Article

Elise A. Span†

David S. Goodsell‡

Ramani Ramchandran§

Margaret A. Franzen and

Tim Herman¶

Daniel S. Sem†*

†From the Department of Pharmaceutical Sciences, ConcordiaUniversity Wisconsin, Mequon, Wisconsin 53097, ‡Department ofMolecular Biology, The Scripps Research Institute, La Jolla, California92037, §Department of Pediatrics, Developmental Vascular BiologyProgram, Children’s Research Institute, Medical College of Wisconsin,Milwaukee, Wisconsin, 53226, ¶MSOE Center for BioMolecularModeling, Milwaukee School of Engineering, Milwaukee, Wisconsin53213

Abstract

A team of students, educators, and researchers has devel-

oped new materials to teach cell signaling within its cellu-

lar context. Two nontraditional modalities are employed:

physical models, to explore the atomic details of several of

the proteins in the angiogenesis signaling cascade, and

illustrations of the proteins in their cellular environment, to

give an intuitive understanding of the cellular context of

the pathway. The experiences of the team underscore the

use of these types of materials as an effective mode for

fostering students’ understanding of the molecular world

and the scientific method used to define it. VC 2013 by The

International Union of Biochemistry and Molecular Biology,

41(4):213–223, 2013

Keywords: visual learning; VEGF signaling; protein structure;

angiogenesis

He said science was going to discover the basic secret oflife someday, the bartender put in. He scratched hishead and frowned. Didn’t I read in the paper the otherday where they’d finally found out what it was?[. . .]What is the secret of life? I asked.I forget, said Sandra.

Protein, the bartender declared. They found out somethingabout protein.

Kurt Vonnegut, Cat’s Cradle

IntroductionTo appreciate the complexity and beauty of the inner work-ings of a living cell, students must have a solid understand-ing of what proteins are: what they look like, what they do,and how the former shapes the latter. Unfortunately, thebeginning student rarely understands protein as anythingmore than something that belongs on a dinner plate.Therefore, a major goal of undergraduate educators in themolecular biosciences is to help their students develop arobust mental model of proteins and their biological activ-ities that is accurate at both the molecular and cellular lev-els. At the molecular level, students need to understand thebasic principles of chemistry and physics that drive a spe-cific sequence of amino acids into their final functionalform. Just as important, students must understand howindividual proteins interact at a cellular level to performthe many processes needed for life.

What is the best way to ensure that our students gobeyond simply memorizing amino acid structures andlearning the basic connections between protein structureand function? How can we help our students truly

hS Additional Supporting Information may be found in the online versionof this article.Abbreviations: CREST, Connecting Researchers, Educators andSTudents; ERK-2, extracellular signal-regulated kinase 2; DUSP-5,dual-specificity phosphatase 5; KEGG, Kyoto Encyclopedia of Genesand Genomes; MAPK, mitogen-activated protein kinase; MSOE,Milwaukee School of Engineering; PDB, Protein Data Bank; BMRB,Biological Magnetic Resonance Bank; VARK, Visual, Aural, Read-Write,Kinesthetic; VEGF, vascular endothelial growth factor

*Address for correspondence to: Department of PharmaceuticalSciences, Concordia University Wisconsin, 12800 N. Lake ShoreDr., Mequon, Wisconsin 53097. E-mail: Daniel.sem@cuw.edu.Received 8 January 2013; Accepted 20 April 2013DOI 10.1002/bmb.20706Published online in Wiley Online Library(wileyonlinelibrary.com)

Protein Structure in Context: The Molecular

Landscape of AngiogenesishS

Biochemistry and Molecular Biology Education 213

understand how proteins function at levels ranging fromatomic to cellular? What would it take for students toappreciate their world as a manifestation of the molecularworld? These were the questions considered by a team ofresearchers, educators and students participating in theCREST project at the MSOE (Milwaukee School of Engineer-ing) Center for BioMolecular Modeling. The premise of theCREST project (Connecting Researchers, Educators andSTudents) is that the collaborative efforts of these threestakeholders will result in innovative instructional toolsthat effectively translate the results of active research proj-ects into the classroom. An important corollary to this out-come is that students will understand that the knowledge intextbooks is the direct product of a series of experimentsperformed by researchers. Speaking as a group of educators,students, and researchers, we believe that it is possible toprovide students with a holistic view of biochemistry that issimultaneously accurate at the molecular level and scalableto the cellular and even macroscopic levels. The approachwe describe here relies on the combined use of two newinstructional tools: physical models of proteins and molecu-lar landscapes, realistic illustrations of portions of cells at amolecular level of detail. The goal of the latter is to put theformer in proper cellular context.

