Elsevier

Brain Research

Volume 1450, 23 April 2012, Pages 174-193
Brain Research

Review
Neuroinformatics analysis of molecular expression patterns and neuron populations in gray matter regions: The rat BST as a rich exemplar

https://doi.org/10.1016/j.brainres.2012.02.034Get rights and content

Abstract

The rat bed nuclei of the stria terminalis (BST) is an important part of the cerebral nuclei, both structurally and functionally. However, the literature is rather scarce and more importantly, often contradictory. In this paper we review the literature related to neuron populations reported in different rat BST parts, and to a set of more than 50 expressed molecules. The information related to neuron populations and molecules detected in the BST was expertly collated manually in a publicly available neuroinformatics system, the Brain Architecture Knowledge Management System (BAMS; http://brancusi.usc.edu/bkms). Using the tools implemented in BAMS, we organized the collated information, and further analyzed it statistically. The result of our analysis over the set of > 50 expressed molecules confirms the BST parcellation scheme proposed by Swanson in 2004, with two exceptions. We present and discuss these results, and propose refined parcellation ventrally in the BST. We also review and discuss the presence of cholinergic neurons in the BST, and of neuron populations that express serotonin receptors. This review is one of the most comprehensive for the rat BST published in the literature, and it was possible only by using neuroinformatics tools.

Introduction

The bed nuclei of the stria terminalis (BST) is a part of the cerebral nuclei (basal ganglia; Bota and Swanson, 2010; for a standard nomenclature see Swanson and Bota, 2010) that was first named as such in the rat forebrain by Johnston (1923) and Gurdjian (1925). The first parcellation of the BST as a stand-alone gray matter region was provided by de Groot (1959). Since then, the BST has been frequently subdivided in different subparts, divisions, or nuclei by different authors (for a review on nomenclatures of the rat BST, see Bota and Swanson, 2010). However, there are currently only two main topographically distinct classification schemas: medial–lateral and anterior–posterior. The classical medial–lateral parcellation schema is based on cytoarchitecture, chemoarchitecture, and input connections from the amygdalar region (de Olmos and Heimer, 1999, de Olmos et al., 1985, de Olmos et al., 2010, Geeraedts et al., 1990, Paxinos and Watson, 1986, Paxinos and Watson, 1998). This schema usually includes BST as part of the “extended amygdala” (de Olmos et al., 2010) and the number of component nuclei is variable. The modern anterior–posterior classification schema is based on a) ontogeny of neuron populations in gray matter regions (Bayer, 1987), as a criterion for demarcation from the preoptic region of the hypothalamus and for distinguishing internal divisions; b) cytoarchitecture and gene expression patterns (Ju and Swanson, 1989, Ju et al., 1989); and most importantly c) complex input connection patterns from the amygdalar region (Dong et al., 2001) and output connection patterns from component nuclei to many parts of the brain (Dong and Swanson, 2003, Dong and Swanson, 2004, Dong and Swanson, 2006, Swanson, 2004). Even though these two dominant parcellation schemes may appear as topographically orthogonal, they also include common nuclei, such as the juxtacapsular nucleus (BSTju; Moga et al., 1989). However the most common results of qualitative (topological) comparison performed by us (Bota and Swanson, 2010) lead to the relation of “partial correspondence” (Swanson and Bota, 2010) between the BST nuclei of these two parcellations.

The set of molecules known to be expressed in the rat BST nuclei increased with each relevant study there or in neighboring regions (de Olmos et al., 2010, Gray and Magnuson, 1987, Hammack et al., 2007, Ju et al., 1989, Moga et al., 1989, Poulin et al., 2006, Puente et al., 2010, Rainnie, 1999, Swanson et al., 1983, Zaborszky et al., 2005). Moreover, some molecular expression patterns are sex-specific (Polston and Simerly, 2003). However, there are very few reviews or surveys of the molecular architecture of rat BST nuclei (de Olmos et al., 2010)—and the same is true for neuron populations, classes, and types in the BST. Although the literature concerning BST neuron classification is more than 30 years old, few studies have used multiple classification schemas, the polythetic approach (see Bota and Swanson, 2010 for a review). Most studies reviewed here classify BST neurons according to one criterion, usually cytoarchitecture (unithetic approach), or several criteria at most. Moreover, the lack of agreement across neuroanatomical nomenclatures, accompanied by equivocal relations between gray matter regions, makes the results of experimental connection analyses difficult to interpret, compare, and integrate.

