Thinking of Biology cell - Oxford Academic

32185_749_1_59.pdf

Thinking of Biology

Toward a theory of cellularity-Speculations on

the nature of the living cell

The modes in which the earlier truths

of science are preserved in its late forms, are indeed various. From being asserted at first as strange discoveries, such truths come at last to be implied as almost self-evident axioms. (Whewell 1872, p. 45) hy do cells exist? Is there an obligate relationship between life and cellular- ity? Certainly there is no exception to the observation that all life on earth is cellular at some stage. Even viruses depend on the machinery of the host cell to complete their life cycles. Evolutionary theory, too, is rooted in the realization that all cells come from pre-existing cells (Virchow

1858), so it is not surprising that the

concept of cellularity lies at the cen- ter of biological thinking. But to what extent do current concepts of the cell limit our ability to understand its role in organismal development?

Biological knowledge is a mosaic

of general concepts or theories, em- pirically derived observations or facts, and a supporting vocabulary.

Although much of this knowledge is

of relatively recent origin (in par- ticular the details of cellular metabo- lism and heredity), many of the con- cepts date back two and three centuries ago, to the first scientific inquiries into the nature of living things. One may suspect, therefore, that what we take for understanding

may sometimes be an illusion pro- duced by the accumulation of layers of terminology, each built on previ- ous usage and therefore circum-

scribed by it.

The word "cell," for instance,

harks back to Robert Hooke's (1665) first use of the term to describe the empty compartments seen in cork (Figure 1). Later, as cells were found in all living things, the word came to by Philip M. Lintilhac embody the universal and fundamen- tal unit of life, which in a modern vernacular might be called the "quan- tum of life" (Sitte 1992). For today's biologist, the word "cell" carries a complex and far-reaching connota- tion, conjuring images of a unit of

cytoplasm that contains a nucleus and a variety of subcellular organelles and is surrounded by a membrane

(Figure 2). In the end, however, al-

though the concept of the cell brings to mind the myriad coupled processes and structures by which the cell con-

trols its own activities, responds to external signals, and maintains conti- nuity between successive generations, we still regard the cell as simply a complex microscopic object. There- fore, our ability to move to a new

level of understanding of the cell and of cellularity may be constrained by the vision that we have inherited from

our predecessors. That there are limits to the under- standing of cellularity is evident from the fact that there is no formal disci- pline of cell theory. The treatment of

the cell in biology courses is largely confined to detailing what goes on within cells and avoids any attempt to infer unifying general principles of cellularity. A theory of the cell

that embodies such unifying prin- ciples might tell us why cells exist at all, why some organisms appear to be released from the constraints of cellularity at certain stages in their life cycles, and what the roles of the cell in multicellular development are

and how these roles evolved. In this article, I begin to construct a model of the cell that provides a

basis for a conceptual approach to cellular interactions in multicellular organisms. I confine my discussion to plant cells and tissues for two reasons. First, questions of nearest neighbor relationships are more eas- ily and unequivocally resolved in plants than in animals because plant cells do not move with respect to each other after division and because differentiation in plant tissues tends to proceed sequentially, on a cell-by- cell basis. Second, I believe that when considered against the full range of developmental mechanisms that have evolved in plants and animals, the devices available to plants represent a more elemental stage in the evolution of the multicellular habit of growth. It therefore behooves us to model the simpler case before proceeding to discuss the more complex one. I have little doubt that the basic proposi- tions that I suggest here may apply to animals as well as to plants, but they may be overlaid by mechanisms of higher order or promoted to domains of control that extend beyond the level of the individual cell. Indeed, these higher-order mechanisms may apply to plants too. However, be- cause of the immobility of the cells and the sequential mode of growth typical of plants, the simpler modes of differentiation and intercellular specialization still remain visible.

Concepts of the cell

The language that we use needs to

reflect and enable the work at hand.

How can we

begin to formalize the relationship between a cell and its environment? A brief inventory re- veals that a spectrum of such relation-

ships is possible. A single, free-living cell "sees" the external world directly across its plasma membrane-a world

of resources and threats over which it has little if any control. It responds to a variety of incoming signals directly, recognizing and processing substrate molecules as they approach and pass through the plasmalemmal interface.

It is functionally self-reliant, requir-

ing the presence of no other cells to remain viable.

For a cell that is part of a multi-

cellular organism, however, the view

January 1999 59

University of California Press

is collaborating with JSTOR to digitize, preserve, and extend access to

BioScience

www.jstor.org

®Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

Figure 1. The earliest con-

cept of the cell, showing cellular compartments in cork. (After Hooke 1665.) across the cell bound- ary is made up of other similar, if not identi- cal, cells (Figure 3). The external world may be evident only as distant signals that have passed through the filter of surrounding tissues, or have been transduced into another form and relayed to the local cell.

