Evolution of cephalopod nervous systems

Giant brains have independently evolved twice on this planet, in vertebrates and in cephalopods (Figure 1A). Thus, the brains and nervous systems of cephalopods provide an important counterpoint to vertebrates in the search for generalities of brain organization and function. Their mere existence disproves various hypotheses proposed to explain the evolution of the mind and the human brain, such as cognition and large brains evolved only in long-lived animals with complex social systems and parental care, none of which is true of cephalopods. Therefore, it is worthwhile to review what is known about the evolution of cephalopod nervous systems to consider how it informs our understanding of general principles of brain evolution.

Figure 1. Cephalopod evolution.

(A) Evolutionary relationships of select animal clades reflecting estimated age. Schematics from phylopic.org, used with Creative Commons licenses. Chiton image used with permission by Noah Schlottman, adapted from a photo by Casey Dunn (CC BY-SA 3.0). (B) Extant cephalopod lineages, relationships based on Anderson and Lindgren, Mol. Phylogenet. Evol. (2021).

Cephalopods are marine molluscs that include octopuses, squid, and cuttlefish (Figure 1B). Many theories have arisen as to the adaptive significance of the intelligence and large brains in cephalopods including predator avoidance, loss of shell, and competition with fish. In addition to large brains, cephalopods have suites of innovations including multiple sucker-lined arms and rapid adaptive camouflage. They also have complex camera-type eyes that are similar to our own but evolved independently. Understanding the developmental and evolutionary events that allowed cephalopods to diverge from their slow and frankly less intelligent cousins — snails, clams, and chitons — can provide insight into how intelligence emerges from neural systems.

Phylogeny

Centralized nervous systems are present in each of the three major clades of bilaterally symmetric animals (bilaterians): the deuterostomes, which include the vertebrates, ecdysozoans, to which insects and spiders belong, and spiralians (also called lophotrochozoans), which includes molluscs and annelids (Figure 1A). Each of these major clades also contain many lineages that lack overt nervous system specialization. It is likely that complex nervous systems (i.e., brains) arose independently at least three times.

Even within extant molluscan diversity, there is a dramatic range of body forms and nervous system complexity (Figure 2): the conchiferans (comprising the monoplacophorans, cephalopods, gastropods, bivalves, and scaphopods) typically have a single shell field, while the aculiferans (chitons and aplacophorans) have multiple shell fields or lack shells entirely. It has been difficult to resolve the deep relationship of cephalopods to other molluscs. Although earlier phylogenies often placed cephalopods as a sister group to gastropods, modern phylogenomic analyses suggest that they branch near the base of the conchiferans, either as a sister group to the enigmatic, deep sea monoplacophorans or as a sister group to the clade containing gastropods, bivalves, and scaphopods. This close placement with the monoplacophorans may support some cephalopod novelties that appear to be elaborations of the ancestral condition, including an early, cord-like appearance of the nervous system that shows similarities to the cord-based chiton nervous system, which is absent in gastropods and bivalves (Figure 2).

Figure 2. Molluscan neuroanatomy.

(A) Schematic anterior view of coleoid anatomy highlighting the nervous system. Ventral is to the left and dorsal to the right. The central brain sits between the eyes and is divided into supraesophageal (red) and subesophageal (purple) masses, and is flanked by the optic lobes (green). The axial nerve cords (ANC, blue) run down the center of each of the sucker-lined arms. The mantle contains visceral organs including two gills, each connected to a branchial heart, a central systemic heart, the hepatopancreas (HP), and posterior salivary glands (PSG). (B) Schematic of an off-axis sagittal section through the central brain of Octopus bimaculoides, illustrating the distribution of lobes in the supraesophageal (red) and subesophageal (blue) masses. The subesophageal mass is divided into anterior, medial, and posterior subesophageal masses (ASM, MSM, PSM, respectively), while the supraesophageal mass contains the vertical lobe (VL), anterior and dorsal basal lobes (aBL, dBL), the frontal lobes (FL), and the buccal lobe (BL). (C) Schematic of generalized gastropod nervous system, where ganglia are connected to each other via commissures and connectives (grey). Illustrated here are the cerebral ganglion (CG, red), pleural ganglion (PlG, purple), pedal ganglion (PeG, blue), buccal ganglion (B, orange), and visceral ganglion (V, royal blue). (D) Schematic of generalized polyplacophoran nervous system, with cerebral cord (CC), lateral cord (LatC), and ventral cord (VC).

