A historical perspective on training students to create standardized maps of novel brain structure: Newly-uncovered resonances between past and present research-based neuroanatomy curricula

Abstract

Recent efforts to reform postsecondary STEM education in the U.S. have resulted in the creation of course-based undergraduate research experiences (CUREs), which, among other outcomes, have successfully retained freshmen in their chosen STEM majors and provided them with a greater sense of identity as scientists by enabling them to experience how research is conducted in a laboratory setting. In 2014, we launched our own laboratory-based CURE, Brain Mapping & Connectomics (BMC). Now in its seventh year, BMC trains University of Texas at El Paso (UTEP) undergraduates to identify and label neuron populations in the rat brain, analyze their cytoarchitecture, and draw their detailed chemoarchitecture onto standardized rat brain atlas maps in stereotaxic space. Significantly, some BMC students produce atlas drawings derived from their coursework or from further independent study after the course that are being presented and/or published in the scientific literature. These maps should prove useful to neuroscientists seeking to experimentally target elusive neuron populations. Here, we review the procedures taught in BMC that have empowered students to learn about the scientific process. We contextualize our efforts with those similarly carried out over a century ago to reform U.S. medical education. Notably, we have uncovered historical records that highlight interesting resonances between our curriculum and that created at the Johns Hopkins University Medical School (JHUMS) in the 1890s. Although the two programs are over a century apart and were created for students of differing career levels, many aspects between them are strikingly similar, including the unique atlas-based brain mapping methods they encouraged students to learn. A notable example of these efforts was the brain atlas maps published by Florence Sabin, a JHUMS student who later became the first woman to be elected to the U.S. National Academy of Sciences. We conclude by discussing how the revitalization of century-old methods and their dissemination to the next generation of scientists in BMC not only provides student benefit and academic development, but also acts to preserve what are increasingly becoming “lost arts” critical for advancing neuroscience – brain histology, cytoarchitectonics, and atlas-based mapping of novel brain structure.

We find that with the research method of teaching we can lead the student much further into the subject than without it. Students do better work when you expect much of them than when you expect little.

– Franklin P. Mall (1896)[1]

In my opinion the list of those suited for scientific work is much longer than generally thought, and contains more than the superior talents, readily adaptable, and keen minds ambitious for reputation and eager to link their names with a major discovery. The list also includes those ordinary intellects thought of as skillful because of the ability and steadiness they display for all manual work, those gifted with artistic talent who appreciate deeply the beauty of Nature’s work, and those who are simply curious, calm and phlegmatic devotees of the religion of detail, willing to dedicate long hours to examining the most insignificant natural phenomena. Science, like an army, needs generals as well as soldiers; plans are conceived by the former, but the latter actually conquer.

– Santiago Ramón y Cajal (1916/1999) [2]

1. Introduction

From the perspective of today’s neuroscientist-educator, it is perhaps surprising that the views excerpted above – articulated over a century ago – embody two aspects of science education that are considered tacitly, if not explicitly, as “innovations” important for the scientific enterprise today (e.g., [3, 4, 5, 6]); namely, engaging students in scientific research as a means to (1) better train them on scientific subjects [7, 8, 9, 10, 11]; and (2) accelerate and better enable our scientific advances through student contributions [12, 13]. As the quotations make clear, these ideas are not new, and, in fact, have likely punctuated discussions since antiquity concerning the best ways by which students learn various subjects.

The reflections of Mall and Ramón y Cajal are also a reminder that current education reform efforts may benefit from examining how similar movements produced successful outcomes in the past. Here, we juxtapose two neuroanatomical education programs that illustrate certain broad parallels between the medical education reforms at the turn of the 20^th century in the United States [14] and those currently underway for undergraduate education [3, 5, 11, 12]. In particular, we examine similarities between the neuroanatomy and histology course curriculum of the Johns Hopkins University Medical School (JHUMS) in the 1890s [15] and the brain mapping laboratory course curriculum we have created at The University of Texas at El Paso (UTEP) for first-year undergraduates [16]. Though the courses were designed over a century apart and for different student cohorts (Figs. 1, 2), they both leverage, to differing extents, discovery-based research to empower students in the scientific process. The student projects from both the JHUMS and UTEP courses are likewise unique in that they both involve the use of brain mapping techniques and brain atlas construction, albeit in different ways. Importantly, both programs also enabled dedicated students to transform their independent research projects into scientific products in the field of neuroanatomy. We offer new insights on previously-published material that we have uncovered of JHUMS graduates that demonstrates how JHUMS and UTEP students have used similar methods to identify novel brain structures and produce original standardized maps of such structures for the scientific community. Our rationale for comparing the JHUMS and UTEP curricula is that, to our knowledge, no other laboratory-based neuroanatomy course (at that time, or at present) has empowered students to perform standardized mapping and annotation of mammalian brain tissues to produce atlas maps of brain structure. Even as calls for standardization of complex neuroanatomical datasets have emerged [e.g., 17, 18], few, if any, courses have this goal in their curricula. Below, we describe each program in turn and then identify the significant research products of both courses, concluding with a discussion of how the resonances between the two programs offer unique lessons as neuroanatomy educators continue to develop courses for the next generation of scientists and clinical practitioners.

Figure 1.

The UTEP Brain Mapping and Connectomics laboratory CURE and its activities. (A): A photograph of a few students with their instructors from the 2016 class within the teaching laboratory. (B–D): Student activities in the course include: studying the cytoarchitecture from images of the rat brain labeled with thionine, a Nissl stain (B); constructing digital overlays of the boundaries determined by Nissl-stained material onto images of immunocytochemically labeled tissue sections of an adjacent tissue series using vector-graphics software (C); and discussions by team members about their mapping determinations with instructors of the course (D). Photographs taken by AMK under authorized UTEP IRB approval.

Figure 2.

