40 years since the Heidelberg genetic screen that revolutionized developmental and cell biology Free

The ability of the animal body to self-assemble based on the information encoded in our genomes is one of the most remarkable miracles of the living world. We now take for granted the idea that shared cellular machinery mediates this and read textbooks laying things out in molecular detail. However, this knowledge did not come down from on high on stone tablets. Instead, a remarkable set of experiments published 40 years ago provided the foundation of our molecular understanding of embryonic development. In three papers in Roux's Archives of Developmental Biology (Jürgens et al., 1984; Nüsslein-Volhard et al., 1984; Wieschaus et al., 1984), two young scientists, Eric Wieschaus and Christiane Nüsslein-Volhard (Fig. 1), unveiled to the world the remarkable ‘Heidelberg screen’ that transformed our knowledge of the molecular basis of development.

To set the stage, we must review what we did and did not know about embryonic development in the early 1980s. Important discoveries via observational and ‘cut-and-paste’ experiments as in those defining the Spemann-Mangold organizer (Spemann and Mangold, 1924) or the localized germ cell determinants by cytoplasmic transplantation (Illmensee and Mahowald, 1974) provided intriguing insights. However, the molecules involved and the mechanisms by which they operated remained merely a matter of speculation. As a PhD student, I heard a remarkable series of lectures in April 1984 by future Nobel Laureate Sydney Brenner. He pioneered using the nematode Caenorhabditis elegans as a model to study development and behavior, and he and his colleagues were screening for mutations that altered behavior and morphology (Brenner, 1974). However, as he noted in the 1974 paper that introduced C. elegans to the world: ‘How genes might specify the complex structures found in higher organisms is a major unsolved problem of biology.’ In the lecture, he could only speculate about development's molecular underpinnings.

To turn these speculations into reality required the right foundation. This rested on the pioneering work from Thomas Hunt Morgan's lab using the fruit fly Drosophila melanogaster, which revealed the principles of genetics. Those following in their footsteps identified a host of genes shaping the morphology of the adult body in many ways, allowing one to link the function of individual genes with specific aspects of the body plan such as eye shape or wing veins. Meanwhile, the foundations of molecular biology were being laid, with the discovery of the central dogma and the analysis of gene regulation in bacteria, a cardinal example of identifying the underlying logic of a process by analyzing mutant phenotypes.

However, little was known about how genes regulated cell fate or embryonic development. One clear example was the homeotic genes of the bithorax and Antennapedia complexes that define segmental identity (Kaufman et al., 1980; Lewis, 1978). Although most of this work focused on their effects on the adult body, Lewis and a few others had analyzed the effects of a handful of genes on the pattern of the larva hatching from the Drosophila egg. It is here that Wieschaus and Nüsslein-Volhard entered the story.

Both were working on Drosophila, Wieschaus beginning as a graduate student and Nüsslein-Volhard as a postdoc. They met in 1975, enjoyed discussing science and, in 1978, became joint junior group leaders at the newly founded European Molecular Biology Laboratory (EMBL) in Heidelberg (https://www.nobelprize.org/prizes/medicine/1995/nusslein-volhard/biographical/; https://www.nobelprize.org/prizes/medicine/1995/wieschaus/biographical/).

It was there that they hatched the plan to screen for embryonic mutations affecting the larval body plan. In retrospect, the logic seems simple. Others had mutagenized the bacterium Escherichia coli to define the full set of proteins crucial for biochemical processes such as DNA replication. Nüsslein-Volhard and Wieschaus applied the same process to defining genes crucial for the embryonic body plan, by removing genes one by one and screening for those in which the body plan was altered. As each described it in their Nobel Lectures (https://www.nobelprize.org/prizes/medicine/1995/nusslein-volhard/lecture/; https://www.nobelprize.org/prizes/medicine/1995/wieschaus/lecture/), this was not the original plan when they arrived at EMBL, but something that gradually emerged from their growing interest in the handful of mutations with embryonic patterning defects that already existed or that they discovered almost by chance. They ‘realized that the screening for embryonic mutants would be very rewarding, and that (they) were the only people in the world who could do it’ (https://www.nobelprize.org/prizes/medicine/1995/nusslein-volhard/biographical/).

