Hum Genet (2005) 116: 172–178 DOI 10.1007/s00439-004-1215-6
O RI GI N AL IN V ES T IG A T IO N
Fei Sun Æ Kiril Trpkov Æ Alfred Rademaker Evelyn Ko Æ Rene´e H. Martin
Variation in meiotic recombination frequencies among human males
Received: 1 September 2004 / Accepted: 13 October 2004 / Published online: 1 December 2004 Springer-Verlag 2004
Abstract Meiotic recombination is essential for the segregation of homologous chromosomes and the formation of normal haploid gametes. Little is known about patterns of meiotic recombination in human germ cells or the mechanisms that control these patterns. Here, newly developed immunofluorescence techniques, based on the detection of MLH1 (a DNA mismatch repair protein) foci on synaptonemal complexes (SCs) at prophase I of meiosis, were used to examine recombination in human spermatocytes. The mean number of MLH1 foci per cell in all donors was 48.0 with range from 21 to 65. Remarkable variation in the recombination frequency was noted among 11 normal individuals: the mean frequencies of chromosomal recombination foci ranged from a low of 42.5 to a high of 55.0 exchanges. Donor age did not contribute to this variation. There was no correlation between this variation and the frequency of gaps (discontinuities) or splits (unpaired chromosome regions) in the SCs. The mean percentage of cells with gaps was 35% (range: 20% to 58%) and with splits was 7% (range: 0% to 37%). Bivalents without a
F. Sun Æ R. H. Martin Department of Medical Genetics, University of Calgary, Calgary, T2N 4N1, Canada F. Sun Æ E. Ko Æ R. H. Martin (&) Department of Genetics, Alberta Children’s Hospital, 1820 Richmond Road S.W., Calgary, Alberta, T2T 5C7, Canada E-mail:
[email protected] Tel.: +1-403-9437369 Fax: +1-403-5439100 K. Trpkov Department of Pathology, Rockyview Hospital, Calgary, T2V 1P9, Canada A. Rademaker Cancer Center, Biometry Section, Northwestern University Medical School, 60611-4402 Chicago, USA
recombination focus were rare, with a frequency of only 0.3%. Thus, achiasmate chromosomes appear to be rare in human male meiosis.
Introduction Meiotic recombination not only generates genetic variation but is also vital for the correct segregation (i.e., disjunction) of homologous chromosomes at the first meiotic division (Hassold et al. 2000). It is well established that this process is under genetic control, with stringent regulation of the number and distribution of exchanges. However, recombination frequencies are not fixed and are subject to intrinsic (e.g., genetic background; Koehler et al. 2002) and extrinsic (e.g., age, environmental effects; Rose and Baillie 1979; Tanzi et al. 1992; Williamson et al. 1970) factors that contribute to significant intra- and inter-individual variability. Thus, variation in the frequency of meiotic recombination has been observed between the sexes and among different strains and/or species in several organisms, including flies, bacteria, fungi, plants, and animals (Broman et al. 1998; Gorlov et al. 1994; Kong et al. 2002; Lawrie et al. 1995; Reeves et al. 1990; True et al. 1996; Zwick et al. 1999; for reviews, see Brooks 1988; Fisher-Lindahl 1991; Robinson 1996). Although little is known about the source of this variability, an understanding of the cause and extent of such variation would have widespread applications from genetic counseling to studies of genome evolution. Recently developed fluorescence immunostaining assays, which can recognize important meiotic proteins, have revolutionized meiotic research (Anderson et al. 1999; Baker et al. 1996; Barlow and Hulte´n 1998; Lynn et al. 2002; Tease et al. 2002). Mammalian MLH1, an ortholog of the Escherichia coli Mut L mismatch repair protein has been found to mark recombination sites. Specifically, MLH1s appear as discrete foci along the synaptonemal complexes (SCs), viz., the proteinaceous
173
structure linking homologous chromosomes in prophase of meiosis I. This distribution parallels the pattern of meiotic recombination events (Anderson et al. 1999; Barlow and Hulte´n 1998) and is in keeping with previous cytogenetic and molecular studies of meiotic recombination in humans and mice (Hulte´n 1974; Lawrie et al. 1995). Using antibodies against SCP1 (SC protein 1; marks transverse elements of the SC), and SCP3 (marks lateral elements of the SC), MLH1, and CREST (calcinosis, Raynaud’s phenomenon, esophageal dysfunction, sclerodactyly, telangiectasia; marks the centromere), meiotic recombination patterns and fidelity of chromosome pairing in pachytene-stage cells can be identified. This direct immunocytological analysis offers a rapid and reliable means of assessing meiotic recombination in human germ cells, provides an overview of recombination across the whole genome, and is, at the same time, able to determine chromosome-specific patterns of crossing over (Oliver-Bonet et al. 2003; Sun et al. 2004b). Recombination maps for individual chromosomes in the male mouse (Froenicke et al. 2002) and in a normal human male (Sun et al. 2004b) have been recently produced by using the combined techniques of immunostaining and multicolor fluorescence in situ hybridization (FISH) on testicular biopsy samples. To date, little information is available on the variation in the recombination frequency in humans, as this has been investigated in only one study (Lynn et al. 2002). In the current study, the extent of this variation in normal men and the underlying causes of this variation, e.g., the effects of age and of gaps (discontinuities) or splits (unpaired chromosome regions) in SCs, have been examined by using newly developed immunostaining techniques.