There are only two clades of extant cephalopods: the Nautilidae and the Coleoidea (Figure 1B). Nautilidae consists of just two living genera: Allonautilus and Nautilus. Many coleoid innovations are absent in the nautiloids, which have a true external shell, as well as a pinhole eye, and dozens of digital tentacles that lack suckers. The earliest known cephalopods date from the Cambrian, and like nautiloids have an external, chambered shell. Nautiloids are found only in the deep waters of the Indo-Pacific Ocean. In contrast, coleoid cephalopods are a diverse group found throughout the world’s oceans. Although the shell has been completely lost in some octopods, the shell gland gives rise to the internal pen of gladius squids, as well as the cuttlebone of cuttlefish.

The extant cephalopods represent only a small fraction of the diversity of extinct cephalopods. The earliest cephalopods arose over 490 million years ago and diversified into three major clades: Orthoceratoidea, Multiceratoidea, and Endoceratodidea. It is agreed that coleoids originated from Orthoceratoidea, but it is not certain whether this is also true for nautiloids. Although molecular data place nautiloids in the Orthoceratoidea clade, fossil evidence suggests that they might belong in Multiceratoidea, which would suggest coleoids and nautiloids diverged early in the evolution of cephalopods, close to 490 million years ago, just 50 million years after the Cambrian explosion. However, recent evidence places the earliest cephalopod fossil 30 million years later. Regardless of the precise date, it is clear that the shared features of the brains of nautilus and coleoids existed 100 million years before the first tetrapods and 230 million years before the first dinosaurs. Moreover, fossils of the first coleoids in the early Carboniferous period (about 350 million years ago) show evidence of a substantial cephalic region between two large eyes.

There are several clades within the coleoids (Figure 1B). The major division of coleoid cephalopods is into the eight-armed octopodiforms (octopuses and vampire squids) and the ten-armed decapodiforms (squid and cuttlefish). The relationships within these clades remain controversial. Many relationships at lower taxonomic levels also are poorly resolved. This is due, in part, to the fact that these soft-bodied, color changing, shape-shifting animals can be difficult to study at morphological levels. More sequencing data are likely to help resolve some of these questions.

Neuroanatomy

Coleoid cephalopod nervous systems rival those of vertebrates in size but have an entirely different organization. The brain sits between the eyes, with two optic lobes flanking the central brain, which is divided into supra- and sub-esophageal masses (Figure 2A,B). The cephalopod supraesophageal, buccal, posterior and anterior subesophageal masses may correspond to the cerebral, buccal, palliovisceral, and pedal ganglia, respectively, in non-cephalopod conchiferans like gastropods, bivalves, and scaphopods.

The large supra- and sub-esophageal masses are partitioned into smaller lobes, each with different functions. The neurons in the lobes are generally arranged in the ‘typical’ invertebrate fashion, with a rind of cell bodies surrounding a central neuropil. Nautiloids also have supra- and sub-esophageal masses divided into 13 lobes. Coleoid supraesophageal and subesophageal masses have 30–40 individual lobes, which suggests that many of these neural elaborations are coleoid, or lineage-specific, innovations (Figure 3). Some of these lobes are found across coleoid cephalopods (e.g. vertical lobe, superior frontal lobe, basal lobes), whereas others are restricted to specific groups.

Figure 3. Schematic of cephalopod brains demonstrates common ground plan and elaborations in coleoids.