The screen's success rested on two key assumptions and some crucial, but simple, technology (Wieschaus and Nüsslein-Volhard, 2016). Their assumptions were: (1) that mutations significantly altering the body plan would lead to embryonic lethality; and (2) that they could simplify things by focusing on a single tissue – the embryonic skin that secretes the larval exoskeleton or ‘cuticle’ (Fig. 2). Cuticles were the key – their features, including the head skeleton, posterior spiracles and segmental denticle belts, provided landmarks to assess whether there were defects in the body plan of mutant embryos. They adapted a simple method to ‘clear’ embryos, dissolving all internal tissues to allow visualizing the cuticle under a standard compound microscope, developed simple methods to analyze progeny of multiple mutant lines simultaneously and used clever Drosophila genetics tricks to reduce fly work.

The screen then began. They focused on generating mutations on the X, second and third chromosomes (for technical reasons each chromosome was the subject of a separate screen). Flies were fed the mutagen, ethyl methane sulfonate, which induces random mutations by alkylating nucleotides, leading to single base pair changes. Individual mutagenized chromosomes were isolated and heterozygous females and males carrying each mutagenized chromosome were crossed to one another. Simple Mendelian segregation meant that 25% of the progeny were homozygous mutant for the new mutation. Eggs were collected from each cross, viable embryos allowed to hatch and cuticle preparations made of dead embryos. They then sat down together at a compound microscope with two sets of eyepieces, scoring cuticles for mutants with penetrant and consistent alterations in the body plan. ‘The bulk of the work was the breeding of flies’ but ‘screening cuticle preparations was fun’ (Wieschaus and Nüsslein-Volhard, 2016). ‘Almost every day we could expect to encounter a new phenotype, a phenotype that would force us to re-evaluate some long-held assumption about embryonic development,’ (https://www.nobelprize.org/prizes/medicine/1995/wieschaus/biographical/).

One key to the screen was its impressive scale: 26,978 crosses were scored (Wieschaus and Nüsslein-Volhard, 2016). All was carried out in the very small spaces they were allotted, some carved out of an old men's room, with Nüsslein-Volhard and Wieschaus joined by their technician, Hildegard Kluding, and later a postdoc, Gerd Jürgens, with visiting scientist Gary Struhl playing a role. About a quarter of the lines were embryonic lethal, but most produced dead embryos with normal cuticles and were discarded. Ultimately, ∼600 mutant lines were kept. The next task was to assign these to genes by complementation testing between mutant lines with similar or overlapping phenotypes, and genetic mapping to particular chromosomal regions (Wieschaus and Nüsslein-Volhard, 2016). At the end, they had 120 genes, many with multiple alleles. Several criteria suggested they had identified most of the genes that could be identified in such a screen, and subsequent work bore this out.

The first surprise was the small number of genes identified – only 120 genes of which zygotic expression is required for correctly specifying the larval body plan among ∼14,000 protein-coding genes in the fly genome. The most surprising, and ultimately most important, result of the screen was that the genes did not have 120 distinct phenotypes. Instead, they fell into phenotypic classes that ultimately revealed much of the underlying logic of development (Fig. 3).