Proteins in the Classroom—AProblem of PerceptionAs proteins are not directly observable, the challenge ofteaching protein structure and function is largely a prob-lem of perception. Representations of proteins must be cre-ated that capture and present the features of interest with-out engendering misconceptions about function in thecellular context. Historically, concepts in biochemistry havebeen simplified and abstracted in order to make them ac-cessible to students. Hence, double-stranded DNA is oftenrepresented as two parallel lines drawn on the white boardor illustrated in a textbook, and proteins such as RNA poly-merase are represented by brightly colored ovals. This isstill the case today, as evidenced by the popularity of car-toon illustrations in biochemistry textbooks [1]. Although ithas long been common practice conceptually to simplifyand abstract proteins in this way—especially those proteinsforming complex signaling cascades—the approach can bejustifiably criticized for several reasons. Perhaps mostimportantly, it omits much of the information about themolecule, so important properties, such as the atomic basisof base pairing in DNA or the molecular driving forces forprotein–ligand interactions, must be taken on faith insteadof observed directly. Such abstracted representations ofproteins are far removed from “real” life; it is thereforeconceivable that the prevalence of this educationalapproach has contributed to public notions about scienceas an overly abstract field—a peculiar conclusion to be

reached about a discipline with such a firm basis in thematerial world. One must conclude that we as educatorsare not adequately bridging the gap between the abstrac-tions we use and the molecular reality in which proteins op-erate within cells. While there is likely to be no simple solu-tion to this problem, we hypothesize that a contributingfactor is educators’ ignorance of the diversity of learningstyles present in their biochemistry classrooms, which entailsthat a successful tool for one type of learner may have lim-ited success with other learning types. Basic two-dimensional(2D) representations of a protein usually require students tohave strong visual-spatial skills that enable them to synthe-size the protein’s intricate features in three dimensions. Stu-dents who are not visual learners may have difficulty withthis task, and stand to benefit from the use of a more diverseset of educational tools for teaching protein structure.

2D representations of proteins, based on their atomiccoordinates (from the Protein Data Bank (PDB) [2]) or otherdata (e.g. chemical shift data from Biological Magnetic Res-onance Bank (BMRB) [3]), are also widely used, and com-puter software for 3D visualization of proteins has becomemore popular in the classroom in recent years [1, 4]. Thetrend toward increasing sophistication of our visual toolshas great potential for the teaching of protein structure–function, with an eye toward providing a holistic molecularpoint of view. As observed by biomedical illustrator LindaNye, “Humanity’s ability to visualize objects and theirtransformations in space is a critical component of ourintelligence. It makes possible the planning of actions andthe anticipation of outcomes, which is the basis of scienceand the scientific method” [5]—this is indeed the abilityrequired for comprehending an enzyme mechanism or pro-tein signaling event. Visualizing the intricate details of bio-molecular phenomena, such as interactions among resi-dues in an enzyme’s catalytic site, requires similarlyintricate illustrations or models. It may seem that with ev-ery new application of realism to a visual model, at thismolecular level of detail, the subject becomes more com-plex and thus more difficult to teach (for the educator) andto comprehend (for the student); however, students do notnecessarily find realistic representations of biomoleculesmore intimidating than their schematic counterparts [4].Conversely, students with lower visual-spatial skills may bedaunted by more realistic representations: for these stu-dents, a 3D software program or high-resolution 2D draw-ing makes a protein appear not more realistic, but as amore complex abstraction [1, 6, 7].

Recently, the use of hand-held, physical models of pro-teins has emerged as a new modality to help students bettergrasp protein structure–function in contemporary biochemis-try courses. Accurately rendered physical models of proteinshave been made possible by recent developments in rapidprototyping technology, but the concept of 3D models in sci-ence is nothing new: Watson and Crick used 3D models (paperand other materials, such as metal) to discover the details of

