On the evolutionary developmental biology of the cell

Abstract

Organisms are complex assemblages of cells – cells that produce light, shoot harpoons, and secrete glue. Identifying the mechanisms that generate novelty at the level of the individual cell is, therefore, essential for understanding how multicellular life evolves. For decades, the field of Evolutionary Developmental Biology (Evo-Devo) has been developing a framework for connecting genetic variation that arises during embryonic development to the emergence of diverse adult forms. With increasing access to new single-cell ‘omics technologies and an array of techniques for manipulating gene expression, we can now extend these inquiries inward to the level of the individual cell. In this essay, I argue that applying an Evo-Devo framework to single cells makes it possible to explore the natural history of cells, where this was once only possible at the organismal level.

An Evo-Devo approach to biodiversity

During embryonic development, a single undifferentiated cell undergoes a series of events that will ultimately give rise to the impressive diversity of cell types that define an adult organism. Depending on the organism, this process might result in cells that make light, like the photophores of a firefly, or cells that eject a venomous, ballistic projectile, like the stinging cells of jellyfish, or even cells that secrete glue, like those in the nectary gland of the carnivorous sundew. An intricate system of temporal and genetic instructions unfolds to transform the single, fertilized egg into cells that perform these specialized – and sometimes bizarre – functions. How does this occur? The process of embryogenesis varies widely across taxa. Some mechanisms known to create variation among embryos include changes to the timing of gene expression during development (heterochrony, see Glossary), changes to the identity of embryonic structures (homeosis), and environmental contingency in the process of development (plasticity). Processes generating variation at the embryonic level result in adults with diverse forms; thus, embryogenesis serves as an important window into the evolution of biodiversity. For example, variation in developmental timing during flower morphogenesis can drive the emergence of distinct reproductive morphologies in plants, changing their interaction with pollinators [1] and homeotic control of abdominal color can promote rapid segregation in bumble bee populations [2]. The field of evolutionary developmental biology (Evo-Devo) arose as a way to explicitly connect the mechanisms driving variation in embryonic development with the evolution of biodiversity and has become the predominant framework for predicting the drivers of morphological change [3]. The advent of tools for precisely manipulating genomes and characterizing the effects on individual cell types across a wide variety of taxa affords us the opportunity to predict how novel cell identities arise. Here, I argue that applying an Evo-Devo framework to the level of the individual cell makes it possible to understand how the extant diversity of cells came to be and, importantly, to predict how mutations that have never occurred naturally might generate cell types that have never actually evolved.

New tools provide new answers to old questions

Early experimental embryologists relied on a limited set of tools to investigate cell fate. From the hot needles once used to ablate individual cells [4] to the baby hair lassos used to separate them [5], these simple manipulations gave rise to some of the first ideas about the diversification of cell identity: while some cells are autonomously programmed to faithfully produce predetermined daughters (e.g., during mosaic development), others produce daughters with identities that are conditional upon cues from their neighbors (e.g., during regulative development). The discovery of autonomous vs conditional specification revealed that embryonic cells could differ dramatically in their fate, highlighting how embryonic development can serve as an agent of evolutionary change [6]. Indeed, many insightful hypotheses about the evolution of cell identity were proposed some time ago. In 1985, Slack [7] suggested homeotic transformation of cell types was likely common during normal tissue development and may be an important mechanism for creating heterogeneity of cell function in organs. At the time, our ability to discriminate different cell types was limited to only those cells that either had distinct morphologies (e.g., secretory vesicles or axons) or reacted to histological dyes in unique ways (e.g., basophilic or eosinophilic cells). Thus, the available tools were inadequate to investigate the mechanisms driving cell type diversity.

