• Campos-Ortega;
  • neurogenesis;
  • neurogenetics;
  • brain;
  • development


  1. Top of page
  2. Abstract

Jose Campos-Ortega stands out as one of the pioneers of developmental-genetic studies of early neurogenesis. He also liked to reflect about the history of science: how one discovery leads to the next, and what role individuals play in the progress of science. He had indeed started to work on a book describing the history of developmental genetics during the last year of his life. His goal in this book was to “explain how developmental genetics originated, how it transformed developmental biology and, while doing so, how it contributed to achieve the biological synthesis.” In the following, I would like to reflect on the origin and growth of the field Campos-Ortega contributed so much. In doing so, it is of particular interest to consider his scientific roots, and the manner in which he entered the stage of developmental genetics. I believe that Campos-Ortega's unusual scientific background influenced in an important manner the way in which he shaped the study of early neurogenesis. Developmental Dynamics 235:2003–2008, 2006. © 2006 Wiley-Liss, Inc.

Campos-Ortega studied medicine in Valencia, Spain. From early on his interests were academic. Following in the footsteps of his compatriot Ramon y Cajal and other neuroscientists of the early 20th century, his driving conviction was that the understanding of neural function rests largely on the detailed knowledge of neural structure. Reconstructing neural connections was the trend in the neuroscience of the 1960s and 1970s, similar to what mutant screens or gene chips represent today. Many journals sprang up around this topic. The reason for this rapid expansion was the introduction of new techniques that allowed one to trace nerve connections. Although detailed microscopic studies on brain structure were started much earlier, the available techniques rarely allowed investigators to determine precisely where a given nerve tract started and where it ended. New techniques were devised to overcome this problem. Among the first was the labeling of preterminal degeneration following discrete lesions. Thus, if one destroys, for example, a small area in the cerebral cortex, the axons that go out from this region will eventually degenerate. At a certain time point, one can visualize these degenerating axons by a special silver impregnation procedure and, thereby, analyze in detail where the axons starting from a defined region of the brain project. Campos-Ortega became engaged with passion in this neuroanatomical research. As a postdoctoral student, he went to Goettingen, Germany, and studied the connections of the visual system in primates in the lab of Paul Glees.

One of the attributes that always characterized Campos-Ortega's approach to science was his awareness of where the field was going, the technical possibilities and the limitations. Still following his original interest, that is to unravel brain structure, he was painfully aware of the staggering difficulties resulting from the number of neurons in a vertebrate brain. This difficulty becomes clear from a glance at Figure 1A, after a diagram from one of Campos-Ortega's last papers on primate neuroanatomy (Campos-Ortega and Hayhow, 1972). Diagrammed are the connections between parts of the visual cortex and the thalamus. Each of the arrows represents millions of axons. This kind of wiring diagram was and still is en vogue in functional neuroanatomy; it spells out how relatively large populations of neurons are interconnected. One might compare this diagram with one that shows how the computers of your laboratory are interconnected. Such a diagram is obviously useful for some purposes, but it will definitely not allow one to understand how a computer works. To do that, one would have to take it apart and study the pattern by which individual transistors and chips are interconnected. In large adult mammalian brains, we are almost as far removed from that elusive goal as we were in the early 1970s. Because of this problem, a new trend among anatomists and physiologists arose during that time: to turn to so-called simple systems, that is, the nervous system of invertebrates. One system that was identified was the visual system of the housefly, Musca domestica. The MPI for Biological Cybernetics in Tuebingen was founded around the task of studying visually controlled motor patterns in the fly. Campos-Ortega switched from the visual system of primates to the one of the fly. Here, at least, one can grasp the connections between different parts of the eye and optic lobe at the level of resolution of individual axons. The diagram in Figure 1B (after Campos-Ortega and Strausfeld, 1972) shows the network of interneurons in the fly lamina, with each element corresponding to a single cell. Campos-Ortega worked for several years in the laboratory of Valentin Braitenberg and made many seminal discoveries of the microcircuitry in the fly optic lobe.

thumbnail image

Figure 1. The resolution achievable with neuroanatomical tracing technologies is inversely proportional to brain size. A: Schematic representation of neural connectivity in the primate visual system (after Campos-Ortega and Hayhow, 1972). B: Connectivity in the fly visual system (after Strausfeld and Campos-Ortega, 1972).

