Ion regulation at gills precedes gas exchange and the origin of vertebrates.

Gas exchange and ion regulation at gills have key roles in the evolution of vertebrates1-4. Gills are hypothesized to have first acquired these important homeostatic functions from the skin in stem vertebrates, facilitating the evolution of larger, more-active modes of life2,3,5. However, this hypothesis lacks functional support in relevant taxa. Here we characterize the function of gills and skin in a vertebrate (lamprey ammocoete; Entosphenus tridentatus), a cephalochordate (amphioxus; Branchiostoma floridae) and a hemichordate (acorn worm; Saccoglossus kowalevskii) with the presumed burrowing, filter-feeding traits of vertebrate ancestors6-9. We provide functional support for a vertebrate origin of gas exchange at the gills with increasing body size and activity, as direct measurements in vivo reveal that gills are the dominant site of gas exchange only in ammocoetes, and only with increasing body size or challenges to oxygen supply and demand. Conversely, gills of all three taxa are implicated in ion regulation. Ammocoete gills are responsible for all ion flux at all body sizes, whereas molecular markers for ion regulation are higher in the gills than in the skin of amphioxus and acorn worms. This suggests that ion regulation at gills has an earlier origin than gas exchange that is unrelated to vertebrate size and activity-perhaps at the very inception of pharyngeal pores in stem deuterostomes.

A pre-vertebrate endodermal origin of calcitonin-producing neuroendocrine cells.

Vertebrate calcitonin-producing cells (C-cells) are neuroendocrine cells that secrete the small peptide hormone calcitonin in response to elevated blood calcium levels. Whereas mouse C-cells reside within the thyroid gland and derive from pharyngeal endoderm, avian C-cells are located within ultimobranchial glands and have been reported to derive from the neural crest. We use a comparative cell lineage tracing approach in a range of vertebrate model systems to resolve the ancestral embryonic origin of vertebrate C-cells. We find, contrary to previous studies, that chick C-cells derive from pharyngeal endoderm, with neural crest-derived cells instead contributing to connective tissue intimately associated with C-cells in the ultimobranchial gland. This endodermal origin of C-cells is conserved in a ray-finned bony fish (zebrafish) and a cartilaginous fish (the little skate, Leucoraja erinacea). Furthermore, we discover putative C-cell homologs within the endodermally-derived pharyngeal epithelium of the ascidian Ciona intestinalis and the amphioxus Branchiostoma lanceolatum, two invertebrate chordates that lack neural crest cells. Our findings point to a conserved endodermal origin of C-cells across vertebrates and to a pre-vertebrate origin of this cell type along the chordate stem.

Fgf signalling is required for gill slit formation in the skate, Leucoraja erinacea.

The gill slits of fishes develop from an iterative series of pharyngeal endodermal pouches that contact and fuse with surface ectoderm on either side of the embryonic head. We find in the skate (Leucoraja erinacea) that all gill slits form via a stereotypical sequence of epithelial interactions: 1) endodermal pouches approach overlying surface ectoderm, with 2) focal degradation of ectodermal basement membranes preceding endoderm-ectoderm contact; 3) endodermal pouches contact and intercalate with overlying surface ectoderm, and finally 4) perforation of a gill slit occurs by epithelial remodelling, without programmed cell death, at the site of endoderm-ectoderm intercalation. Skate embryos express Fgf8 and Fgf3 within developing pharyngeal epithelia during gill slit formation. When we inhibit Fgf signalling by treating skate embryos with the Fgf receptor inhibitor SU5402 we find that endodermal pouch formation, basement membrane degradation and endodermal-ectodermal intercalation are unaffected, but that epithelial remodelling and gill slit perforation fail to occur. These findings point to a role for Fgf signalling in epithelial remodelling during gill slit formation in the skate and, more broadly, to an ancestral role for Fgf signalling during pharyngeal pouch epithelial morphogenesis in vertebrate embryos.

The pseudobranch of jawed vertebrates is a mandibular arch-derived gill.

The pseudobranch is a gill-like epithelial elaboration that sits behind the jaw of most fishes. This structure was classically regarded as a vestige of the ancestral gill arch-like condition of the gnathostome jaw. However, more recently, hypotheses of jaw evolution by transformation of a gill arch have been challenged, and the pseudobranch has alternatively been considered a specialised derivative of the second (hyoid) pharyngeal arch. Here, we demonstrate in the skate (Leucoraja erinacea) that the pseudobranch does, in fact, derive from the mandibular arch, and that it shares gene expression features and cell types with gills. We also show that the skate mandibular arch pseudobranch is supported by a spiracular cartilage that is patterned by a shh-expressing epithelial signalling centre. This closely parallels the condition seen in the gill arches, where cartilaginous appendages called branchial rays, which support the respiratory lamellae of the gills, are patterned by a shh-expressing gill arch epithelial ridge. Together with similar discoveries in zebrafish, our findings support serial homology of the pseudobranch and gills, and an ancestral origin of gill arch-like anatomical features from the gnathostome mandibular arch.

