| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Topical Review |
1 Genes to Cognition programme, Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridgeshire CB10 1SA, UK
 |
Abstract |
|---|
 Top Abstract IntroductionReferences |
|---|
1011 neurons in the brain having around 103–104 synapses. Proteomic studies of the synapse have revealed that the postsynaptic density is the most complex multiprotein structure yet identified, with 103 different proteins. Such studies, however, use brain tissue with many different regions and therefore different cell types, and there is clear potential for heterogeneity of protein content at different synapses within and between brain regions. Although large-scale mRNA-based assays are in progress to map this sort of complexity at the cellular level, and indeed all brain-expressed genes, analysis of protein distribution (at synapses and other structures) is still in the very early stages. We review existing large-scale protein expression studies and the specific technical obstacles that need to be overcome before applying the scaling used in nucleic acid based approaches.
(Received 18 May 2006;
accepted after revision 16 June 2006;
first published online 22 June 2006)
Corresponding author S. G. N. Grant: Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, Cambridgeshire CB10 1SA, UK. Email: sg3{at}sanger.ac.uk
|
Introduction |
|---|
|
TopAbstractIntroductionReferences |
|---|
There is now a growing interest in large-scale approaches to studying gene expression and a greater understanding of the benefits of carrying out such studies. Not only do these projects provide very useful resources for the scientific community when looking for information on individual genes, but the potential for meta-analysis of large amounts of data is becoming increasingly realized. For example, analysis of microarray data for around 24 different neural tissues, performed by Zapala et al. (2005), has revealed that different regions of the brain have transcriptomes that differ according to each tissue's region of origin in the early embryonic neural tube/brain. Microarray-based methods as well as other large-scale techniques are also proving useful in the study of neurological diseases (reviewed in Baranzini, 2004; Galvin & Ginsberg, 2004; Kannanayakal & Eberwine, 2005)
The study of gene expression in brain regions by microarray (or by protein-based electrophoretic methods) gives values for average regional expression levels, but cellular and subcellular detail, such as that provided by in situ hybridization (ISH) or immunohistochemistry (IHC), is desirable so that a more in-depth analysis can be performed. However, all expression data sets can be analysed to search for common and exclusive patterns of gene expression, particularly with regard to functionally related genes or gene lists (Zapala and colleagues found 192 regionally enriched or uniquely expressed genes). Focusing on the coexpression, or lack of expression of genes in brain regions associated with a disease or a particular phenotype may give us insights into signalling pathways utilized in these different parts of the brain. Co-expression studies will likely increase predictive power for further studies and this may in turn help identify new potential targets for therapeutic intervention. Studying the promoter regions of genes that share similar expression patterns will help us to define more fully the regional and/or global transcriptional control in the brain. We should also be able to gain useful insights into the regulation of gene expression by differential methylation according to brain region.
RNA expression – in situ hybridization
Large scale expression studies focusing on mRNA are well underway and Sunkin (2006) has written an extensive review of the various projects in this field. As a brief summary of ISH for adult mouse brain, the largest projects are the Allen Brain Atlas (http://www.brainatlas.org/aba which uses a colorimetric ISH protocol, includes a wide range of parasagittal sections and has detail which can be seen at the cellular level – coverage > 10 000 genes) and the Brain Gene Expression Map (Magdaleno et al. 2006 and http://www.stjudebgem.org, which has lower coverage in terms of numbers of genes (3000) and numbers of different sections, but since it uses a radioactive ISH method, it is more quantitative in its readout, although the level of resolution is lower). The Gene Paint project (http://www.genepaint.org) also has a database of adult mouse brain ISH, but numbers are significantly lower than the previous studies as the main focus is on developmental gene expression. The issues that arise for studies of these types so that the information can be used to its full are ones of image capture, annotation and the ability to integrate data from different data sets, although these issues are understood within the field.
Single Cell mRNA analysis
Analysis of mRNA expression in single cells is now possible with the use of laser capture microdissection and linear amplification of mRNA (reviewed in Bohm et al. 2005; Kannanayakal & Eberwine, 2005). As an example, these kinds of analyses have been performed on different subregions of the hippocampus and have demonstrated clear expression differences between subregions and even within subregions (Bonaventure et al. 2002; Kamme et al. 2003; Torres-Munoz et al. 2004; Ginsberg & Che, 2005). These experiments are complemented by the studies where individual subregions of the hippocampus were dissected and subjected to microarray anlaysis (Zhao et al. 2001; Lein et al. 2004).