The advent of single-cell mRNA sequencing (scRNA-Seq) has led to the development of high-resolution approaches for discriminating cell types based on the unique combination of genes they express [8]. Application of scRNA-Seq across stages of embryonic development reveals how transcriptional changes relate to the appearance of distinct cell types [9]. Because this technique can be adapted for use in nearly any tissue in any organism, scRNA-Seq has become a powerful tool for comparing cell identities across wide evolutionary distances [10–12]. The continuous proliferation of single-cell ‘omics technologies provides increasingly deeper access to intracellular phenotypes. Single-cell Assay for Transposase-Accessible Chromatin sequencing (scATAC-Seq) can identify heterogeneity in the regulatory response of individual cells responding to injury [13] and single-cell chromatin immunoprecipitation sequencing (scChIP-Seq) has revealed the sequence of events required to push a cell from quiescence to proliferation [14]. The recent development of single-cell ribosome sequencing (scRibo-Seq) makes it possible to identify only mRNAs that have been loaded onto a ribosome, revealing how translation efficiency can generate cell-type specific temporal variation in protein abundance [15]. When combined with simple manipulations of cell environment (e.g., single-cell ablation with a hot needle), these techniques can be used to understand how each remaining cell in an embryo responds to the loss of its neighbor, shedding new light on old questions about how cell identity is generated (Figure 1, Key Figure). Specifically, it becomes possible to determine which genes are uniquely upregulated in each remaining cell (scRNA-Seq), how access to transcription factor binding sites is changing in each cell (scATAC-Seq), and how two cells upregulating the same suite of genes might still exhibit variation in the daughters they produce (scRibo-Seq). Furthermore, the development of new cell cycle reporters including cell cycle timers – genetically encoded fluorescent proteins that indicate transit time through the cell cycle – means we can now visualize how long each cell type spends resting or proliferating [16,17]. Combined with the ever-increasing precision of CRISPR-based genome editing [18,19], the tools described here will revolutionize our ability to link genomic heterogeneity to variation in cell phenotype. It seems the time is right to revisit the insights of our predecessors and apply the concepts they developed to a level of biological organization they could not access.

Figure 1.

New methods shed light on old questions about cell identity. (A) Early experiments using single-cell ablation (tunicate [4]) and/or isolation of individual embryonic cells (sea urchin, [62]) led to the discovery of mosaic and regulative modes of development. Cell ablation results in loss of specific tissues in tunicates. In sea urchins, isolated embryonic cells can make complete adults, but they are slightly smaller than unmanipulated adults. (B) Embryonic fate maps were developed using taxa that had natural variation in cell color or shape (tunicate, [63]; sea urchin, [64]). Combined with ablation/isolation results, these experiments revealed two distinct modes of cell fate specification: autonomous (where each cell controls the fate of a subset of daughters) and conditional (where each cell gives rise to a subset of daughters while retaining the capacity to compensate for other lost lineages.) (C) The development of fluorescent dyes for tracing cell lineages means we are no longer restricted to following cell fate in organisms that have naturally pigmented cells. Each cell in the early embryo can be labeled and/or ablated to follow the impact on development. Integration of classical methods in experimental embryology with new methods for interrogating intracellular processes will reveal deeper insights about the control of cell fate specification. For example: (D) single-cell mRNA sequencing (scRNA-Seq) in ablated embryos can reveal how the remaining cells transcriptionally compensate for the loss of one cell during conditional fate specification and (E) single-cell ATAC sequencing (scATAC-Seq) allows for the identification of cis-regulatory regions associated with these transcriptional changes. (F) Pairing ablation with single-cell ribosome sequencing (scRibo-Seq) can reveal post-transcriptional processes generating variation during compensation.

What is cell identity?

Cell types can be defined by many different attributes: their location (e.g., ectoderm or endoderm), their behavior (e.g., secretory vs. non-secretory), their morphology (e.g., mononucleate, multinucleate, or anucleate), their potency (e.g., stem or differentiated), and, increasingly, by the suite of genes they express (their “transcriptome”). The morphological and functional traits that define cell types are encoded by networks of interacting gene products called gene regulatory networks, which have become a valuable tool for linking genotype to phenotype in EvoDevo [20]. To highlight the modular nature of these networks, I will refer to the parts of these networks that control individual aspects of cell phenotype as “gene modules”. Distinct cell types must express gene modules for establishing basic cell properties like intracellular osmolarity, membrane integrity, and mitochondrial function, in addition to the modules that underlie a cell’s unique capacities. Identifying the gene modules encoding the traits that define each cell’s unique attributes is generally considered an important way to conceptualize cell type diversity [21,22]. Novel cell types arise either with the advent of new modules (e.g., genes encoding truly novel cell functions, like the synthesis of spider silk) or through the shuffling of existing modules into new spatial or temporal relationships (e.g., through gene co-option) [23]. What follows is a series of examples of cells that have evolved novel phenotypes and some suggestions from the field of Evo-Devo about the mechanisms that might be generating these phenotypes.

