Analysis of global gene expression in complex biological systems using microarray technology

Abstract

The general aim of this thesis has been to explore the use of Affymetrix microarray technology in studies of global gene expression in complex biological systems, particularly with regard to the dynamics and heterogeneity of gene expression. The effects and systems studied are: 1. Delayed effects on gene expression in bulk cultures and individual cell clones of primary human lymphocytes exposed to ionising radiation (study I). 2. Changes in gene expression in relation to genetic markers, disease progression and prognosis in chronic lymphocytic leukemia (B-CLL) (study II and III). 3. The effects of genetically engineered over-expression of a single gene that is important in recombinational DNA repair, HsRad51, on the gene expression in a human fibrosarcoma cell line (study IV).

The results from study I showed that: A notable variability and progressive differentiation of gene expression was observed during long-term culture. In bulk cultures, the number of genes which displayed a change of gene expression after irradiation increased dramatically with increasing culture time. The changes of gene expression showed a significantly more diversified pattern in irradiated Tcell clones than in non-irradiated clones. Finally, by combining gene expression profiles from bulk cultures and cell clones we were able to sort out a set of genes whose change of expression correlates with radiation exposure regardless of cell origin and culture time.

In study II, the presence or the lack of a genetic marker (VH3-21) was used to distinguish between two BCLL populations. The result showed a distinct expression profile for VH3-21+ B-CLL vs. non- VH3-21+ B-CLL. Our novel finding of a specific 'VH3-21 profile' strengthens the suggestion that this group comprises a separate subgroup of B-CLL and may also give insights into the pathogenesis Of VH3-21+ B-CLL.

In study III B-CLL patients were classified according to clinical criteria in stable or progressive disease. No distinction based on expression profile was obtained between these groups. Nevertheless, it was possible to identify a set of discriminatory genes that might be of importance for the progression of the disease. In addition, with an unsupervised clustering approach we also identified subgroups among the samples from patients with progressive disease.

In study IV we used an in vitro system to study how overexpression of one single gene affects the global gene expression. The most prominent findings were the over-representation of mismatch repair (MMR) genes which play a role both directly in recombination& repair and in the regulation of the repair process. These genes were transcriptionally downregulated in response to increased HsRad51 protein levels and transcriptionally upregulated in response to decreased HsRad51 levels.

The overall conclusions from these results are that global gene expression in complex biological systems is subject to dynamic change and may display considerable variation during cell proliferation, stress response and disease progression. Nevertheless, the combination of global, subtle changes in many genes may be useful as a marker for biological change in complex cell systems. Subgroup classification based on clinical observations (e.g. stable and progressive B-CLL) might represent multiple causes and mechanisms for a disease and provide less distinction in gene expression profiles, compared to more concrete sub-classification based on a common genetic marker (e.g. VH3-21). If single marker genes do not give a clear distinction between two clinically separated patient subgroups a combination of genes might give a more clear class distinction.

The results of this thesis support the notion that the global gene expression in normal cells is in a state of robust dynamic equilibrium. Since the expression of genes form interactive 'networks', the response to either an acute stress affecting many genes (e.g. ionising radiation) or to a change of a single gene's expression (e.g. genetically engineered upregulation) will depend on how many pathways and downstream genes that are affected. Eventually, the global gene expression may be restored at the same or slightly different level, or the balance may move to a distinctly different state that may be visualized as a persistent change in the global expression profile. The understanding of the dynamic interactions in these networks and connections is a major challenge for the future studies of the regulation of gene expression in complex biological systems.