Biochemistry andMolecular Biology Education

214 Protein Structure in Context: The Landscape of Angiogenesis

base pairing in DNA, and Linus Pauling discovered the basicsecondary structures of proteins by experimenting with mod-els. Assessments suggest that protein models facilitate mas-tery of basic protein structure–function, and students reportfeeling more connected to the subject when they use a tactilemodel [1, 6, 7]. It is also thought that students with low visual-spatial skills may learn more easily via kinesthetic interactionwith the subject matter [6, 8]. Our own experience indicatesthat the biochemistry classroom is composed of students witha diversity of learning styles: we have quantified this observa-tion using the VARK learning assessment (Fig. 1), a surveythat measures students’ preferences among Visual, Aural,Read/write, and Kinesthetic learning styles [8]. This being so,our VARK data do not indicate that any particular learningstyle preference puts students at an advantage for overallexam scores (see Supporting Information Fig. S1), although itappears that students who are weaker visual learners mayimprove their molecular visualization and reasoning skills af-ter taking a course that employs multiple teaching modalities,including the use of protein physical models to complementtraditional lectures (based on the molecular visualizationassessment in Supporting Information Fig. S2). Our own

preliminary data (from the course described, below) suggestthis outcome; further studies in other, larger student groupsare currently under way.

Studying Trees—Without Seeing theForestBuilding on our experiences and previously reported suc-cesses with using physical models in teaching [1, 6], weincorporated them into a university-level biochemistrycourse that is tailored to chemistry majors (with a classsize of 15). These are more advanced students, who havealready taken two semesters of organic chemistry. Lecturesand homework assignments were centered on protein bio-chemistry and metabolic pathways with emphasis on mech-anism and the nature of intermolecular interactions thatlead to ligand binding (e.g. substrate, drug or protein part-ner). To complement the teaching of traditional biochemis-try content, students worked in small groups (three to fivestudents per group) on a semester-long, structure-baseddrug design project focused on vascular endothelial growthfactor (VEGF) receptor signaling, a signaling cascade thatleads to angiogenesis and is an important target for the de-velopment of anticancer drugs (Fig. 2, see Supporting Infor-mation for detailed pathway information).

Each group was assigned a different protein in this sig-naling pathway and then visited an academic research labinvestigating that protein. After learning about the researchproject, groups used molecular visualization software (Dis-covery Studio Visualizer (Accelrys) and Jmol) and a rapidprototyping platform (at the MSOE Center for BioMolecularModeling) to design and build 3D physical models of theirproteins with features emphasizing structure and function,such as identifying and highlighting residues involved inligand binding. Each group also designed a set of smallmolecules with shape and electronic complementarity tothe active site pocket that in theory might inhibit the pro-tein’s activity. The purpose of this exercise is to help stu-dents: (a) learn how drugs are designed, and (b) betterunderstand the intermolecular forces that lead to ligandbinding. Typically, students started with a known ligand(substrate or inhibitor), and designed changes to it thatretained the shape complementarity with the active site,and properly positioned hydrogen bond donors and accept-ors (and charged groups, if present). If there was an opencavity in the active site proximal to the bound ligand, thenstudents were told that additional functionality (e.g. methyl,ethyl, or even phenyl groups) should be added to their mol-ecule. If there was an unsatisfied active site hydrogen bonddonor (e.g. threonine OH) or acceptor (e.g. asparagine car-bonyl) proximal to the bound ligand, then students wouldbe counseled to add a complementary acceptor (e.g. fluo-rine; carbonyl) or donor (e.g. OH), respectively, to theirmolecule. As students are not used to being creative with

Biochemistry classroom VARK data. The average

scores for student learning styles are illustrated by

the dark blue plot, and the range of scores by the

light blue lines on each axis. Students taking the

VARK questionnaire are scored for their learning

preferences in categories of Visual, Auditory, Read-

write, or Kinesthetic. The possible range for each

style is 0–16. The data indicate that the class is

populated with students equally representing all

learning styles, on average. Class size 5 15 stu-

dents. [Color figure can be viewed in the online

issue, which is available at wileyonlinelibrary.com.]

FIG 1

Span et al. 215

regard to organic chemistry, this required iterative roundsof the students making attempts at designing new ligands,and then getting feedback from the instructor on how tobetter approach the problem. The ligands that studentsdesigned were first drawn on paper, then using Spartansoftware (Wavefunction) and minimized using AM1 calcula-tions. The molecules they designed were then manually

positioned into the active site using Discovery Studio Visu-alizer software, using the original ligand to orient theirligand (i.e. overlay them), after which the original ligandwas then deleted. Students then assessed the fit of theirnewly designed ligand into the active site. In effect, stu-dents were learning how to do rational drug design, withsequential rounds of guidance from the instructor. Students

Schematic of the VEGF pathways to be depicted in the landscape. The DUSP-5/ERK-2 interaction that occurs in the nu-

cleus (bottom right) is represented using physical models in Fig. 3. [Color figure can be viewed in the online issue,

which is available at wileyonlinelibrary.com.]

