Artophagy

Abstract

Science informs art, and art informs science. Both processes involve creativity and imagination, and collaboration between scientists and artists often leads to new insights in both fields. We took advantage of the power of artistic imagery to demonstrate a dynamic cellular process, autophagy. In particular, we depicted the cytoplasm to vacuole targeting pathway, which involves dynamic membrane rearrangements to sequester a specific cargo via an autophagy-related process. By depicting this event in the context of a crowded cellular milieu, we hoped to stimulate researchers to consider aspects of the process that might be overlooked in the overly simplistic schematic drawing that typify most scientific models.

Illustrations often provide the impetus to integrate the knowledge of a field at a given point in time, creating a coherent image of what is known. Since science is a continually growing discipline, there are bound to be aspects that are well determined, and other aspects that are still unknown or controversial. In these cases, some artistic license is often necessary.¹

In 2006, we collaborated in the creation of an image of autophagy. As with previous work,^2,3 the goal was to create a cross section, showing the process in the context of a living cell. The collaboration involved a steady dialogue of ideas, identifying the scientific facts and concepts that would be included, and designing an illustration that presented these as accurately and clearly as possible. After several rounds of sketches and critiques, the final painting was created and presented as the cover image that accompanied a research article⁴ with the credit: “Original painting by David S. Goodsell based on the scientific design of Daniel J. Klionsky,” underscoring the collaborative nature of the project.

Scientific Considerations

One of the great challenges of the project was developing a scene that showed a significant portion of the process of autophagy. Many of the structures of autophagy are far too large to illustrate if we also want to show the individual molecules that are involved. Instead, we chose to illustrate the formation of a cytoplasm to vacuole targeting (Cvt) vesicle, which is of a scale that is appropriate to the 1,000,000× magnification needed to see significant detail in the individual molecules. Through discussion, we decided to include a portion of the mitochondrion based on data that suggested this may be one of the membrane sources for the expanding phagophore,⁵ a Cvt complex in the process of sequestration, and a Cvt vesicle fusing with the vacuole.

A variety of sources were used to support the structure of the Cvt complex, and the form and interaction of proteins on the surface of the vesicle.^4,6–8 The proteins in the vesicle membrane required the most dialogue, since this is an area of active research in the author’s laboratory. The first step was to create a list of players with molecular weights (since atomic structures are not known in most cases), to ensure that the size of each protein was appropriate. Then several rounds of sketches and corrections refined the connectivity and placement of proteins in the vesicle; for example, the interactions of the components in the Atg1 kinase complex were revised based on published and (at the time) unpublished data (Fig. 1).^9–12 The model represented the current molecular details for several additional characterized Atg proteins. For example, Atg9 is an integral membrane protein¹³ that cycles from peripheral sites that include or are near the mitochondria to the PAS,⁵ whereas Atg8 is present on both sides of the phagophore¹⁴ and interacts with Atg19 on the Cvt complex.¹⁵ Atg11 self-interacts and may be a scaffold for the assembly of several Atg proteins,¹⁶ including the Atg1 kinase complex.¹⁷

Figure 1.

The process of collaboration. The collaboration involved a steady dialog of published papers, ideas, and sketches as the autophagy image was designed. These sketches show a point late in the collaboration, as the molecular details were being refined. (A) A sketch from Goodsell of the portions of the painting involved in autophagy. (B) A sketch from Goodsell showing the Cvt complex being surrounded by a phagophore. (C and D) Schematics from Klionsky correcting the interactions of the kinase complex (C) and the Atg12–Atg5-Atg16 components (D). (E) A revised sketch from Goodsell.

The Cvt complex is composed primarily of precursor aminopeptidase I, a large multimeric protein,^18,19 bound to its receptor, Atg19. We have speculatively drawn this particle as a symmetrical capsid-like assembly, with Atg19 bound on the surface. Molecules in the vacuole were taken from Klionsky et al. 1990,²⁰ including a vacuolar H⁺-ATPase modeled after ATP synthase and the SNARE proteins involved in vacuolar fusion.²¹ The vacuolar proteases were taken from Van den Hazel 1996.²² Other molecules in the scene, including the cytoplasm and mitochondrion, were taken from previous work.^23–25

Aesthetic Considerations

The overall layout was designed to tell the story of autophagy, using the popular artifice of capturing the temporal aspect of the process in the spatial arrangement from top to bottom (Fig. 2). The mitochondrion is shown at the top with a small vesicle budding from its outer membrane, the Cvt complex is shown at the center in the process of sequestration, and the vacuole is shown at the bottom in the process of fusion.

Figure 2.

The painting “Autophagy 1,000,000X.” This figure was previously published on the cover accompanying ref. 4, and is reproduced by permission of the American Society for Cell Biology, copyright 2006.

A cross-sectional metaphor was chosen for the image to allow display of a large portion of the cell.² This also allows display of a cross section through the Cvt vesicle, revealing both the arrangement of proteins in the phagophore membrane and its interaction with the Cvt complex. In all cases, the cross section is taken to the intersect approximately perpendicular to the membrane—this typically provides for a clearer presentation of the geometry of the membrane that intersects at other angles. In early sketches, we also tested cross sections through the Cvt complex, but these proved to add unwanted complexity to the image, and we chose to show the unclipped Cvt complex.

The colors were chosen to separate the molecules involved in autophagy and those involved in other cellular processes. The molecules of autophagy are shown in yellow and orange, and the other molecules are shown in complementary blues, greens and purples (Fig. 2). In some respects, these color choices can be jarring, but given the strong intermingling of molecules in the cytoplasmic compartment, strong color differences were necessary to allow easy recognition of the molecules of autophagy from the surrounding molecules.

Purpose

One goal of this collaboration was to generate an image of selective autophagy in the context of the cellular environment. That is, even though we tend to make schematic images that only show the compartments and proteins we are referring to at the moment, showing the cell as mostly empty space, this is not realistic. Rather, the cell is a complex and crowded environment including an abundance of cytosolic proteins, ribosomes, and cytoskeletal elements. In the context of selective autophagy this raises many questions. For example, if the phagophore expands by vesicular addition as shown here, how are the vesicles targeted to the correct site? Also, how is cargo recognized and selectively sequestered? We hope that this image will foster additional thought about these and other questions relating to the membrane dynamics of autophagy.