Single-cell heterochrony

In its simplest conception, heterochrony describes how morphological differences can arise from altered timing of the same process in two embryos. Originally applied to ancestor and descendant, modern usage frequently invokes heterochrony to explain the generation of diverse morphologies across a clade. For example, changes in the rate of cell proliferation can explain the development of short- and long-faced species of bats [24] or digit reduction in lizards [25]. This is important because it explicitly links an embryonic mechanism (length of the proliferative stage of organ development) to variation in adult form (longer or shorter bones). Heterochrony can be invoked to describe changes in the rate of a process (e.g., acceleration or deceleration), a shift in the onset of a process (e.g., earlier or later in development), or even rearrangements in the timing of discrete modules expressed during development (“sequence heterochrony” sensu [26]). The examples below suggest heterochrony may also generate diversity at the level of the individual cell.

Amoebas (Acanthamoeba castellanii) reproduce by mitotic cell division, following duplication of their organelles. Amoebas are found with mono- and multinucleate phenotypes and it was recently discovered that preventing amoebas from adhering to substrate will inhibit cytokinesis without affecting organelle duplication [21]. The free-swimming multinucleate amoebas are both larger and capable of eating more than their adherent, mononucleate kin, suggesting the multinucleate phenotype may have a competitive advantage in a new environment. Thus, uncoupling cytokinesis from organelle replication provides a pathway to the establishment of distinct ecological strategies and novel phenotypes for unicellular organisms. Likewise, a delay in the progression of the cell cycle may have also been important for the evolution of land plants from aquatic ancestors [28]. During the production of reproductive spores, aquatic plants undergo multiple rounds of meiosis, each of which is followed by cytokinesis and deposition of a specialized cell wall (sporoderm). In the ancestor of land plants, cell wall deposition was delayed until after several rounds of nuclear replication, enabling the emergence of desiccation-resistant spore clusters that could be transmitted aerially (a prerequisite to the diversification of terrestrial taxa). Finally, sequence heterochrony is a predictor of phenotype in the lineage of cells that gives rise to blood cells in mammals (hematopoietic stem cells) [29]. Changing the order in which two key transcription factors (C/EBPα and GATA) are active during lineage commitment in the hematopoietic stem cell population shifts the identity of their daughters from eosinophils (C/EBPα before GATA) to basophils (GATA before C/EBPα). This study demonstrates how sequence heterochrony can translate the same sets of expressed genes into two distinct cell types.

Viewing these examples as evidence of single-cell heterochrony provides a testable hypothesis to pursue to understand how cells with these phenotypes may have evolved. Are multinucleate cells stuck in mitosis or do they progress through their cell cycle despite having failed to undergo cytokinesis? The recent development of a CRISPR knock-in strategy for generating transgenic amoebas [30] means it should now be possible to examine the dynamics of cell cycle progression in mono- vs multinucleate cells using cell cycle reporters (Figure 2A). Studies like this could reveal how the amoebas tolerate the loss of cytokinesis and could generate predictions for understanding the emergence of many other naturally occurring multinucleate cell types [31]. Similarly, viewing the rearrangement of transcription factors as an example of sequence heterochrony provides a new platform for predicting cell phenotype. Translational pausing can rearrange the sequence of transcription factors that have access to a regulatory site during cell type specification (Figure 2B). One study using modeling to predict the evolution of cell type diversity demonstrated that the greatest diversity was achieved when the order, not just the identity, of expressed transcription factors was considered [32]. Future studies pairing scRNA-Seq with scRibo-Seq may reveal the presence of more cell types than would be predicted based on scRNA-Seq alone.