The cell acts in concert

with its cohorts and neighbors to stabilize the internal environ- ment and to provide consistent resource availability at the local level. Furthermore, different cells and groups of cells may exhibit functional specialization due to a division of labor within the organ- ism. Viability and self-sufficiency are thus inherent in the whole organism rather than in the individual cell.

To understand relationships at

work in the environment of a cell in a multicellular organism, it helps to assign entities and functions to dif- ferent compartments, such as "soil," "rhizosphere," "atmosphere," "rest- of-the-organism," or "other organs."

Similarly, the internal workings of

the cell are commonly dealt with in terms of intracellular compartments vlmttllR tt fr /,<|(lQiS t BI; Gr \'aa 1Ki \%^^^g8, such as "mitochondria," "chloro- plasts," "cytoplasm," or "nucleus."

Whatever the nature of the cellular

environment, however, it is the cor- respondence between events occur- ring within the cell and events taking place outside of the cell that consti- tutes the essential commerce of cel- lularity. How can the cell be defined in a way that usefully reflects this com- merce? I begin by suggesting that the cell is more than just a class of ob- jects that can be observed under the microscope. To study the cell as a discrete object, a self-contained regu- latory unit, or a volume of complex biochemical processes is to ignore half of its nature. Upon reflection, it is clear that there can be no cell without an exter- nal environment, which leads directly to the following concept ^1" \ of the cell: The cell is a boundary that estab- lishes a relationship 4 ~ between an internal universe and an exter- nal universe.

1 / According to this

definition, the internal (j) ?0/ universe includes the

Figure 2. The modern con-

cept of the eukaryotic cell: an organelle-filled, mem- brane-bounded compart- ment. sum of all the information systems, biochemical pathways, and activi- ties occurring within the boundary of the cell. The external universe is defined as the sum of all the biotic and abiotic factors and activities oc- curring outside the boundary of the cell. This broad definition includes, within the internal universe of the cell, all minor housekeeping func- tions, sequestered materials, inac- tive pathways, and redundant con- trol systems, whether or not they contribute to the performance of the cell in its environment. Likewise, the external universe includes the entire universe of remote events and ob- jects, most of which have no bearing on the life of the individual cell. Such an all-inclusive definition is so broad that it is not useful.

To make this definition of the cell

more tractable, the description of the external universe can be restricted so that it does not have to encompass the entire external world but includes only that part of it that the cell sees directly across its boundary. This limited definition of the cell's envi- ronment can be called the "ex- ternum." In this way, the entire uni- verse of external influences can be legitimately represented as a constel- lation of inputs at the cell surface. In the case of an internal cell in a mul- ticellular organism, these inputs re- sult from the percolation of distant signals through the tissues of the organism, as well as from the imme- diate inputs of more proximal cellu- lar neighbors that interface directly on the cell and make up part of its externum. Likewise, and perhaps more surprising, a description of the internal universe of the cell does not have to include all of the biochemi- cal and metabolic pathways and

structures that occur within the cell, nor does it have to detail genetic information or hereditary mechan-

ics. It need only describe the net effect of these activities at the cell surface. This limited definition of the internal universe of the cell can be called the "internum."

A thought experiment. To illustrate

this restricted definition of internum and externum, consider the conse- quences of a thought experiment in which we remove a cell from within a living tissue and replace it with a

BioScience Vol. 49 No. 1 60 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

tiny automaton. The automaton is carefully constructed to be, in all respects, a faithfully functioning stand-in for the living cell that it replaces. Clearly, if the automaton is constructed with sufficient skill and ingenuity, it will behave exactly like a real cell: processing substrate, react- ing to external stimuli, and even divid- ing. It will, in fact, perform seamlessly in the biology of the tissue as a whole and consequently will be seen by its neighbors as just another cell. What are the minimum requirements for such an automaton if it is to perform the same as the original cell in the context of the surrounding tissue?

As long as it behaves superficially

exactly like a real cell, it will play a normal role in the development of the organism. That is, as long as the automaton "looks" like a real cell to its neighbors, it does not matter what contrivance we put inside it to achieve that end (Figure 4). Should we conclude from this metaphorical experiment that the inner workings of the cell are irrel- evant to an understanding of cellu- larity? They are certainly not irrel- evant to an understanding of the biology of the organism as a whole.

But for the purpose of elaborating a

theory of the cell and of cellular interaction, ignoring the details of cellular metabolism would seem to

be a potentially powerful simplifica- tion from which it might be possible to formulate a basic equation of cel-

lularity that relates the inner work- ings of the cell to its environment.

Therefore, in this simplified view

the cell can be defined as the internum, the externum, and the sur- face that separates them. The exter- nal face of the plasmalemma, be- cause it is the logical candidate for the cell boundary, is thus a two-way surface-it is not simply the outer surface of the cell but also the inner surface of the cell's environment.

The plasmalemma is, in fact, the

surface on which the internal and external universes of the cell are pro- jected, as internum and externum, respectively. The challenge is to dis- cover the rules that govern the corre- spondence between these two pro- jected universes.