Protein expression – projects and resources
While being very useful, gene expression studies are limited to information about the cells that express a particular gene; they provide little if any information about subcellular localization of the protein product, whether to axons or dendrites, etc. This is important since projections of axons and arborizations of dendrites are a central feature of brain anatomy.
However, protein expression studies are currently far behind those focusing on mRNA. Several efforts are being made in this area though, but coverage, both anatomically and in number of proteins, is low. The first is based in Sweden and is part of a cancer project (Uhlen et al. 2005; Nilsson et al. 2005; http://www.proteinatlas.org) – the group now has in excess of 700 rabbit polyclonal antibodies which have been used in tissue microarrays, using cores of various human tissues including cerebellar cortex and cerebellum from the brain. These tissue cores have all been annotated at a cellular level (e.g. neuronal versus non-neuronal in cortex, but purkinje/molecular layer/granular layer annotations for the cerebellum), but a core of cortex and of cerebellum is not as informative in terms of the brain as most neuroscientists would like. A similar project is underway for protein expression in the mouse at the Wellcome Trust Sanger Institute (Warford, 2004; Warford et al. 2004 Atlas of Protein Expression http://www.sanger.ac.uk/Teams/Team86/) using in-house generated phage-based antibodies. As well as tissue cores, parasagittal adult brain and mouse embryo (E14.5) sections are being tested. Another venture based in University of California at Davis is seeking to generate up to 500 quality-assured monoclonal antibodies against neuronal proteins, starting with membrane and synaptic proteins, and the plan is to provide antibodies cheaply to the research community (http://www.neuromab.com). However, this project is not seeking to analyse and compare multiple expression patterns.
The most obvious route for the detection of protein expression is to use antibodies, as there is no other reporter system for mammals that is currently scalable, but there are non-trivial potential problems with antibody technology. Perhaps the main problem is the availability of target-specific antibodies and especially those that work in immunohistochemistry. The specificity of each antibody needs to be thoroughly tested, since antibodies may react with other proteins than the one they were created against. Ideal testing would involve the use of knockout tissue, but if several antibodies are made to different regions of a particular protein, and they all give the same expression pattern, then one can have reasonable confidence as to the validity of the staining patterns. The Protein Atlas group also add the criterion of an antibody reacting with only the correct size band(s) on a Western blot. As with all large-scale studies, thought should also be given to the minimum necessary information required to be published with IHC data and an example of the development of such standards for ISH and IHC can be found at http://scgap.systemsbiology.net/standards/misfishie/.
When large numbers of antibodies are used and the data analysed, we would expect that the effect of some false-positive/non-specific binding would not mar the general messages obtained. There is a strong case therefore that antibodies can be used in immunohistochemistry combined with the technology for scaling that is used in mRNA studies. It is nevertheless important to validate the use of antibodies in obtaining protein expression data on a larger scale and to see how these data compare with those obtained from the corresponding mRNA studies. To do this we carried out a pilot study of protein expression in the 10–100 scale comparing the data, where possible, with mRNA results. To maximize the possibility of finding functionally significant results, we focused on a set of proteins defined by their physical interaction at the postsynaptic density of excitatory synapses. It will therefore be necessary to describe briefly the current understanding of the postsynaptic compartment from a proteomics perspective.
Proteomic studies
Changing protein localization and composition at synapses are likely to be the key features of learning and memory at the cellular level, with the main proteins regulating these changes also being present at the synapse. Proteomic studies of the synapse typically consist of biochemically purifying different synaptic components and then defining protein composition by mass spectrometry. One such study by Collins et al. (2006) (and see references therein) reviewed the literature and added to it to provide a picture of current synapse proteomics and in particular the postsynaptic density (PSD) of excitatory synapses. Over 1100 proteins have been identified as being potentially present at the PSD with just under half being identified by two or more studies. This makes the PSD the most complex cellular structure known in terms of protein numbers.
In the PSD there appear to be important subsets of proteins, associated with different classes of glutamate receptors, which have particular physiological significance for signalling. Affinity purification of the NMDA receptor, or using the C-terminus of an NMDA receptor subunit to purify interacting proteins, has led to the identification of a complex of 186 proteins (a subset of the PSD) that are closely associated with the NMDA receptor (Husi et al. 2000; Collins & Husi, 2006). The NMDA receptor complex (NRC) is also known as the (membrane-associated guanylate kinase or) MAGUK-associated signalling complex, MASC. The significance of the NRC/MASC has been shown by the finding that a third of these proteins have been implicated in synaptic plasticity or behaviour in rodents, or in mental disorders in humans (Pocklington et al. 2006).
