Environment, Genes, and Cancer

In January, comedian George Burns turned 100 years old. In recent appearances in the media, he still seems sharp as a tack, and is still seen smoking his trademark cigars. Others of us, however, were never very funny, and would die of cancer at age 60 if we continuously smoked cigars or cigarettes. Burns presents a common but perplexing paradox; some people are able to tolerate at least moderate exposure to toxins such as cigarette smoke with little adverse affect, while others develop cancer, emphysema, or heart disease. New studies support the idea that there is an interaction between genes and the environment, and that this interaction may be an important determinant of cancer risk. To understand such risks, it is essential to look at both an individual`s genetic makeup and environmental exposures. Such studies require the collaboration of molecular epidemiologists and molecular biologists. At the NIEHS, Jack A. Taylor, a lead clinical investigator in the Epidemiology Branch, and Douglas A. Bell, an investigator with the Genetic Risk Group of the Laboratory of Biochemical Risk Analysis, have worked together and with other scientists to uncover new information in this area.


Mde4 Protein Sequences
The Schizosaccharomyces pombe Mde4 protein sequence was derived from the NCBI protein database (gij2894279jembjCAA 17047.1), the Mde4 fragment from Schizosaccharomyces japonicus was translated via S. pombe Mde4 and genomic sequences deposited by the Whitehead Institute for Biomedical Research at the NCBI trace archive (gnljtij955794238 and gnljtij1044634558).
The motif within the coiled-coil region was defined by the Gibbs recursive motif sampler [S2].
Tandem Affinity Purification TAP purifications from mitotic and meiotic cells and mass spectrometry were performed as previously described [S5].

ChIP-chip Method
ChIP-chip analyses were carried out as previously described [S6] except for DNA amplification step. We noticed that PCR-based amplification of pombe genomic DNA is susceptible for biased amplification, because of more complex genomic sequence of S. pombe than S. cerevisiae. Therfore, we adopted T7-based in vitro transcription linear amplification method (IVT amplification method) as previously described [S7]. After addition of poly T tails to the ends of DNA fragments to be amplified by terminal transferase, strand synthesis was carried out with a T7 polyA primer adaptor to produce doublestranded templates suitable for IVT. After IVT was performed, amplification products was transformed into cDNA and used for labeling and hybridization to S. pombe whole-genome tiling array as previously described [S6]. S. pombe whole-genome tiling array was commercially available from Affymetrix (S. pombe 1.0FR array, P/N 900647). The S. pombe tiling array consists of 6 million probe pairs tiled through the complete S. pombe genome. Probes are tiled for both strands of the genome at an average of 20 base pair resolution, as measured from the central position of adjacent 25-mer oligos, creating an overlap of approximately 5 base pairs on adjacent probes. The information of oligo probes on the arrays (sequence and location on the chip) is available from Affymetrix. For the primary analysis of tiling chip data, a domestic software was constructed that exactly follows the statistical algorithm used for Affymetrix GeneChip Operating Software. The detailed information for the algorithm used can be downloaded from the Affymetrix web site at http://www.affymetrix.com/support/technical/technotes/ statistical_reference_guide.pdf. The analysis is available from K.S. on request. In all cases, one unit for analysis (locus) was set to 100 bp with the step size of 50 bp. Fold change value, change p value, and detection p value for each locus were obtained by primary analysis. For the discrimination of positive and negative signals for the binding, we used three criteria as follows. First, the reliability of the signal strength was judged by detection p value of each locus (p value % 0.001). Second, reliability of binding ratio was judged by change p value (p value % 0.001). Third, clusters consisting of at least 500 bp contiguous loci that satisfied the above two criteria were selected, because it is known that a single site of protein-DNA interaction resulted in immunoprecipitation of DNA fragments that hybridized not only to the locus of the actual binding site but also to its neighbors [S6, S8].
All the array data are available at GEO database (http://www.ncbi. nlm.nih.gov/projects/geo/) through the accession number of GSE5257.

Strains, Media, and Growth Conditions
The genotypes of S. pombe strains used in this study are listed in the Table S4. S. pombe media and growth conditions were as described in [S9].

Microscopy
The immunofluorescence and microscopy techniques were as described in [S9]. Chromatin binding assay was performed as described in [S10].