Evolution of the new head by gradual acquisition of neural crest regulatory circuits.

The neural crest, an embryonic stem-cell population, is a vertebrate innovation that has been proposed to be a key component of the 'new head', which imbued vertebrates with predatory behaviour1,2. Here, to investigate how the evolution of neural crest cells affected the vertebrate body plan, we examined the molecular circuits that control neural crest development along the anteroposterior axis of a jawless vertebrate, the sea lamprey. Gene expression analysis showed that the cranial subpopulation of the neural crest of the lamprey lacks most components of a transcriptional circuit that is specific to the cranial neural crest in amniotes and confers the ability to form craniofacial cartilage onto non-cranial neural crest subpopulations3. Consistent with this, hierarchical clustering analysis revealed that the transcriptional profile of the lamprey cranial neural crest is more similar to the trunk neural crest of amniotes. Notably, analysis of the cranial neural crest in little skate and zebrafish embryos demonstrated that the transcriptional circuit that is specific to the cranial neural crest emerged via the gradual addition of network components to the neural crest of gnathostomes, which subsequently became restricted to the cephalic region. Our results indicate that the ancestral neural crest at the base of the vertebrate lineage possessed a trunk-like identity. We propose that the emergence of the cranial neural crest, by progressive assembly of an axial-specific regulatory circuit, allowed the elaboration of the new head during vertebrate evolution.

hox gene expression predicts tetrapod-like axial regionalization in the skate, Leucoraja erinacea.

The axial skeleton of tetrapods is organized into distinct anteroposterior regions of the vertebral column (cervical, trunk, sacral, and caudal), and transitions between these regions are determined by colinear anterior expression boundaries of Hox5/6, -9, -10, and -11 paralogy group genes within embryonic paraxial mesoderm. Fishes, conversely, exhibit little in the way of discrete axial regionalization, and this has led to scenarios of an origin of Hox-mediated axial skeletal complexity with the evolutionary transition to land in tetrapods. Here, combining geometric morphometric analysis of vertebral column morphology with cell lineage tracing of hox gene expression boundaries in developing embryos, we recover evidence of at least five distinct regions in the vertebral skeleton of a cartilaginous fish, the little skate (Leucoraja erinacea). We find that skate embryos exhibit tetrapod-like anteroposterior nesting of hox gene expression in their paraxial mesoderm, and we show that anterior expression boundaries of hox5/6, hox9, hox10, and hox11 paralogy group genes predict regional transitions in the differentiated skate axial skeleton. Our findings suggest that hox-based axial skeletal regionalization did not originate with tetrapods but rather has a much deeper evolutionary history than was previously appreciated.

Distinct proliferative and middle ear skeletal-patterning functions for SHH-expressing epithelia in the chick hyoid arch.

During chick craniofacial development, the second (hyoid) pharyngeal arch expands to close the neck and gives rise to skeletal elements, including the columella of the middle ear (a homologue of the mammalian stapes). Sonic hedgehog (SHH) signalling has been implicated in hyoid arch expansion and columella formation, but spatial and temporal aspects of these signalling interactions within the hyoid arch remain poorly understood. Here, we show that SHH is initially expressed in the posterior endoderm of the hyoid arch, and that this domain subsequently splits into a distal domain at the site of arch expansion (the posterior epithelial margin, PEM), and a proximal domain that lines the foregut (the proximal hyoid epithelium, PHE). Pharmacological manipulations and heterotopic grafting experiments demonstrate that SHH signalling is required for hyoid arch expansion and skeletogenesis, and reveal distinct roles for the PEM and PHE in these processes. The PEM promotes mesenchymal cell proliferation during arch expansion but is not sufficient to repattern the columella. Conversely, the PHE promotes mesenchymal cell survival, and PHE grafts induce partial duplication of the columella. This work demonstrates crucial and distinct roles for endodermal SHH signalling in hyoid arch morphogenesis and patterning of the middle ear skeleton.

Evolution of vertebrate gill covers via shifts in an ancient Pou3f3 enhancer.