(A): Photograph, dated May of 1913, showing, from left to right, Lewellys F. Barker, Professor of Clinical Medicine at JHUMS, William Osler, founding director and former Professor of Clinical Medicine of JHUMS, and Frank J. Sladen, chief medical resident at Johns Hopkins Hospital. Osler posed for this photograph during his final visit to North America in April–May of 1913, when he visited Johns Hopkins University and Yale University. Barker succeeded Osler as director of the JHUMS when Osler was recruited to the University of Oxford. (B): Photograph, dated 1893, of the early staff of the Johns Hopkins Hospital. Lewellys Barker is seated in a chair in the second row, second from left. William Osler is seated in the same row, third from right. (C): Photograph, dated 1900, of the Johns Hopkins School of Medicine, Class of 1900. Florence Sabin is standing in the second row on the far-left side. (D): Portrait, dated 1911, of Franklin P. Mall, Sabin’s mentor and Director of the Anatomical Laboratory, sitting at a desk at Johns Hopkins University. (E): Portion of the table of tissue preparations published [15] for the 1895 Normal Histology and Anatomy laboratory course at JHUMS taught by Lewellys F. Barker and his assistant, Charles R. Bardeen. (F): Photograph, dated between 1925–1938, of Florence Sabin, sitting at a lab bench with a microscope at the Rockefeller Institute for Research, where she was a full member and directed her own laboratory. Photo credits: (A), (C), (D), (F): U.S. National Library of Medicine, History of Medicine Division, Profiles in Science Collection. (B): Photo reproduced from [56]. (E): Digital photograph reproduced from [15]. All photos are in the public domain.

2. The Brain Mapping & Connectomics Laboratory at UTEP (est. 2015)

2.1: Background.

As the science, technology, engineering, and mathematics (STEM) higher education landscape has shifted toward providing greater access to students over the past decades, there is a need to revise instructional strategies to serve a student population with a more diverse educational background and discipline-specific proficiency [3, 6, 8, 11]. The resulting consensus among biology educators is to move toward “student-centered” pedagogies that focus on natural discovery through student engagement in mentored research [6, 11]. Course-based undergraduate research experiences (CUREs) have gained great traction over the last decade as one effective mechanism to achieve that goal. They serve to replace existing traditional laboratory courses by providing students with the ability to collaboratively explore novel research questions in an environment that fosters high student autonomy [4, 5]. Over the past decade, assessment of the CURE model has repeatedly demonstrated significant, positive impacts on student retention in STEM coursework and degree completion, student enjoyment, and increased identity as a scientist [19, 20, 21, 22, 23]. CUREs have facilitated greater reasoning skills development and increased related-content exam scores among underrepresented versus well-represented student peers [19, 24]. Among first-year students, the access to early research acculturation builds an appreciation for the relevance of current and future course material, enhances student autonomy, and offers students the chance to offset peer imbalances in college preparation status, all of which prime students for the remainder of their academic careers [7, 10, 19, 21, 25].

Undergraduate neuroscience programs in the United States have proliferated over the past decade, and parallel to this, curricular interventions ranging from full-semester courses to stand-alone modules have been developed to address competencies among a wide range of neuroscience sub-disciplines including molecular histology, electrophysiology, and behavioral physiology (see also Appendix I in D’Arcy et al. [16] for highlighted active learning and CURE curricula examples) [26, 27]. However, with limited exceptions, the CURE format has not been extensively applied to translational neuroanatomy instruction. Moreover, published curricula for neuroanatomy-specific or neuroanatomy-supportive CUREs directed toward introductory-year students are less frequently identified within extant literature compared to upper-division offerings [16].

2.2: Introducing the course.

Capitalizing upon the synergy of first-year experiences and positive CURE outcomes among students, and with the aid of a grant from the Howard Hughes Medical Institute, we developed Brain Mapping and Connectomics, a freshman-level, two-semester CURE sequence for general biology students, as described in [16]. Enrollment in this sequence is open to any student who has met the necessary course prerequisites. Student demography is, on average, 79% Hispanic and 80% female, with 20% of individuals self-identifying as first-generation college attendees and as many as 50% indicating that English is their second language (approximately 5% self-identified as Asian, 5% as Black/African-American, and 12% as White). Over 60% of the majors represented in the sequence are within biology tracks (including the pre-medical track) offered at our institution, with physics, psychology, forensic science, kinesiology, biochemistry, and engineering also frequently represented in the course population. These data underscore a diverse representation of >100 first-year students served by this CURE since its inception.

2.3: Learning objectives.

In structuring the course sequence, we identified four student learning outcomes: 1) to understand terminology pertaining to nervous system structures; 2) to be able to ascribe functional salience to the structures of the nervous system; 3) to be able to predict or observe the functional, physiological, or behavioral outcomes from a manipulation of or alteration to a component of the nervous system; and 4) to be able to translate objectives 1–3 across multiple anatomical scales: molecular, mesoscopic, and macroscopic.

2.4: Overview of procedures and workflows of students.

Detailed elsewhere [16], the BMC course incorporates a blend of modern-day tools (immunofluorescence microscopy with digital image capture, software to manipulate vector graphics) and classical neuroanatomical approaches (Nissl staining, cytoarchitectural study) carried out in a model organism (rat) to engage students in identifying chemical phenotypes of neurons in a functionally-complex region of the rat brain (hypothalamus). The course is designed to provide mentorship to the students and requires that students demonstrate an increasing degree of autonomy in carrying out the research itself while receiving regular formative feedback from course instructors.

Specifically, students collaborate to generate sets of high-spatial-resolution chemoarchitectural maps of hypothalamic neurons and their axonal projections using a standardized reference atlas of the brain [28]. Students learn how to carry out a sequence of processing steps, some of which are shown in Fig. 1: Nissl staining, mounting Nissl-stained tissue, immunocytochemistry, mounting immunolabeled tissue, parcellation of Nissl-delimited regions within the tissue sections, transferring of those boundaries to images of the immunolabeled series, drawing structures in a vector-graphics platform, and transferring those drawings to brain atlas templates in vector-graphics format.