When we regard the cell in this

way, it becomes clear that the living

Figure 3. The cell embed-

ded in its universe of other cells and environmental signals.

be understood in isola- / tion. Rather, the cell must be studied as an interface phenomenon, such that the cell mem- brane provides the de-

fining surface on which V patterns of internal and o external complexity are projected. It is also the \ : . surface on which the il- lusion of cellularity is projected by our au- tomaton, the illusion then being either ac- cepted or rejected by the surrounding cells. One can propose, then, that the relation- ship between internum and externum is the fundamental relationship of

cellularity and that it defines the cell not simply as the boundary itself, but as the correspondence implied in

the projection of internum and externum on each other at that boundary. This view of cellularity simplifies the task of describing the external environment of the individual cell and frees us from having to specify every detail of the internal physiol-

ogy and genetic structure of the cell before attempting to model interac- tions of the cell with its environ- ment. It also refocuses the concept of

stimulus as a change in the pattern of cell is not a thing in itself and cannot an environmental signal at the externum. Similarly, cellular behav- ior can be defined as a change in the pattern of cellular response mani- festing at the internum.

Sands' principle of equivalence.1 It

should be clear from this discussion that the internum and externum can represent the projections of two very different worlds. In the case of a

single free-living cell, for instance, the externum represents the projec- tion of a more or less dilute, and for

the most part poorly coupled, as- semblage of mineral ions, substrate molecules, waste products, and chemical or physical signals. The internum, by contrast, represents the projection of a concen-

1I coined the term "Sands'

Principle" to denote the gen- eral notion that, in some sense, the internal world of the cell must be equal to its external world. To my knowledge this idea was first stated by David C. Sands in

October 1968 over a glass of

beer at La Val's Cantina in

Berkeley, California.

Figure 4. The metaphor of

M / the "virtual cell." Just as a

dancer behind a screen is indistinguishable from a projected image (center), so an automaton that projects a perfectly cell-like repre- sentation on its internum surface is indistinguishable from a "real" cell.

January 1999 61 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

trated and highly coupled biochemi- cal engine (the automaton), a com- plex system whose description re- quires the synthesis of the most rarified concepts and terminologies (Kauffman 1993). At the same time, it is apparent that there must be a reconciliation between the two at the cell surface so that irrespective of differences in complexity and order between processes at work in the interior of a cell and those at work in its environment, the two are finally brought into some kind of equiva- lence at the boundary. This reconciliation between inter- num and externum would seem to be one of the prerequisites for the viabil- ity of the system. An incoming signal whose appearance at the externum is not reflected in some reconfiguration of the internum is nonexistent as far as the cell is concerned and can be ignored. I have termed this corre- spondence Sands' Principle of Equiva- lence (Figure 5), which implies that some form of equivalence or comple- mentarity between internum and externum is a sine qua non of the

living state. The ultimate challenge is therefore to define the compo- nents of this equivalence, thereby

establishing a basis for modeling and characterizing the interactions be- tween the cell and its environment, the cell and neighboring cells, and the cell and the whole organism.

Differentiation. As biologists, we

have inherited another terminology from our standard bearers-that as- sociated with the idea of "differen- tiation." This term encompasses some of the most uni- versally accepted notions in all of developmental biology, and yet it is one that is often used without be- ing clearly defined. What is cellu- lar differentia- tion? A quick an- swer is that it is any process by which a cell changes in some respect and becomes different. . But now we must ask: different from a what? By differen- tiation, do we

simply mean that a cell becomes dif- ferent from what it was-namely, some "undiffer- Figure 6. Autodiffere entiated" starting in which two vital cc condition? Or do (shaded and open), ar we mean that two level and a thick, resi

initially identical becomes relatively in cells diverge in num and externum i; their capabilities so that they become different from each other? Clearly we can mean either, and yet it is also clear that these two alternatives can imply quite differ- ent cellular processes. In fact, there are three basic types of differen- tiation, each of which / has different implica- tions in terms of cellu- lar process (Gurdon , \ 1974). * Autodifferentiation.

Perhaps the most basic

kind of differentiation is that which is charac- teristic of the prokary- Figure 5. Sands' Principle. The correspondence be- tween internum and ex- ternum is absolute. (See text for details.)

ntiation. The formation of a resistant spore ellular functions, represented by two boxes re down-regulated to the minimal sustainable

stant spore wall is deposited. The boundary npermeable. Correspondence between inter- s reduced. otes, in which a cell differentiates to form a resistant structure that is ca- pable of surviving adverse conditions or serving as a propagule that can be carried some distance without the need for a substrate. In this type of differentiation, a single cell, initially fully active metabolically, secretes a resistant wall, accumulates storage products, slows its cellular processes, and essentially shuts itself off from the outside world (Figure 6). In this state of dormancy, it is clearly differ- ent from what it was at the start, and so the term "differentiation" may be appropriate. In a more general sense, the term "autodifferentiation" im- plies a kind of differentiation that can proceed without the presence of any other cell. * Eudifferentiation. Now consider the case of two adjoining cells, ini- tially derived from the same mer- istematic or blastemic cell line, which gradually diverge in function so that