It should be noted, however, that all the proteomic studies referred to so far have used whole brain or whole forebrain for collection of PSDs. Since even the forebrain preparations contain several functionally and morphologically distinct brain regions, the question arises as to the degree of heterogeneity of synapses between (and even within) brain regions. As yet we have relatively little knowledge of how these protein complexes vary in terms of synapses in different parts of the brain, let alone how all these differing combinations of proteins might interact to modify synaptic function. A more recent proteomic study compared PSDs from forebrain with those obtained from cerebellum (Cheng et al. 2006). Relative (and absolute) quantification of proteins in the PSD indicated that upwards of 40 proteins showed significant differences in levels between the two sets of PSDs. Immunohistochemical analysis of some of the proteins identified as being different between the two PSD sets correlated well with the results of the proteomic studies.
The use of proteomics after laser capture microdissection is also being developed, but a problem with these types of protein studies is that there are no amplification techniques for proteins obtained from a handful of cells as there are for mRNA (Kunz & Chan, 2004). This means that hundreds of or several thousand cells need to be captured for adequate amounts of protein for reproducible results in subsequent mass spectrometry or two-dimensional gel electrophoresis. The advantage of being able to take single cells is less clear in a neuronal context, as cell bodies could be captured, but interest may often lie in protein expression in the dendritic/synaptic fields, which cannot be distinguished on a cell by cell basis. However, one study has used laser capture microdissection to isolate neurofibrillary tangles from the CA1 subregion of the hippocampus and identified over 70 proteins associated with the tangles, many of which were considered novel components (Wang et al. 2005).
Immunohistochemistry – a complementary necessity
In comparison with proteomic methods studying the PSD, IHC needs less tissue and no specific dissection to monitor protein expression in specific regions or subregions of the brain, although the level of spatial resolution provided is less (individual postsynaptic densities cannot be seen). Brain IHC is therefore a complementary method for determining the relative contributions of different proteins in different areas of the brain. When this was done for the four MAGUK family members PSD95, PSD93/Chapsyn 110, SAP97 and SAP102 (all NRC/MASC compononents), it was clear from the literature that mRNA and protein expression varies considerably between the four members, particularly between SAP97 and the other family members (Muller et al. 1995; Fukaya et al. 1999; Fukaya & Watanabe, 2000 – see also Fig. 1). The relative contributions to synapses of each of the MAGUK family members are likely therefore to be different in different regions. It is also important to note that levels of mRNA expression do not always correlate with those of the protein (for the example of MAGUKs, SAP102 protein is much lower than expected in the cerebellum when compared to the relative levels of mNRA – see references above and BGEM database for SAP102). Large-scale protein studies are therefore required to give information on the localization of proteins and the quantification of their levels.
|
We therefore propose a model whereby there are some core proteins at glutamatergic synapses, such as PSD95 (Petersen et al. 2003), and other additional proteins whose levels at the synapse are much more variable depending on the function and/or location of the synapse within the brain. Thus regional differences in synapse composition could account for regional differences in the response to particular plasticity-inducing trains of action potentials.
In our own lab we have looked at the expression profiles of over 75 proteins (the majority were components of the NRC) using Western blots and/or immunohistochemistry, and we have found that there are indeed widespread variations in the levels of these proteins in different regions of the brain (unpublished observations). This strengthens the idea of variable NRC/PSD composition depending on/influencing the function of the synapse.
Recommendations
Given the above data a larger scale immunohistochemical study seems not only sensible but in fact greatly needed, and the resources are also now being developed. An IHC-based approach is complementary to mRNA-based and mass spectrometry/proteomics-based approaches and is by no means redundant, due to its provision of protein localization information at a cellular level. Once large amounts of data are available, an integrative systems biology approach is almost certain to yield very interesting results. For example, protein interaction data can be refined by expression data to tell us which proteins are likely to interact in which regions of the brain (and not just which proteins can interact in a particular assay). Disease or behaviour phenotypes can be mapped onto this information to enhance predictions of significant signalling pathways involved in a disease/behaviour. This in turn could lead to better predictive ability for therapy strategies. Thought needs to be given, however, to how exactly protein expression data can be integrated with other forms of expression data and with phenotype or disease databases, for example how to compare annotation nomenclature from protein-based projects with those from the ISH projects. Brain IHC is a scalable process, but not easily scalable in a short time frame, and therefore we suggest the prioritization of synaptic proteins and those proteins associated with disease.
