Quantitative mass spectrometry to define subcellular fractions: insights to peroxisome biology.

David Goodlett
University of Washington
Medicinal Chemistry


Joint work by David R. Goodlett1, Marcello Marelli, Eugene C. Yi,
Sam M. Donohoe, Alexey I. Nesvizhskii, Jennifer J. Smith, Ruedi Aebersold,
Richard A. Rachubinski2 and John D. Aitchison.


University of Washington1, Institute for Systems Biology, and
University of Alberta2


Introduction.
What are peroxisomes,what is their function and how can
proteomics help answer these questions?
Peroxisomes are organelles bound by a single unit membrane that house a
number of enzymes that carry out a number of oxidative reactions, including
the b-oxidation of fatty acids.  In addition, in higher eukaryotes
peroxisomes contain the enzymes superoxide dismutase and catalase that
detoxify peroxides and superoxides.   Peroxisomes are also involved
in the synthesis of ether-linked glycerolipid (plasmalogens, abundant
constituents of nerve myelin); the interconversion of cholesterol into bile
acids; and glyoxylate transamination (reviewed in Kunau and Hartig, 1992; van
den Bosch et al., 1992; Subramani, 1993).  In yeast their most notable
function is in the b-oxidation of fatty acids. Why study Peroxisomes?
Peroxisomes are essential for human survival. Peroxisomes have a central
role in metabolism, and their function has been linked to a number of human
health concerns including aging, cancer, heart disease, obesity, and diabetes
(Clarke et al., 1999). Several genetic diseases are caused by peroxisomal
dysfunction: amiotrophic lateral sclerosis (ALS, Lou Gehrig’s disease),
X-linked Adrenoleukodystrophy, (ALD; popularized in the film Lorenzo’s Oil)
are a result of the dysfunction of a single enzyme pathway in peroxisomes. 
Zellweger syndrome, neonatal adrenoleukodystrophy, and infantile Refsum’s
disease are leukodystrophies that result from a deficiency in the overall
biogenesis of peroxisomes and are fatal in early infancy (Lazarow and Moser,
1994).


Method.
In previous work we used gas phase fractionation of isolated peroxisomes
to identify >95% of the known peroxisomal proteins (Yi et al. 2002). 
This unprecedented level of coverage indicated that we could “see” most of the
proteins known to reside in the peroxisome. To discriminate between those
components that are bone fide components of the peroxisome and those
that are contaminants we are using quantitative mass spectrometry to help us
define which components occupy the peroxisomal space by comparing a highly
purified sample of affinity purified peroxisomes to a fraction of isolated
peroxisomes. Peroxisomal membranes isolated from a yeast strain expressing a
protein tagged with the IgG binding domain of the Staphylococcus protein A
(Pex11-pA); (This fraction contains many contaminants). IgG coated
beads were used to affinity purify intact peroxisomes from the fraction of
isolated peroxisomes. Equal amounts of affinity purified peroxisomes (heavy
ICAT) and isolated peroxisomes (light ICAT) were differentially labeled with
appropriate ICAT reagent and the samples combined and analyzed by uLC-MS/MS.
Since we label equal amounts of both samples we expect to see an enrichment of
peroxisome specific proteins in the affinity purified fraction as per Ranish
et al. 2003. We thus used quantitative mass spectrometry to compare a highly
purified peroxisome preparation with a fraction of isolated peroxisomes to
define components that enrich with the peroxisomal membranes.


Summary.
The complete and accurate characterization of an organellar
proteome necessitates: 1) Thorough coverage of the sample-> achieved through
gas phase fractionation technique; 2) large quantities of relatively “simple”
protein mixtures->biochemical approaches were used to reduce the dynamic range
of proteins in the sample and enrich for membrane bound proteins; 3) the
ability to differentiate between bona fide components and contaminants from
other cellular sources at the “front end” of the analysis-> Quantitative (ICAT)
MS was used to identify proteins that enriched with an affinity purified
peroxisome fraction and 4) Statistical models to compare MS data-sets and
identify potentially interesting candidates. Candidates identified by this
method will be functionally characterized and their role in peroxisome biology
investigated. All aspects of this work will be discussed including: sample
preparation, statistical validation of proteins identified by MS/MS and
corroboration of “apparent” peroxisomal proteins by co-localization
experiments using fluorescently tagged proteins.


References.
Kunau and Hartig A Antonie Van Leeuwenhoek. 1992, 62(1-2):63; van den Bosch H, et al.
Annu Rev Biochem.
1992, 61:157; Subramani S.  Annu Rev Cell Biol. 1993, 9:445;
Clarke SD,
et al. Am J Clin Nutr. 1999; 70(4):566; Yi EC, et al.
Electrophoresis. 2002, 23(18) 3205. J.A. Ranish, et al. Nat. Genetics
2003, 33, 349.



Back to Workshop II: Medical Applications and Protein Networks