Physical models of DUSP-5 and ERK-2, demonstrating the binding interaction known to occur between these proteins.

DUSP-5 in comprised of two domains, an ERK-binding domain and a phosphatase domain, connected by an unstruc-

tured linker region (plastic tubing in the model). The ERK-2 protein binds between these two domains of DUSP-5. [Color

figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

FIG 2

FIG 3

Biochemistry andMolecular Biology Education

216 Protein Structure in Context: The Landscape of Angiogenesis

then prepared and presented posters along with their phys-ical models, showing their work and demonstrating theirunderstanding of their protein’s function, acquired basedon discussions with their research mentors and their read-ing of the primary literature.

The drug design component of the final project encour-aged students to think at atomic and molecular levels ofinteraction (protein–ligand) and narrowed their focus to asingle protein within the signaling pathway. Lecturesincluded an overview of the pathway and instruction on theinteraction between two proteins in the pathway, phospho-rylated extracellular signal-regulated kinase 2 (pERK-2)and dual-specificity phosphatase 5 (DUSP-5), which dephos-phorylates pERK-2 (Fig. 2). This pairwise interaction wasexplored via molecular animations (e.g. using PyMOL soft-ware) and physical models (Fig. 3) and encouraged stu-dents to think about the molecular interactions (protein–protein) that constitute a signaling cascade. When studentspresented their projects, they demonstrated overall masteryof protein structure and function with regard to their pro-tein member of the angiogenesis signaling cascade; how-ever, their explanations of the signaling cascade itselftended to be qualitatively weak. Despite our best efforts touse all the learning tools available to us (computer visual-ization, animation, lectures, physical models, active learn-ing via presentations), students still did not demonstrate asatisfactory appreciation of the network of protein–proteininteractions that are behind signaling cascades, nor of thecellular context of these interactions. Although the physicalmodels appeared to promote students’ awareness ofatomic-level function of the individual proteins, the “bigpicture” was lost. We decided to continue studying theeffects of 3D models on student learning in different stu-dent populations, but also that subsequent work on theCREST project should focus on identifying and correctingthis shortfall, by reinforcing the connection between all theproteins in the pathway. The further CREST work resultedin the design of a novel educational tool that in this articlewe propose for widespread use.

How Can Students Come toAppreciate the Forest?To identify ways to help students better appreciate the cel-lular context in which individual proteins function in theVEGF-angiogenesis signaling pathway, we investigated theprocess by which most students learned about the pathway.A popular starting place for students trying to learn thepathway was an online pathway database, such as theKyoto Encyclopedia of Genes and Genomes (KEGG) [9],which describes the networks of interacting proteins in sig-naling cascades and metabolic pathways. We found thatstudents had trouble with two aspects of these diagrams.First, the diagrams have been developed primarily for the

research community, and the conventions of the displaymay be difficult to comprehend by novices. They are oftenlinear, glossing over details such as multiprotein complexesand subcellular compartmentalization; they often ignorethe scale and population of the different protein players;and they often use coded descriptions for protein activitiessuch as phosphorylation that may not be obvious to thoseoutside the field. Second, and perhaps a greater problem,students were discouraged by the inconsistency amongpathway diagrams from different sites, which is a naturaloutcome of research that is current and ongoing in devel-oping fields, such as angiogenesis. Specifically, it is oftendifficult to discern between a pathway event that is stronglysupported by the primary literature and an event that isless well-defined, perhaps because of ongoing scientific dis-putes. In short, the clarity and certainty with which text-books present biochemistry leaves students ill-equipped todeal with the sometimes uncertain state of our understand-ing of many cellular processes, which the scientific processseeks to clarify; to borrow the phraseology of DonaldRumsfeld (February 12, 2002), science textbooks teach stu-dents that science is full of “known knowns,” leaving themunprepared for the “known unknowns” and “unknownunknowns” that are pursued in active research projects.

Although the first problem may be addressed withimproved diagrams, such as those described below, thesecond problem poses a greater conceptual challenge. Wehave found that the best way for students to appreciate theprocess of scientific discovery is to have a dialog withexperts in the field. Exposure to active researchers in thefirst phase of the CREST project (before preparing posters)was a step in that direction, but that single conversationappears to have been insufficient to empower students tointerpret the sometimes-conflicting literature. The need forstudents to have ongoing interactions with primaryresearchers must be balanced, of course, with the manyother demands on the researchers’ time. One of theresearchers involved in the CREST project, and a coauthorof this article (RR), suggested that meetings held betweenstudents and researchers approximately every two weeksmight strike a good balance between teaching the scientificprocess and respecting the time constraints of an activeresearch program.