Whereas the gill chambers of jawless vertebrates open directly into the environment, jawed vertebrates evolved skeletal appendages that drive oxygenated water unidirectionally over the gills. A major anatomical difference between the two jawed vertebrate lineages is the presence of a single large gill cover in bony fishes versus separate covers for each gill chamber in cartilaginous fishes. Here, we find that these divergent patterns correlate with the pharyngeal arch expression of Pou3f3 orthologs. We identify a deeply conserved Pou3f3 arch enhancer present in humans through sharks but undetectable in jawless fish. Minor differences between the bony and cartilaginous fish enhancers account for their restricted versus pan-arch expression patterns. In zebrafish, mutation of Pou3f3 or the conserved enhancer disrupts gill cover formation, whereas ectopic pan-arch Pou3f3b expression generates ectopic skeletal elements resembling the multimeric covers of cartilaginous fishes. Emergence of this Pou3f3 arch enhancer >430 Mya and subsequent modifications may thus have contributed to the acquisition and diversification of gill covers and respiratory strategies during gnathostome evolution.

Conserved and unique transcriptional features of pharyngeal arches in the skate (Leucoraja erinacea) and evolution of the jaw.

The origin of the jaw is a long-standing problem in vertebrate evolutionary biology. Classical hypotheses of serial homology propose that the upper and lower jaw evolved through modifications of dorsal and ventral gill arch skeletal elements, respectively. If the jaw and gill arches are derived members of a primitive branchial series, we predict that they would share common developmental patterning mechanisms. Using candidate and RNAseq/differential gene expression analyses, we find broad conservation of dorsoventral (DV) patterning mechanisms within the developing mandibular, hyoid, and gill arches of a cartilaginous fish, the skate (Leucoraja erinacea). Shared features include expression of genes encoding members of the ventralizing BMP and endothelin signaling pathways and their effectors, the joint markers nkx3.2 and gdf5 and prochondrogenic transcription factor barx1, and the dorsal territory marker pou3f3. Additionally, we find that mesenchymal expression of eya1/six1 is an ancestral feature of the mandibular arch of jawed vertebrates, whereas differences in notch signaling distinguish the mandibular and gill arches in skate. Comparative transcriptomic analyses of mandibular and gill arch tissues reveal additional genes differentially expressed along the DV axis of the pharyngeal arches, including scamp5 as a novel marker of the dorsal mandibular arch, as well as distinct transcriptional features of mandibular and gill arch muscle progenitors and developing gill buds. Taken together, our findings reveal conserved patterning mechanisms in the pharyngeal arches of jawed vertebrates, consistent with serial homology of their skeletal derivatives, as well as unique transcriptional features that may underpin distinct jaw and gill arch morphologies.

Trunk neural crest origin of dermal denticles in a cartilaginous fish.

Cartilaginous fishes (e.g., sharks and skates) possess a postcranial dermal skeleton consisting of tooth-like "denticles" embedded within their skin. As with teeth, the principal skeletal tissue of dermal denticles is dentine. In the head, cranial neural crest cells give rise to the dentine-producing cells (odontoblasts) of teeth. However, trunk neural crest cells are generally regarded as nonskeletogenic, and so the embryonic origin of trunk denticle odontoblasts remains unresolved. Here, we use expression of FoxD3 to pinpoint the specification and emigration of trunk neural crest cells in embryos of a cartilaginous fish, the little skate (Leucoraja erinacea). Using cell lineage tracing, we further demonstrate that trunk neural crest cells do, in fact, give rise to odontoblasts of trunk dermal denticles. These findings expand the repertoire of vertebrate trunk neural crest cell fates during normal development, highlight the likely primitive skeletogenic potential of this cell population, and point to a neural crest origin of dentine throughout the ancestral vertebrate dermal skeleton.

A shared role for sonic hedgehog signalling in patterning chondrichthyan gill arch appendages and tetrapod limbs.

Chondrichthyans (sharks, skates, rays and holocephalans) possess paired appendages that project laterally from their gill arches, known as branchial rays. This led Carl Gegenbaur to propose that paired fins (and hence tetrapod limbs) originally evolved via transformation of gill arches. Tetrapod limbs are patterned by asonic hedgehog(Shh)-expressing signalling centre known as the zone of polarising activity, which establishes the anteroposterior axis of the limb bud and maintains proliferative expansion of limb endoskeletal progenitors. Here, we use loss-of-function, label-retention and fate-mapping approaches in the little skate to demonstrate that Shh secretion from a signalling centre in the developing gill arches establishes gill arch anteroposterior polarity and maintains the proliferative expansion of branchial ray endoskeletal progenitor cells. These findings highlight striking parallels in the axial patterning mechanisms employed by chondrichthyan branchial rays and paired fins/limbs, and provide mechanistic insight into the anatomical foundation of Gegenbaur's gill arch hypothesis.