This was immediately apparent when considering genes affecting cell fate choice along the anterior-posterior axis (Wieschaus and Nüsslein-Volhard, 2016). Nüsslein-Volhard and Wieschaus were rapidly able to divide these into three broad categories. Mutations in gap genes led to deletion of multiple body segments (e.g. Kruppel mutants lack thorax and anterior abdominal segments). Most striking were the pair-rule mutants, in which every other body segment was cleanly deleted (e.g. even skipped mutants lacked even-numbered abdominal segments). We now know that these genes encode transcription factors for which sequential activation and patterned expression turn a maternally contributed anterior-posterior gradient of the transcription factor, Bicoid, into individual cell identities of each row of cells along the anterior-posterior axis, all within the first 3 h of development. A third category of mutants, the segment polarity genes, affected patterning within each body segment, with a portion of the segment deleted and replaced by a mirror-image duplication of the remaining portion. We now know that these genes encode many components of two key developmental signaling pathways – those for Wnt and Hedgehog ligands. These mutants were striking enough that they were published before the screen's completion, in a paper profiling ‘Mutations affecting segment number and polarity in Drosophila’ (Nüsslein-Volhard and Wieschaus, 1980).

The names of the gap, pair-rule and segment polarity genes also point up another daunting challenge: naming 120 new genes. They followed the existing convention of naming genes based on mutant phenotype. However, although this was straightforward for naming a gene in which the wild-type red eye color was changed to white (i.e. the white gene), with 120 genes in hand the task required quite a bit more creativity. Mutants in which denticles were replaced by naked cuticle were christened after animals or fruits that were (or were imagined to be) uniformly hairy: hedgehog, armadillo and gooseberry. Mutants that failed in dorsal closure and thus had an open dorsal cuticle were named after boats: punt, kayak and canoe. Mutants in which the cuticle was broken into pieces were named bazooka, shotgun, stardust and crumbs. Everyone has their favorite names (my spouse's is shaven baby, also known as ovo, which lacked denticles).

This initial set of mutants was eye-opening, but the relatively small number of genes identified – only 120 – was a surprise. However, Nüsslein-Volhard and Wieschaus already knew that they had missed genes and why, setting the stage for the next set of mutant screens. As in all animals, Drosophila packs mRNAs and proteins into the egg that the embryo will need in early embryogenesis before zygotic gene expression ramps up. The rapidity of Drosophila embryogenesis means that, for some genes, the maternal contribution is sufficient for the embryo to hatch as a larva even if it is zygotically mutant. Thus, to see the phenotype of these genes, the mother (or later at least her germline) had to be mutant. They had already begun to examine a handful of ‘maternal effect’ mutants affecting the pattern of the embryo in dramatic ways (Wieschaus and Nüsslein-Volhard, 2016): in bicaudal mutants, the head was replaced by a mirror image of the tail; whereas in dorsal mutants, ventral denticle belts were replaced by dorsal structures.

The success of the zygotic screen stimulated similar screens for female sterile mutations. Some mutants produced ‘normal’ eggs, but the embryos died, sometimes with body plan alterations such as those seen in the zygotic screen. Other mutations altered oogenesis, some visibly affecting anterior-posterior or dorsal-ventral polarity of the eggshell. Madeleine Gans and colleagues had carried out a screen for genes on the X-chromosome in the 1970s (Gans et al., 1975), though characterization of their phenotypes was limited. In the 1980s, parallel screens were carried out by Trudi Schüpbach and Eric Wieschaus on the second chromosome (Schüpbach and Wieschaus, 1986, 1989, 1991) and by the Nüsslein-Volhard lab (including Kathryn Anderson, Gerd Jürgens, Ruth Lehmann and Hans Georg Frohnhöfer) on the third chromosome (Anderson et al., 1985; Nüsslein-Volhard et al., 1987). Norbert Perrimon and colleagues added an important extension. Many important maternally contributed factors are also required for zygotic viability, and thus homozygous mutants die before adulthood. They developed approaches to make the female germline homozygous mutant for these genes, thus uncovering additional genes required for oogenesis and embryonic patterning (Perrimon et al., 1989; Perrimon and Mahowald, 1987).