Dorsal views of Nautilus brain (top), octopus brain (middle), and Sepia brain (bottom). (Top) In Nautilus, the supraesophageal cerebral cord (CC) and subesophageal palliovisceral cord (PVC) and pedal cord (PC) are flanked by two optic lobes (OL). The buccal ganglia (BuG) are situated outside of the central brain. (Based on Young (1965) and Ponder et al. (2019).) (Middle) The octopus brain has elaborations in the supraesophageal brain, colored in shades of red, including the vertical lobe (VL), above the buccal lobe (BuL), which is merged into the central brain below the anterior subesophageal mass (ASM). The posterior subesophageal mass (PSM) is at the top of the image. Based on O. bimaculoides. (Bottom) Similar to the octopus brain, the Sepia brain has large optic lobes and a large vertical lobe. In these animals, the buccal ganglion sits outside of the central brain, closer to the buccal mass. (Based on 3D rendering of Sepia bandensis on cuttlebase.org.)

Functional studies provided early indications that the lobes have distinct roles in mediating cephalopod behavior. Within the supraesophageal brain, the vertical lobe system and the frontal lobe system play roles in visual and tactile learning and memory, respectively, whereas the basal lobes play roles in higher motor control. The subesophageal mass contains lower motor centers controlling the arms, eyes, mantle, and funnel, while the optic lobes play a role in visual processing.

The brain of Octopus vulgaris is estimated to comprise approximately 500 million neurons — six times the number in the mouse brain. Of these, approximately 200 million are located in the optic lobes and the central supraesophageal and subesophageal masses. The remaining 300 million neurons are distributed throughout the axial nerve cords in the arms, which lie at the center of each arm. The neurons in the axial nerve cords are both within and between ganglia. Each ganglion is associated with a sucker. It is important to note that the brain and the arms continue to add neurons as the octopus grows, increasing thousands fold as an octopus grows from less than a milligram at hatching with just a few suckers on each arm to almost two kilograms with hundreds of suckers by eight months of age.

The axial nerve cords also contain ascending and descending axonal projections conveying information between the periphery and the central brain. The axial nerve cords are flanked by smaller intramuscular nerve cords that provide proprioceptive information and coordination between the arms. Interbrachial connectives carry axons of these coordinating neurons between the arms.

Neurogenomics

Cephalopod genomes have yielded insights into the molecular basis of cephalopod brain evolution and development. At first glance, cephalopod genomes largely resemble those found in other sequenced invertebrates: they possess most pan-bilaterian genes in single copy, suggesting that cephalopod evolution did not undergo a parallel phase of whole genome duplication, as happened in vertebrates. However, cephalopod genomes encode massive expansions of a handful of gene families, including the protocadherins and the C2H2 zinc finger transcription factors.

The protocadherins are a family of cell adhesion molecules that play a key role in the development of neuronal wiring in vertebrate brains. Vertebrate genomes typically encode dozens of protocadherins (50–70 in mammals, 100+ in teleosts). Experimentally decreasing protocadherin diversity in the mouse brain causes defects in neuronal wiring. In contrast to ecdysozoan genomes (including Drosophila melanogaster and Caenorhabditis elegans), which lack any protocadherins, cephalopod genomes encode even more protocadherins than any vertebrate genome sequenced to date; estimates range from 100 to over 300. Moreover, small clusters of protocadherin genes have now been identified in other molluscan genomes. The vast majority of octopus and squid protocadherins are encoded on a single chromosome in a 40–50 Mb supercluster made up of tight subclusters of densely packed protocadherin genes, all oriented in the same transcriptional direction. The intercluster space encodes still more protocadherins that are more evenly distributed. Transcriptome sequencing in both squid and octopus indicates that many of these genes are expressed in the nervous system, suggesting a molecular parallel between the brains of coleoid cephalopods and vertebrates.

The repertoire of C2H2 zinc finger transcription factors has also been massively expanded in cephalopod genomes. This family is typically the most abundant transcription factor in many animal genomes: most invertebrate genomes encode 200–400 genes with C2H2 zinc finger motifs. The human genome has more than 700 of these genes, thanks in part to an expansion of the KRAB-motif-containing genes. In contrast, thousands of C2H2-domain-containing genes have been identified in cephalopod genomes, often in clustered arrays As with the protocadherins, many of these C2H2-domain-containing genes are expressed in the nervous system, though what role they play in cephalopod brain evolution and development has yet to be explored.