2.5: Essential features.

By examining a Nissl-stained preparation of a reference series of coronal-plane brain tissue sections (Fig. 1B), students are allowed to intuitively explore key fiducial structures (white and gray matter, ventricles) for use as registration marks to identify the corresponding atlas level and how the tissue’s plane of section may deviate from it. Furthermore, by assigning boundaries (parcellation) to the observable structures in their reference series (Figs. 1C; 3B) and using these boundaries to identify corresponding structures in their immunocytochemically-stained series (Figs. 1C; 3D), students can gain a sense of the visual “texture” unique to each subregion arising from the cytoarchitecture. This provides additional cueing for the recall of neuroanatomical terms and further enables students to intuitively connect names with places in the brain [29, 30, 31, 32]. Through immunohistochemical preparations (Fig. 3D), students are able to visualize the distributions of neuronal populations expressing neuropeptides and are able to build a working knowledge of neuropeptide function. This enables them to interpret the relevance of their observations in a broader physiological context and to hypothesize relationships among neuronal phenotypes for future exploration. Moreover, they emerge with a skillset that preserves the classic approaches to neuroanatomy while simultaneously providing them with a greater appreciation for the processes of discovery and observation that are the roots of scientific advancement.

Figure 3.

(A): A drawing reproduced from Dr. Florence Sabin’s 1897 study [72], published while she was still a medical school student. The drawing was made by L. Schmidt of a transverse section through a human infant brainstem, prepared by John Hewetson in the laboratory of Wilhelm His, Sr., at the University of Leipzig. This section corresponds approximately to line 4 in the flatmap shown in (C). (B): Photomicrograph of a Nissl-stained transverse section of the rat brain (Experiment #18-012), prepared and stained by Mr. Kenichiro Negishi and Ms. Vanessa I. Navarro, graduate student research assistants in the UTEP Systems Neuroscience Laboratory (USNL). The anatomical parcellations (boundary line overlays) were prepared by Ms. Diana Sotelo (USNL undergraduate student research assistant) with the guidance of Mr. Eduardo Peru (USNL graduate student research assistant). (C): A diagram, reproduced from Dr. Florence Sabin’s 1897 study [72], representing “flat reconstruction of the nuclei of reception of the cochlear and vestibular nerves” that she completed from tissue slides prepared by John Hewetson, which were visualized for myelinated fibers (Weigert-Pal stain). As noted in her study [72], the line a, a represents the lateral wall of the ventricle; the line b corresponds to the lateral outline of the corpus restiforme; the lines d₁ to d₄, d₁ to d₃, and e, e, e correspond to sulci in the floor of the fourth ventricle. (D): Photomicrograph of an immunocytochemically-stained transverse section of the rat brain (Experiment #18-012) processed by Ms. Navarro (see description for B, above). The labeling consists of a dual stain, one for an antibody recognizing α-melanocyte-stimulating hormone (α-MSH; black), and the other for an antibody recognizing the neuropeptide, hypocretin/orexin (H/O; light brown). The boundary overlays produced by Ms. Sotelo in (B) have been superimposed on this image. (E): Representations of the two immunocytochemical labels, shown in (D), mapped by Ms. Sotelo onto a digital atlas template (level 30) of the Swanson rat brain atlas [28]. The portion of the atlas template shown is for subregions of the hypothalamus. Pale blue lines represent axonal fibers containing the neuropeptide signal, α-MSH, and red filled circles represent neuronal cell bodies that contain the neuropeptide signal, H/O. The scale bar in (B) denotes 500 μm, and also applies to the photomicrograph in (D). Please consult the abbreviations list at the end of this article for an explanation of the abbreviations in these diagrams. Photo credits: (A), (C): Reproduced from Ref. [72], these photos are in the public domain. (B), (D), (E): Previously unpublished photographs and map prepared by members of the USNL, as noted above.

2.6: Learning outcomes: Scientific products and professional development.

Each team produces, as a final output, a map representing the patterns of chemoarchitecture that they have observed in each stained section, and an evaluation rubric (published in Appendix II of Ref. [16]) is used to provide feedback to each team. As detailed in our study [16], students were scored on a binary scale in four skills: using vector-graphics software, comparing the planes of section of their experimental tissue and atlas sections, correctly identifying gray and white matter regions, and accurately cataloging immunostained nerve cell bodies and axonal fibers. Four student cohorts graded across these categories scored between 10 and 14 out of 16 points, with an average score of 13.

A few enterprising students, upon completion of the course, are encouraged to join as undergraduate volunteers in the UTEP Systems Neuroscience Laboratory (directed by AMK), where they continue their mapping work, either by refining the maps produced in the class or working on a new mapping project altogether. An example of a final, refined map produced in this fashion is shown in Fig. 3E, which was prepared by Ms. Diana Sotelo, a former BMC student and undergraduate research assistant in the UTEP Systems Neuroscience Laboratory.

Thus, two sets of maps emerge from the activities of the laboratory course: those produced by the class and those refined or mapped later by students who have continued the work as research assistants. After these atlas maps of hypothalamic chemoarchitecture were thoroughly checked by graduate student teaching assistants and faculty supervising the course, they were submitted for presentation at the Annual Meetings of the Society for Neuroscience [33, 34, 35, 36, 37, 38], a Keystone Symposium [39], and a Janelia Farm Workshop [40]. To date, 16 course undergraduates have been presenters among 97 fellow co-authors on conference abstracts/posters. Several of these students are also co-authors on the papers reporting this work, which are currently in final preparation for submission to peer-reviewed journals. Others have gone on to join neuroscience laboratories at UTEP and are published co-authors as undergraduate students on studies with their faculty supervisors (e.g., [41, 42, 43]). Importantly, BMC students have not only successfully graduated as STEM majors, but several have now secured medical or graduate school admissions around the U.S.

