home | lab | find me | science | publications | software | toolbox | blog | site map


Whole brains


Our lab is interested in understanding how the brain works as a whole. We start from the realization that every area of the brain relays information to many others. Therefore, understanding any one aspect of brain function primarily associated with a particular area requires acquiring some understanding first about many other areas.

This impossible conundrum posed by the pervasive interconnectivity can be approached by first obtaining a map of all neuronal connections, as well as identifying individual neurons capable of eliciting or disrupting specific behaviors (Vogelstein et al. 2014). The synaptic wiring diagram of small brains, such as that of the fruit fly larva, can be mapped with relative ease from serial electron microscopy, partly thanks to technology that we and others have developed (e.g. CATMAID, TrakEM2). The circuits maps that we have obtained so far have proven extraordinarily useful in formulating hypotheses of circuit function and constructing computational models that can reproduce observed behaviors and account for experimental alterations of neural activity. The neural-behavior maps generated by our collaborators in the Zlatic lab enables us to prioritize specific neurons and areas for reconstruction.

Clear handles into the yarn ball that is the brain are its inputs and outputs, that is, its sensory and motor systems. As a first approximation, everything in between can be thought of as a black box that implements a history-dependent sensorimotor transformation. But in acquiring some structural and functional understanding of the first-order networks for sensation and motor control, the second-order neuronal layer becomes approachable. Therefore we concentrated first in mapping the wiring diagram of the optic, olfactory (Berck, Khandelwal et al., 2016) and somatosensory systems (Ohyama, Schneider-Mizell et al., 2015), among others, as well as the motor systems (Schneider-Mizell, Gerhard et al., 2016, Fushiki et al., 2016), and are now studying deeper areas of the nervous system such as the mushroom bodies, known to mediate associative memories, and the central complex, known to mediate spatial learning and motor planning.

In the context of known circuitry, and thanks to the genetic tools for the targeted manipulation and monitoring of neural function in Drosophila, we are unveiling the contribution of higher-order neurons to specific functions, one layer and one identified neuron at a time. In acquiring an understanding of some areas of the brain we complete the inputs and outputs of deeper areas, therefore enabling the study of their contributions to specific behaviors. In summary, by mapping the wiring diagram, observing behavior, monitoring and altering neural activity with electrophysiology and optophysiology, and modeling circuit function, we pry open the black box and acquire an understanding of how the nervous system works.

Neuronal circuit analysis from serial section TEM


The reconstruction of entire nervous systems jump started with the completion of the colossal work by Sydney Brenner and collaborators in describing the entire nervous system of the nematode Caenorhabditis elegans (1986). Their work in reconstructing every single neuron in the worm from thousands of 50 nm serial sections was possible not just by their excellent craft and painstaking dedication, but also from their wisdom in choosing an organism of the appropriate dimensions, considering the technological means of the time.

Since then, any further attempts to reconstruct neuronal tissue has used small stacks of serial sections (20-50 sections), and has focused in specific dendrites or axons, i.e. parts of neurons. It was difficult or impossible to determine in how far the characters observed for a specific set of terminal neurites in such small samples can be generalized to other regions of the brain, or what type of neurons gave rise to these profiles. The main innovation with which these problems can be overcome now is the progress in digital image handling: in acquisition speed, in storage capacity, and in processing power for their assembly and analysis.

In my lab, we are examining once again very large volumes of neuronal tissue. We focus on an organism, the fruit fly Drosophila melanogaster, with an enormous genetic toolkit for the targeted manipulation and labeling of its cells and tissues. In particular, the generation of genetic constructs specific for all neural classes is well underway (Pfeiffer et al, 2008).

In the fruit fly Drosophila, we have the opportunity to reconstruct at least complete functional units of the nervous system, and learn important general principles about wiring patterns and the function of individual components. Our model system of choice is the first order sensory processing centers in the abdominal segments of the ventral nerve cord, particularly well characterized in the larva at the optical microscopy level. Each abdominal hemisegment receives a total of 42 sensory axons, of stereotyped projection (Merrit and Whittington, 1995). By reconstructing the arborization and synapses of the sensory axons and their interneuron targets, we aim at elucidating important principles about their function and capabilities.