In addition to expansions of genes found in other animals, cephalopod genomes also encode suites of taxonomically restricted novel genes, many of which are also found in clustered expansions in the genome. Many of these gene families are associated with cephalopod novelties, such as the suckerins, which are a secreted structural protein found in the sucker ring teeth in squid, or the reflectins, which are associated with the structural color produced in the iridocytes in the skin of coleoids.

Cephalopod genomes also demonstrate dramatic rearrangements: many genes that tend to be found together on animal chromosomes, called ancestral linkage groups, are disrupted in cephalopod genomes. These rearrangements present a unique system to study the evolution of gene regulation in nove environments and may have provided the substrate underlying the evolution of cephalopod brain and body innovations.

Nervous system development

Cephalopod development is itself an evolutionary novelty. Cephalopod embryos are large and yolky, and early cleavages are bilaterally symmetric and superficial, resembling the early development of teleost fish, rather than the spiral cleavage shared by other molluscs with members of Spiralia, named for this cleavage pattern. Moreover, cephalopod embryos develop directly, without undergoing intermediate larval stages, which is another exception to the widely retained developmental program found across other molluscs and spiralians. These departures from the ancestral state are further underappreciated examples of convergence with teleost fish.

Although many events in early cephalopod embryogenesis remain undescribed using modern molecular techniques, nervous system development has begun to be characterized in several species spanning cephalopod diversity, including Nautilus, octopuses, squid, and cuttlefish. In each case, it has been found that neural progenitors and neurons appear early, as the body plan is set up, and are widespread throughout the embryonic territory — indeed, much of the early octopus embryo appears to be neurectodermal (Figure 4). These neuronal progenitors then migrate from the neurectoderm into the developing brain, giving rise to the nascent supraesophageal and subesophageal masses, optic lobes, and axial nerve cords. This widespread neurogenic territory contrasts both with neurodevelopment in other molluscan embryos, which is restricted to discrete territories, and with vertebrate neurodevelopment, which restricts central nervous system neurogenesis to the neural tube — a structure that inverts from the neurectoderm. Recent studies are beginning to characterize how this widespread neurogenic domain eventually produces cephalopod brains. Whether cephalopod brain development requires similar or novel mechanisms to brain elaboration in vertebrates remains an area of open investigation.

Figure 4. Neurogenesis is widespread in cephalopod embryos.

Neurogenic territories (marked in blue) in coleoid (octopus), molluscan trochophore, Drosophila, and zebrafish (Danio rerio) embryos. a, arm; AO, apical organ; BR, brain; CC, cerebral cord; OC, optic cord; PVC, palliovisceral cord; PC, pedal cord; SC, spinal cord; VNC, ventral nerve cord.

It remains to be determined whether the molecular mechanisms for defining cephalopod brain regions are similar to those in other animals. Similar suites of genes have been described as being deployed in the development of animal brains. Although a focus on select model species with highly centralized nervous systems indicated a common molecular signature, recent studies in a wider diversity of animals find these ‘brain’-deployed genes in similar patterns along the ectoderm rather than the nervous system. Developmental studies in a variety of cephalopods demonstrate the expression of developmental transcription factors and signaling ligands in the developing brain and eye. Studies find both similar and novel molecular signatures in the development of these structures to what has been described in other animals. Given the broad distribution of the neurectoderm in cephalopod embryos, additional approaches (e.g. functional studies, lineage tracing, or co-expression) may be required to establish the roles of these core developmental genes in patterning the brain and body. Recent technological advancements in testing gene function (e.g. CRISPR-mediated gene knockouts) in cephalopods will surely open new avenues for studying these developmental mechanisms and understanding their contribution to the evolution of cephalopod nervous systems.

Conclusions

There is still much to be learned about cephalopod nervous system evolution. Beyond the insights gained into cephalopod neurobiology, understanding the evolutionary relationship between cephalopod brain regions to those of vertebrates, insects, annelids, and other animals provides a direct test of important hypotheses on brain evolution. Alternative hypotheses such as whether all animal central nervous systems are organized on the same basic plan or whether there are alternative ways to develop and organize a nervous system need to be tested by examining the most distantly related giant brain. Thanks to the development of new molecular tools for the study of non-traditional organisms, including for cephalopods, we are now in a place to address these longstanding questions.