The core of this CURE, and others like it, centers on students engaging with subject content via diverse modalities (reading, discussion, application) at a natural pace as they progress through the activities of their research [30, 31, 32, 44, 45, 46]. Salient information is sought out by students as they encounter a need to know more [30, 45] (Fig. 1D). Moreover, through iteration and refinement of their products, students gain in-depth exposure to neuroanatomical terminology and have constant reinforcement of the anatomical boundaries, characteristics, and juxtaposition of neuroanatomical structures.

Students take great strides toward understanding the broader organization and interconnectedness of the brain by engaging with their own atlas maps of brain chemoarchitecture. Within the same model space, they superimpose multiple datasets as overlays (Fig. 1C) with the use of vector-graphics programs, such as Adobe Illustrator, in lieu of onion skins, acetate, or, in the case of JHUMS-sponsored investigations [47] in the 1900s, “paraffined wrapping paper” (p. 86). In vector-graphics space, each neuronal and axonal distribution pattern for a given neurotransmitter or neuropeptide can be given a unique color representation and its own data layer, and visibility can be toggled freely for multiple data layers in various combinations (Figs. 1C, 3E). Superimposing multiple patterns of distribution of entities in the same map space has historically provided important scientific insights that are otherwise not possible to obtain, much as novel perceptions are created by listening to an elegantly-composed polyphonic music texture containing multiple melodic lines (e.g., [48]; p. 36 of Ref. [49]; also see Section 5 in [50]; [51]). In parallel, students derive a thorough and contextualized exploration of neuroanatomy through experimentation and practice in the careful drawing of diverse datasets onto digital vector-graphics atlas templates.

Yet, stripping away the tools of modernity from this process, we see the underlying practice of teaching neuroanatomy through neuroanatomical mapping. In fact, a simple camera lucida microscope and printed copies of the atlas templates, which are freely downloadable [28], are all that are needed to map the locations of neurons in stained sections mounted on slides. The basic exercise of drawing boundaries and patterns of brain structures to create atlas maps was also present within the curricular activities of JHUMS medical students over a century past, suggesting that this exercise is a reliable means to empower students across diverse career levels in the scientific process. In the next section, we pivot to the past and describe details of the JHUMS course, first providing historical context regarding the larger medical education reform movement within which it was created. In Section 4, we compare the UTEP and JHUMS courses. Readers are invited to consult Table 1 to see a full comparison of the two courses.

Table 1:

Comparison of JHUMS and UTEP curricula in the student teaching laboratory

Parameter	Johns Hopkins (est. 1894) [15]	UTEP (est. 2015) [16]
Student cohort
Level	College graduate	High school graduate
Prerequisite training	1–2 years of college physics, chemistry	None
Language prerequisite	German, French	English
Career path	Physician	Variable
Pedagogy
Type of instruction	Medical school histology and microscopic anatomy laboratory	Freshman biology laboratory
Subject of instruction	General histology of the whole body	Histology of brain tissue sections, immunocytochemistry, atlas-based mapping
Format of instruction	lab techniques, lectures, demonstrations	lab techniques, discussions, team projects
Duration of instruction	~October 1 – March 15, 15 hrs/week ~330 hours total	~August 25 – May 5, 6 hrs/week ~210 hours total
Supplementary texts	textbooks and scientific papers in English, German and French	scientific papers in English, atlas of the rat brain in English [28]
Audiovisual aids	darkroom projections, charts, models, light/dissecting microscopes	Internet resources, Powerpoint slides, Adobe Illustrator digital atlas templates, iMac workstations, Wacom drawing tablets, light/dissecting microscopes
Techniques/Research
Tissue sectioning	Yes	No
Tissue mounting	Yes	Yes
Histological staining	Yes	Yes
Immunocytochemistry	No (not yet invented!)	Yes
Drawing of tissue section details	Yes	Yes
Features being drawn	Myeloarchitecture	Cytoarchitecture
Mapping anatomical relationships	For some advanced students	For all students

3. The Normal Histology & Microscopic Anatomy Course at JHUMS (est. 1894)

3.1: Background.

The establishment of the Histology and Anatomy Laboratory at JHUMS [1], and within it, the Normal Histology and Microscopic Anatomy Course [15], occurred as part of larger shifts, not in undergraduate, but in medical education reform [14]. Many leaders of this movement received post-graduate training in Europe, particularly in Germany, where they perceived an education system that was highly organized, producing well-trained and highly-skilled medical practitioners [52] (but see [53]). Thus, these medical educators felt that the standards for entry into U.S. medical schools needed greater rigor and that the curricula within these schools should be aligned closely with laboratory work so that students could appreciate the scientific process [52]. To transmit this expertise from the continent to the U.S. became the clarion call for these early medical education reformers (also see [54]).

In 1889, Johns Hopkins Hospital was established in association with the newly-created Johns Hopkins University (founded in 1876), with an aim to have clinical practice associated with the university. A medical school was established in 1893, with eighteen students admitted as the first freshman class ([55], I, pp 388–389). Per its mandate, JHUMS had more stringent admissions requirements than other contemporary schools, requiring that medical students have college undergraduate degrees, including two years of pre-medical science training and a reading knowledge of German and French [15]; (also: [55], I, pp 388–389). The Canadian physician, William Osler, was recruited to be the first director of the medical school (Fig. 2A, B). Among Osler’s pioneering reforms that transformed U.S. clinical practice was the hiring of physician-scientists, including Franklin P. Mall (quoted in the Introduction; Fig. 2D), to establish “pre-clinical” departments, such as Anatomy and Pathology.