Proposed functions of the rat BST range from modulation of basic autonomic responses (Moga et al., 1989, Swanson and Sharpe, 1973), large sets of motor behaviors (Dong and Swanson, 2003, Dong and Swanson, 2006), and circadian rhythmicity (Amir et al., 2004), to stress- and anxiety-related behaviors (Choi et al., 2009, Levita et al., 2004, Walker et al., 2003), and behaviors related to drugs of abuse (Aston-Jones and Harris, 2004). This large set of functions can be explained only by complex patterns of connections, and a large set of expressed molecules. This in turn implies multiple neuron populations in the rat BST that can be distinguished using the multiple criteria (polythetic) approach for their classification (Bota and Swanson, 2010). Hence, a survey of molecules differentially expressed in the rat BST nuclei, and of identified neuron types, is not only valuable for a comprehensive description of this cerebral nuclei region, but is also necessary for establishing the structure–function relationships of each of the BST nuclei.

In this paper we survey comprehensively the published literature on molecular expression patterns and neuron populations, classes, and types identified in the rat BST. It can be viewed as a companion to an earlier paper (Bota and Swanson, 2010) where we proposed a methodology for collating and curating neuroanatomical nomenclatures in online neuroinformatic systems, and reviewed the literature on rat BST parcellation schemes. Because large scale efforts are greatly facilitated by neuroinformatics systems, we collated the molecular expression pattern and neuron population information in the Brain Architecture Knowledge Management System (BAMS). This allowed us to manage the inserted data in tabular formats and to analyze it statistically. Moreover, all the collated data and metadata is publicly available to the neuroscience community.

First, we briefly describe BAMS and its new interfaces and inference engine, created to integrate the results of different experiments and data types. This is accompanied by a description of the methodology used to map and collate data from the published literature. Second, we present the results of our survey of molecular expression data collated in BAMS, and the results of relevant statistical analyses. We then review the published literature on neuron populations identified in the BST, and on relations between them. Finally, we discuss the results of our survey and analysis, and propose several possible future experiments. This series of papers is meant to provide a general approach to collating information about the organization of neural systems from the legacy neuroscience literature and extracting knowledge from this information.

Section snippets

Results

The BAMS user interface includes different neuron population and molecule search options, described in detail in Bota and Swanson (2006), and Bota and Swanson (2010). The neuron populations and molecules associated with a gray matter region in BAMS can be directly accessed from the Menu of the system, along with information associated with the results of online searches about gray matter regions. Typical results of a gray matter region search that returns associated neuron populations and

Discussion

The rat BST is a very complex part of the cerebral nuclei (basal ganglia), in terms both of molecular expression patterns and of neuron populations, classes, and types. It is therefore not surprising that functions assigned to the various BST nuclei also cover a broad spectrum, and the associated literature is far too extensive to review here. In addition, these nuclei have not yet been analyzed and described in as much detail as many other regions of the mammalian CNS, and the integration of

BAMS presentation

BAMS (http://brancusi.usc.edu/bkms) has been under development since 2002 (Bota et al., 2003) and is an online neuroinformatics workbench for handling neuroanatomical information as collated from the published literature. The structure of BAMS is modular (Bota et al., 2003, Bota et al., 2005) and its five inter-related modules handle information related to four levels of nervous system organization. This structure allows insertion of data and metadata associated with the nervous system of any

Acknowledgments

M.B. and L.W.S. are supported by supported by the National Institutes of Health Grant NS050792. O.S. is supported by the James S. McDonnell Foundation.