Embryonic origin of the gnathostome vertebral skeleton.

The vertebral column is a key component of the jawed vertebrate (gnathostome) body plan, but the primitive embryonic origin of this skeleton remains unclear. In tetrapods, all vertebral components (neural arches, haemal arches and centra) derive from paraxial mesoderm (somites). However, in teleost fishes, vertebrae have a dual embryonic origin, with arches derived from somites, but centra formed, in part, by secretion of bone matrix from the notochord. Here, we test the embryonic origin of the vertebral skeleton in a cartilaginous fish (the skate, Leucoraja erinacea) which serves as an outgroup to tetrapods and teleosts. We demonstrate, by cell lineage tracing, that both arches and centra are somite-derived. We find no evidence of cellular or matrix contribution from the notochord to the skate vertebral skeleton. These findings indicate that the earliest gnathostome vertebral skeleton was exclusively of somitic origin, with a notochord contribution arising secondarily in teleosts.

Positive and negative forms of replicability in gene network analysis.

MOTIVATION: Gene networks have become a central tool in the analysis of genomic data but are widely regarded as hard to interpret. This has motivated a great deal of comparative evaluation and research into best practices. We explore the possibility that this may lead to overfitting in the field as a whole. RESULTS: We construct a model of 'research communities' sampling from real gene network data and machine learning methods to characterize performance trends. Our analysis reveals an important principle limiting the value of replication, namely that targeting it directly causes 'easy' or uninformative replication to dominate analyses. We find that when sampling across network data and algorithms with similar variability, the relationship between replicability and accuracy is positive (Spearman's correlation, rs ∼0.33) but where no such constraint is imposed, the relationship becomes negative for a given gene function (rs ∼ -0.13). We predict factors driving replicability in some prior analyses of gene networks and show that they are unconnected with the correctness of the original result, instead reflecting replicable biases. Without these biases, the original results also vanish replicably. We show these effects can occur quite far upstream in network data and that there is a strong tendency within protein-protein interaction data for highly replicable interactions to be associated with poor quality control. AVAILABILITY AND IMPLEMENTATION: Algorithms, network data and a guide to the code available at: https://github.com/wimverleyen/AggregateGeneFunctionPrediction CONTACT: jgillis@cshl.edu SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Patterns of late-life depressive symptoms and subsequent declines in cognitive domains.

BACKGROUND: Depression frequently co-occurs with cognitive decline, but the nature of this association is unclear. We examined relations of late-life depressive symptom patterns to subsequent domain-specific cognitive changes. METHODS: Depressive symptoms were measured at up to 3 timepoints among 11,675 Nurses' Health Study participants prior to cognitive testing. Depressive symptom patterns were categorized as non-depressed, variable or persistent, based on published severity cutpoints. Outcomes were global, verbal, and executive function-attention composite scores. RESULTS: Participants with persistent depressive symptoms had worse executive function-attention decline compared with non-depressed participants (multivariable-adjusted mean difference = -0.03 units/year, 95% CI: -0.05, -0.01; p = 0.003); this difference was comparable with 8 years of aging. However, being in the persistent versus non-depressed group was not significantly related to verbal (p = 0.71) or global score (p = 0.09) decline. By contrast, compared with the non-depressed group, those with variable depressive symptoms had worse verbal memory decline (multivariable-adjusted mean difference = -0.01 units/year, 95% CI: -0.02, -0.002; p = 0.03); this group showed no differences for global or executive function-attention decline. CONCLUSIONS: A variable pattern of depressive symptom severity related to subsequent decline in verbal memory, while a persistent pattern related to decline in executive function-attention. Findings could signal differences in underlying neuropathologic processes among persons with differing depression patterns and late-life cognitive decline. Copyright © 2016 John Wiley & Sons, Ltd.

Guidance for RNA-seq co-expression network construction and analysis: safety in numbers.