Arriving at JHUMS, Mall urged Lewellys Barker (Fig. 2A, B), his newly appointed Associate in Anatomy, to gain an appreciation for anatomical research in German laboratories. Accordingly, Barker and his colleague, John Hewetson, visited Leipzig in 1895 [56]. Barker’s training in Leipzig, especially from Fleschig’s lectures [56], led him to become the strongest stateside proponent of the neuron doctrine, which he popularized [57, 58] and incorporated in his teaching curriculum. Convinced of the need for medical students to participate in research activities, he embodied this philosophy in his course [56] and modeled the first-year medical school curriculum of anatomy and histology along the lines of what he and Mall perceived was the practical laboratory training in German university research laboratories. This was extended later to develop an intellectual culture at JHUMS that focused on the current research coming out of European laboratories (e.g., [59]). Barker wrote a neuroanatomy textbook centered upon the neuron doctrine [60] in which he detailed the findings from the major European research centers and extensively reproduced their histological drawings, especially Ramón y Cajal’s, whose carefully-prepared tissue slides he also later examined firsthand in Ramón y Cajal’s laboratory [56, 61]. The JHUMS training program was unique at the time, in part, because its students were among the first U.S. medical students to receive training on the neuron doctrine and the cellular organization of the nervous system. As we shall see, this training had a lasting impact on the next generation of medical scientists and practitioners.

3.2: Introducing the course.

Barker, together with his assistant Charles Bardeen, a JHUMS medical student at the time, taught the Normal Histology and Microscopic Anatomy Course. The 22-week course served first-year medical students, with three full days per week of laboratory instruction, followed by time on research projects for interested students (Table 1; [1, 15]). The course, or perhaps an early version of it, was first established in 1894, and in Bardeen’s view, was “the first good course offered in microscopic neurology given in a medical school in this country.” (p. 97 in Ref. [62]). A larger-scale and more focused course in central nervous system anatomy was later taught to sophomore medical students [63]. The main activities centered upon supervised practical work, alongside sixty lectures that were accompanied by demonstrations and the use of charts, models, and lantern-slide projections of images in a darkroom, often derived from exemplar specimens generated by the students themselves [15, 64]. The published account [15] that Barker and Bardeen give is a snapshot of the 1895–1896 course during its relatively formative stages.

3.3: Learning objectives.

From their early account [15], it can be gathered that the course was designed for medical students to focus on the human body and its multiple systems and organs, supplemented by studies of other animal tissue samples where appropriate and necessary. Students first participated in dissection and gross anatomical investigations of animal (especially human) tissues, drawing and studying tissue samples at a macroscopic level. They moved to microscopic examinations of various tissues to prepare them for later microscopy within specialty subjects (surgery, pathology, obstetrics, neurology, etc.). They were also required to participate in supervised self-study, where they gained an in-depth appreciation of histological techniques by preparing their own tissue samples [15].

Often, students transformed their course efforts into a broader, independent project that was usually finished after the course was completed, an approach that reflected the larger culture within the medical school for students to contribute original and independent projects for scientific dissemination [65] (transcript, p 74). Although the specific projects derived from Barker and Bardeen’s course are detailed in Section 3.6, it is worth noting here that this self-study and independent generation of published experimental data appears to have been unique to JHUMS medical students, insofar as the generation of novel atlas-based mapping of the brain was concerned. It is clear that the course was among the first to apply the neuron doctrine to the study of serial human brain tissue sections, and Bardeen’s own thorough 1905 account [62] of the history of neuroanatomical education in the U.S. highlights the uniqueness of the JHUMS course.

3.4: Details of procedures and workflows of students.

Course activities, apart from gross anatomical dissections, centered upon teaching general histology to the students, who were encouraged to isolate, manipulate, and dissociate tissues before they began detailed microscopic studies of them. In the course, coverage of the nervous system occurred toward the curriculum’s conclusion. It was in this unit that Barker’s enthusiasm for incorporating aspects of the neuron doctrine as part of the instruction made itself felt.

…[W]e have arranged the subjects so that especial attention should be given to those parts of the body in which at present most advance is being made. Thus the sense organs and nervous system, it will be seen, have been examined in considerable detail, and as far as possible the students are made conversant with the newer ideas concerning the minute anatomy and histogenesis of the peripheral and central portions of the nervous system…. A series of sections at various levels of the spinal cord and brain are …carefully studied and drawn, and finally several days are devoted to the study of tracts, thus bringing the student’s knowledge together in a more or less orderly fashion. [15], p 102.

The “series of sections” Barker refers to include the set of slides of two human infant brains, sectioned in two planes, that Barker had brought from Leipzig, and which were prepared there by his colleague, Hewetson. Students used these and similarly-prepared slides of adult human brain from which to construct drawings and determine boundaries for various gray matter regions and fiber tracts [15]; ([56], p 60) (Fig. 2E). From these slides, students began learning the boundaries and spaces through which fibers traveled in the brain, the gray matter regions associated with them, and how to visualize these pathways in three dimensions from their two-dimensional drawings. Although it is not clear whether drawings from the students have survived from iterations of this course during those early years, we have a detailed understanding of what these tissue sections looked like in both planes, because professional drawings of them were published in Barker’s textbook [60] (see example in Fig. 3A). Importantly, these tissue slides were also used by a few of Barker’s students to publish original research, as discussed in Section 3.6.

3.5: Essential features.

The JHUMS course included microscopic studies of serial tissue sections of the nervous system stained for cellular features that were still being discussed (and debated) by scientists of that time. The students were trained to carefully draw two-dimensional microscopic gray matter regions and fiber systems from tissue sections and infer their three-dimensional positions. Lectures were frequent, and although they appeared to be formal, a medical student at that time who attended a related laboratory course (pathological histology) has recalled how the lectures of Barker and other instructors for that course were delivered in an informal style [66]. Finally, self-study was encouraged outside of normal class hours, and some students availed themselves of such time more than others to pursue independent projects that led to the publication of novel reports.

3.6: Scientific products and professional development.