BioScience Vol. 49 No. 1

v

62 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

Figure 7. Eudifferentiation. Cell specialization is achieved through communication and negotiated division of labor. At least two cells are required. The boundary remains fully permeable. Cor-

respondence between internum and externum is fully enabled. Figure 8. Teleodifferentiation (apoptosis). Contraction of cell function proceeds until cytoplasmic function ceases and the cell contents are consumed. The boundary ceases to exist. Internum and externum lose their meaning. they become different specialized cells. Neither is dormant or dead, but the two function together in a complementary way to maintain a viable whole (Figure 7). This kind of differentiation is distinguished by the fact that it requires the participation of at least two cells. * Teleodifferentiation. In this third type of differentiation, the cell con- tributes only its nonliving skeleton to the mature organism (Figure 8).

This terminal self-sacrifice is a form

of apoptosis or programmed cell death. Clearly, these three examples rep- resent quite distinct processes, and yet they are all commonly under- stood to qualify as legitimate ex- amples of differentiation. In the case of autodifferentiation, for which the ready example is bacterial sporula- tion, the cell erects a barrier to ex- change with the external world, selec- tively restricting the number of functions represented at the externum and internum, and thereby restrict- ing the flow of information and ma- terials. This kind of differentiation, which is usually triggered by some environmental change, can, at least in principle, proceed autonomously, without input from neighboring cells.

At the other extreme, in the case

of eudifferentiation, the two cells maintain open boundaries but relin- quish functions to each other so that in the final differentiated state they are completely dependent on each other; neither maintains all of the functions necessary for viability, but together they complement one an- other and remain viable. Perhaps the best example of eudifferentiation is that quintessentially botanical ex-

ample of differentiation, the phloem sieve tube-companion cell complex. At maturity, the sieve tube is a highly

specialized, anucleate conductive cell that is closely attended by a nor- mally nucleate companion cell, which presumably provides the nuclear function. The two cells have differ- entiated from one another, and both are active metabolically, although in specialized ways. The sieve tube is dependent on the companion cell for nuclear products, and the compan- ion cell is dependant on the sieve tube for nutrition. No impermeable wall is formed between them.

In the case of teleodifferentiation,

a good example is that of a water- conducting cell, that is, a xylem ves- sel element. The xylem vessel ele- ment starts as a fully active metabolic cell, derived from cambium, but it subsequently undergoes a controlled process of growth and enlargement, elaborating a specialized and usually thick extracellular wall. It then dies before it achieves its final functional role as a tubular element in an exten- sive water-conducting system.

Focusing for a moment on the

process of eudifferentiation, in which two identical cells gradually diverge in function, becoming specialized and therefore interdependent, it is ap- parent that the very first steps in this

January 1999 63 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

process, beginning with two virtu- ally identical postmitotic cells, must involve a mutually negotiated stepwise relinquishing of essential function. This negotiation means that the down-regulation of a particular vital function by one cell of a eudifferentiating pair must be ac- companied by the up-regulation of that same function by its partner. Without such communication, dif- ferentiating cells risk being left with- out a full complement of necessary functions, and the viability of the individual cells will be compromised. Therefore, eudifferentiation requires that informational channels between the two cells remain open and that communication be ongoing. It fur- ther implies that eudifferentiation is analogous to a contractual negotia- tion: A series of agreements must be struck between the two differentiat- ing cells, ensuring that neither one relinquishes vital functions in an untimely manner.

In the current literature on the

topic of differentiation, the example of bacterial sporulation is frequently used as a model for eukaryotic dif-

ferentiation (Losick 1995). However, this analogy would appear to be lim- ited, because, like other forms of autodifferentiation, sporulation is achieved at the cost of communica- tion with the external world, so that the way in which the necessities of

Sands' Principle are met by a sporu-

lating cell may be different from the

way in which they would be met by a eudifferentiating cell. Although more complex forms of prokaryotic differentiation may be possible, it is

clear that eudifferentiation, besides being intrinsically impossible in single-cell systems, requires more formal intercellular communication protocols and more sophisticated controls over cellular metabolism and intracellular resource management.

Thus, eudifferentiation would ap-

pear to be characteristic of multicel- lular eukaryotic organisms, occur- ring to only a very limited extent in the prokaryotes. The most spectacular example of intercellular dependency-namely, that between cellular endosymbionts (chloroplasts and mitochondria) and their host cells-is also strikingly different from eudifferentiation. Al- though these endosymbiotic relation- ships represent a high degree of in- terdependency and demonstrate true functional specialization, they are not capable of arising de novo from originally identical partners. In other words, the process by which this interdependency was established is not ongoing; instead, the basic rela- tionship was established hundreds of millions of years ago and is main- tained in the continued dependency between host cell and symbiont. The merits of eudifferentiation then be- come apparent in that the functional relationships necessary for the more complex types of biological develop- ment do not have to be carried for- ward from generation to generation in the form of a cytoplasmically inher- ited homunculus (i.e., the proplastid or promitochondrion) that contains the seeds of the differentiated cell types. The necessary relationships can instead arise epigenetically, provid- ing that the differentiating partners are appropriately equipped and free to communicate with each other.