Download figure to PowerPoint

The study of circuits in the fly and other invertebrates is still a vibrant and productive field, but Campos-Ortega moved his focus once again after opening his own lab and research group in Freiburg, Germany. Another quantum leap in resolution can at least in part be considered as his motivation to direct his approach to genetics and development. Therefore, he moved into Drosophila. For half a century, the study of genetics and the causal analysis of development had been two separate disciplines. Drosophila geneticists had amassed many mutations and studied their interactions, but there was little interest in studying how the mutation affected development. That gradually changed. By the late 1960s, the first mutations had been characterized embryonically (reviewed in Wright, 1970). The importance of mutant screens was appreciated. Interesting new techniques were developed, such as the induction of clones by somatic recombination, and of gynandromorphs (Garcia-Bellido and Merriam, 1969, 1971). These techniques could be used to analyze development; for example, to reconstruct fate maps. Of course, ultimately, developmental-genetic techniques looked at individual gene products, rather than whole cells.

In his first years in Freiburg, Campos-Ortega combined light and electron microscopy and the analysis of clones to study the development of the Drosophila eye and optic lobe. Among many other topics, the nature of the phenotype of the eye mutation sevenless, discovered from a behavioral screen in Seymour Benzer's lab a few years earlier, was analyzed (Campos-Ortega et al., 1979). Gradually, however, the research focus in Campos-Ortega's lab shifted away from the developing eye, in which he applied genetics as a merely “instrumental” approach, toward another neurobiological question, early neurogenesis, where his work initiated the discovery of new genes, in addition to establishing new concepts and descriptions of a fundamental developmental process. The pattern and development of sensory organs, or bristles, had attracted attention of some early pioneers of developmental genetics, such as C. Waddington and Curt Stern (e.g., Lees and Waddington, 1942; Stern, 1954). A gene complex, called the achaete-scute complex, was known since the 1940s to be involved in bristle patterning. Another gene, E(spl), also contributed to bristle pattern; at the same time, it was known that it had something to do with the gene Notch, because split is an allele of Notch. And Notch was the gene that limited the size of the central nervous system, as was known from the work of Donald Poulson in the 1930s (Poulson, 1940). Could there be a relationship between the pattern of bristles and of neuroblasts in the early embryo?

This question prompted the undertaking of a deficiency screen that included deficiencies uncovering the achaete-scute complex and that led to the discovery of the proneural phenotype (lack of neuroblasts) in the embryo (Jimenez and Campos-Ortega, 1979; see below). A decisive event for many Drosophila biologists, including Campos-Ortega, was the big screen for embryonic lethal mutations with discrete pattern defects that was initiated by Christiane Nuesslein-Volhard and Eric Wieschaus (Nuesslein-Volhard and Wieschaus, 1980). At the time of the screen, collaborations among the groups of Nuesslein-Volhard and Campos-Ortega took shape, and eventually the focus of Campos-Ortega's research shifted entirely toward early neurogenesis in the embryo. Early neurogenesis comprises the phase during embryogenesis when a specialized region, the neurectoderm, appears within the larger context of the ectoderm. The neurectoderm produces progenitor cells that give rise to the neurons and glial cells of the nervous system. Up until the 1970s, many basic concepts and questions concerning early neurogenesis had been formulated, mostly based on research on vertebrate embryos. Ideas about cell fate, fate determinants, induction, signaling, and morphogens were debated. However, to address such problems, only the tools of “classical embryology” existed. For example, to address mechanisms controlling cell fate in the vertebrate neural tube, parts of this structure were explanted, or rotated, and the effect on the remaining cells was analyzed (Watterson, 1965). This type of experiment provided evidence for intrinsic, as well as extrinsic, fate determinants. However, the interpretation of these experimental results had to remain at a fairly abstract level, because techniques to isolate the underlying molecular mechanisms had not yet emerged.

