Scaling Up How Cells Build an Organism

A single cell can develop, more or less on its own, into a full adult organism. This basic phenomenon was noted nearly as long ago in the history of Western thought as we can go back, by Aristotle. In On the Soul, he wrote, "It is a fact of observation that plants and certain insects go on living when divided into segments; this means that each of the segments has a soul in it identical in species, though not numerically identical in the different segments, for both of the segments for a time possess the power of sensation and local movement." William Bateson (Bateson 1913) likened organismal growth to the basic process of cell division, and Bonner (Bonner 2000) suggests that the beginning of complex organisms was the failure of cell division products to separate, presumably in the early ocean.

A single cell somehow contains within it all the requisite "information" to go beyond its first division. Here, we will describe how a repertoire of a relatively few basic and logically simple processes is widely used to produce organismal diversity. Much of this chapter may seem like gene name-dropping. Listing genes is not very informative, and we are most likely to have found only what we know how to find, not every gene involved in the trait or process of interest, but the results do make sense. Not only are the same processes used and reused, but very similar genes or genetic mechanisms are involved over and over again to bring those processes about. These repeatedly used processes include what we would expect: signaling factors and receptors, transcription factors, activating and inhibiting effects. By mixing the order, frequency, and timing of their use—that is, the repeated contex-tually varying expression of subsets of the same set of genes—an organism can bring about its own transformation from a simple beginning to an adult.

HOW CELLS MAKE ORGANISMS: BUILDING MORE WITH LESS In a way, cell division was a fundamental trick in making more of life. The mechanisms by which the constituents of a cell replicate and a cell divides are related to

Genetics and the Logic of Evolution, by Kenneth M. Weiss and Anne V. Buchanan. ISBN 0-471-23805-8 Copyright © 2004 John Wiley & Sons, Inc.

the way cells adhere, move, or cooperate (e.g., Bonner 1998; 2000). Thus materials in spindle fibers are similar to those used in flagella or cilia, and various speculations can be made about how both cell division, with its rules for the distribution of materials into daughter cells, and other multicellular functions evolved. Much of what happened in the rest of evolution is a use and reuse of the many things that were needed for basic cell biology and division.

One ultimate question that we touched on earlier is why cell division has led to the aggregation of cells, and their dependence on each other, to the point that only some of them are even able to copy themselves into new colonies (individuals). We can imagine biological challenges or potential opportunities that single cells could not answer very well. Autonomy can be liberating in many ways, but we can imagine (and observe) circumstances in which large aggregates of cells can better withstand environmental changes or acquire resources. Bacterial aggregates appear to have this effect: bacteria in biofilms are much more difficult to destroy with antibiotics than planktonic bacteria, for example, due to a number of antibiotic resistance mechanisms that have recently begun to be elucidated (e.g., Stewart 2002; Stewart and Costerton 2001). Directed mobility enables organisms to pursue potential food if there is a local shortage (or avoid becoming somebody else's food).

On Being Big

It has been said that there is usually an open ecological niche at the top of the size range (e.g., Bonner 1998). Several times in evolutionary history a new branch of organisms has evolved a subset of lineages of ever-larger species, a premise sometimes known as Cope's law (see, e.g., Bonner 1988). Thus the first fish, land creatures, dinosaurs, birds, etc. may have been small, but larger species ensued. Yet, to reiterate a point we have made before, small is beautiful, too.

Large species have experienced several major die-offs, so it may be precarious to sit atop the size pyramid even if it gains you an empty niche when you first move in, but there are still a lot of large creatures around. Perhaps a kind of arms race develops among aggregates of cells (that eventually are reproductively isolated enough to become different species). Multicellularity has arisen many times in many different ways. Even when "bigger" was evolving, small members of the same lineages also typically did very well. Tiny reptiles scurried through the giant legs of dinosaurs. Single-celled and other simple organisms are thriving still and may be what is left standing at biological Armageddon.

Being large and complex implies a division of labor, which in turn requires specialized cells or tissue structures that can communicate with each other, but must be kept chemically sequestered so they can be specialized.This means carrying messages and nutrients to them and porting waste and messages from them, which usually means formalizing communication. It means protection against micro- and macro-attack and processing information from the outer world as well as among parts of the inner world. As a result, we see specialization for respiration, digestion, neural control, immune resistance, sensation, and the like (Bonner 1988). A large organism must also be sufficiently supported physically. Water can do this (internally or externally), and large organisms like seaweed or octopuses have very little skeletal structure. But most large creatures, land and sea, plant and animal, have skeletal structures of some kind (and so do most small creatures). Plants have rigid cell walls, and animals have external (arthropods) or internal (vertebrates, some gastropods) skeletons.