Understanding the cell

is different from understanding what goes on inside the cell

In the case of terminal differentia-

tion, or teleodifferentiation, the cell may or may not maintain open boundaries, but it commits all of its resources to the construction of a complex cell wall. In the process, the cell consumes itself, destroying the boundary. In this case, as with eudifferentiation and autodifferenti- ation, the end result is clearly a dif- ferentiated structure, but the process terminates the living entity itself, a sacrifice that stretches the meaning of the word differentiation in a new direction.

Most differentiating systems are a

blend of the three basic types. In vessel element differentiation, for instance, there may be a period of eudifferentiation of the cambial de-

rivatives, which is followed by a pe- riod of autodifferentiation that pre- cedes teleodifferentiation. Many

examples of eudifferentiation may also blend in an element of the exclusion that is characteristic of auto- differentiation, such as the differentia- tion of thick-walled idioblastic cells in plants. In general, most examples of eukaryotic differentiation show some elements of eudifferentiation, whereas differentiation in the prokaryotes tends to conform more strictly to the autodifferentiation model.

What distinguishes the three cel-

lular processes that are included un- der the rubric of differentiation is that each is characterized by a par- ticular behavior at the cell bound- ary. In autodifferentiation, the boundary itself changes and becomes increasingly impermeable, so that communication with the externum is reduced. Fewer and fewer functions correspond across the boundary, and the cell becomes insensitive to many environmental changes. In eudif- ferentiation, the boundary remains open and the cells remain viable, even though they have relinquished essential functions to each other. In- dividual cells become increasingly invested in their environments and neighbors. In teleodifferentiation, a period of auto- or eudifferentiation, either singly or in combination, is followed by a period during which the cell consumes itself, so that the boundary that defines internum and externum eventually ceases to exist and the living cell comes to an end, leaving only a nonliving extracellu- lar skeleton in its place. The significance of these distinc- tions lies in the fact that differences in

the mechanics of differentiation imply differences in behavior and, ultimately, in the complexity and interconnect-

edness of the structures produced. Eudifferentiation may be a paradigm for all multicellular differentiation

in that the controlled assignment of what would normally be internal cellular functions to external enti- ties, whether they are neighboring cells or distant organs, must entail a similar sequence of "agreements" between internum and externum.

In all three forms of differentia-

tion, the requirements of Sands' Prin- ciple must be met. Internum and externum must be reconciled at the cell surface, whatever the specialized nature of the cell, and at the same time essential functions must be pro- vided for as long as the cell is alive.

BioScience Vol. 49 No. 1 64 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

One consequence of the notion

that internally generated behaviors and externally generated stimuli are brought into correspondence at the cell boundary is that the cell repre- sents an equilibrium between internum and externum. This notion distills the fundamental postulate

deriving from Sands' Principle: Internum and externum have equal roles in the determination of cellular

behavior. Differentiation and the problem of resource management

A key question that confronts us in

trying to understand cellular eudif- ferentiation is the nature of the rela- tionship between cellular resource management and the negotiation of contracts between eudifferentiating cells, that is, between individual cell interna and their externa. In other words, what does "contractual ne- gotiation" mean in terms of Sands'

Principle and the allocation of cellu-

lar resources?

First, two cells must remain essen-

tially identical as long as their externa remain identical. Because the cells are clonal, there is no internal com- mand that can be invoked by one that will not necessarily be invoked simultaneously by the other. Conse- quently, the primary event in eudifferentiation must always be an external signal that affects the two cells differently, causing one to "see" an externum different from that seen by its companion. In plant tissues, such an external signal might be mechanical, resulting from differen- tial growth stresses and transmitted from cell to cell through the continu- ous cell wall system, finally resolv- ing itself as different stress mechani- cal environments at the locations of the two target cells.

One can visualize, however, that

as soon as such a signal arrives at the pair of cells, a cascade of events

occurs that results in a multiplier effect. The first cell will interpret and respond to the change in the

externum with a change in its own internum. This change in the in- ternum will be projected outward, where it will result in a change in the externum of the neighboring cell.

The second cell will then respond

with a change in its own internum, which will then be resubmitted to the first cell. Each iteration of the process is akin to a proposal that is submitted, reviewed, and resubmit- ted. This incremental negotiation of mutually acceptable changes eventu- ally cascades into the final meta- stable, interdependent, differentiated end state.

The net result of this process is

that where interdependencies arise between cells due to eudifferentiation and functional specialization, differ- ent subroutines in effect become con- tractually assigned to separate cells.

