. 2009 Mar 6;109(2):193–205. doi: 10.1093/toxsci/kfp050

TABLE 2.

Expression of Genes that Contain Putative CAREs^a,^b

Gene^c		Time point (weeks) in B6C3F1	Microarray analysis^d	qRT-PCR analysis	Methylation change^e	No. of putative CAREs
Exhibited expression changes based on microarray analysis^f	*Abcc3*	2	↑	C^g		5
	*Bax*	2	↑	A.C.^h		6
	Casp3	2	↑	n.c.ⁱ		2
	*Cbs*	2	↓	C^g		3
	*Cebpb*	2	↑	C^g		2
	*Gadd45b*	2	↑	A.C.^h		4
	Id3	2	↓	n.c.ⁱ		1
	Lcn2	2	↑	n.c.ⁱ		1
	*Prdm2*	2	↓	A.C.^h		1
	*Shmt1*	2	↑	A.C.^h		2
	*Tnfsf10*	2	↑	C^g		1
	Ephb4	4	↑	n.c.ⁱ		5
	*Gadd45a*	4	↑	C^g		2
	*Map3k5*	4	↑	A.C.^h		4
	*Mtrr*	4	↑	C^g		1
	*Plk3*	4	↑	A.C.^h		1
	Ppargc1a	4	↑	n.c.ⁱ		1
	*Wee1*	4	↑	C^g		5
DNA methyltransferases^j	Dnmt1	2	N.D.^k	↓^l		3
	Dnmt1	4	N.D.^k	↓^l		3
	Dnmt3a	2	N.D.^k	↓^l		2
	Dnmt3a	4	N.D.^k	↓^l		2
	Dnmt3b	2	↓	C^g		1
	Dnmt3b	4	N.D.^k	↓^l		1
Exhibited unique regions of altered DNA methylation (RAMs)^m	Anapc7	4	N.D.^k	↓^l	HypoM	2
	Bcat2	4	N.D.^k	↓^l	NewM	2
	Bcl2l13	4	N.D.^k	↓^l	NewM	2
	Ddx54	2	N.D.^k	n.a.ⁿ	HypoM	1
		4	N.D.^k	↓^l	HypoM
	Ncor2	4	N.D.^k	↓^l	NewM	2
	Plekhf1	2	N.D.^k	n.a.ⁿ	NewM	3
		4	N.D.^k	↓^l	HypoM
	Prickle2	2	N.D.^k	n.a.ⁿ	HypoM and/or NewM^∧	1
		4	N.D.^k	n.s.^o	HypoM
	Srms	2	N.D.^k	↓^l	HypoM or HyperM, and/or NewM^∧	4
	Tgfbr2	4	N.D.^k	n.s.^o	HypoM	1

Base sequences of putative CAREs, located 10 kb upstream of the transcriptional start site, are listed in Supplementary Table S10.

We used the same position weight matrix and 80% stringency cut-off (as described in “Methods”) on 25,000 randomly constructed sequences of 10 kb in length and found that, on average, there were only 0.561 CAREs found per sequence (across all 25,000 sequences). The probability of a gene having more than 1 CARE is 10.8%, whereas the probability of having more than 2, 3, or 4 CAREs is 2.2, 0.38, and 0.044%, respectively. The probability of not having any CAREs is 57.4%. We view the informatic approach for identification of CAREs as hypothesis generating.

Genes that exhibited unique PB-elicited changes in expression (Table 1, Supplementary Tables S3G–I and S4–S5) and/or methylation (Phillips and Goodman, 2008a) in B6C3F1 mice at 2 and/or 4 weeks of PB treatment.

Expression changes (↑, upregulation and ↓, downregulation) of uniquely active genes that were identified based on microarray analysis (as described in the Methods). For genes and time points listed in bolded font, the expression change was confirmed, or appeared to be confirmed, by qRT-PCR analysis. The expression of genes listed in normal black font was evaluated but was not confirmed.

PB-induced decreases in methylation (Hypomethylation, HypoM) and increases in methylation (Hypermethylation, HyperM and new methylations, NewM), as determined by Phillips and Goodman (2008). For certain genes (^∧), the methylation statuses were ambiguous (i.e., not definitively determined).

Detailed confirmatory qRT-PCR data are found in Supplementary Tables S4 (2 weeks) and S5 (4 weeks).

The expression change, based on microarray analysis, was confirmed (C), as described in the “Methods.”

The expression change, based on microarray analysis, was apparently confirmed (A.C.), as described in the “Methods.”

ⁱ

The expression change, based on microarray analysis, was not confirmed (n.c.).

Detailed qRT-PCR data are found in Table 1.

N.D. (not detected) indicates that the gene was not active.

Direction (↑, upregulation and ↓, downregulation) of the statistically significant (Student's t-test, p < 0.05) fold change.

Detailed qRT-PCR data are found in Supplementary Table S9.

ⁿ

Expression was not evaluated.

Fold change was not significant, n.s. (Student's t-test, p < 0.05).