3.5. Functional Validation Using Marker Genes
Based on the results of the functional analysis, we examined the expression levels of genes involved in cell cycle regulation, DNA metabolic processes and chromosome organization. We selected genes with well-described functions in order to best characterize the corresponding biological responses.
microarrays-04-00025-t002_Table 2 Table 2  Biological processes based on GO terms that are differentially represented in the PpIXT, XT and PpIX-XT groups, with reference to the NT group. The data were generated using the functional annotation chart provided in Database for Annotation, Visualization and Integrated Discovery (DAVID).   We first examined marker genes involved in cell cycle regulation, because the functional analysis identified 15 GO terms related to this process. Figure 5a shows the heat map of marker gene expressions for cell cycle regulation. Each row represents a gene, and each column represents the ratio of the average fold change in gene expression relative to the NT group, as well as associated p-values. The color intensity depicts the degree of expression; the blue color indicates highly negative expression, while red indicates highly positive expression. Cyclins, cyclin-dependent kinases and cyclin-dependent kinase inhibitors, among other genes, were selected as markers of the cell cycle. CCNE1, CCNE2 and CDK4 were downregulated, whereas Cdkn1a was upregulated in the XT and PpIX-XT groups. Furthermore, upregulation of ATM, CHK2 and CDKN1A indicated inhibition of DNA replication. Figure 5b shows the heat map of the selected genes for the DNA metabolic process (GO:0006259). Key genes involved in DNA replication, such as PCNA, the MCM complex (MCM5, MCM8 and MCM10), LIG1 and DNA polymerase (POLA1, POLD3), were downregulated relative to the NT group. Therefore, XT and PpIX-XT treatment arrested the tumor cells at the G1 transition of the cell cycle and inhibited DNA replication, whereas PpIX treatment showed no effect on the cell cycle or on the DNA replication compared to the NO group.
Figure 5  Heat map of marker genes representing (a) cell cycle regulation and (b) DNA metabolic processes. Each row represents a gene, and each column represents the average fold change in gene expression relative to the NT group (* p < 0.01). The colors indicate the intensity based on the log expression ratio from blue (highly negative) to red (highly positive).   Ionizing radiation typically induces direct and/or indirect damage to DNA, triggering various cellular responses, including cell-cycle arrest, transformation and cell death [20,21,22]. The results of our study suggest a similar pattern in response to DNA damage. The PpIX-XT group differed from the XT group in terms of quantitative gene expression, but qualitative gene expression was the same in both groups.
In the case of genes selected for chromosome organization, we found that several core histones (HIST1H2BB, HIST1H2BD, HIST1H2BE, HIST1H2BH, HIST1H2BJ, HIST1H2BL, HIST1H2BO, HIST2H2BE and HIST3H2BB) were upregulated in both the XT and PpIX-XT groups. A significant amount of the histone synthesis occurs during the S phase. The synthesis of the histone proteins is tightly coupled to DNA synthesis [23,24]. Histone synthesis in cycling tissue culture cells can be separated into basal synthesis and S phase synthesis. The upregulation of histone gene expressions in this study was caused by the basal synthesis or by transitional response. The genes responsible for cellular responses to stress were identical to the genes involved in DNA metabolic processes.
In our previous in vivo study, treatment with 3-Gy irradiation at 10 intervals significantly suppressed tumor growth. However, in this study, we irradiated the cells with a single dose (3 Gy) following administration of PpIX. This dose appeared to be weak and ineffective at suppressing tumor growth. Therefore, we evaluated microarray data, which revealed systematic changes in cellular biological responses, as well as differences between the PpIX-XT and XT treatments.