The term "hypercellular" can then

be used to describe functions that are provided by one cell in excess of its own needs and that therefore come to be overrepresented at its surface.

These hypercellular functions are

therefore available for one or more neighbors. Similarly, the term "hypo- cellular" could be applied to cellular functions that have contracted to the point at which they are no longer ca- pable of supporting even a single cell and must therefore be supplemented by outside sources and made avail- able at the cell's externum. The terms hypercellular and hypocellular are the cellular analogies of biochemical up-regulation and down-regulation.

The logic of multicellularity. These

seemingly necessary relationships can be used to construct a basic logic of multicellularity. For instance, it would follow from the above discus- sion that in a two-cell eudifferenti- ating system, when one cell has be- come hypocellular with regard to a particular function, the other cell must be hypercellular with respect to that same function for the pair to remain viable. Similarly, in a viable three-cell system in which the summed interna of any two cells are hypocellular for a given function, the makeup of the internum of the third can be inferred to be hyper- cellular with respect to that func- tion. And so, too, in multicellular systems, tissue- and organ-wide de- ficiencies must in the end be rec- onciled at the level of individual cell interna and externa. Finally, what is the relationship between growth (meaning the con- struction of new cytoplasmic compo- nents and cells) and differentitiion?

Clearly, in the case of teleodifferenti-

ation, further growth becomes im- possible as the cell approaches its terminal end state and disappears.

Similarly, in those examples of

autodifferentiation in which the dif- ferentiated end state is a dormant cell or resistant spore, the process of differentiation eliminates any possi- bility of growth, not simply because the dormant cell cannot grow by virtue of physical restraint, but be- cause fully enabled anabolic pro- cesses must be reflected in full corre- spondence across the cell boundary.

What, then, of eudifferentiation and

growth?

The growth dilemma. In general, for

prokaryotes and eukaryotes alike, abundant substrate and favorable conditions will up-regulate metabo- lism and propel the cells into growth mode. But for eukaryotic cells, gen- eral metabolic up-regulation would appear to be intrinsically at odds with differentiation because, on the one hand, growth and division in- volves the replication of the entire machinery of the cell, and, on the other hand, eudifferentation neces- sarily results in selective representa- tion of some functions: hypocellular (down-regulated) in some cases, and hypercellular (up-regulated) in others.

How, then, can eudifferentiation

occur in a substrate-rich environ- ment? That is, how can contractual relationships between cells be ar- ranged when the cells in question are committed to rapid logarithmic growth (Figure 9)? For a cell to en- gage in stepwise abdication of cellu- lar functions to surrounding entities, the commitment to maximum repro- ductive capacity must be broken.

Eudifferentiating cells must there-

fore find some way to slow anabolic metabolism and growth while main- taining open boundaries, even in a substrate-rich environment. This is not to say that cell growth and divi- sion and eudifferentiation are neces- sarily mutually exclusive. Procam- bial cells, for example, appear to be able to divide and differentiate at the same time. What do appear to be mutually exclusive, however, are fully enabled exponential growth and eudifferentiation.

Most eudifferentiating cells divide

rarely, if at all, which raises the fol- lowing question: Which came first,

January 1999 65 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

I\ / /\ ////\ // / \ /\

A, AC- I A XIPAA X

ELI ELILIl moi E ZTI:] L I

A A A A A A A A A A A A

A A A A A/A AO n AAAA AAA

?C3 ~mn mo 31 1Sr 13 mm B~ 130 .- .n I a a a 0 0M 0 t E 3so

3e3CTjaE3BBBBBBB BaSSU CDsX tB I _ a I Bmml 0 r iaBBm B DEa

Figure 9. Eudifferentiation and rapid exponential growth (the growth dilemma). Is there a basic conflict between the subtle demands of eudifferentiation and the necessities of rapid cell proliferation?

the ability to down-regulate cellular metabolism in a substrate-rich envi- ronment or the ability to enter into complex intercellular separations of function? Part of my hypothesis is that the two are inseparably inter- twined and that the key to eukary- otic eudifferentation therefore lies in the ability to uncouple the linkage between substrate abundance and rapid exponential growth while maintaining the open boundary con-

dition necessary for the development of true intercellular dependencies. Apart from the universal cellular attribute of growth and division,

eukaryotic cells would appear to have three ways to deal with excess en- ergy in times of substrate abundance.

First, energy can be sequestered as

starch, glycogen, or other forms that remove it from the active energy bud- get of the cell. Second, energy can be rejected at the cell boundary by some mechanism that causes incoming pathways to close. Third, energy and materials can be burned off by means of catabolic mechanisms that con- sume cellular resources without con- verting them into new protoplasm. This last mechanism brings up an interesting question: Did cellular motility (cyclosis) arise not as a means of locomotion but as a cellular mechanism for fine-tuning the cell's energy budget? Such a mechanism could allow the cell, even in times of substrate abundance, to free itself from the necessities of endless repro- ductive growth and begin to indulge in the delicate negotiations that must be part of eudifferentiation and that are the hallmark of eukaryotic growth.