Progress entered the study of early neurogenesis from the invertebrate side. A true pioneer of the developmental genetic approach was Donald Poulson, followed by Thomas Wright, who described the embryonic phenotype of Drosophila mutations, notably Notch (Poulson, 1940; Wright, 1970). Poulson also did the first analysis of early neurogenesis in normal embryos. He recognized that the role of Notch must be to restrict the number of ectoderm cells that become neuroblasts. Another important novelty that gave a new flavor to the study of early neurogenesis was the single cell approach: rather than describing and experimenting with large chunks of neuroepithelium, as one did in vertebrate studies, the focus was on individual, recognizable cells. It was observed that, in insects, identifiable neurons not only existed in adult brains but also in embryos. A good example is the neuroblasts that can be recognized shortly after gastrulation. In 1976 appeared the first precise neuroblast map, not for Drosophila, but for grasshopper (Bate, 1976). Insect neuroblasts form fixed lineages, by budding off ganglion mother cells one after another. This peculiar “stem cell mode” of division was recognized more than 100 years ago (Wheeler, 1893). But only in the 1970s did the implications strike several developmental biologists that (to speak in Michael Bate's [1976] words) “neuroblasts have restricted access to the genetic programme which commit them to alternative and sometimes unique pathways of differentiation.”

To reiterate: Important concepts and ideas about early neurogenesis, as well as a framework of morphological observations, had been around for a long time. But it was the technological revolution of molecular biology, that made it feasible to isolate and clone genes, that provided the fuel for the rapid expansion of the field (Fig. 2). These new technologies attracted established scientists and students to the field. In addition, scientists with ambitions and expertise in certain fields, such as genetics, anatomy, or embryology, were able to integrate the technological scientific energy with the available concepts, and thereby shape the fledgling field of “modern developmental biology” in a decisive way. Campos-Ortega, with his background in neuroanatomy and strong preference for a single cell approach, as well as his ambition to adopt new technologies in his research, acted as one of the shapers of the field.

thumbnail image

Figure 2. The synergy between old concepts, original experimental approaches, and the new technology of molecular biology created modern developmental biology. [Color figure can be viewed in the online issue, which is available at]

Download figure to PowerPoint

During the first decade after he entered the field, many of the cornerstones were laid. First and foremost was the identifications of two gene families, the proneural and neurogenic genes, that determine the pattern of neural progenitors in Drosophila. Several groups were working on the achaete-scute complex (AS-C), which was involved in the patterning of bristles. In one of Campos-Ortega's first papers in the field of Drosophila neurogenesis, it was demonstrated directly that the AS-C controlled the formation of the embryonic nervous system (Jimenez and Campos-Ortega, 1979), confirming earlier findings by Garcia-Bellido and Santamaria (1978) that were based on the analysis of gynandromorphs. Deficiencies of the AS-C show a loss of neurons in late embryos. The histological technique used for these early characterizations of gene function involved in early neurogenesis was the fuchsin whole-mount. It should be pointed out that the technology to fix and stain fly embryos was published only in the late 1970s (Zalokar and Erk, 1977); one would like to think that without this seemingly trivial feat, the discovery of mutations might have taken a very different path.

Probably the most decisive discovery for the field of early neurogenesis was that an entire family of genes, the neurogenic genes, collaborated during early neurogenesis in restricting the number of neuroblasts. In addition to Notch, which had been around for half a century, a sizeable group of other genes was identified. These included big brain and Delta mutants, in which hyperplasia of the embryonic nervous system was visualized in fuchsin whole-mounts, histological sections, and cuticle preparations (Lehmann et al., 1981).

Aside from the phenotypic characterization of genetic mutations, progress was made on the analysis of normal neurogenesis in Drosophila. A preliminary neuroblast map (based solely on histology, because specific molecular markers were not yet available) was established for the fly embryo (Hartenstein and Campos-Ortega, 1984). Two important aspects of this map were, first, that neuroblasts do not segregate all at once, but arise in distinct generations, or waves; and second, that neuroblasts and epidermoblasts emerge from the same region in the ectoderm. Poulson, in his fate map, assumed that neuroblasts and epidermoblasts originated in separate regions: all neuroblasts in his map formed a continuous region near the midline, whereas the epidermoblasts laid more laterally. However, the new analysis showed that that was not the case. Neuroblasts and epidermoblasts are intermingled in a large neurogenic ectoderm. This discovery proved to be important for our understanding of the mechanism of lateral inhibition between neighboring cells.

