Supplemental Figure 1. Gene targeting of MIWI2.
(A) Schematic diagram of the genomic structure, target vector, and disrupted genomic structure of Miwi2. The targeting vector includes the PGK-neo gene (neo) and the thymidine kinase (TK) gene. (B) Southern blot analysis of the Miwi2 gene in tail DNA from Wild-type, heterozygous and homozygous mice. The wild-type allele produces a 4.5-kb KpnI and BamHI product, while the disrupted allele creates a 7.9-kb KpnI and XhoIfragment, with the 5′-end hybridization probe. (C) Comparison of the testes from 9-week-old mice. (D) Hematoxylin and eosin-stained sections of testes from 9-week-old mice. (E) Merged image of synaptonemal complexes stained with the B antiserum (Moens, P.B. andSpyropoilos, B. Chromosoma 104, 175-182. 1995), which recognizes both COR1 and SYN1 (green), and stained for DNA with DAPI (blue).
Supplemental Figure 2. Microarray analysis of MILI-deficient and control testes.
The gene expression profiles of Day-10 MILI–/– and MILI+/– testes were compared by Affymetrix GeneChip (U74Av2) (Affymetrix, Santa Clara, CA) microarray hybridization. Each dot indicates an individual gene. The right panel is a magnified version of the left panel, to enable visualization of the genes with significantly high expression levels. Five genes are up-regulated more than threefold in the MILI–/–testes, and all of these genes belong to the IAP retrotransposon (circled). The NCBI database accession number for the microarray data is GSE4742.
Supplemental Figure 3. Methylation-sensitive Southern blot analysis of the Line-1 promoter region in 2-week-old testes. Whole-testis DNA was extracted from 2-week-old control (+) and null (–) mice and digested with KpnI and the methylation-sensitive restriction endonuclease HpaII or the methylation-insensitiveenzyme MspI. The type Gf Line-1 5′-noncoding region was used as a probe.
Supplemental Figure 4. Expression of Dnmt3L and Dnmt3a2 in MILI-deficient testes
(A) Semi-quantitative RT-PCR analysis ofDnmt3L and Dnmt3a2 in the MILI- deficient fetal testis. Total RNA samples were prepared from E16.5 testes of MILI-deficient and control mice. Equivalent amounts of total RNAwere subjected to reverse transcription and different numbers of PCR cycles (28, 30, and 32 cycles for Dnmt3L, Dnmt3a2, and Mili; 24, 26, and 28 cycles for G3pdh; from left to right, respectively). The primer sequences (Sakai et al. 2004)are shown in Supplemental Table S6.
(B) Immunohistochemical analysis of Dnmt3L and Dnmt3a2 expression in the MILI-deficient fetal testes. Testes were dissected from E16.5 embryos and fixed with 4% paraformaldehyde. The fixed testes were washed with PBS that contained 10% sucrose, embedded in Tissue-Tek OCT compound (Pelco, Redding, CA), and cryosectioned. Sections of E16.5 embryonic male gonads from MILI+/- and MILI-/- mice were stained with the anti-Dnmt3L and anti-Dnmt3a2 antibodies (Sakai et al. 2004) and with DAPI.
Supplemental Figure 5. Expression of MIWI2 in developing gonads.
(A) Expression of Miwi2 in the developing gonads was analyzed by RT-PCR. The data for whole embryos from E7.5 to E10.5, gonads of E11.5, male gonads from E12.5 to 3 weeks postpartum, and MILI-deficient testes are shown. (B) Semi-quantitative RT-PCR analysis on the representative date. Equivalent amounts of total RNAwere subjected to reverse transcription and different numbers of PCR cycles (28, 30, and 32 cycles for Miwi2; 24, 26, and 28 cycles for G3pdh; from left to right, respectively).
Supplemental Table 1: Annotation of embryonic testes small RNAs
Supplemental Table 2: Annotation of repeat small RNAs
Supplemental Table 3: piRNA clusters on the embryonic testes
Supplemental Table 4: Characterization of repeat associated small RNAs
Supplemental Table 5: Target sequences for microarray
Supplemental Table 6: Primers used in this study