To help students better integrate molecular and cellularconcepts, we have developed an approach to producingmore realistic renderings of cellular processes. We refer tothe resulting images as molecular landscapes. The goal ofthe approach is to create an illustration that accurately sim-ulates a portion of a cell, using visualization methods thatproduce an intuitively comprehensible image. The illustra-tion and its educational goals might draw a comparison withthe Harvard BioVisions project, which has developed highlyeffective multimedia presentations of cellular scenes thatbeautifully convey concepts of dynamics and scale as theyapply to cellular physiology [10]. According to the project’s

Span et al. 217

illustrator, David Bolinsky, the animations were designed sothat Harvard students “would have an internalized view ofwhat a cell really is, in all of its truth and beauty, and beable to study with this view in mind” [11]. Although static,2D molecular landscapes cannot convey molecular motionsto the same extent as 3D molecular animations, they havethe advantage of accurately portraying cellular crowding,something that in its entirety “would prevent the vista fromhappening” [11] and limit the intended effect of a 3D anima-tion. In this sense, molecular landscapes and animations arecomplementary. Molecular landscapes attempt to simulatethe environment of molecules inside living cells, with illus-trations based on atomic structures of individual biomole-cules, electron micrographs of cells, and biochemical studiesof molecular concentrations, locations, and interactions [12–15]. These molecular landscapes provide a framework forintuitive understanding that prepares the student’s mind toreceive the onslaught of facts we provide about pathways,interactions, cellular context, protein dynamics, and soforth.

As part of one of the CREST team projects, we havecreated a molecular landscape of VEGF-receptor signalingat two million times magnification. In the landscape, all ofthe proteins are present at approximately accurate relativebiological concentrations, with shapes that attempt to rep-resent their observed domain architectures. Although thelandscape presented here was made possible by collabora-tion with a skilled molecular artist and co-author of this ar-ticle (DG), our team believes that any creative approach—regardless of skill level—to constructing a molecular land-scape will be useful in very similar, if not the same, educa-tional ways.

Construction of the MolecularLandscapeTo integrate the cellular context effectively, it was neces-sary to go beyond the proteins connected with the classproject in order to ensure that we had identified all the keyprotein players in the VEGF-angiogenesis pathway. To doso, we performed an initial KEGG inquiry into the VEGF-angiogenesis pathway, followed by primary literaturesearches. This work was coupled with meetings with thesame researchers who participated in the in-class phase ofthe project. The meetings helped to sort out inconsistenciesin the literature and to provide a realistic reflection ofresearchers’ current understanding of the pathway. Fromthe information gathered, we generated a pathway dia-gram (Fig. 2).

We then compiled characteristic information abouteach of the proteins involved in the pathway, includingsize, domain arrangement/shape, biological assembly, mod-ifications, and their interactions/functions. These character-istics, along with sequence information from UniProt [16]

and PDB ID numbers (where available), were used to gen-erate a protein description table (Fig. 4). Based on the com-piled information, a series of sketches were produced bythe molecular artist (Fig. 5). Several revisions were made,based on our pedagogical goals and the constraints of pro-ducing the painting. During this process, our team metwith researchers actively studying angiogenesis and VEGFpathways to iteratively refine the sketch of the signalingcascade. Meetings with the researchers over the course ofthe CREST project played a key role, serving the dual pur-pose of: (a) ensuring an accurate rendering of the proteinsin the cascade, with proper relative scale, subcellular loca-tion, and interacting partners, and (b) teaching studentsthe scientific process, including how much is unknown andwhere the gaps in our knowledge exist—in short, teachingwhy science is exciting, relevant, and dynamic in a waythat is not well-conveyed in textbooks.

Some of the features that developed with successivesketches, literature searches, and discussions included thefollowing:

1. In the pathway, molecules are shown in several states togive an idea of the process in action. This includes theVEGF receptor in monomeric unbound states and di-meric active states, various combinations of signalingproteins, and monomeric and heterodimeric versions ofC-fos and Jun. Initial sketches of the pathway had manyof the proteins assembled into a large complex; but, dis-cussions with active researchers suggested that thesecomplexes should be broken into smaller units.