The tools for carrying out the zygotic and maternal-effect screens had been available for decades, but the timing of the screens meant their impact was immediate and broad for two key reasons. First, as Eric noted in his Nobel Lecture: ‘One historical coincidence that contributed to the significance of the Heidelberg screens was the almost simultaneous development of molecular techniques that allowed genes to be cloned based on their genetic position’ (https://www.nobelprize.org/prizes/medicine/1995/wieschaus/lecture/). Pioneering work in the Hogness, Bender, Gehring and Kaufmann labs had developed the necessary tools, and the year before the screen was published they described cloning the Hox genes of the bithorax and Antennapedia complexes (Bender et al., 1983; Garber et al., 1983; Scott et al., 1983). Second, as Eric also noted, this led to the totally unexpected ‘discovery of homology between key players in development throughout the animal kingdom’, both the Hox genes and then ‘the members of the conserved signaling pathways such as Notch, Hedgehog, Wnt, EGF and BMP – (this) underscored the usefulness of studying Drosophila as a model for development and even human disease’ (https://www.nobelprize.org/prizes/medicine/1995/wieschaus/lecture/). This led to a golden age in studying mouse development, as scientists cloned mouse homologs of their favorite Drosophila developmental regulators and then used the new embryonic stem cell gene knockout technology to eliminate their function and examine phenotypes. We now take for granted the fact that all animals use related cellular machinery to carry out important processes, but those of us around at the time remember the tremendous excitement this realization generated.

It's hard to overemphasize the importance of the screens for our understanding of the molecular basis of development. They revealed mutants that illuminated many key cellular processes, from transcription to RNA localization, cell signaling, the basics of cell polarity and the cytoskeleton. The gap and pair-rule genes were only the tip of the iceberg in terms of key families of conserved transcription factor families identified in the screens. A few, such as the Hox genes, serve similar roles in all animals, but most are used for distinct processes in different animal lineages, and many are used in multiple tissues in the same animal. Often, they form families with related DNA-binding domains, such as the Pax genes related to the pair-rule gene paired, or the Krüppel-like factors (KLFs). Most have broader roles outside the epidermis: twist and snail and their mammalian relatives are key regulators of the epithelial-to-mesenchymal transition, whereas orthodenticle (oc) and its vertebrate Otx relatives pattern the nervous system. The collection also included histone-modifying complexes such as Polycomb.

One of the screen's most important impacts was on our understanding of cell-cell signaling. Among the dozens of signaling pathways encoded in our genomes, five play inordinately important roles in embryonic development and adult homeostasis, and also are mutated in many human cancers. These are the Wnt, Hedgehog, Notch, BMP/TGFβ and Receptor tyrosine kinase (RTK) pathways, three of which are named for mutants identified or whose role in embryonic development were defined in the screen. Between the zygotic and maternal-effect screens, most proteins in these pathways, from transmembrane receptor to the nucleus, were identified; parallel work with lineage (lin) mutants identified in C. elegans complemented this (Kayne and Sternberg, 1995). Genetic analysis also allowed ordering these genes in their pathways, and cloning the genes empowered molecular and biochemical analyses. Chemical inhibitors of proteins identified in these screens, such as the EGF-Receptor (Egfr; identified as faint little ball in the zygotic screen and as torpedo in the maternal-effect screen), Raf [identified in the maternal-effect germline screen as l(1)pole hole] or the Hedgehog pathway component and G-protein coupled receptor, Smoothened, are now in the clinic to treat different cancers.

However, the impact of the screens on cell biology went well beyond this. They identified key components of the cadherin-catenin complex mediating cell-cell adhesion (shotgun is E-cadherin and armadillo is β-catenin) and its linkage to the actin cytoskeleton (canoe), as well as the integrins mediating cell-matrix adhesion (myospheroid and scab). Epithelial cells also have apical-basal polarity and many key regulators were first identified in the screens. The Crumbs complex, including crumbs and stardust, defines the most apical domain. Next is the Par complex, including bazooka (Par3) plus the other Par complex proteins identified in C. elegans. The cytoskeleton and its regulators were also prominent. These included the non-muscle myosin heavy chain (zipper), the non-conventional myosin VIIa (crinkled), and many actin cytoskeleton regulators including the formin cappuccino, scraps (encoding Anillin) and sponge (encoding a DOCK family RhoGEF).