In the future, our goal is to reconstruct the complete central nervous system of this model organism, whose neurons number in the tens of thousands.

White JG, Southgate JN, Thomson JN and Brenner FRS. 1986. The structure of the nervous system of the nematode Caenorhabditis elegans. Phil. Trans. R. S. Lond. B 314:1-340.

Pfeiffer BD, Jenett A, Hammonds AS, Ngo TT, Misra S, Murphy C, Scully A, Carlson JW, Wan KH, Laverty TR, Mungall C, Svirskas R, Kadonaga JT, Doe CQ, Eisen MB, Celniker SE, Rubin GM. 2008. Tools for neuroanatomy and neurogenetics in Drosophila. PNAS 105(28):9715-20.

Merritt DJ, Whitington PM. 1995. Central projections of sensory neurons in the Drosophila embryo correlate with sensory modality, soma position, and proneural gene function. J Neurosci. 15(3 Pt 1):1755-67.

Serial section TEM data set of the ventral nerve cord of the first instar Drosophila larva

Complete neural arbor with polyadic synapses

Neural arbor detail. Pre-synaptic sites in yellow; post-synaptic sites in magenta

Drosophila brain: connectivity in the first instar larval brain


Why I am working with fruit flies? It's mostly about a person, a man, a professor at UCLA. His name is Volker Hartenstein. His enthusiasm can move mountains. His students are usually well-natured and just love what they do. He manages to attract this kind. I came to Los Angeles blindly, looking for help in my planarian research for my PhD. What I found was someone from which I could learn the profession of scientist.

Over the years the Hartenstein's lab has gathered considerable knowledge on the celullar and molecular details of fly brains. In his lab, there is unique knowledge in the anatomy of and gene expression patterns in the embryonic and larval brains which is nowhere else to be found.

First instar brain, dorsal view  Lateral view of two brain neurons projecting to the ventral nerve cord  Bolwig's nerve  First instar brain lobe montage

Fly brains are beautiful. Bilaterally symmetric, sleek and extremely ordered. From fly to fly one can identify the same exact neurons, the same exact projections. And track them over larval and pupal developmental stages. Such properties make the fly brain amenable to genetic dissection, using mutants, and bring feasibility to the idea of mapping all brain axons and dendrites.

My specific focus: the microcircuitry within the brain neuropile, and the descending and ascending projections to and from the ventral nerve cord. About the former, my colleagues at the Max Plank Institute for Molecular Cell Biology and Genetics are hosting 400 TEM serial sections of the first instar neuropile that I imaged there on Fall 2006 in Pavel Tomancak's group [1]. I have a couple hundred more sections acquired at the the Automated Molecular Imaging group in San Diego, using Leginon, still unreleased. Of course the whole 600 gigabyte dataset is viewed, modeled and annotated using TrakEM2, and rendered in Blender.

[1] Special thanks to Quentin De Robillard (TEM help and excellent rock climbing partner), Stephan Preibisch (FFT king of image registration and stitching) and Stephan Saalfeld (feature extraction, image registration and an impressive large-scale image web viewer; see the paper CATMAID and support data).

Planarians: the embryonic development of Schmidtea polychroa


My PhD Thesis was a journey of discovery into the egg capsule of planarians, by the hand of Rafael Romero (University of Barcelona) and Volker Hartenstein (UCLA). Planarians, least you don't know it, are fascinating non-parasitic slimy little flatworms with an idiosyncratic embryonic development. Reference papers dated from the early XX century, plus a dozen published between the fifties and early seventies. My thesis work was to bring triclad embryonic development to the molecular era, and study the relationships between embryogenesis and regeneration.

Given the staggering lack of literature and the non-standardness of the model, planarians are rather hard to work with. Below I have collected three figures from my PhD Thesis which may help you in the hard task of identifying developmental stages of planarian embryos when dissected (left), in raw paraffin sections (center) and in epon, methylene blue-counterstained sections (right). The species in particular is Schmidtea polychroa.

Dissected planarian embryos  Guide to embryonic stage identification of planarian embryos in paraffin sections  Planarian embryonic development in 8 stages


Last updated: 2016-05-22 11:10 Zurich time. Copyright Albert Cardona.