructurally smooth thick partitioning and volume stitching for large-scale connectomics.” Nature Methods 12:319-322.
- Helmstaedter M et al. (2013) “Connectomic reconstruction of the inner plexiform layer in the mouse retina.” Nature 500:168-174.
- Kasthuri N et al. (2015) “Saturated reconstruction of a volume of neocortex.” Cell 162(3):648-661.
- Knott G et al. (2008) “Serial section scanning electron microscopy of adult brain tissue using focused ion beam milling.” Journal of Neuroscience 28(12):2959-2964.
- MICrONS Consortium (2021) “Functional connectomics spanning multiple areas of mouse visual cortex.” bioRxiv.
- Motta A et al. (2019) “Dense connectomic reconstruction in layer 4 of the somatosensory cortex.” Science 366(6469):eaay3134.
- Scheffer LK et al. (2020) “A connectome and analysis of the adult Drosophila central brain.” eLife 9:e57443.
- Shapson-Coe A et al. (2024) “A petavoxel fragment of human cerebral cortex.” Science 384(6696):eadk4858.
- Sporns O, Tononi G, Kötter R (2005) “The human connectome: a structural description of the human brain.” PLoS Computational Biology 1(4):e42.
- Van Essen DC et al. (2013) “The WU-Minn Human Connectome Project: an overview.” NeuroImage 80:62-79.
- Varshney LR et al. (2011) “Structural properties of the Caenorhabditis elegans neuronal network.” PLoS Computational Biology 7(2):e1001066.
- White JG et al. (1986) “The structure of the nervous system of C. elegans.” Philosophical Transactions of the Royal Society B 314:1-340.
- Winding M et al. (2023) “The connectome of an insect brain.” Science 379(6636):eadd9330.
- Witvliet D et al. (2021) “Connectomes across development reveal principles of brain maturation.” Nature 596:257-261.
- Zheng Z et al. (2018) “A complete electron microscopy volume of the brain of adult Drosophila melanogaster.” Cell 174(3):730-743.