A novel method pursued by JHUMS investigators involved the use of wax reconstructions of serially-sectioned brain tissues to create three-dimensional structures of the brain, a laborious process that is somewhat analogous to 3-D printing technology today. This method was devised by Gustav Born [67] and later adapted for use in Mall’s Anatomical Laboratory [68]. Mall and Barker encouraged a few medical students, including Florence R. Sabin and Harry A. Fowler (in the graduating classes of 1900 and 1901, respectively) [69] (see Fig. 2C for a photograph of Sabin with her graduating class); to take Hewetson’s microscope slides and utilize the Born method to create three-dimensional models of the human infant brain. Fowler published a brief report of a three-dimensional model of the cerebellum in the year he graduated from JHUMS [70], whereas Sabin’s work on Hewetson’s slides was showcased in a textbook [60], articles [71, 72, 73], and as an atlas [74]. Of Sabin’s datasets, three were published either before graduation or during her graduation year [60, 71, 72], and the remainder afterwards, indicative of her deep investment in independent study while still a student.

It is the drawings of Sabin’s that are of interest here. Specifically, Sabin traced the myelinated fiber systems and their associated gray matter regions from Hewetson’s Weigert-Pal-stained tissue sections to construct a “flat reconstruction on millimetre paper” ([72], p 253) – or, what is now termed a flatmap – of their entry points and courses along the anteroposterior axis of the human infant brain (Fig. 3B). Sabin published two such flatmaps as a student in 1897 [72], which, to our knowledge, are certainly among the earliest, if not the first, flatmap representations of the brain found in stateside scientific reports. Sabin consolidated her earlier flatmaps into an expanded flatmap for Barker’s textbook in 1899 [60], where it was inserted into the large volume at multiple locations (Figs. 325, 407, 593) to illustrate various aspects of the human brainstem and cranial nerves at a mesoscopic scale. Her 1897 flatmaps were also reproduced at various locations in Barker’s book (Figs. 335, 367, 411), along with two of her 3-D models (Figs. 610, 611). As discussed in the next section, the process by which these maps were produced by Sabin bear striking similarities to the methods used by our BMC students at UTEP.

Florence Sabin’s student work has been cited in peer-reviewed studies by several neuroanatomists and embryologists studying the hindbrain and was also referenced later in leading textbooks and chapters by influential scientists in these fields [75, 76, 77]. Her brainstem flatmaps prefigure a later flatmap construction of this structure devised using different methods [78] and also anticipate later, more complete, flatmap representations of the full nervous systems in various species (reviewed on pp. 35–36 in Ref. [79]; also see [80]). Sabin, by all accounts, was an exemplar JHUMS student and went on to pursue what continued to be a remarkable scientific career. She secured a prestigious JHUMS internship and became the first female assistant professor and, later, full professor at JHUMS. In 1925, the Rockefeller Institute recruited her to direct her own laboratory (Fig. 2F) with a focus on the lymphatic and circulatory systems and the cellular pathologies associated with tuberculosis. In that same year, she became the first woman to be elected to the U.S. National Academy of Sciences [81].

4. A comparison of the JHUMS and UTEP courses

Having described each course in turn, it is worth expanding on their similarities and differences (Table 1). Beginning with the technical skills taught to students, both classes emphasize careful observation and drawing of histological material. Whereas JHUMS students used tissue sections stained to mainly visualize myeloarchitecture, or the white matter fiber tracts of the brain, UTEP BMC students study tissue sections stained for cytoarchitecture of gray matter regions. Both groups draw the general outlines and landmarks of their tissue sections, and boundaries demarcated by the stained structures (parcellation), to infer the 3-D organization of subregions from their 2-D preparations. In both courses, a few students go on to develop high-quality mapped representations of this organization across several scales, and, in some instances, assemble such maps as an atlas. Thus, Florence Sabin’s human infant hindbrain atlas was a de novo assembly of professional drawings of Hewetson’s prepared slides, her own careful descriptions of their neuroanatomy, and the 3-D structures she constructed from those drawings [71, 73, 74]. For UTEP BMC students, atlas-based mapping involves migrating experimentally-derived chemoarchitectural data to Swanson rat brain atlas templates in digital space through drawings [28]. Whereas the goal of the JHUMS effort is to understand the 3-D structure of the human brain at a macro- to mesoscale, the UTEP effort has contributed new findings to an existing model of the rat brain at a meso- to microscale. Despite differences in scale and application, the core pedagogy of student-driven, independent research to produce mapped representations of neuroanatomical substructures for the scientific community is, to our knowledge, uniquely shared by these programs.

While the program developers in each case apprehended the import of research to student scholarship and comprehension, the programs rapidly diverge to serve two very different populations. Reflective of modern educational reform initiatives like the AAAS Vision and Change report [3], BMC introduces first-year college students to the scientific process through engagement with hands-on original research activities, where they gain a working understanding of neuroanatomy and basic concepts in neuroscience. Students at this stage have nascent and diverse career outlooks, with their interests still relatively undifferentiated. Accordingly, the UTEP course does not place competitive or rigorous gates to admission and capitalizes on the observation that students can successfully grapple with complex information without extensive background preparation or learning through extensive lectures. In contrast, JHUMS students were already college-trained, intent upon completing medical school, and were being trained to gain a deep functional understanding of human anatomy in order to become physicians steeped in scholarly practice rooted in the latest scientific knowledge.

There are also differences in equity in the educational opportunities afforded UTEP and JHUMS students. It is clear from accounts of JHUMS students themselves in the 1890s that conditions were difficult for female medical students at the medical school [82], where they faced discrimination despite the mandate for JHUMS to accept women on an equal footing with men [83]. Similarly, medical education reform triggered by the Flexner report [14] resulted in inequities to large demographic groups through the loss of many historically black colleges and universities [14, 84]. We, as a country, continue to grapple with the repercussions of actions like this, both in the medical community and in STEM research fields alike. Though, presently, at UTEP, we continue to pursue the “Access and Excellence” model fostered by the University’s recently-retired president, Diana Natalicio. Introductory-level courses like BMC, upper-division independent research sections, and a faculty culture that seeks to provide our undergraduates with access to research opportunities have improved the outlooks of Hispanics in STEM training (reviewed in [85]) for our campus, with 43.5% of graduating seniors reporting having participated in on-campus research with UTEP faculty at the time of this writing [86].