Immunofluorescence and Live-Cell Imaging
Cultures were grown at 25 C and then shifted to 18 C for 2-4 hr (to increase the frequency of lagging chromosomes). Cells were fixed for 7 min with 3.7% formaldehyde at room temperature before processing for immunofluorescence [S11]. Sheep anti-Cnp1 serum was used at 1:3000; rabbit anti-GFP (Molecular Probes) was used at 1:3000. Alexa-488, -568, or -594 coupled secondary antibodies (Molecular Probes) were used at 1:1000. Live imaging was performed as described previously [S12]. Images were captured with a Zeiss Axioplan imaging 2 microscope and 1003 Plan Neofluar 1.3NA, 1003 PlanApochromat 1.3NA, or 1003 Fluar 1.3NA objectives, Chroma 86000 and 84000 filter sets, Photometrics Cool-SNAP-HQ camera, with Metamorph Software (Universal Imaging). Images presented have been autoscaled (Metamorph). Further adjustments to brightness and contrast have been made to the cen2-GFP image in Figure 5 (green channel only) to highlight the cen2-GFP spots. Filming was performed at 21 C-25 C. 1 s exposures were taken every 30 s. Quicktime movies are speeded up 1803. Brightness and contrast have been altered in the selected stills shown in Figures 6 and S8.

Classification and Measurements of Lagging Chromosomes
To determine the percentage of lagging kinetochores that were stretched versus normal, cells containing lagging chromosomes were identified in the DAPI channel, and then kinetochore and DAPI images were captured. Lagging kinetochores were classified on the basis of morphology (see Figure 4) as normal (dot-like), elongated, bilobed shape, or split. It was not possible to categorize some lagging kinetochores, e.g., because the pattern of DAPI and CENP-A Cnp1 staining was too complex, and these kinetochores were not included in the totals. Kinetochore dimensions were measured by Metamorph software. Kinetochores were classified as elongated if their ratio of length (in the plane of the spindle) to width was 1.5 or greater; those with a length:width ratio of less than 1.5 were classified as dot-like. Thus, the elongated class is probably underestimated; this may be reflected in the measurement data presented in Table S2, which indicates that the ''dot-like'' lagging kinetochores are slightly elongated in the plane of the spindle.