We cannot answer the general question of why cells originally began to aggregate, but we can address the question of how that aggregation led to something more than the sum of its parts: a differentiated organism. In this chapter we will first look at some of the general differentiation "strategies" and then at the repertoire of their use in the development of complex organisms. Much of the discussion will relate to morphology, but physiological systems can be similarly characterized in terms of their biochemical pathways.

Differentiation Is Hierarchical

Differentiation is a contingent process, in that what a cell does next depends on its current state. Because of the sequestration of gene expression patterns within cells, and the mitotic inheritance of those patterns, an organism develops a hierarchy of different parts. Cells and their descendants in this sense are partially autonomous, in a nested way, over time. However, they are not completely sequestered, and they may be directed to change their state by circulating signals from headquarters, or from the environment. Hierarchical differentiation means that if cells of a given type are needed in a particular context, the local precursor cells must provide them by division and changing gene expression, and if signaling is to be a part of this process, the cells must be prepared to be able to detect the appropriate signal.

It is a kind of developmental dogma that differentiated cells rarely revert to undifferentiated cells. In vertebrate organ systems, there are beds of stem cells that are committed to a particular organ but have not yet undergone terminal differentiation. When the time is right they differentiate into a stem cell and a descendant cell that undergoes this final state change process. These decisions are made during development but also throughout life as part of homeostasis or response to circumstances. In some cell lineages, like those of most vertebrate neurons, the final differentiation is thought generally to occur only once (mature neurons cannot undergo further mitosis). Some stem cells give rise to the panoply of different types of descendants, as in the generation of the diverse blood cell types. The epithelia that line most vertebrate organs, a layer of stem cells, undergo terminal differentiation repeatedly as needed. For example, cells in crypts between the villi that line the gut continually divide to replace the villus cells as they slough off into the lumen. Figure 9-1 shows an example of hierarchical differentiation of cells that form the central nervous system, a cellular rather than tissue fate-map.

The idea in animals generally is that differentiated cells within an organism have made too many gene expression commitments—often meaning modification of their DNA or its packaging—to go backward. This is analogous to Dollo's "law" for species: evolution does not reverse itself (Marshall et al. 1994). But unlike species, the cells in an animal retain the genome and under some circumstances can reverse their differentiation, whereas once mutations have occurred in a species it can't really regenerate its prior state. The recent engineering of cloned individuals derived from chromosomes obtained from differentiated somatic cells demonstrates that this can be done under artificial conditions (such as somatic chromosomes being placed in enucleated host undifferentiated cells to be stripped of the changes that differentiation had already made in it). Of course, like species, somatically derived chromosomes may bear some somatic mutations, a minor caveat to their

Radial Radial Morphological Ependymal progenitor cell glial cell transformation cells of radial glial cells

Figure 9-1. Cell lineages in the developing central nervous system, showing hierarchical cytological differentiation from stem cells. Modified after (Carlson 1999).

Radial Radial Morphological Ependymal progenitor cell glial cell transformation cells of radial glial cells

Figure 9-1. Cell lineages in the developing central nervous system, showing hierarchical cytological differentiation from stem cells. Modified after (Carlson 1999).

differentiation capabilities for most genes (though not in all, as will be seen in future chapters).

Cloning is one example of a growing realization that animal cells may not be as wholly committed as had been thought, but the generalization about normal context-dependent differentiation, to whatever state, seems generally to hold. And as we also noted in Chapter 8, plants are much less rigidly committed to their cell fates. Nonetheless, the use of hierarchical differentiation to produce sequestered, differentiated cells is a fundamental aspect of the development of complex plants and animals.

Organisms are Built by Duplication and Repetition

William Bateson (Bateson 1894; 1913) wrote extensively and prominently on the problem of reproduction and replication, in particular the meristic, or repetitive, nature so common to biological traits (Carroll et al. 2001; Carroll 2001). As he said (Bateson 1886), "greater or less repetition of various structures is one of the chief factors in the composition of animal forms." We have cited Bateson many times in this book because of his interest in repetitive traits. He coined the term "genetics" and was an early advocate of Mendel's work on heredity, but he did not think repetitive traits could be accounted for in mendelian terms, one of the reasons that he did not think natural selection could explain evolution. We now know that genes can indeed be responsible for such traits, and we are finding the genes, as this chapter will illustrate.

