The boundaries of the silenced HMR domain in Saccharomyces cerevisiae

Abstract

The chromosomes of eukaryotes are organized into structurally and functionally discrete domains that provide a mechanism to compact the DNA as well as delineate independent units of gene activity. It is believed that insulator/boundary elements separate these domains. Here we report the identification and characterization of boundary elements that flank the transcriptionally repressed HMR locus in the yeast Saccharomyces cerevisiae. Deletion of these boundary elements led to the spread of silenced chromatin, whereas the ectopic insertion of these elements between a silencer and a promoter blocked the repressive effects of the silencer on that promoter at HMR and at telomeres. Sequence analysis indicated that the boundary element contained a TY1 LTR, and a tRNA gene and mutational analysis has implicated the Smc proteins, which encode structural components of chromosomes, in boundary element function.

Figure 1

The silenced domain emanates bi-directionally from the silencers. (A) A schematic representation of the 10-kb BamHI fragment encompassing the HMR locus with the sites of insertion of the URA3 gene shown. All coordinates used in this study are based on the Saccharomyces Genome Database (SGD) coordinates. Strain numbers are shown in parentheses. (B) Three sir3Δ strains, ROY656, ROY834, and ROY836, in which the URA3 gene was inserted at SGD coordinates 292140 (∼640 bp to the right of HMR-E), 290921 (∼475 bp to the left of HMR-E), and 288489 (∼2840 bp to the left of HMR-E), respectively, were generated. Strains ROY648, ROY508, and ROY513 are Sir⁺ derivatives of ROY656, ROY834, and ROY836, respectively. Cells were grown in liquid media, and 3 μl of 10-fold serial dilutions were spotted on YPD plates (complete), on supplemented YMD plates lacking uracil (−Ura), or on supplemented YMD plates containing 1 mg/ml 5-FOA to assay for URA3 expression.

Figure 2

(A) The 3.5-kb silenced HMR domain can be expanded to 7 kb. sirΔ strains carrying insertions (at SGD coordinate 293032) of either (1) 1-kb of the TRP1 gene (ROY49 and ROY1075) or (2) 1 kb of the TRP1 gene plus 2.5 kb of the HMG2 coding sequence (ROY55 and ROY1080) were generated. Strains ROY803, ROY84, ROY1076, and ROY 1079 are Sir⁺ derivatives of strains ROY49, ROY55, ROY1075, and ROY1080, respectively. All cells were grown in liquid media and 3 μl of 10-fold serial dilutions were spotted on either YPD plates (+Trp) or on YMD plates lacking tryptophan (−Trp) to assay for TRP1 expression. Strain numbers are shown in parentheses. (B) Differential restriction endonuclease digestion analysis of the expanded silenced HMR domain. Nuclei isolated from wild-type and sirΔ strains were digested with various restriction endonucleases. The DNA following purification was digested with a second restriction endonuclease and analyzed by DNA blot hybridization. For each site tested the band corresponding to wild-type cells is present on the left and the sirΔ strain is on the right.

Figure 2

(A) The 3.5-kb silenced HMR domain can be expanded to 7 kb. sirΔ strains carrying insertions (at SGD coordinate 293032) of either (1) 1-kb of the TRP1 gene (ROY49 and ROY1075) or (2) 1 kb of the TRP1 gene plus 2.5 kb of the HMG2 coding sequence (ROY55 and ROY1080) were generated. Strains ROY803, ROY84, ROY1076, and ROY 1079 are Sir⁺ derivatives of strains ROY49, ROY55, ROY1075, and ROY1080, respectively. All cells were grown in liquid media and 3 μl of 10-fold serial dilutions were spotted on either YPD plates (+Trp) or on YMD plates lacking tryptophan (−Trp) to assay for TRP1 expression. Strain numbers are shown in parentheses. (B) Differential restriction endonuclease digestion analysis of the expanded silenced HMR domain. Nuclei isolated from wild-type and sirΔ strains were digested with various restriction endonucleases. The DNA following purification was digested with a second restriction endonuclease and analyzed by DNA blot hybridization. For each site tested the band corresponding to wild-type cells is present on the left and the sirΔ strain is on the right.