|
References |
|---|
|
TopAbstractIntroductionReferences |
|---|
Bohm C, Newrzella D & Sorgenfrei O (2005). Laser microdissection in CNS research. Drug Discov Today 10, 1167–1174.[CrossRef][Medline]
Bonaventure P, Guo H, Tian B, Liu X, Bittner A, Roland B, Salunga R, Ma XJ, Kamme F, Meurers B, Bakker M, Jurzak M, Leysen JE & Erlander MG (2002). Nuclei and subnuclei gene expression profiling in mammalian brain. Brain Res 943, 38–47.[CrossRef][Medline]
Cheng D, Hoogenraad CC, Rush J, Ramm E, Schlager MA, Duong DM, Xu P, Rukshan S, Hanfelt J, Nakagawa T, Sheng M & Peng J (2006). Relative and absolute quantification of postsynaptic density proteome isolated from rat forebrain and cerebellum. Mol Cell Proteomics (in press).
Collins MO, Husi H, Yu L, Brandon JM, Anderson CNG, Blackstock WP, Choudhary JS & Grant SGN (2006). Molecular characterization and comparison of the components and multiprotein complexes in the postsynaptic proteome. J Neurochem 97, 16–23.[CrossRef][Medline]
Fukaya M, Ueda H, Yamauchi K, Inoue Y & Watanabe M (1999). Distinct spatiotemporal expression of mRNAs for the PSD-95/SAP90 protein family in the mouse brain. Neurosci Res 33, 111–118.[CrossRef][Medline]
Fukaya M & Watanabe M (2000). Improved immunohistochemical detection of postsynaptically located PSD-95/SAP90 protein family by protease section pretreatment: a study in the adult mouse brain. J Comp Neurol 426, 572–586.[CrossRef][Medline]
Galvin JE & Ginsberg SD (2004). Expression profiling and pharmacotherapeutic development in the central nervous system. Alzheimer Dis Assoc Disord 18, 264–269.[Medline]
Garry EM, Moss A, Delaney A, O'Neill F, Blakemore J, Bowen J, Husi H, Mitchell R, Grant SG & Fleetwood-Walker SM (2003). Neuropathic sensitization of behavioral reflexes and spinal NMDA receptor/CaM kinase II interactions are disrupted in PSD-95 mutant mice. Curr Biol 13, 321–328.[CrossRef][Medline]
Ginsberg SD & Che S (2005). Expression profile analysis within the human hippocampus: comparison of CA1 and CA3 pyramidal neurons. J Comp Neurol 487, 107–118.[CrossRef][Medline]
Hatten ME & Heintz N (2005). Large-scale genomic approaches to brain development and circuitry. Annu Rev Neurosci 28, 89–108.[CrossRef][Medline]
Husi H, Ward MA, Choudhary JS, Blackstock WP & Grant SG (2000). Proteomic analysis of NMDA receptor-adhesion protein signaling complexes. Nat Neurosci 3, 661–669.[CrossRef][Medline]
Kamme
F
&
Salunga
R, Yu
J, Tran
DT, Zhu
J, Luo
L, Bittner
A, Guo
HQ, Miller
N, Wan
J
&
Erlander
M (2003). Single-cell microarray analysis in hippocampus CA1: demonstration and validation of cellular heterogeneity. J Neurosci
23, 3607–3615.
Kannanayakal TJ & Eberwine J (2005). mRNA methods used in dissecting gene expression of the brain. Ageing Res Rev 4, 513–528.[Medline]
Komiyama
NH, Watabe
AM, Carlisle
HJ, Porter
K, Charlesworth
P, Monti
J, Strathdee
DJ, O'Carroll
CM, Martin
SJ, Morris
RG, O'Dell
TJ
&
Grant
SG (2002). SynGAP regulates ERK/MAPK signaling, synaptic plasticity, and learning in the complex with postsynaptic density 95 and NMDA receptor. J Neurosci
22, 9721–9732.
Kunz GM Jr & Chan DW (2004). The use of laser capture microscopy in proteomics research – a review. Dis Markers 20, 155–160.[Medline]
Lein
ES, Zhao
X
&
Gage
FH (2004). Defining a molecular atlas of the hippocampus using DNA microarrays and high-throughput in situ hybridization. J Neurosci
24, 3879–3889.