Although the zygotic screens targeted epidermal development, the mutants also include those affecting more global processes. Several key cell cycle regulators were identified including string, the CDC25 phosphatase, cortex and fizzy (an activator of and a component of the anaphase-promoting complex, respectively), and pebble (encoding a Rho guanine nucleotide exchange factor essential for cytokinesis). faint sausage (currently known as fandango) encodes part of the PRP-19 complex that activates the spliceosome, whereas the Halloween genes such as spook and phantom encode the enzymes carrying out the complex biosynthesis of the steroid hormone ecdysone. The maternal-effect screens also identified key cellular machines with germline-specific functions. These included the germline determinants, Oskar and Vasa, as well as the RNA-binding proteins and cytoskeletal machinery required to localize them to the posterior end of the egg, such as the dynein-dynactin cargo adapter, BicaudalD. They identified machinery required for meiosis, such as the helicase encoded by okra, as well as the key players in piRNA generation or function (e.g. squash, zucchini and aubergine). There are few areas of biology that were not included in these remarkable mutant collections.

The Heidelberg screen was carried out quite early in the careers of the participants, and for each it was just a prelude to many important scientific contributions. Eric Wieschaus was the only one to continue working on Drosophila for the rest of his career. He was a leader in bringing cell biology into the study of development. His lab made many contributions, including identifying key components of the Wnt and G-protein coupled signaling pathways. They led efforts to explore how the cytoskeleton is regulated, how cell junctions assemble and polarize, and how junction-cytoskeletal connections regulate the ability of cells to change shape and move. This latter topic was a deep interest; Eric was already filming gastrulating wild-type and mutant embryos in the early 1980s using video cameras designed for surveillance. He pioneered bringing math and physics tools into the field, in long-term collaborations that had impacts in many areas of biology, from morphogen gradients to the biophysical properties of membranes and cytoplasm to systems biology.

Christiane Nüsslein-Volhard took a different path. Her lab initially extended work from the zygotic and maternal screens, identifying the Bicoid morphogen gradient and defining the Toll signaling pathway and its role in dorsal-ventral patterning. However, she became intrigued by George Streisinger's work on zebrafish and, by 1988, pioneered using them for genetic analysis of vertebrate development. Her lab and those of Wolfgang Driever (her former PhD student) and Mark Fishman embarked on parallel genetic screens for mutations altering the body plan in diverse ways, affecting many of the body's organs. In December 1996, this remarkable effort was jointly published in a special issue of Development that presented in 37 papers the results of these large screens (Nüsslein-Volhard, 2012). Subsequent work by her lab and others made zebrafish a premier model animal.

Gerd Jürgens made his impact even further from his roots in Drosophila. After postdoctoral work with Nüsslein-Volhard and then Herbert Jäckle, where he made important contributions to our understanding of developmental patterning, he made the bold decision as a young, untenured principal investigator to switch to working on plants, becoming one of those who made Arabidopsis a powerful model for studying plant development and, ultimately, all things plant related. In his own group, they pioneered the genetic dissection of plant embryogenesis.

In addition to the remarkable work of all three scientists, each trained many postdocs and graduate students who went on to further their respective fields – I was fortunate to benefit from Eric's mentorship. We all stand on the shoulders of those who came before us. It's worth knowing about the experiments that empowered us and learning the stories behind the experiments. To learn more about the screen and the scientists behind it, their Nobel autobiographies and their personal accounts of the screen are informative and entertaining (https://www.nobelprize.org/prizes/medicine/1995/nusslein-volhard/biographical/; https://www.nobelprize.org/prizes/medicine/1995/wieschaus/biographical/; Wieschaus and Nüsslein-Volhard, 2016).

I thank Gerd Jürgens for reading an earlier draft.