MOTIVATION: RNA-seq co-expression analysis is in its infancy and reasonable practices remain poorly defined. We assessed a variety of RNA-seq expression data to determine factors affecting functional connectivity and topology in co-expression networks. RESULTS: We examine RNA-seq co-expression data generated from 1970 RNA-seq samples using a Guilt-By-Association framework, in which genes are assessed for the tendency of co-expression to reflect shared function. Minimal experimental criteria to obtain performance on par with microarrays were >20 samples with read depth >10 M per sample. While the aggregate network constructed shows good performance (area under the receiver operator characteristic curve ∼0.71), the dependency on number of experiments used is nearly identical to that present in microarrays, suggesting thousands of samples are required to obtain 'gold-standard' co-expression. We find a major topological difference between RNA-seq and microarray co-expression in the form of low overlaps between hub-like genes from each network due to changes in the correlation of expression noise within each technology. CONTACT: jgillis@cshl.edu or sballouz@cshl.edu SUPPLEMENTARY INFORMATION: Networks are available at: http://gillislab.labsites.cshl.edu/supplements/rna-seq-networks/ and supplementary data are available at Bioinformatics online.

Measuring the wisdom of the crowds in network-based gene function inference.

MOTIVATION: Network-based gene function inference methods have proliferated in recent years, but measurable progress remains elusive. We wished to better explore performance trends by controlling data and algorithm implementation, with a particular focus on the performance of aggregate predictions. RESULTS: Hypothesizing that popular methods would perform well without hand-tuning, we used well-characterized algorithms to produce verifiably 'untweaked' results. We find that most state-of-the-art machine learning methods obtain 'gold standard' performance as measured in critical assessments in defined tasks. Across a broad range of tests, we see close alignment in algorithm performances after controlling for the underlying data being used. We find that algorithm aggregation provides only modest benefits, with a 17% increase in area under the ROC (AUROC) above the mean AUROC. In contrast, data aggregation gains are enormous with an 88% improvement in mean AUROC. Altogether, we find substantial evidence to support the view that additional algorithm development has little to offer for gene function prediction. AVAILABILITY AND IMPLEMENTATION: The supplementary information contains a description of the algorithms, the network data parsed from different biological data resources and a guide to the source code (available at: http://gillislab.cshl.edu/supplements/).

Electrosensory ampullary organs are derived from lateral line placodes in cartilaginous fishes.

Ampullary organ electroreceptors excited by weak cathodal electric fields are used for hunting by both cartilaginous and non-teleost bony fishes. Despite similarities of neurophysiology and innervation, their embryonic origins remain controversial: bony fish ampullary organs are derived from lateral line placodes, whereas a neural crest origin has been proposed for cartilaginous fish electroreceptors. This calls into question the homology of electroreceptors and ampullary organs in the two lineages of jawed vertebrates. Here, we test the hypothesis that lateral line placodes form electroreceptors in cartilaginous fishes by undertaking the first long-term in vivo fate-mapping study in any cartilaginous fish. Using DiI tracing for up to 70 days in the little skate, Leucoraja erinacea, we show that lateral line placodes form both ampullary electroreceptors and mechanosensory neuromasts. These data confirm the homology of electroreceptors and ampullary organs in cartilaginous and non-teleost bony fishes, and indicate that jawed vertebrates primitively possessed a lateral line placode-derived system of electrosensory ampullary organs and mechanosensory neuromasts.

A randomized, placebo-controlled, double-blind study of sapropterin to treat ADHD symptoms and executive function impairment in children and adults with sapropterin-responsive phenylketonuria.

Symptoms of attention deficit-hyperactivity disorder (ADHD), particularly inattention, and impairments in executive functioning have been reported in early and continuously treated children, adolescents, and adults with phenylketonuria (PKU). In addition, higher blood phenylalanine (Phe) levels have been correlated with the presence of ADHD symptoms and executive functioning impairment. The placebo-controlled PKU ASCEND study evaluated the effects of sapropterin therapy on PKU-associated symptoms of ADHD and executive and global functioning in individuals who had a therapeutic blood Phe response to sapropterin therapy. The presence of ADHD inattentive symptoms and executive functioning deficits was confirmed in this large cohort of 206 children and adults with PKU, of whom 118 responded to sapropterin therapy. In the 38 individuals with sapropterin-responsive PKU and ADHD symptoms at baseline, sapropterin therapy resulted in a significant improvement in ADHD inattentive symptoms in the first 4 weeks of treatment, and improvements were maintained throughout the 26 weeks of treatment. Sapropterin was well-tolerated with a favorable safety profile. The improvements in ADHD inattentive symptoms and aspects of executive functioning in response to sapropterin therapy noted in a large cohort of individuals with PKU indicate that these symptoms are potentially reversible when blood Phe levels are reduced.