Genetic studies revealed further that the neurogenic region, from where all neuroblasts segregate, is a genetically specialized part of the ectoderm. A study published in 1983 (Campos-Ortega, 1983) looked at double mutants of Notch and a hypomorphic dorsal allele. The main finding was that dorsal is epistatic to Notch. In wild-type embryos, the entire ventral half of the ectoderm gives rise to both neuroblasts and epidermoblasts. In Notch mutants, this ventral half turns completely into neural tissue. However, when dorsal function is simultaneously reduced, only a narrow band of ectoderm right next to the midline produces neuroblasts. In this background, all that Notch does is to turn the narrow stripe into neural tissue. In other words: dorsal acts first and subdivides the ectoderm into a dorsal non-neurogenic and a ventral neurogenic region. Notch acts exclusively on the neurogenic region. If this domain is decreased in a dorsal reduction of function mutation, many fewer cells become neuralized.

An approach that was also initiated in Campos-Ortega's lab was the transplantation of labeled cells, both wild-type and mutant. Labeled mutant cells were transplanted into wild-type host embryos, and the resulting clones were visualized in late embryos (Technau and Campos-Ortega, 1987). Using this approach with individual neuroblasts lacking any one member of the neurogenic genes, it was possible to order these genes into autonomously and nonautonomously acting factors.

By the late 1980s, early neurogenesis had become a rapidly growing field. In particular, molecular progress was made by several different groups. The proneural genes and neurogenic genes were cloned and sequenced, and their expression pattern was described. As a result, after one decade of frenzied research, one could look back at an impressive advance of our understanding of early neurogenesis, as reflected by the flurry of reviews that appeared within a short period of time during the late 1980s and early 1990s on the topic of proneural and neurogenic genes (Artavanis Tsakonas, 1988; Knust and Campos-Ortega, 1989; Ghysen and Dambly-Chaudiere, 1989; Artavanis-Tsakonas et al., 1990; Giangrande and Palka, 1990; Greenspan, 1990; Muskavitch and Hoffmann, 1990; Simpson, 1990a, b; Campuzano and Modolell, 1992). What is particularly noteworthy is that all of the available technical approaches, including genetics, molecular biology, biochemistry and experimental embryology, were brought together. Every reader is probably familiar with one or another version of the model published by Campos-Ortega and Knust (1990), according to which neurogenesis in Drosophila is viewed as a two-step process, with proneural clusters being defined by the expression of proneural genes in the first step, followed by individual neuroblasts being selected by a lateral inhibitory mechanism controlled by the neurogenic genes.

Since the late 1980s, the field spawned by the analysis of early neurogenesis in Drosophila exploded. A simple experiment to illustrate this notion is to type a set of keywords, such as “Notch pathway” into a search engine for publications. For the 5 years between 1985 and 1989, three papers are logged; 22 between 1990 and 1994; 227 from 1995 and 1999; and 637 between 2000 and 2004. One major reason for this scientific explosion is that the signaling pathways and transcriptional regulators involved in early neurogenesis, such as the Notch pathway and the family of proneural genes, are so promiscuous: they are found in embryos and mature animals; in the gut, hair, brains and teeth; and probably in every single type of cell one time or another.

The modern study of early neurogenesis approaches this problem at three main levels: the molecular, the cellular, and the developmental levels. Molecularly, we want to know about the structure of the proteins involved, to find new such molecules, and to study their interactions in the cell. Cell biologically, we are interested in how these molecules play out in the ultrastructural context of the cell. What does activation of Notch does to adhesion or the cytoskeleton; how does it affect the mitotic spindle or cell polarity. And developmentally, we want to understand the embryological context in which the signaling pathways function. What kind of cells produce a signal; where are they located relative to the cells receiving the signal; what morphogenetic events are triggered by the signal; and what cell types result from the interaction. Work in Campos-Ortega's lab in the 1990s and the first years of the present century continued to address early neurogenesis along each of these three levels.