2. The initial concept was to produce two panels, one ofthe cell surface and one of the nucleus, as the portion ofthe cell shown in a typical painting is not sufficient toshow both. Fortunately, we found an electron micro-graph of capillary endothelial cells, and these cells oftenhave the nucleus tightly pressed to the cell surface nearthe junctions between cells. So, it was possible to de-velop a scientifically accurate sketch that showed theadherens junction, cell surface, and nucleus all in onepanel (Fig. 6).

3. The initial sketches focused on a process where Hsp27modifies the polymerization of actin. Discussion withthose actively pursuing angiogenesis research, however,determined that this is a minor biological effect, so thefocus was then changed to show two effects: (a) the dis-assembly of tight junctions through the Src pathway andthe role of the released alpha-catenin in actin reorgan-ization, and (b) the control of transcription through theMAPK (mitogen-activated protein kinase) pathway andphosphorylation of C-fos by pERK-2.

Once these details were finalized, a full sketch was cre-ated, filling in all of the other molecules (Fig. 7). This drewlargely from the cellular panorama illustration from theMachinery of Life [17]. The cadherin/catenin structure of

Biochemistry andMolecular Biology Education

218 Protein Structure in Context: The Landscape of Angiogenesis

the adherens junction was adapted from an earlier paperin the Oncologist on the same subject [18] and a recent pa-per on the process of VEGF signaling [19]. The color designof the painting was one of the challenging aspects of the

project, as there were two conflicting goals that needed tobe addressed with color: 1) to highlight the molecules inthe two pathways, showing the order of signaling, and 2) tomake the various compartments of the cell apparent. The

Summary of the proteins in the VEGF-mediated angiogenesis signaling cascade (Fig. 2), including relevant structural in-

formation needed for drawing the molecular landscape.FIG 4

Span et al. 219

final color scheme uses a rainbow sequence of bright red-yellow-orange-magenta to convey the sequential action ofmolecules in the signaling cascade; blues, purples, andgreens were used for the other molecules in the cell. Thisscheme may also be observed in the landscape’s legends(Fig. 8).

Impact of the New Materials onResearchers, Educators, and StudentsThe collaborative creation of a series of physical models ofinteracting proteins and the molecular landscape of VEGFsignaling described here has had a significant impact onevery member of the team:

For students working on the project, the physical modelsof a protein facilitated meaningful conversations with theresearchers regarding specific details of a protein’s structurethat were being investigated in the lab. At the same time, theplanning of the landscape required a sophisticated synthesisof results from multiple research papers, leading to anunderstanding of how the destabilization of adherens junc-tions is a necessary prerequisite to angiogenesis, or to thetranscriptional activation of a specific set of genes in the nu-cleus. In addition to this factual knowledge, the studentswere active participants in a modeling project in which theresults of a wide range of experiments in different laborato-ries were synthesized into a conceptual model of how thissignaling pathway leads to angiogenesis. This experienceserved as a clear example of how the concepts that aretaught in a classroom are the direct product of experimentsperformed in a research lab. The students also had the rare

experience of working closely with an educator outside thenormal environment of a classroom. In the collaborativeenvironment of this project, the “teacher” was transformedin the eyes of the students into a “scientist” who was work-ing productively with a “researcher” to arrive at a deeperunderstanding of a molecular process.

(a) An initial sketch for the molecular landscape painting depicting angiogenesis signaling, and (b) a sketch after three

rounds of revisions, with markings suggesting further alterations. Relative protein sizes and shapes are based on the

data compiled from Fig. 4. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Simulated electron micrograph (based on ref. 20])

showing two cells from the vascular endothelium.

The size and location of the portion of the vascu-

lar endothelium that is to be depicted in the mo-

lecular landscape painting is shown in color. This

is the cellular context within which the angiogene-

sis signaling cascade is to be presented. [Color

figure can be viewed in the online issue, which is

available at wileyonlinelibrary.com.]

FIG 5

FIG 6

Biochemistry andMolecular Biology Education

220 Protein Structure in Context: The Landscape of Angiogenesis

The final molecular landscape depicting VEGF signaling. Blood plasma is shown in tan at the upper left. The adherens

junction between two cells is in green at left, with surface VEGF receptors shown in yellow. The cytoplasmic proteins

are in turquoise, and the nuclear pore is at the center in green. The nucleus is at the right, with proteins shown in blues

and purples.