In fact, the living world surrounds us with duplicate structures. Vertebrae and fingers are examples of this, but the same applies to leaves and petals, starfish arms, insect antennae and legs, worm segments, feathers, hair, and scales; branches in nerves and arteries; pancreatic islets, intestinal villi, and taste buds. Examples of meristic structures (that have repeated segmental elements) can be seen in Figure 9-2. Duplication of structures is clearly a fundamental aspect of complex organisms, and we have already stressed the repetitive nature of the underlying genetic and biochemical systems themselves.

Morphogenesis is Stimulated by Inductive Signaling

Because an organism is not just a simple linear array of different structures, but a complex three-dimensional structure assembled from a simple beginning, each component can only begin developing when the proper context exists. Organs are typically built when cells capable of the appropriate differentiation cascade are induced to begin differentiating by some form of context-specific signaling. We introduced examples of this in Chapter 8.

The Spemann organizer is an area on an early vertebrate embryo that patterns axial polarity, and an area between the future mid- and hindbrain is an organizer for brain regionalization. Epithelial-mesenchymal interaction (EMI) involving overlying epithelium and migrating NC cells was mentioned in Chapter 8 as an important characteristic of vertebrates. EMI is responsible for inducing sites where teeth will form along vertebrate jaws; subsequently, local spots of signaling factor (SF) release called enamel knots serve as organizers for cusp development within teeth. These are examples in which the Fgf signal spreads from its source to induce cell growth in surrounding cells.

All the cells in a tissue field may be capable of differentiating in some particular ways, such as to initiate tooth development, but only those that receive the SF signal will be induced to respond. Surrounding areas may then be inhibited from forming the structure. In arthropods, corresponding to vertebrate organizers, the imaginal disk precursors of wings and other structures are organizers for those structures. In plants, cell differentiation of pluripotent meristem cells is induced by cell-cell interactions and hormones in response to external conditions such as day length, temperature, water and so forth. Shoot and root meristems in plants can be viewed as containing organizers, and some of the signaling and indicator genes that maintain the source of stem cells will be described below (e.g., Benfey 1999; Laux 2003;Weigel and Jurgens 2002).

Inductive signals may come from adjacent to distant cells, as communication from one organism to another via pheromones or behavioral cues such as displays, or simply the presence of conspecifics, or from environmental information (olfaction, vision, immunity). The use of terminology for interorganismal communication that

nested repetition of structures in plants; segmental nature of tetrapod limb; structural repetition within an individual organ (primate kidney).

is similar to that for cell-to-cell communication is justified because similar genes or at least information transfer mechanisms are involved.

Position in a Region Can Be Specified by Signaling Gradients

Local differentiation can be specified by concentration gradients of extracellular SFs (in this context known as morphogens). Morphogens can be protein products, such as those from the Hedgehog gene family, or other small, diffusible molecules like retinoic acid. The cells across an otherwise undifferentiated tissue field inter pret their position by sampling the local concentration of the morphogen (see, e.g., Tabata 2001; Teleman et al. 2001; Wolpert 1969; 1981; Wolpert et al. 1998), using appropriate receptor mechanisms. Gradients of concentration can occur if a substance is produced in a localized source and physically diffuses away from it along the tissue.

A concentration gradient can form either simply as a passive ink-in-water process or by the binding of diffusing signal molecules to extracellular receptors removing them from the flow, by being diminished by active degradation by a factor produced at another source, and probably in other ways. Signals may move in various ways, including by passive diffusion or even by being relayed cell to cell (see, e.g., Figure 9-3). Gradients can induce gene expression changes if the signal concentration exceeds a threshold (Figure 9-3A), for example, by binding a large enough number of cell-surface receptors or by leading to the binding of enough REs flanking a target gene to induce its expression.

A classic example of gradient signaling involves establishing anterior-posterior (AP) polarity by the diffusion of transcription factors (TFs) within the early Drosophila egg (Figure 9-3B). Maternally deposited Bicoid mRNA diffuses from the anterior, and Nanos mRNA from the posterior ends (the early stage of the embryo is essentially a single cell with many nuclei on its inner side). Initially, mRNA from the genes Caudal and Hunchback are distributed rather uniformly throughout the egg and are translated in the cytoplasm. At the anterior end Bicoid message is translated, and Bicoid protein diffuses toward the posterior end. As described in Chapter 7, Bicoid protein binds to and inhibits the translation of Caudal mRNA, generating a gradient of increasing Caudal protein posteriorly establishing the first gradient. A corresponding Hunchback protein gradient is established because Bicoid protein binds to REs upstream of Hunchback, inducing expression of that gene. At the same time posteriorly, Hunchback mRNA is bound by Nanos protein, which is concentrated there.