5. Future directions

By reviewing our newly-uncovered knowledge of the activities that students performed in the JHUMS course over 125 years ago, we can consider refining our own course. For example, the Born method of modeling three-dimensional brain structure could be applied in modern form to the Swanson rat brain atlas model by our students (also see [87]). We could also gain a sense of a student’s baseline level of conceptualization of the nervous system by evaluating drawings of their work in class before and after they have received training, as has been performed in other courses for assessment or evaluation [88, 89]. As has been noted by others [90, 91], the addition to our curriculum of materials that emphasize the history of the histological processing of brain tissue and atlas-mapping/drawing techniques could further enhance student appreciation of the research and discovery process [e.g., 92, 93].

Finally, an additional resonance we have discovered between our courses is how histology, observation, drawing, and mapping can create dissatisfaction in some students. Well-documented and much discussed [94, 95, 96, 97, 98] is the declared ‘boredom’ of the famous writer and patron of the arts, Gertrude Stein, who was also a JHUMS student studying with Florence Sabin, but who failed her medical school courses during the clinical portion of her training. Trained by Franklin Mall himself, she thought favorably of his instruction [94] and contributed important neuroanatomical descriptions of a midbrain region known as the nucleus of Darkschewitsch [94, 95, 96] to Barker’s textbook (see pp. 721, 725, and 875 in Ref. [60]).

Whereas Stein was not motivated enough to connect her neuroanatomical training to her clinical training, some BMC students do not have the motivation necessary to carry out the drawing and mapping tasks themselves, which require intense concentration and a disregard for their tediousness. Even a graduate student teaching assistant of the course has attested to the tediousness of mapping, a feeling that becomes magnified in students who are not confident about their skills in performing this task (see p. 102 in Ref. [99]). Writing of this very problem, Ramón y Cajal himself has noted that “most people who lack self-confidence are unaware of the power of intense concentration” ([2], p. 33). As the social scientist, Sarah de Rijcke, has observed:

Cajal’s drawings were material articulations of an expert selection process. But from his own perspective, there was also another side to drawing. To Cajal, drawing was much more than a technique for reproduction; it was an essential tool in disciplining the eye.…The crux of the matter was a difference between ordinary and expert observation. To a degree, this difference depended on talent. But, more importantly, it was a matter of discipline and willpower. [100], p. 295

As Ramón y Cajal and de Rijcke remind us, teachers can learn lessons from the student experiences of motivation and its loss during research. We could strive to instill discipline and willpower within students, while simultaneously affording them the expectation that they may lose their motivation and patience as they struggle to contend with the process of scientific research.

In fact, we have found that when the students’ expectations are informed by knowledge that the process of mapping is iterative, and that mistakes are part of the process, their own motivation to persist is bolstered. Despite the rigors and challenges posed by intensive neuroanatomical mapping efforts, the self-efficacy engendered within students while rising to meet those challenges gives testament now, as then, to the benefits of learning by doing. In our own experience, within one semester, first-year students are able to advance from rank beginners to exhibit behaviors expected of well-trained graduate students. These skillsets serve them well in future pursuits, as demonstrated by the entry of several BMC undergraduate alumni into the research laboratory of one of us (AMK). These students go on to complete either independent or team-based projects, present them at scientific conferences, and co-author scientific papers. As one former BMC student, Ms. Monica Ponce, recently stated:

I was impressed that, as freshmen, we were considered real researchers…. The lab inspired me to keep going in the science field and made me feel I was contributing something to the scientific community. If I hadn’t taken the [BMC] lab, I wouldn’t have changed my major to neuroscience. I love the brain now. [101]

Clearly, then, the lessons of brain mapping learned 125 years ago by JHUMS medical students can readily be taught, in a similar fashion, to junior-level students just starting their college careers. In this respect, we do not have to mimic the past by restricting these technical skills to advanced and more senior career-level students; instead, we are making these tools and skills more broadly accessible, thereby helping to ensure their wider dissemination, preservation and use.

6. Concluding Remarks

The students of neuroanatomy past and present have effectively demonstrated the value in using atlas-based mapping to advance understanding in the topic. In bringing these records back into view, we are reminded of the role the UTEP BMC course has played and continues to play in preserving “lost arts.” Traditionally, only a tiny fraction of neuroscientists has been formally trained in the methods needed to migrate their datasets into an atlas model. However, the UTEP BMC course provides a “crowd-training” mechanism by which these skills can be preserved and more broadly disseminated. More importantly, the ultimate goal of erasing, or at least blurring, the cultural lines that separate expert from novice, and instructor from student, are also being realized, as we discover from students how our own mapping process can be refined and improved upon.

If the examples of the JHUMS students, especially of Florence Sabin, and those of our BMC students, thus far, are any indication, the common strategies employed for the two courses provide a powerful convergence of evidence suggesting that an emphasis on conducting original neuroanatomical research, as part of basic neuroanatomy training and education, is an effective means of empowering students in the scientific process at early stages of their careers. Formal personalized assessments of learning outcomes are, of course, not possible to conduct now for the students of the JHUMS course. However, our qualitative comparison here of the two courses, as well as previously-reported outcomes for the UTEP BMC course [16], suggest that student-conducted research to produce novel maps of the brain is a bellwether for future student success in becoming basic or medical research scientists.

Highlights.

• Course-based undergraduate research experiences (CUREs) train students on research.

• A CURE on brain mapping (BMC) is taught at The University of Texas at El Paso (UTEP).

