Mar 14, 2023 / Developmental Biology / Neuroscience

Octopod optics: Mapping out the octopus visual system

Josh Azevedo

Convergent evolution is when species independently evolve similar features. It also happens to be a good metaphor for life: we may wind up at the same destination, but the paths we travel to get there are rarely similar.

Take the octopus, for example. Their view of the world is similar to ours. Literally. It isn’t just because of their intelligence (they can solve complex puzzles and actually unscrew baby-proof jars) but because—like humans, but unlike most invertebrates—they have a remarkably complex camera-style eye. Over 500 million years of evolutionary divergence raise the question: what does the underlying circuitry of their visual system look like?

Digging deeper into that question requires two key pieces of information: the specific cell type(s) that make up the octopus’ optic lobe, and the organization of these cells in the context of the visual system. In a recent paper (1), Songco-Casey et al. sought to provide answers by combining single cell RNA sequencing (scRNA-seq) and fluorescence in situ hybridization (FISH) methods. With these tools, they provided the best map to date of the octopus’ optic lobe and gave us a glimpse into how the octopus sees the world.

Visualizing octopus vision

While the octopus’ camera-style eye is remarkably similar to ours, the underlying visual “hardware” is extremely different. The eyes sit immediately adjacent to the optic lobes, which comprise over 60% of their central brain. Moving inward, the optic lobe is then further organized into the outer granular layer (OGL), plexiform layer (PL), inner granular layer (IGL), and medulla (MED).

Assembling a parts list

“A driving goal of this project was to identify the ‘parts list’ of the octopus optic lobe and create an integrated model of cell-type organization within their visual system.” – Songco-Casey et al.

Form dictates function in biological systems, and the distinct organization of the optic lobe implies different cellular functions in each layer. To get at the underlying circuitry, researchers used 10x Genomics Chromium Single Cell 3’ Gene Expression technology to characterize the cell types that make up the optic lobe. 

Their initial analysis yielded 41 clusters of cells, 33 of which appeared to be neuronal. Using prior work that identified glutamate, dopamine, and acetylcholine as the primary neurotransmitters in the octopus visual system (2), the researchers then narrowed down their data to identify markers of neurotransmission.

Of the 33 neuronal clusters, 24 were resolved into 6 overarching cell types based on their complement of neurotransmitters. The two fewest cell types were unusual and were characterized by the expression of either orcokinin or octopamine expression (neuropeptides found primarily in invertebrates). The four major cell types were dopaminergic, glutamatergic, dopaminergic + glutamatergic, or cholinergic neurons, with each having a characteristic distribution within the various layers of the optic lobe.

While conserved markers were useful for identifying these 24 clusters of neuronal cells, the remaining 9—representing immature neurons—were characterized by a novel gene (obimac0019980). The group further segregated immature neurons into three subgroups with both unique patterns of expression and widespread expression in the medulla, consistent with migration of neurons into the optic lobe from the medulla as they mature.

Putting it all together

A complete map of the octopus visual system requires both a knowledge of the cellular composition and their anatomical distributions. So, using multiple iterations of FISH guided by scRNA-seq, the authors delineated the specific cell (sub)types that comprise each layer of the visual lobe.

It’s striking that, while neurons can be grouped into broad classifications based on their specific complement of neurotransmitters, the spatial organization tells a more nuanced story. For example, glutamatergic neurons are present in each layer of the visual lobe; however, OGL neurons are highly organized and positive for the axon-patterning hedgehog (hh) gene, while IGL neurons are labeled by the calcium channel gene cacng and distributed throughout the IGL. These findings emphasize the intricacy of neural circuitry and underscore the importance of understanding both function and place in complex systems.

Looking ahead

While the human and octopus eyes are visually similar at a gross anatomical scale, this work from Songco-Casey et al. highlights how the underlying machinery could not be more different and provides a road map for other researchers going forward. While the authors acknowledge several limitations to their work—particularly the fact that there is almost certainly additional cellular diversity—their accomplishments are impressive, and we appreciate them giving us a deeper glimpse into the way these fascinating creatures see the world.

In situ methods made this paper possible: but what could you do if you had the capability to generate in situ data on 100s-1,000s of genes simultaneously at subcellular resolution? Explore our new Xenium In Situ platform and take your first look into a bigger world.


  1. Songco-Casey J, et al. Cell types and molecular architecture of the Octopus bimaculoides visual system. Curr Biol 32(23): 5031–5044 (2022). doi: 10.1016/j.cub.2022.10.015
  2. Messenger JB. Neurotransmitters of cephalopods. Invertebrate Neuroscience 2: 95–114 (1996). doi: 10.1007/BF02214113