Quantification of DAPI Fluorescence Intensity
With the known sizes of fission yeast chromosomes [S13], the theoretical relative masses of all chromosomes (or combinations) were (B) Proteins associated with TAP-tagged Pcs1 were isolated by tandem affinity purification from diploid S. pombe cells induced to enter synchronous meiosis by inactivation of Pat1 (K12524) [S14] and harvested around metaphase I. calculated and expressed as percentages of the total DNA in a mitotic cell (2n DNA content), as shown in Table S3. Images of cells with lagging chromosomes were inspected and those that fit the following criteria analyzed: the laggard was distinct from the other DNA, and the DNA at both poles was also in focus. Cells with very complex segregation patterns were not included in the analysis. Rare cells (w1%-5% of cells with laggards) with very small DAPIstained objects between the poles that might represent chromosome fragments were not included. Quantification of DAPI fluorescence intensity was carried out with Metamorph software. Regions of interest (ROI) were drawn around each DAPI-stained object in the image (typically three), and identical ROIs were placed at DAPI-negative regions of the image to measure background fluorescence (i.e., 3 DNA ROIs and 3 background ROIs per image). For each DNA ROI, DAPI fluorescence intensity was integrated, and the appropriate integrated background fluorescence was subtracted. The relative mass of each DAPI object was expressed as a percentage of the total DAPI fluorescence in the cell. Three sets of cells were analyzed. (1) The method was shown to be valid and accurate by analysis of clr4D cen2-GFP cells with single chromosome 2 laggards (i.e., one cen2-GFP spot on the laggard, one at the pole). The theoretical relative masses are: 50% (chromosomes 1+2+3 at one pole): 33.4% (chromosomes 1+3 at the other pole): 16.7% (chromosome 2 laggard). The relative masses calculated from DAPI fluorescence intensity measurements (49.0% 6 2.1%: 33.0% 6 2.2%: 17.8% 6 1.6%, n = 16) were very similar to theoretical values. In addition, three cells in which the cen2-GFP spots were very close or coincident (i.e., adjacent or unseparated chromosome 2) were analyzed.
(2) clr4D and pcs1D cells with cen2-GFP that were stained with anti-CENP-A Cnp1 antibodies. Because it was critical to establish that the laggards with stretched/split CENP-A Cnp1 morphology are  indeed single chromatids, the analysis focused on these cells. Therefore, only a small number of cells displaying a ''dot'' CENP-A Cnp1 signal on laggards were analyzed for the purposes of comparison (although they account for w70% of cells with lagging chromosomes). Cells with unseparated/adjacent cen2-GFP were also analyzed.
(3) clr4D, pcs1D, and mde4D cells with unmarked chromosomes (cells used for Figure 4 and Tables S1 and S2). All cells with split/stretched CENP-A Cnp1 morphology that were suitable for DAPI fluorescence quantification (see above) were analyzed. A few cells with a large lagging mass of DNA, which had been interpreted as two adjacent laggards (e.g., ''dot'' plus ''stretched'' laggards), were also analyzed.  Wild-type (K11248), pcs1D (K14820), mde4D (K14818), and pcs1D mde4D (K14822) cells were serially diluted and spotted onto YES medium containing 10 or 20 mg/ml of thiabendazole (TBZ). Cells were grown for 2 days at 32 C. Figure S5. Pcs1 and Mde4 Are Dispensable for Chromosome Segregation during Meiosis I but Are Required for Meiosis II A wild-type h 90 lys1-GFP strain carrying lacO sequences integrated near the lys1 locus and expressing the LacI-GFP fusion protein, which binds to lacO (wt) (K11248), and h 90 lys1-GFP strains carrying either the knockout allele of pcs1 (pcs1D) (K11420), or the knockout allele of mde4 (mde4D) (K12495), or a double knockout (pcs1D mde4D) (K12539) were sporulated. Unfixed cells were stained with Hoechst and examined under the fluorescence microscope. Segregation of chromosome I was scored in at least 80 asci. Spore viability was determined by dissecting spores from at least 60 asci and scoring the ability to form colonies. Figure S6. Analysis of Sister-Chromatid Segregation during Meiosis I and Meiosis II (A) The wild-type strain h 2 lys1-GFP (wt) (K11338) or strain h 2 lys1-GFP strain carrying a knockout allele of pcs1 (K14820), mde4 (K14818), or a double knockout of pcs1 and mde4 (K14822) were crossed to h + strains that were of the same genotype and lacked lys1-GFP (K11339, K14821, K14819, and K14823, respectively). Cells were sporulated, fixed, stained with Hoechst and antibodies against tubulin and GFP, and examined under the fluorescence microscope. Chromosome I was visualized by lys1-GFP (heterozygous lys1-GFP). Anaphase I cells were identified according to tubulin staining of the spindle and number of nuclei. At least 100 cells were counted. Segregation of chromosome I and lagging chromosomes were scored in at least 100 anaphase I cells.

Supplemental References
(B) Cells were processed and analyzed as described in (A). Segregation of chromosome I, visualized by lys1-GFP (heterozygous lys1-GFP), and lagging chromosomes were scored in at least 90 anaphase II cells. The CENP-A Cnp1 staining pattern on lagging chromosomes was classified as dot-like (i.e., of normal appearance), split, bilobed, or elongated, examples of which are shown in Figure 4. Numbers are given as percentages. Ambiguous cases that could not be classified were excluded from the analysis. The elongated class is composed of kinetochores for which the ratio of length (along the spindle axis) to width was at least 1.5; those with a length:width ratio of less than 1.5 were included in the ''dot'' category.  CENP-A Cnp1 -stained kinetochores were measured in the indicated strains. Measurements are in microns (mean 6 SD): top numbers indicate length of kinetochore (i.e., along plane of spindle); lower numbers indicate width. Lagging kinetochores in anaphase were measured, except for wild-type in which kinetochores in early mitosis were measured. The sizes of S. pombe chromosomes are shown, along with the percentage contribution each chromosome or combination of chromosomes makes to the 2n DNA content of a mitotic cell. h + pcs1D::ura4 + leu1 ura4 ade6-210 lys1 his7 RA6419 pcs1D::ura4 + mis12-GFP-LEU2 + leu1-32 ade6-210 ura4 RA6428 pcs1D::ura4 + nuf2-GFP-KAN leu1-32 ade6-210 ura4 RA8618 = K14819 h + mde4D::KANMX4 lys1 his7 leu1 ura4 ade6-210 FY10533 pcs1D::ClonNat cen2(D107)::KAN-ura4 + -lacO his7 + ::lacI-GFP Figure S7. Lagging Chromosomes in clr4D Cells Are Single Chromatids clr4D cells in which cen2 is marked by GFP were fixed and processed for immunofluorescence with anti-GFP antibody. Those with a lagging cen2-GFP signal were scored (n = 100).