Biology Reference
In-Depth Information
10.2. Description of Data
To d efine abnormalities in the early phases of autoimmunity, we have conducted
comprehensive studies of gene expression in young NOD mice and mice from
control strains (NON and C57BL/6) that do not develop diabetes or autoimmunity
to beta cells [13, 14]. Genes encoded in DNA are transcribed into mRNA, which
is then translated into proteins that are the major determinants of a cell's activa-
tion and function. To gain a comprehensive picture of how the genetic differences
between our strains can affect the development of autoimmunity, we evaluated
gene expression at the mRNA level using Affymetrix MOE430A/B arrays, and at
the protein level using 2D-gel electrophoresis. We collected mononuclear spleen
leukocytes from each of the three strains at both two and four weeks of age. This
is a critical window for our analysis, because it represents the prepathology stage
before leukocytes begin to infiltrate the islets of Langerhans, which typically oc-
curs in NOD pups when they are about five weeks old. See Fig. 10.1. From each
of the six strain/age groups, we collected five independent samples for a total of
30 samples in the complete dataset. Experimental details regarding the analysis
of mRNA and protein expression levels have been published previously [14].
Fig. 10.1. At birth, NOD mice have normal blood glucose levels, with no indication of destruction of
insulin-producing beta cells (located in the islets of Langerhans). At five weeks of age, the first signs
of pathology occur with leukocytes invading the space around the islets. This infiltration progresses
to involve additional leukocytes with more invasive and destructive character. By twelve weeks of age
or later, so many beta-cells have been destroyed that insulin production capacity has been severely
diminished. Because blood glucose can no longer be regulated normally, the mice become diabetic.
We sampled leukocytes in the early and late prepathology stage to evaluate defects at the molecular
level associated with initiation of the pathology.
Because the data is biological, it has a fairly high level of noise. At the time of
sample collection, individual mice may or may not have just eaten, been fighting,
been scared, been sleeping, etc. These biological parameters can be difficult to
control, and can have an influence on expression levels of some genes. In addition
to this biologically derived noise, there are also technical sources of noise to be
considered. The mRNA gene expression arrays have a very effective normaliza-
tion and scaling process and very good technical reproducibility on identical sam-
Search WWH ::
Custom Search