Figure 2.

Single-cell heterochrony as a mechanism for generating novel cell types. (A) Schematic representation of a mononucleate cell passing through various stages of the cell cycle to produce two mononucleate daughters compared to the same cell cycle after transgenic insertion of a cell cycle reporter (e.g., PCNA-RFP [17]). (B) This tool could be used to evaluate two hypotheses about the production of a multinucleate daughter cell from a mononucleate progenitor. Null hypothesis (H0) – wildtype cell division of a mononucleate progenitor. Hypothesis 1 (H1) – cell skips cytokinesis; multinucleate daughter returns to G1. Hypothesis 2 (H2) – cell stalls in mitosis, never entering cytokinesis or progressing to G1. (B) Translational pausing can change the order in which transcription factors appear in a cell. Proteins are generally translated within minutes after expression of mRNA (left). Often the order of expression of mRNAs is used as an indicator of the order of appearance of proteins (middle). Translational pausing can elongate the interval between mRNA and protein, changing the order in which proteins are available to the cell (right).

Single-cell homeosis

The transformation of one part of the developing organism into another is referred to as a homeotic transformation [33]. This phenomenon is exemplified by the shift from stamen to petal in developing flowers [34] and the conversion of an abdominal segment into a thoracic segment in flies [35]. The idea that significant morphological variation could be generated by homeotic transformation has a long history with embryology, from formalization of the idea of homeosis [34] to the identification of regulatory genes that drive such transformations [36]. While the evolutionary value of these mutations was slow to gain acceptance [37], recent studies connecting homeotic transformations to environmental perturbations [38] have revived the value of this class of mutations for producing important, selectable variation in nature. Single-cell homeosis is a significantly newer concept. First invoked to describe induced transformations in neural cell identity in a variety of animal taxa [39], single-cell homeosis can now be assessed broadly using scRNA-Seq. A recent study comparing wildtype and mutant Arabidopsis plants (lacking a single transcription factor) identified a transformation from hair cells to non-hair cells in the roots of the mutants [40]. These studies demonstrate how a large phenotypic effect (cell identity) can be coordinated through a single switch (homeotic gene), providing a mechanistic explanation for saltatory evolution of phenotype. Indeed, homeotic transformation of cell identity could generate secretory cells in a previously non-secretory epithelium, explaining the origin of novel organs like the toxin-secreting tergal gland of rove beetles [41].

A special type of homeotic transformation is the reemergence of an ancestral trait after its evolutionary loss. This type of homeosis, termed atavism, is exemplified by mutations that result in the restoration of reptile-like teeth in birds [42]. Applied at a single-cell level, atavism results in the transformation of a contemporary cell type into an ancestral one, and can be viewed as an example of temporal, rather than spatial, homeosis [43]. The atavism theory of cancer frames the transformation of healthy, quiescent cells into actively dividing tumor cells as an ancestral reversion to the uncontrolled proliferative behavior of our unicellular ancestors [44]. Consistent with this reversion, cancer cells tend to express genes that are phylogenetically older than those expressed in the healthy cells of the same tissue [45]. A recent study of single-cell atavism in sea anemones used phylogenetic inference to demonstrate the restoration of an ancestral type of stinging cell in response to genome editing [46]. Knockout of a single Sox-family transcription factor in a burrowing sea anemone resulted in the restoration of a type of stinging cell normally found only in closely related sea anemones. In this study, manipulation of a single gene was sufficient to restore an ancestral cell type, demonstrating that the developmental program required for patterning a trait can be maintained in the genome even without expression of that trait. If such “hidden” phenotypic potential is released under changing environmental conditions with novel selection pressures, atavism can serve as a source for adaptive evolution and the origin of cellular novelty.

