The past, present, and future of cell lineage studies

Developmental biologists seek to understand how multicellular organisms arise from a single cell. One way of unpacking this complex question is to look at embryonic cell lineage. These investigations have historically fallen into two categories: 1) studies that seek to uncover which structures or regions in the adult body individual or small groups of cells in the early embryo give rise to, and 2) studies that describe the step-by-step process of how one developmental stage relates to the next. In the first case, investigators are interested in understanding the end point, or product, of those cells' descendants, whereas in the second case, investigators make close observation of cell shape change, growth, and positioning as well as gene expression patterns and cell-cell communication.

The study of cell lineage is highly dependent on the technique used and the species being studied. Each technique has its benefits and limitations, particularly with regards to the manner in which and how long development can be traced. The size of the egg, the size of the subsequent blastomeres, the yolk content of the cells, developmental temporality, and the accessibility of the embryos are all significant considerations. 

Marine invertebrate embryos are particularly amenable to cellular investigation of development. They are easily accessible along coastal shorelines, can be collected in large numbers, have large transparent cells, and develop relatively slowly, which allows more time for each stage to be observed or imaged. In is in large part because of this that some of the first lineage studies were carried out at the Marine Biological Laboratory (MBL). In the last few decades of the nineteenth and into the early twentieth century, American embryologists such as Charles Otis Whitman, Edmund Beecher Wilson, and Edwin Grant Conklin investigated cell lineages in several species of marine annelids and mollusks. Whitman, who was the founding director of the MBL, was the first to study cell lineage in Clepsine, a species of leech, the 1870s. This study was published as "The Embryology of Clepsine" in 1878. The Embryo Journal Club logo, which is Figure 22 from Whitman's paper, is an homage to this groundbreaking study.

Top row: The first four figures of E.B. Wilson's "The Cell-Lineage of Nereis" (1892). Bottom row: The corresponding stages in the same species, Nereis limbata. Collected from Eel Pond in Woods Hole and imaged by Beatrice Steinert in 2015. 

Whitman, Wilson, and Conklin embarked on these time-consuming endeavors largely in response to the widely debated embryological theories of their time. An especially dominant one was Haeckel's biogenetic law, which postulated that a species' development reveals its evolutionary history, an idea encapsulated in Haeckel's well-known phrase "ontogeny recapitulates phylogeny." Central to the biogenetic law was the hypothesized "Gastrea," an imagined organism, made up of two germ layers, that Haeckel proposed to be the common ancestor of all multicellular forms. These ideas caused a great deal of confusion and debate over whether embryonic primary germ layers could be considered homologous across all species (a foundational tenant of the biogenetic law), as well as what constitutes homology. The cell lineagists sought to clarify some of this confusion by meticulously describing the earliest stages of development, an endeavor many of their contemporaries thought to be pointless in light of their belief that those stages were merely a replication of cellular material that held no significance for later development.

The first cell lineagists relied on a light microscopes and camera lucidas to observe the divisions and movements of cells. Working from both live or fixed and stained embryos, they spent countless hours looking through the microscope. Drawing hundreds of specimens with the aid of the camera lucida was necessary for visualizing whole embryos and provided a means of recording various stages. With this method, they were able to trace many lineages far enough through development to determine the fates of individual cells in the early embryo.

Plate 3 from "The Embryology of Crepidula" (Conklin 1897). The lithograph illustrations shown here were produced from hundreds of camera lucida drawings. 

But at a certain point the cells became too small to see. In the 1880-90s Wilson and Conklin did not having the ability to label individual cells in the embryo so that they could more easily follow their divisions. It wasn't until the latter part of the twentieth century that researchers devised a way to do this by microinjecting fluorescent dyes, such as DiI or dextran-FITC, either onto the cell surface or into the cytoplasm. After injection, they would let the embryo develop and image the larva or young adult body to see what regions the dye ended up in. (see Lyons et al. 2012 as an example) With the advent of fluorescence confocal microscopy, the ability to live image these labeled embryos also became relatively easy to do.

Figure 2 from Lyons et al. 2012. The Crepidula "4d" cell visualized here is the same cell colored in pink in Conklin's 1897 illustrations above. 4d was injected with rhodamine green dextran and then live imaged with a confocal microscope.

While this technique allowed investigators to see beyond what the first wave of lineage studies could, it also had its limitations. After several rounds of cell division, the boundaries of individual cells become hard to see and the dye can become dilute depending on how much particular clones proliferate. This method primarily gives you information about the ultimate product of particular blastomeres in the larval or adult body, not detailed information about the process of development.

More recently, developmental biologists have begun microinjecting mRNAs coding for fluorescent reporter proteins. For many model systems reporter lines have been around for decades, and provide robust labeling of particular cell types, but very few of these exist for marine invertebrate species. After injection, the mRNAs get translated into proteins which can localize to particular parts of the cell such as the nucleus or cell membrane. These labeled embryos can be live imaged and then that data entered into cell tracking programs such as Bitplane Imaris or MaMuT, a free and open access ImageJ plug-in. When the user tracks cells through time, these programs can create detailed lineage trees or other lineage visualizations (see Özpolat et al. 2017 as an example). However, like fluorescent dyes, the mRNAs become diluted over time and eventually get degraded. Dilution coupled with cellular stress that results from fluorescence imaging phototoxicity substantially limits the imaging duration of studies using this approach.

Time-lapse dataset of a Platynereis dumerilli embryo injected with fluorescent cell membrane and cell cycle mRNA reporter constructs (Özpolat et al. 2017). Platynereis dumerilli, a marine annelid, is closely related to Nereis limbata, the species E.B. Wilson primarily studied in the 1880s. 

The same Platynereis dataset as above with cell lineages highlighted in blue, pink, orange, and yellow (Özpolat et al. 2017). The dividing cells were manually tracked with the image analysis software Bitplane Imaris. 

The limited temporal resolution of fluorescence live imaging has been addressed in the last few years by the emergence of light sheet microscopy. By only illuminating the part of the tissue being imaged, light sheet microscopes significantly reduce phototoxicity. This allows specimens to be imaged for longer and from multiple perspectives when using a microscope that has more than one objective. These datasets can also be loaded into cell tracking programs and analyzed in a variety of ways (see Wolff et al. 2018 as an example). Especially when using an organism with a transgenic reporter, lineages can be tracked with remarkable detail all the way through development.

Presently, the two major bottle-necks on embryonic cell lineage studies are the enormous size of the imaging files (which can be several terabytes per dataset) and our ability to process and analyze this four-dimensional data. Although several challenges still remain, technological developments over the past few decades have significantly pushed lineage studies forward. Along with other emerging approaches to understanding lineage, such as single cell sequencing and transcriptome analysis, sophisticated imaging and data analysis programs are allowing us to see developing embryos and address old problems like never before.

As we move forward and continue to identify what gaps in our knowledge or technologies still remain, it will be important to keep in mind what the aims of our studies are. We now have the ability to generate massive amounts of data relatively quickly and easily. But how do we determine what information we want to extract from it? Powerful computers and sophisticated analysis programs can certainly do a lot of the work for us. But if we rely entirely on computers and don't spend the time looking at and interacting with embryos, our own understanding of development will be diminished. We must certainly embrace the enormously promising computational future. We must also, however, continue to hone our observational skills to be as sharp as those of the very first cell lineagists.


This post is based on our June 1st kick-off meeting titled "Exploring the past, present, and future of embryonic cell lineage studies." Presented and written by Beatrice Steinert.