The evolution of differentiation. It

makes sense to place the differentia- tion process itself in an evolutionary context. In simple cells, we find the simplest kinds of differentiative changes. Plant cells afford the op- portunity to easily observe eukary- otic differentiation because cell lin- eages are preserved in the patterns of tissues. In plants, then, in which cells maintain their relative positions to one another, we can still see the expression of eudifferentiation, in which the negotiation of individual cell differences has proceeded on a cell-by-cell basis. In animals, in which cell movement plays a major role in embryonic development, the role of individual bilateral or multilateral contract negotiations may be super- seded by the ability of cells to place themselves in different environments by moving to different places. Fur- thermore, the whole question of func- tional specialization may be trans- ferred to the tissue and organ level rather than proceeding on a cell-by- cell basis.

Any meaningful discussion of cel-

lularity and its role in functional differentiation will eventually have to encompass a variety of solutions.

But in plant cellular differentiation,

it may be possible to study an inter- mediate stage in the evolution of differentiative mechanisms that lies somewhere between the simpler types of cellular change available to the prokaryotes and the more widely coordinated and complex changes available to animals.

BioScience Vol. 49 No. 1 66 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

Conclusion

As to the notion of cellularity and

the suggestion that it has only mini-

mal significance in any general theory of plant development (Kaplan 1992), I propose that the seeming lack of

relevance of individual cellular func- tion to the overall pattern of organismal development is an arti- fact resulting from our incomplete understanding of the necessities of cell-cell interactions. Furthermore, this gap in our understanding of the contractual nature of cell interac- tions is in part a consequence of an inappropriate definition of the cell, which has led to a very limited view of the cell as a structural object- that is, as an encoded instruction set within a microscopic bag of enzymes that embodies a complex biochemi- cal processor. When we define the basic unit of life not as the individual histologi- cally identifiable cell but as the boundary or "event surface"2 that separates internum from externum (Figure 10), then the concept of the cell can expand or contract to in- clude syncytial organisms, solitary unicells, differentiated cells within a multicellular domain, or, indeed, whole organs wherein individual cell boundaries have broken down, giv- ing way to larger units of reactivity with their own boundaries. Such su- per cells of specialized function have interna that collectively negotiate a common relationship with the rest of the organism and externa that represent a genuine projection sur- face on which the interests of the rest of the organism are recorded. The obligate relationship between life and cellularity thus resides in the neces- sary requirement for a clearly de- fined event surface, rather than in the contents of the cell, the struc- tural attributes of the extracellular matrix, or the size constraints that cellularity imposes on the internal workings of the cell.

The power of Sands' Principle re-

sides in its ability to reduce the entire universe of external influences to a

2The term "event surface" is meant to evoke the

concept that the essential defining feature of the cell is not the machinery that is carried within it, but the boundary at which internal and external events are necessarily reconciled with each other. Figure 10. The cell as an event surface. The irreduc- ible element in cellularity is the boundary itself. constellation of stimuli . ' that are represented at '..: the cell externum. Simi- * larly, the complex world . . of the cell interior is rep- * ' f resented at the cell sur- . face by the limited con- ' . a stellation of factors that constitute the inter- num, whose spatial and * ' ' temporal pattern can be* . ':: interpreted as cellular responses or behaviors.

Sands' Principle makes

it possible to contem- plate expressing the re- lationship of a cell to its environment in a single equation.

Cellularity in all its manifestations

then becomes a set of formal rela- tionships between locally defined interna and externa. It has been shown that properties such as shape (Green 1984), mechani- cal stress pattern (Lintilhac 1974), and kinematic flow (Silk and Erickson

1978), which emerge from whole

organ structure, can be derived from a summation of individual cell char- acteristics. These properties may con- stitute a form of genuine organismal information. Thus, organismal con- trol of the developmental process is real. These organismal factors may appear to have organ-wide organo- genetic consequences that are inde- pendent of individual cell behavior, but they have their effect by means of coordinated chemical, mechani- cal, or electrical signals that appear in some form at individual cell externa. The broad similarity of the target cells, plus the multicellular nature of the eliciting stimulus, en- train whole regions of cells into simi- lar behavior patterns.

Therefore, given the conceptual model and associated structure pro- vided by a reconsideration of the

nature of the cell and of cellularity, the cell is defined as a balanced rela- tionship between an internum and an externum that complement each other across a common boundary.

Increasingly complex multicellular

differentiated structures are built up by codifying intercellular relation- ships in terms of contractual agree- ments that can, if necessary, be ex- tended over large multicellular do- mains, eventually emerging as large-scale organismal influences that can again percolate down to the level of the single-cell externum, where they influence individual cell behavior.