At the whole-organism level, atavism appears to be quite rare. One previous study examining the restoration of ancestral teeth in mutant chickens demonstrated only partial restoration of tooth rudiments [47]. Further investigation suggested many of the components of tooth enamel had been eroded (e.g., through pseudogenization) from the gene module that once controlled this trait [48]. So how could ancestral identity maintained at the single-cell level? The answer may be obvious in the case of cancer: positive selection to maintain highly proliferative cells is important for both embryonic development and adult tissue dynamics in healthy organisms [49]; cancer emerges when the proliferative module is activated in the wrong cell type. The case with stinging cells is not so clear. One hypothesis suggests the ancestral cell type is maintained because it is a phenotypic foundation upon which the contemporary cell type is built [46]. If true, this would suggest that cell identity can evolve by addition of new modules to old cell types (Figure 3A). A recent study of homeotic transformation of neurons in Drosophila provides evidence of such additive regulation of cell identity and demonstrates how this transformation can be controlled by individual transcription factors [50]. These examples suggest that additive, single-gene control of cell identity may well be widespread and, therefore, co-option of individual regulatory genes could be a common pathway to the evolution of new cell phenotypes [51]. With new tools for manipulating genes in a wide variety of taxa and assessing the phenotypic outcome at the single-cell level, this hypothesis can now be explicitly tested (Figure 3B–D).

Figure 3.

Modular control of cell identity and assessment of single-cell homeosis. (A) Individual cells are composed of modular phenotypes. In stinging cells, gene modules control the type of harpoon, nature of the capsule wall, and type of synapse controlling harpoon discharge, among other traits (not indicated). New traits can be added on top of an existing phenotype by addition of a new module (Mnew) downstream of the transcription factor (TF) that controls stinging cell development. (B) An approach for testing modularity in cell identity could start with knockout of a gene of interest (KeyTF) and labeling of the mutated locus with a fluorescent tag (e.g., green fluorescent protein, GFP). Probing dissociated cells from wildtype (WT) and mutant (KO) tissues with a combination of scRNA-Seq and scATAC-Seq will enable identification of the target genes and the regions of chromatin regulating the expression of these genes, leading to assembly of the cell identity module controlled by the KeyTF. Genes expressed in this cell that are not controlled by the KeyTF (e.g., geneA, geneD) are part of another cell gene module. (C) To infer single-cell homeosis, compare cell phenotypes before and after mutation to identify changes in morphology, function, and/or transcriptional identity. If loss of the KeyTF shifts cell identity to an alternative, existing cell phenotype (e.g., a shift from pink to purple) this would indicate homeosis (black arrows). (D) To infer single-cell atavism, the mutant phenotype must be analyzed in a phylogenetic context. Loss of the KeyTF would result in a phenotype not normally found in the study organism (e.g., a shift from pink to white) but that is present in closely related taxa. This would demonstrate that gain of the KeyTF ancestrally led to the gain of a new cell phenotype module, changing the identity of the cell. Images in A were adapted from images initially published in Babonis et al. [46] under a CC BY 4.0 license (https://creativecommons.org/licenses/by/4.0/).

Single-cell plasticity

In embryology, plasticity refers to the phenotypic changes that arise because of environmental influence on the developing embryo [52]. Famously exemplified by the development of wings in the offspring of wingless pea aphids experiencing crowded conditions [53], plasticity has come to be viewed as an important (if not necessary) precursor to the evolution of new traits [54]. Might single-cell plasticity likewise lead to the evolution of new cell types? In fact, the existing data on this topic suggest the opposite may be true. A particularly well-characterized example of single-cell plasticity can be found in the development of the larval skeleton of sea urchins [55]. Ablation of the cells normally responsible for skeleton secretion (the primary mesenchyme cells) enables secreted growth factors (VEGF) to reach the adjacent blastocoelar cells, triggering them to secrete skeleton [56]. Thus, upon induction by environmental cues (growth factors), the blastocoelar cells transdifferentiate into skeletogenic cells. Unlike sea urchins, sea stars lack a larval skeleton (and primary mesenchyme cells). Ancestral state reconstruction suggests a larval skeleton was present in the last common ancestor of modern echinoderms [5]; thus, the loss of the larval skeleton in sea stars may have resulted from loss of plasticity in the blastocoelar cells.