We thank Dr. Clifford Saper for his valuable suggestions and feedback.

References (66)

  • A.J. McDonald

    Neurons of the bed nucleus of the stria terminalis: a Golgi study in the rat

    Brain Res. Bull.

    (1983)
  • J.-F. Poulin et al.

    Neuroanatomical characterization of endogenous opioids in the bed nucleus of the stria terminalis

    Prog. Neuro-Pyschopharmacol.

    (2009)
  • D.L. Walker et al.

    Role of the bed nucleus of the stria terminalis versus the amygdala in fear, stress, and anxiety

    Eur. J. Pharmacol.

    (2003)
  • L. Zaborszky et al.

    Three-dimensional chemoarchitecture of the basal forebrain: spatially specific association of cholinergic and calcium binding protein-containing neurons

    Neuroscience

    (2005)
  • H. Akil et al.

    Challenges and opportunities in mining neuroscience data

    Science

    (2011)
  • S. Amir et al.

    A circadian rhythm in the expression of PERIOD2 protein reveals a novel SCN-controlled oscillator in the oval nucleus of the bed nucleus of the stria terminalis

    J. Neurosci.

    (2004)
  • D.M. Armstrong et al.

    Distribution of cholinergic neurons in rat brain: demonstrated by the immunocytochemical localization of choline acetyltransferase

    J. Neurosci.

    (1983)
  • S.A. Bayer

    Neurogenetic and morphogenetic heterogeneity in the bed nucleus of the stria terminalis

    J. Comp. Neurol.

    (1987)
  • R. Bleier et al.

    A cytoarchitectonic atlas of the hypothalamus and hypothalamic third ventricle of the rat

  • M. Bota et al.

    Integrating databases and expert systems for the analysis of brain structures, connections, and homologies

    Neuroinformatics

    (2004)
  • M. Bota et al.

    A new module for manipulation and display and manipulations of molecular information in the Brain Architecture Management System

    Neuroinformatics

    (2006)
  • M. Bota et al.

    The neuron classification problem

    Brain Res. Rev.

    (2007)
  • M. Bota et al.

    Collating and curating neuroanatomical nomenclatures: principles of use of the Brain Architecture Management System (BAMS)

    Front. Neuroinformatics

    (2010)
  • M. Bota et al.

    From gene networks to brain networks

    Nat. Neurosci.

    (2003)
  • M. Bota et al.

    Brain architecture management system

    Neuroinformatics

    (2005)
  • G.A.P.C. Burns

    Knowledge management of the neuroscientific literature: the data model and underlying strategy of the NeuroScholar system

    Philos. Trans. R. Soc. Lond. B Biol. Sci.

    (2001)
  • D.C. Choi et al.

    Bed nucleus of the stria terminalis subregions differentially regulate hypothalamic–pituitary–adrenal axis specificity: implications for the integration of limbic inputs

    J. Neurosci.

    (2009)
  • W.E. Cullinan et al.

    Ventral subicular interaction with the hypothalamic paraventricular nucleus: evidence for a relay in the bed nucleus of the stria terminalis

    J. Comp. Neurol.

    (1993)
  • J. de Groot

    The rat hypothalamus in stereotaxic coordinate

    J. Comp. Neurol.

    (1959)
  • J.S. de Olmos et al.

    The concepts of the ventral striatopallidal system and the extended amygdala

  • J.S. de Olmos et al.

    Amygdala

  • J. de Olmos et al.

    Amygdala and the extended amygdala of the rat: a cytoarchitectonical, fibroarchitectonical, and chemoarchitectonical survey

  • H.-W. Dong et al.

    Projections from the rhomboid nucleus of the bed nuclei of the stria terminalis: implications for cerebral hemisphere regulation of ingestive behaviors

    J. Comp. Neurol.

    (2003)
  • Cited by (0)

    View full text