• BMC mirrors a course once taught at Johns Hopkins University Medical School (JHUMS).

• UTEP/JHUMS course curricula are unique in teaching similar brain mapping methods.

• Atlas-based brain mapping by BMC students prevents it from becoming a “lost art”

Abbreviations

Swanson atlas nomenclature (see Ref. [28] for explanation of naming conventions for each term) (Fig. 3B, D, E):

ARH

arcuate hypothalamic nucleus (>1840)

COApm

cortical amygdalar area, posterior part, medial zone (>1840)

DMHa

dorsomedial hypothalamic nucleus, anterior part (>1840)

DMHp

dorsomedial hypothalamic nucleus, posterior part (>1840)

DMHv

dorsomedial hypothalamic nucleus, ventral part (>1840)

I

internuclear hypothalamic area (Swanson, 2004) [102]

int

internal capsule (Burdach, 1922) [103]

lem

lateral eminence (>1840)

LHAd

lateral hypothalamic area, middle group, lateral tier, dorsal region (Swanson, 2004) [102]

LHAjd

lateral hypothalamic area, middle group, medial tier, juxtadorsomedial region (Swanson, 2004) [102]

LHAjvv

lateral hypothalamic area, middle group, medial tier, juxtaventromedial region, ventral zone (Swanson, 2004) [102]

LHAma

lateral hypothalamic area, middle group, lateral tier, ventral region, magnocellular nucleus (Paxinos & Watson, 1986) [104]

LHAs

lateral hypothalamic area, middle group, perifornical tier, suprafornical region (Swanson, 2004) [102]

LHAsfp

lateral hypothalamic area, middle group, perifornical tier, subfornical region, posterior zone (Swanson, 2004) [102]

LHAvl

lateral hypothalamic area, ventral region, lateral zone (Swanson, 2004) [102]

LHAvm

lateral hypothalamic area, middle group, lateral tier, ventral region, lateral zone (Swanson, 2004) [102]

ME

median eminence (Tilney, 1936) [105]

MEex

median eminence external lamina (>1840)

mtth

hypothalamic mammillothalamic tract (Swanson, 2018) [28]

opth

hypothalamic optic tract (Swanson, 2015) [106]

PH

posterior hypothalamic nucleus (>1840)

pofh

hypothalamic postcommissural fornix (Swanson, 2015) [106]

PR

perireuniens nucleus (Brittain, 1988) [107]

PVi

periventricular hypothalamic nucleus anterior part intermediate zone (Swanson, 2018) [28]

REcp

nucleus reuniens, caudal division, posterior part (Risold et al., 1997) [108]

STN

subthalamic nucleus (>1840)

tis

tuberoinfundibular sulcus (>1840)

TUi

lateral hypothalamic area, middle group, lateral tier, tuberal nucleus, intermediate part (Swanson, 2004) [102]

TUl

lateral hypothalamic area, middle group, lateral tier, tuberal nucleus, lateral part (Swanson, 2004) [102]

TUsv

lateral hypothalamic area, middle group, lateral tier, tuberal nucleus, subventromedial part (Swanson, 2004) [102]

TUte

lateral hypothalamic area, middle group, lateral tier, tuberal nucleus, terete part (Petrovich et al., 2001) [109]

V3h

hypothalamic part of third ventricle, principal part (Swanson, 2018) [28]

vlt

ventrolateral hypothalamic tract (Swanson, 2004) [102]

VM

ventral medial thalamic nucleus (>1840)

VMHc

ventromedial hypothalamic nucleus central part (>1840)

VMHvl

ventromedial hypothalamic nucleus ventrolateral part (>1840)

ZI

zona incerta (>1840)

Sabin’s drawing and map (Fig. 3A, C):

Br. Conj.

brachium conjunctivum

C. d.

nucleus nervi cochlearis dorsalis

C. r.

corpus restiforme

C. v.

nucleus nervi cochlearis ventralis

F. G.

fasciculus ventro-lateralis (Gowersi)

F. l. m.

fasciculus longitudinalis medialis

Floc.

flocculus

K. VII

knee of nervus facialis

L.

medial portion of nucleus nervi vestibuli lateralis (Deiters)

L₁

medial portion of nucleus nervi vestibuli lateralis (Deiters)

N. c.

nervus cochleæ

N. IX and X

nervus glossopharyngeus et vagus

N. c.

root bundle of nervus cochleæ

N. m. p. V.

nucleus motorius princeps nervi trigemini

N. o. s.

nucleus olivaris superior

N. s. V.

nucleus nervi trigemini (sensory)

N. vest.

root bundle of nervus vestibuli

Nuc. XII

nucleus nervi hypoglossi

Nuc. VI

nucleus nervi abducentis

Nu. n. c. v.

nucleus nervi cochlearis ventralis

Nu. n. v. l.

nucleus nervi vestibularis lateralis (Deiters)

Nu. n. v. m.

nucleus nervi vestibularis medialis (Schwalbe)

Nu. n. v. s.

nucleus nervi vestibularis superior (Bechterew)

Nu. o. i.

nucleus olivaris inferior

Nu. y.

nucleus y (anterior-lateral portion of nucleus nervi vestibularis medialis)

P. f.

pedunculus flocculi

Py.

pyramis

R. d. n. vest.

radix descendens nervi vestibuli

S.

nucleus nervi vestibuli superior (Bechterew)

St. i. l.

stratum interolivare lemnisci

Tr. s. n. t.

tractus spinalis nervi trigemini

Y

see Nu. y

z

decussation nervi trigemini

Other abbreviations:

α-MSH

α-melanocyte-stimulating hormone

BMC

Brain Mapping and Connectomics laboratory course

CURE

course-based undergraduate research experience

H/O

hypocretin/orexin

JHUMS

Johns Hopkins University Medical School

USNL

UTEP Systems Neuroscience Laboratory

UTEP

University of Texas at El Paso