Efforts to induce alternative identities in a variety of differentiated cell types have revealed that not all cells exhibit the same capacity for plasticity, and it is the “failures” that may hold key information for understanding the relationship between plasticity and the evolution of new cell types. Dedifferentiation, the reversion of a differentiated cell back to a pluripotent state, is another form of plasticity manifested at the single-cell level and is the target of many recent efforts to develop pluripotent cells for therapeutic purposes. Induced dedifferentiation (often called reprogramming) requires at least two steps: clearing out the markers of differentiated cell identity (typically encoded in the epigenome) and activating the gene module that confers pluripotent cell identity [58]. A recent study demonstrated that failure to achieve pluripotency in cultures of certain adult cell types was a consequence of incomplete activation of the DNA replication machinery [59]. In contrast to induced cells from an embryonic origin, these adult cells activated fewer initiation sites during DNA replication (replication forks) and exhibited biased replication in one direction (3’ to 5’), rather than proceeding equally from both 5’ and 3’ ends. These irregularities resulted in genomic instability and cell death, revealing one mechanism by which cells may lose plasticity.

Why would any cell give up the opportunity to change identity in the first place? One idea is that cells that escape the obligation to re-enter the cell cycle are free to evolve phenotypes that might be incompatible with pluripotency [60]. Mutations that prevent the possibility of re-entering the cell cycle could be beneficial for the origin of novelty if they make available intracellular resources (e.g., space, enzymes, mitochondria) that can be allocated to another function. In this context, it would be exciting to find an example whereby a novel cell phenotype evolved by co-option of a gene product normally used for successful re-entry into the cell cycle. Such a co-option event would necessarily be cell type specific but a result like this would significantly expand our understanding of the constraints that limit the evolution of specialized cell identity and offer an opportunity to elucidate the tradeoffs between canalization and the maintenance of cellular plasticity (Figure 4A). The anucleate red blood cells of mammals exhibit an extreme example of canalized cell type; by giving up their nucleus, these cells sacrifice the ability to take on an alternative phenotype but also achieve the highest capacity for deformation, allowing them to fit through small-diameter capillaries. Neurons may well be another example of a canalized cell type. Several differentiated cell types that can be induced to take on a neural fate [61]; but, thus far, the opposite does not appear to be true. Plasticity in neurons appears to be restricted to the nature of their connections (e.g., synaptic pruning), rather than the identity of the differentiated cell. Challenging various differentiated cell types with cues to induce dedifferentiation or transdifferentiation and examining the outcome with scATAC-Seq, should make it possible to determine if evolution has resulted in common mechanisms driving the loss of plasticity (Figure 4B).

Figure 4.

Investigating single-cell plasticity with scATAC-Seq. (A) Cells have the potential to take on many possible fates. Some cells retain the capacity to activate modules that cause dedifferentiation and re-entry into the cell cycle; others may transdifferentiate from one differentiated cell type into another, without access to a dedifferentiation module. Cells that maintain their phenotype irrespective of how they are cued are canalized. (B) Single-cell ATAC sequencing can be used to determine how different cell types respond to plasticity cues. Embryonic cells have an epigenomic signature (grey ATAC peak); successful dedifferentiation to pluripotency is associated with loss of this signature (dotted circle) and gain of open regions of chromatin associated with activation of pluripotency (pink peaks). Successful transdifferentiation involves loss of the epigenomic signature of cell type 1 and gain of signature for cell type 2. Comparing the ATAC-Seq signatures of different adult cell populations can reveal whether they fail to achieve dedifferentiation because they fail to clear their differentiated identity (blue peak) or fail to fully activate a pluripotent identity.