After studying flies for more than a decade, Campos-Ortega switched to zebrafish to analyze the genetic mechanism of neurogenesis in vertebrates. A question that stirred discussions among many workers in the field at the time when proneural and neurogenic genes were searched for in vertebrates asked: what, if any, role do these genes play in vertebrate neurulation? At a first glance, one might argue that there is no need for this gene cassette at all. The specific mechanism carried out by proneural and neurogenic genes in flies is to determine, on a cell-by-cell basis, what ectodermal cells will become neuroblasts and which cells will become epidermoblasts. This decision apparently does not occur in vertebrates: once induced, the entire neural ectoderm becomes neural. However, it was discovered soon that neurogenic and proneural genes are expressed in the vertebrate neural tube. What do they control, if not the number of neural vs. epidermal progenitors?

The results of genetic studies, but also the careful analysis of expression patterns of markers for different cell types and classic embryological studies in both Xenopus and Zebrafish (to which Campos-Ortega contributed decisively; Bierkamp and Campos-Ortega, 1992; Bayer and Campos-Ortega, 1992; Papan and Campos-Ortega, 1994, 1997, 1999; Dornseifer et al., 1997) revealed that the morphogenetic process controlled by proneural genes and neurogenic genes in vertebrates is astonishingly similar to the one in flies. True, eventually the entire neural plate adopts a neural fate. However, neurons are born in a long succession of coordinated delamination events. Notably, there is a population of early differentiating neurons, called primary neurons, that delaminate from the neural epithelium. The other cells remain epithelial, continue to divide, and give rise to neurons and glial cells later. Primary neurons are derived from specialized stripe-like domains of the neural plate that can be visualized by the expression of proneural or neurogenic genes. In embryos where neurogenic gene function is defective, many more primary neurons segregate from these regions. In other words, there is a decision similar to the one between neuroblast and epidermoblast in flies. This is the decision between primary neuron, which like a neuroblast segregates from the neuroepithelium, and undifferentiated neural precursor, which remains epithelial.

The second line of investigation studied intensely by Campos-Ortega and his collaborators during the past decade is that of epithelial morphogenesis, a direct offspring of the study of early neurogenesis. After it was discovered that Notch and Delta had long stretches of epidermal growth factor–like repeats in their extracellular domain, more genes with this characteristic were screened (Knust et al., 1987). Other screens searched for mutants with phenotypes similar to the one of the “classic” neurogenic genes. Several genes were discovered, among them crumbs, encoding a membrane protein restricted to the apical surface of epithelia; stardust and bazooka, with a phenotype similar to that of crumbs; or shotgun, the Drosophila E-cadherin homolog (Tepass et al., 1990, 1996; Knust et al., 1993). With the discovery of these genes, one could start looking at epithelial cell morphogenesis: how is the apicobasal polarity of epithelial cells controlled; how does this polarity affect the cytoskeleton and adhesion.

Finally, I would like to mention a direction of research, currently hotly pursued, which also branched off from the study of early neurogenesis: asymmetric cell division. In 1996, Campos-Ortega and his collaborators uncovered a gene, inscuteable, which localized to the apical pole of neurectodermal cells (Kraut and Campos-Ortega, 1996). In the following years, it was found that Inscuteable and other proteins form a complex that controls the orientation of the spindle of the dividing neuroblast (Jan and Jan, 2001; Betschinger and Knoblich, 2004). More importantly, the same complex is essential to channel cell fate determinants, such as Numb, into the progeny of the neuroblast. Asymmetric divisions were also found in the vertebrate neural tube and were deemed important for the asymmetric division that distinguishes between the undifferentiated progenitor, that continues to divide, and the cell that becomes postmitotic and differentiates as a neuron (Betschinger and Knoblich, 2004). The importance and role of asymmetric divisions in the vertebrate neural tube remains quite controversial. In one of his last papers, Campos-Ortgega and his collaborators analyzed spindle orientation in the zebrafish neural plate and neural tube (in this special issue). Here, only symmetric divisions in the plane of the epithelium are evident.

Campos-Ortega's extraordinarily active scientific career historically coincided with a phase in developmental neurobiology during which the field took a quantum leap forward, made possible largely by technological advances in molecular biology introduced into genetics. However, technological progress is most effective if scientists rooted deeply in the vast body of knowledge about the structure of the nervous system formulate decisive questions and are farsighted and daring enough to use new technologies to address these questions. Jose Campos-Ortega ranks prominently among these scientists.


  1. Top of page
  2. Abstract