Magdaleno S, Jensen P, Brumwell CL, Seal A, Lehman K, Asbury A, Cheung T, Cornelius T, Batten DM, Eden C, Norland SM, Rice DS, Dosooye N, Shakya S, Mehta P & Curran T (2006). BGEM: an in situ hybridization database of gene expression in the embryonic and adult mouse nervous system. PLoS Biol 4, e86.[CrossRef][Medline]
Migaud M, Charlesworth P, Dempster M, Webster LC, Watabe AM, Makhinson M, He Y, Ramsay MF, Morris RG, Morrison JH, O'Dell TJ & Grant SG (1998). Enhanced long-term potentiation and impaired learning in mice with mutant postsynaptic density-95 protein. Nature 396, 433–439.
Muller BM, Kistner U, Veh RW, Cases-Langhoff C, Becker B, Gundelfinger ED & Garner CC (1995). Molecular characterization and spatial distribution of SAP97, a novel presynaptic protein homologous to SAP90 and the Drosophila discs-large tumor suppressor protein. J Neurosci 15, 2354–2366.[Abstract]
Nilsson P, Paavilainen L, Larsson K, Odling J, Sundberg M, Andersson AC, Kampf C, Persson A, Al-Khalili Szigyarto C, Ottosson J, Bjorling E, Hober S, Wernerus H, Wester K, Ponten F & Uhlen M (2005). Towards a human proteome atlas: high-throughput generation of mono-specific antibodies for tissue profiling. Proteomics 5, 4327–4337.[CrossRef][Medline]
Petersen
JD, Chen
X, Vinade
L, Dosemeci
A, Lisman
JE
&
Reese
TS (2003). Distribution of postsynaptic density (PSD)-95 and Ca2+/calmodulin-dependent protein kinase II at the PSD. J Neurosci
23, 11270–11278.
Pocklington AJ, Cumiskey M, Armstrong JD & Grant SGN (2006). The proteomes of neurotransmitter receptor complexes form modular networks with distributed functionality underlying plasticity and behaviour. Mol Syst Biol 2, E1–E14.
Porter K, Komiyama NH, Vitalis T, Kind PC & Grant SG (2005). Differential expression of two NMDA receptor interacting proteins, PSD-95 and SynGAP during mouse development. Eur J Neurosci 21, 351–362.[CrossRef][Medline]
Sunkin SM (2006). Towards the integration of spatially and temporally resolved murine gene expression databases. Trends Genet 22, 211–217.[CrossRef][Medline]
Torres-Munoz JE, Van Waveren C, Keegan MG, Bookman RJ & Petito CK (2004). Gene expression profiles in microdissected neurons from human hippocampal subregions. Brain Res Mol Brain Res 127, 105–114.[Medline]
Uhlen
M, Bjorling
E, Agaton
C
et al. (2005). A human protein atlas for normal and cancer tissues based on antibody proteomics. Mol Cell Proteomics
4, 1920–1932.
Wang
Q, Woltjer
RL, Cimino
PJ, Pan
C, Montine
KS, Zhang
J
&
Montine
TJ (2005). Proteomic analysis of neurofibrillary tangles in Alzheimer disease identifies GAPDH as a detergent-insoluble paired helical filament tau binding protein. FASEB J
19, 869–871.
Warford A (2004). Tissue microarrays: fast-tracking protein expression at the cellular level. Expert Rev Proteomics 1, 283–292.[CrossRef][Medline]
Warford A, Howat W & McCafferty J (2004). Expression profiling by high-throughput immunohistochemistry. J Immunol Meth 290, 81–92.[CrossRef][Medline]
Zapala
MA, Hovatta
I, Ellison
JA, Wodicka
L, Del Rio
JA, Tennant
R, Tynan
W, Broide
RS, Helton
R, Stoveken
BS, Winrow
C, Lockhart
DJ, Reilly
JF, Young
WG, Bloom
FE
&
Barlow
C (2005). Adult mouse brain gene expression patterns bear an embryologic imprint. Proc Natl Acad Sci U S A
102, 10357–10362.
Zhao X, Lein ES, He A, Smith SC, Aston C & Gage FH (2001). Transcriptional profiling reveals strict boundaries between hippocampal subregions. J Comp Neurol 441, 187–196.[CrossRef][Medline]
|
Acknowledgements |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. M. Sunkin and J. G. Hohmann Insights from spatially mapped gene expression in the mouse brain Hum. Mol. Genet., October 15, 2007; 16(R2): R209 - R219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Cherubini, S. Gustincich, and H. Robinson The mammalian transcriptome and the cellular complexity of the brain J. Physiol., September 1, 2006; 575(2): 319 - 320. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