Expanded and color-coded views of the molecular landscape from Fig. 7. (a) VEGF signaling, across the cell membrane

into the cytosol. VEGF-A (1) in blood plasma binds to VEGFR (2) and causes it to dimerize. This activates the VEGFR tyrosine

kinase domains inside the cell. Two downstream pathways are shown. VEGF Pathway-1: C-src (3) is phosphorylated, caus-

ing it to open up and phosphorylate cadherins (4) in the adherens junctions, releasing alpha-catenin (6), which dimerizes

and bundles actin. Beta-catenin (5), which is involved in the adherens junction structure, is also shown]. VEGF Pathway-2:

VEGFR dimerization initiates a cascade of phosphorylation reactions through PLC-gamma (7), PKC (8), Raf-1 (9), MEK (10),

and ultimately ERK-2 (11). Phosphorylated ERK-2 (pERK-2) is transported through the nuclear pore (the large structure at the

center of the full painting). (b) Continuing the signaling cascade into the nucleus. pERK-2 (11) phosphorylates C-fos (12),

causing it to form a heterodimer with Jun (13) and becoming active as a transcription factor effecting transcription of pro-

teins needed for blood vessel growth. It is shown here in an enhancer binding to the transcription mediator (14) and RNA po-

lymerase (15). Finally, DUSP-5 (16) terminates the process by dephosphorylating pERK-2. This final interaction is depicted

with the physical models in Fig. 3.

FIG 7

FIG 8

The educators in this project also reported multiplebeneficial outcomes, including new insights into the molec-ular mechanism of a specific signaling pathway, and a spe-cific example they can use in their classrooms to demon-strate how reductionist biochemistry (details of interactingprotein structures involved in ligand binding and subse-quent activation of phosphorylation domains) combineswith cell biology to explain the molecular process of howangiogenesis is activated in tumors. In addition, the educa-tors reported a new appreciation for the talents and skillsof the students involved in this project. In this and othermodeling projects in the CREST program, educators reportthat the instructional materials that result from these col-laborations simply would not have been developed withoutthe efforts of students who readily embraced the use ofnew technologies in exploring the topic.

The researchers benefit directly from the final productof the project—they receive both a physical model of a pro-tein and a molecular landscape they can use to communicatetheir research to others. The landscape can be used to intro-duce new students to the “big picture” of the lab’s work,while the physical model of a protein can be used to discussthe specific details of the protein’s structure that are believedto play a critical role in the process. These same tools canalso be used to explain their work to colleagues visiting fromother institutions. Researchers also report satisfaction fromworking with undergraduate students and contributingdirectly in the training of a future generation of researchers.A researcher’s experience in working with a CREST model-ing team is often the focus of the “broader impact” statementthat is required of some research proposals.

It cannot be ignored that this work has had at its dis-posal cutting-edge resources and a network of innovative,skilled team members. This being so, the educationalmodalities we have created or applied are not constrictedto groups with identical resources; in fact, we argue thatthe same educational goals, and perhaps others, may beattained through diverse means. Some may see a lack ofaccess to rapid prototyping technology as a barrier to con-structing physical protein models, but this is less and lessthe case: many schools, even some high schools, have 3Dprinters in their art and engineering departments, and inmany cases these resources may be shared. If teachers donot have access to 3D printers, 3D Molecular Designs, LLC(affiliated with the MSOE Center for Biomolecular Model-ing) can print models for a fee (cost varies depending onsize and complexity, typically ranging from $100 to $400).In addition, the proficiency in Jmol (or other protein visual-ization software tools) required to design the models priorto 3D printing may be accessed through online tutorials.Finally, existing molecular landscapes may be used as toolsin themselves; for example, they may be published onlinein clickable formats for student-users to actively learn theprinciples and content of signal transduction.

The process of constructing a molecular landscapearguably has even more value than using an existing land-scape, and the process can take multiple forms. The artisticquality of the final painting is of secondary importance—the salient matter is the process of assembling the informa-tion needed to construct the image. For example, the land-scape could be prepared as a crude sketch with pencil andpaper; alternatively, established graphics design softwarecould be used to build protein shapes approximating rela-tive scale, cellular concentration, and location. One goal ofthe ongoing CREST project is to create an online LandscapeCreator software package with which students can con-struct a landscape using a palette of images of commonlyencountered proteins, membranes, and nucleic acids.