As cell walls develop in the syncytium forming the cellular blastoderm, the intra-cellular environments become more sequestered, autonomous, and precisely controllable. Hunchback, Bicoid, and Caudal proteins are involved in expressing the "gap" genes that are the first zygotic genes to be expressed in an axially segmenting pattern, as the larva develops (of which more below).

A second example of a gradient patterning mechanism controls dorsoventral (DV) patterning. There, Bmp-class SFs (Dpp and Screw) are produced along with the proteins Tolloid in the fly blastoderm (Srinivasan, Rashka et al. 2002) (Figure 9-3C) (terminology and nomenclature can be complicated; Dpp is a fly relative of the Bmp class of TGFfiSF genes, Bmp being the name coined for vertebrate genes). Sog protein diffuses dorsally from a ventrolateral source in the embryo which at this stage consists of a layer of cells surrounding an inner extracellular space. Tolloid (Tol) and Tolkin (Tok) proteins bind and degrade the Sog protein as it diffuses dor-sally, and another protein, Dynamin (Dyn) retrieves undegraded Sog molecules.

Together, these extracellular interactions establish a Sog concentration gradient, decreasing from the lateral source to the dorsal sink for this protein. Since Sog protein binds Bmps so they cannot serve as receptor ligands, and there is more active Sog laterally than dorsally, there is an inverse DV Bmp gradient. Its source is dorsal, and Sog serves as a sink. The resulting Bmp gradient affects gene expression differences that compartmentalize the dorsal embryo, as we will see below in a broader context.

A third example of morphogenic gradients patterns the vertebrate limb. At specified sites along the side of the embryo, diffusible Fgf8 protein induces (or at least is associated with) the initial outgrowth of limb buds. Anterior-posterior (AP) limb polarity is initially established by the induction of the SF Shh (a vertebrate Hedgehog family gene product) in a posterior (caudal, or tailward) part of the future limb known as the zone of proliferating activity (ZPA). ZPA-derived Fgf4 diffuses to induce the asymmetries associated with the posterior side of the limb. The organization of the distal part of the limb begins in the Fgf4-negative part of the limb bud. As the limb develops, regional variation in signaling molecules helps establish combinations of Hox gene expression, in roughly summed-sequential ways, along different limb axes, as will be discussed below.

These are tidy stories, but similar phenomena need not be due to the same mechanism, even within the same animal, even if a similar logical means is used. For example, gradients of Dpp activity are also used in AP patterning of the Drosophila wing imaginal disk. But in this case, unlike DV patterning in the early embryo, the gradients are established not by interaction with degrading proteins, but because a separate protein, called Thick-Veins (Tkv) inhibits the rate of Bmp diffusion away from its source (Srinivasan, Rashka et al. 2002) (Figure 9-3D). It does this because Bmp molecules are tightly binding ligands to Tkv and this stops their further

Figure 9-3. Gradient patterning mechanisms. (A) Multiple concentration thresholds (T1, T2, T3) established across a tissue field affect local cell gene expression, inducing regional differentiation (bottom) of previously identical cells (middle, induction indicated by arrow); (B) AP patterning in the Drosophila syncytium; (C) DV patterning in the fly cellular blastoderm: product of the Sog gene (+) diffuses dorsally intercepting Dpp (gray) diffusing laterally from its dorsal source. Tol, Tok, and Dyn gene products inactivate or intercept Sog molecules dorsally; (D) AP patterning in the Drosophila wing imaginal disk: Dpp (gray) gradient spreads asymmetrically from a source because the concentration of its ligand Tkv (+) is greater posteriorly; cell-to-cell movement of Dpp is aided by Dyn, not shown. For C & D see (Srinivasan et al. 2002).

Figure 9-3. Gradient patterning mechanisms. (A) Multiple concentration thresholds (T1, T2, T3) established across a tissue field affect local cell gene expression, inducing regional differentiation (bottom) of previously identical cells (middle, induction indicated by arrow); (B) AP patterning in the Drosophila syncytium; (C) DV patterning in the fly cellular blastoderm: product of the Sog gene (+) diffuses dorsally intercepting Dpp (gray) diffusing laterally from its dorsal source. Tol, Tok, and Dyn gene products inactivate or intercept Sog molecules dorsally; (D) AP patterning in the Drosophila wing imaginal disk: Dpp (gray) gradient spreads asymmetrically from a source because the concentration of its ligand Tkv (+) is greater posteriorly; cell-to-cell movement of Dpp is aided by Dyn, not shown. For C & D see (Srinivasan et al. 2002).

Oocyte mRNAs hunchback mRNA

hunchback mRNA

Oocyte mRNAs

Early cleavage embryo proteins

Posterior B

Posterior B

0 0

Post a comment