Figure 3

Deletion of the right boundary leads to a spread of silencing. (A) A schematic depiction of the URA3 insertion at HMR in strains with (ROY687 and ROY853) or without (ROY852 and ROY850) the right boundary element. (B) ROY853 is a sir3Δ strain carrying a URA3 insertion between SGD coordinates 295027 and 295277 (∼1420 bp to the right of HMR-I) with an intact right boundary element. ROY687 is isogenic to ROY853 except that it is Sir⁺. In strain ROY850 (sir3Δ), the putative right boundary element (between nucleotides 293957 and 294977) was deleted and the URA3 gene with a 1-kb fragment of pUC18 stuffer DNA inserted between nucleotide 295027 and 295277. ROY852 is a Sir⁺ version of ROY850. All the strains were also transformed with SIR3 on a 2μ-based plasmid (pRO146). The cells were grown in liquid medium and 3 μl of 10-fold serial dilutions was spotted on either supplemented YMD plates lacking tryptophan (−Trp) to select for the plasmid, or supplemented YMD plates lacking tryptophan and uracil (−Trp−Ura), or on supplemented YMD plates lacking tryptophan but containing 1 mg/ml 5-FOA to assay for URA3 expression. Strain numbers are shown in parentheses.

Figure 4

Boundary elements block the spread of silenced chromatin. (A) (Top) A schematic representation of boundary element insertions at the EcoNI site of the MATa2 gene at nucleotide 292140 of wild-type HMR. (B) ROY113 (MATα HMRΔ, nucleotide 288980–295350) was transformed with a 10-kb fragment of wild-type HMR carrying either (1) a 1 kb stuffer fragment of the TRP1 gene, or (2) a 1.6-kb fragment of the putative left boundary element (from nucleotide 289256 to 290846), or (3) a 1-kb fragment of the putative right boundary element (from nucleotide 293695 to 294686) at the EcoNI site of the MATa2 gene (at nucleotide 292140). Expression of the MATa1 gene was monitored by patch-mating assays using JRY19 as the mating tester lawn. All matings were performed under conditions that constantly selected for the plasmid. (B) (Top) A schematic representation of boundary element insertions at the EcoNI site of the MATa2 gene at nucleotide 292140 of HMRΔI. (Bottom) ROY113 (MATα HMRΔ) transformants carrying an 8-kb fragment of HMRΔI with the same set of putative boundary insertions as described in A (bottom) were used to monitor expression of MATa1.

Figure 5

(A) The HMR boundary element blocked the spread of telomere position effect. ROY783 contained a synthetic telomere on chromosome VR with an URA3 insertion in close proximity to the telomere. Strains ROY791 and ROY787 carried 1-kb insertions of the right boundary element, whereas ROY838 and ROY844 carried 1-kb insertions of the TRP1 gene in both orientations. All strains were transformed with SIR3 on a 2μ-based plasmid (pRO329). The strains were grown in liquid media lacking leucine, and 3 μl of 10-fold serial dilutions were spotted on supplemented YMD plates lacking leucine (−Leu) to select for the plasmid, on supplemented YMD plates lacking leucine and uracil (−Leu−Ura), or on supplemented YMD plates lacking leucine but containing 1 mg/ml 5-FOA to assay for URA3 expression. (B) The right boundary element does not function in trans. Diploid strains bearing different versions of the HMR-URA3 locus were analyzed for expression of the URA3 gene as described in Fig. 3.

Figure 5

(A) The HMR boundary element blocked the spread of telomere position effect. ROY783 contained a synthetic telomere on chromosome VR with an URA3 insertion in close proximity to the telomere. Strains ROY791 and ROY787 carried 1-kb insertions of the right boundary element, whereas ROY838 and ROY844 carried 1-kb insertions of the TRP1 gene in both orientations. All strains were transformed with SIR3 on a 2μ-based plasmid (pRO329). The strains were grown in liquid media lacking leucine, and 3 μl of 10-fold serial dilutions were spotted on supplemented YMD plates lacking leucine (−Leu) to select for the plasmid, on supplemented YMD plates lacking leucine and uracil (−Leu−Ura), or on supplemented YMD plates lacking leucine but containing 1 mg/ml 5-FOA to assay for URA3 expression. (B) The right boundary element does not function in trans. Diploid strains bearing different versions of the HMR-URA3 locus were analyzed for expression of the URA3 gene as described in Fig. 3.

Figure 6

A deletion analysis of the boundary element. ROY113 (MATα hmrΔ) transformants carrying an 8-kb fragment of HMRΔI with a series of right boundary element deletions inserted at the EcoNI site of the MATa2 gene at nucleotide 292140 were used to monitor expression of the MATa1 gene as described in Fig. 4.

Figure 7

Extragenic mutations that affect boundary function. MATα strains, with either HMRΔI or HMRΔI with the right boundary element inserted between HMR-E and the MATa1 promoter at nucleotide 292140 and carrying mutations in various genes as indicated, were generated. The strains were monitored for expression of the MATa1 gene using a patch-mating assay as described in Fig. 4.