Other landscape modalities may succeed in othercapacities that our team did not choose to focus on, suchas the capacity for team-building, or for group discussion/learning: as an example, entire classes (small college orhigh school) could use an online landscape creator tool tobuild a chosen signaling pathway. In this setting, decision-making would have to occur as a group and could fuelinteresting and valuable dialogue. Regardless of how thelandscape is constructed, students will be confronted withdesign decisions that require them to learn not only specificinformation about the protein players and their interactingpartners, but also the skill of critically engaging with scien-tific literature and databases. Besides learning difficult con-tent, they will begin to develop the skills set that will makethem strong scientific thinkers. This dual instructionalmode endows molecular landscape projects with high edu-cational value and makes them decidedly promising for fur-ther classroom use.

AcknowledgmentsThe authors thank Dr. Qing (Robert) Miao and Dr. GeorgeWilkinson at the Medical College of Wisconsin and Dr. Mag-dalena Chrzanowska-Wodnicka at the Blood ResearchInstitute in Milwaukee, WI, for mentoring students and forhelpful discussions on the VEGF-angiogenesis pathway, aswell as Dr. Terrence Neumann at Concordia University forcreative input. Disclosures: One of the co-authors (TH) hasan ownership interest in 3D Molecular Designs. This workwas supported in part by NIH research grants HL102745-01A1 and HL112639-01, as well as by NSF CREST grantCCLI #1022793.

References

[1] Harris, M. A., Peck, R. F., Colton, S., Morris, J., Chaibub Neto E., Kallio, J.

(2009) A combination of hand-held models and computer imaging pro-

grams helps students answer oral questions about molecular structure

and function: A controlled investigation of student learning. CBE Life Sci.

Educ. 8, 29–43.

[2] Available at: http://www.rcsb.org/pdb/ (accessed August 24, 2012).

[3] Available at: http://www.bmrb.wisc.edu/ (accessed August 24, 2012).

Biochemistry andMolecular Biology Education

222 Protein Structure in Context: The Landscape of Angiogenesis

[4] Kramer, I. M., Dahmani H. R., Delouche, P., Bidabe, M., Schneeberger,

P. (2012) Education catching up with science: Preparing students for

three-dimensional literacy in cell biology. CBE Life Sci. Educ. 11, 437–

47.

[5] Nye, L. S. (2004) The mind’s eye-biomedical visualization: The most

powerful tool in science. Biochem. Mol. Biol. Educ. 32, 123–131.

[6] Herman, T., Morris, J., Colton, S., Batiza, A., Patrick, M., Franzen, M.,

Goodsell, D. S. (2006) Tactile teaching—Exploring protein structure/

function using physical models. Biochem. Mol. Biol. Educ. 34, 247–

254.

[7] Roberts, J. R., Hagedorn, E., Dillenburg, P., Patrick, M., Herman, T.

(2005) Physical models enhance molecular three-dimensional literacy in

an introductory biochemistry course. Biochem. Mol. Biol. Educ. 33, 105–

110.

[8] Fleming, N. D. and Mills, C. (1992) Not another inventory, rather a cata-

lyst for reflection. Improve Acad. 11, 137–146.

[9] Available at: http://www.genome.jp/kegg/ (accessed August 24, 2012).

[10] (2007) About BioVisions, BioVisions at Harvard University, Avail-

able at: http://multimedia.mcb.harvard.edu/ (accessed August 24,

2012).

[11] Bolinsky, D. (2007) Visualizing the wonder of a living cell [Video file].

Available at: http://www.ted.com/talks/david_bolinsky_animates_a_

cell.html (accessed December 30, 2012).

[12] Goodsell, D. S. (1991) Inside a living cell. Trends Biochem. Sci. 16, 203–

206.

[13] Goodsell, D. S. (2009) The Machinery of Life, 2nd ed., Springer, New

York.

[14] Goodsell, D. S. (2009) Illustrating the machinery of life: Escherichia coli.

Biochem. Mol. Biol. Educ. 37, 325–332.

[15] Goodsell, D. S. (2012) Putting proteins in context. Bioessays 34, 718–

720.

[16] Available at: http://www.uniprot.org/.

[17] Goodsell, D. S. (2011) Eukaryotic cell panorama. Biochem. Mol. Biol.

Educ. 39, 91–101.

[18] Goodsell, D. S. (2002) The molecular perspective: Cadherin. Oncologist

7, 467–468.

[19] Dejana, E., Orsenigo, F., and Lampugnani, M. G. (2008) The role of

adherens junctions and VE-cadherin in the control of vascular perme-

ability. J. Cell. Sci. 121, 2115–2122.

[20] Available at: http://www.histology.leeds.ac.uk/circulatory/capillaries.php.

Span et al. 223