Lewis Wolpert has raised the ques-

tion of whether we are approaching the end of the major discovery era of developmental biology, which is to say, have all the basic principles of developmental biology been de- scribed? Given what we know in principle, then, is it possible to "com- pute" the developmental fate of an egg cell (Wolpert 1994)? I believe that there are whole areas of multi- cellular interaction for which no good theoretical framework exists on which to hang our extensive descrip- tive capabilities; therefore, we can- not at present contemplate "com- puting the egg." The underlying logic of cellularity is still a mystery, a fact that reveals itself in the nature of the questions that are still being raised about the role of the cell in develop- ment (Kaplan and Hagemann 1991,

Kaplan 1992). One of the immediate needs of

developmental biologists is to for- malize the relationship between what

I have called internum and externum

in terms of a basic equation of cellu- larity that can be built into a model of the most elemental two-cell dif- ferentiating systems and then ex- tended into more complex models of

January 1999 67 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023

multicellular organization. The ap- parent conflict between the cellular and organismal points of view will then evaporate in what will be a general theory of cellularity and mul- ticellularity. In this article, I have attempted to present a stripped-down view of cel- lularity, in which compartments on both sides of the boundary disap- pear from view so that we are deal- ing only with the end result of cellu- lar activities as they manifest in patterns of boundary interactions.

Perhaps this simplification can be

used to develop a symbolic charac- terization of cellular activities.

Although I cannot claim to have

succeeded in developing a truly rig- orous framework for understanding cellularity, I believe it is a worthy goal that holds the promise of relat- ing all cell types in a precise tax- onomy of relationships that inte- grates individual cell equilibria and higher-order boundary relationships throughout the organism to produce the emergent characteristics that we see as organismal properties. Thus, as developmental biologists we can envision a theory of cellularity that distinguishes prokaryotes and eu- karyotes on the basis of their func- tional characteristics while resolv- ing intermediate forms and that en- compasses all eukaryotic organisms, from giant characean algal cells to the most complex differentiated mul- ticellular systems.

Acknowledgments

I would like to thank the National

Aeronautics and Space Administration

for research support during the time when I was writing this article, and

Krista Hassert for the illustrations.

References cited

Green P. 1984. Analysis of axis extension.

Pages 53-75 in Barlow PW, Carr DJ, eds.

Positional Controls in Plant Development.

Cambridge (UK): Cambridge University Press.

Gurdon JB. 1974. The Control of Gene Ex-

pression in Animal Development. Cam- bridge (MA): Harvard University Press.

Hooke R. 1665. Micrographia: Or Some

Physiological Descriptions of Minute Bod-

ies Made by Magnifying Glasses with Observations and Inquiries Thereupon. London: Royal Society of London.

Kaplan DR. 1992. The relationship of cells to

organisms in plants: Problem and impli- cations of an organismal perspective. In- ternational Journal of Plant Science (Spe- cial Issue: Katherine Esau International

Symposium) S153: 28-37.

Kaplan DR, Hagemann W. 1991. The rela-

tionship of cell and organism in vascular plants. BioScience 41: 693-703.

Kauffman S. 1993. The Origins of Order:

Self-Organization and Selection in Evolu-

tion. New York: Oxford University Press.

Lintilhac PM. 1974. Differentiation, organo-

genesis, and the tectonics of cell wall ori- entation. II. Separation of stresses in a two dimensional model. American Jour- nal of Botany 61: 135-140.

Losick R. 1995. Differentiation and cell fate

in a simple organism. BioScience 45: 400- 405.

Silk WK, Erickson RO. 1978. Kinematics of

hypocotyl curvature. American Journal of

Botany 65: 310-319.

Sitte P. 1992. A modern concept of the 'Cell

Theory': A perspective on competing hy-

potheses of structure. International Jour- nal of Plant Sciences (Special Issue: Katherine Esau International Symposium) S153: 1-6.

Virchow R. 1858. Die Cellular Pathologie in

ihrer Begrundung auf Physiologische und

Pathologische Gewebelehre. Berlin: Au-

gust Hirschwald.

Whewell W. 1872. History of the Inductive

Sciences, From the Earliest to the Present

Time. 3rd ed. New York: D. Appleton and

Company.

Wolpert L. 1994. Do we understand develop- ment? Science 266: 571-574.

Philip M. Lintilhac (e-mail: plintilh@zoo.

uvm.edu) is a professor in the department

of Botany and Agricultural Biochemistry at the University of Vermont, Burlington, VT 05405-0086. ? 1999 American Insti-

tute of Biological Sciences. ?...-..,-.....*.-*** ................ _ .. _* .... ?* :. Membership in AIBS... : *: a better value than ever! .: - * 1999 membership rates -* i are unchanged from 1998 * -*. NEW membership category for 1999 * * Just $20 for undergraduate * and K-12 students

BioScience Vol. 49 No. 1 68 Downloaded from https://academic.oup.com/bioscience/article/49/1/59/292302 by guest on 12 June 2023