Concluding remarks

The mechanisms presented here are largely unexplored at the single-cell level, making this a truly exciting time to be a biologist interested in the diversification of cells. We now have tools that allow us to access the secrets of the most peculiar cells in the strangest of organisms, opening up new possibilities for probing the evolution of novel cell function (see Outstanding Questions). No matter how the ‘omics revolution continues to unfold, we will still need the steady hands and sharp needles that gave Evo-Devo its start; but future studies coupling the classical methods of experimental embryology with techniques for mixing-and-matching gene modules from different cells in different organisms will provide an opportunity to link genotype to phenotype on a per-cell basis. These approaches will allow us to make cells that defy our current understanding of cellular biodiversity and push the boundaries of cell type evolution.

Outstanding Questions.

• How might incorporating the idea of developmental constraint influence our understanding of cell type evolution? Is cell identity limited by ancestry, either phylogenetically and at the level of the cell lineage?

• What are the limitations to employing an Evo-Devo framework to studies of cell type biodiversity? What do we miss by taking this approach?

• What are the limitations of focusing on the individual cell? Might the modular control of cell identity differ in an isolated cell and a cell measured in a tissue?

• How common is additive phenotypic evolution as a mechanism of cell type innovation?

• What additional tools will significantly improve our ability to investigate the evolution of cell identity?

• Research is, necessarily, limited to a small set of taxa (those which can be reared in a lab, manipulated, and analyzed). Are there key taxa that should be prioritized as important comparative models to further progress toward a natural history of cells?

Highlights.

• Organismal diversification can be modeled as the gain of new cell types.

• Just as variation that arises during embryonic development can drive the emergence of new adult forms, so, too, can intracellular variation serve as a driver of cell type biodiversity.

• Examining phenomena like heterochrony, homeosis, and plasticity in individual cells encourages the development of testable, mechanistic hypotheses regarding the emergence of new cell types.

• Application of new sequencing technologies to organisms from across the tree of life provides increased resolution for evaluating hypotheses about the relationship between genotype and phenotype at the individual cell level.

• The development of new tools for manipulating genomes frees us from the limited phenotypes that evolution has produced, allowing us to generate new cell phenotypes and explore the boundaries of cellular novelty.

Glossary

Ablate

to eliminate one or more cells from an embryo

Canalization

loss of plasticity; results in cells that have lost the ability to re-enter the cell cycle

Cell cycle reporter

genetically encoded fluorescent proteins that vary in brightness throughout the cell cycle

CRISPR-based genome editing

use of an enzyme to cut a specific region of DNA and induce a repair; sloppy repairs can induce mutations through excision, conversion, or insertion of nucleotides

Dedifferentiation

a change in cell identity from a differentiated cell type to a pluripotent cell type; requires re-entry into the cell cycle

Epigenome

modifications to the genome that influence when/where it becomes activated; the specific modifications vary by cell

Gene module

a network of interacting gene products that controls a component of cellular phenotype

Heterochrony

a relative change in the timing of a developmental process across embryos (or cells)

Homeosis/homeotic transformation

a phenomenon in which one part of the organism assumes the fate of another part of the organism during development

Mosaic development

a type of embryonic development in which cells faithfully produce their programmed daughters

Novelty

a cell with a unique phenotype and a phylogenetically restricted distribution; phenotype can be assessed morphologically, by gene product expression, or by the function/behavior of the cell

Plasticity

the capacity of a cell to re-enter the cell cycle and to renew/replace itself or generate differentiated daughter cells in response to an environmental cue

Pseudogenization

the accumulation of mutations in a gene sequence or the surrounding regulatory region that cause the gene to lose function

Regulative development

a type of embryonic development in which embryonic cells consistently produce a subset of daughter cells but retain the capacity to make other daughters under certain conditions

scATAC-Seq

a method for identifying regions of open chromatin in a cell

scChIP-Seq

a method for identifying sites in the DNA that are bound by transcription factors or other DNA binding proteins

scRibo-Seq

a method for discriminating mRNAs that are actively undergoing translation (loaded in the ribosome) from other mRNAs in the cell

scRNA-Seq

a method for identifying all mRNA transcripts in a cell

Transdifferentiation

a change in cell identity from one differentiated cell type to another, without re-entry into the cell cycle