Materials and methods Testicular sample collection An ideal study population for the identification of interindividual variation among human males would consist of volunteers of known fertility, but difficulties in acquiring testicular samples makes this impractical. Instead, samples were analyzed from urology patients who were ascertained for reasons unrelated to meiotic defects. Specifically, testicular samples were obtained from orchiectomies in men who had testicular cancer (n=2), epididymis cancer (n=1), prostate cancer (n=4), Leydig cell cancer (n=1), epididymitis (n=1), differentiated liposarcoma (n=1), and a non-reproductive cancer (patient had normal testes; n=1). In all cases, histological examination showed normal spermatogenesis. Lynn et al (2002) did not find any difference in recombination frequencies based on patient status (e.g., cancers, cystic fibrosis, or previous vasectomy). Testicular material was kept in phosphate-buffered saline (PBS, pH 7.4) until used in experiments. This study received ethical approval from the University of Calgary institutional review board.
Fluorescence immunostaining Specimens were processed for analysis by a modification of the technique detailed by Barlow and Hulte´n (1998). Briefly, testicular tissue was shredded, and the released pachytene cells were spread evenly over microscope slides layered with paraformaldehyde solution at pH 9.2 (Fisher Scientific, Edmonton, AB, Canada) and TritonX (Sigma, Oakville, ON, Canada). Slides were dried for approximately 24 h at room temperature in a humid chamber; drying was completed on the bench for approximately 30 min. Dried slides were washed for 4 min in 0.04% Photo-Flo (Kodak, Rochester, N.Y., USA), drained, and air-dried. Air-dried slides were soaked for 30 min in 1 · ADB (1% normal donkey serum, Jackson ImmunoResearch, West Grove, Pa., USA; 0.3% bovine serum albumin, Sigma, Oakville, ON, Canada; 0.005% Triton-X in PBS), drained but not allowed to dry, overlaid with a cocktail of primary antibodies, viz., human CREST (a gift from M. Fritzler, University of Calgary), rabbit MLH1 (Oncogene, San Diego, Calif., USA), goat SCP3 (a gift from T. Ashley, Yale University), and mouse SYN1 (a gift from P. Moens, York University), then cover-slipped, and incubated overnight at 37C. Cover slips were removed, and the slides soaked 10 min in 1 · ADB. After a second 1· ADB wash at 4C for 48 h, slides were incubated with a cocktail of secondary antibodies, viz., AMCA-conjugated donkey anti-human IgG (Jackson Immunoresearch), Alexa 488-conjugated donkey anti-rabbit IgG (Molecular Probes, Eugene, Ore., USA), Alexa 555conjugated donkey anti-goat IgG (Molecular Probes), and Cy3-conjugated donkey anti-mouse IgG (Jackson Immunoresearch), for 90 min at 37C. Slides were washed in three changes of PBS for 10, 20, and 30 min, respectively, and mounted in antifade. Slides were examined on a Zeiss Axiophot epifluorescence microscope equipped with propidium iodide, fluorescein isothiocynate, and 4,6-diamidino-2-phenylindole filters and a cooled charge-coupled device camera. Three images (red, green, and blue) of the SC, MLH1 sites, and CREST locations, respectively, were captured by using Applied Imaging Cytovision 2.81 software (Applied Imaging Corporation, Santa Clara, Calif., USA). We analyzed 100 pachytene-stage cells per individual, scoring the number of MLH1 foci per autosomal bivalent and the total number of foci per autosomal complement from the prints of the captured images. The XY bivalent was excluded from MLH1 scoring, as it desynapses earlier than the autosomes. Data analysis The extent of variation in the recombination frequency among individuals was assessed by analysis of variance (ANOVA). Individual variation in the number of cells with gaps/splits was assessed by the Chi-square test. The relationships between the mean number of MLH1 foci
174
per cell and donor age, the percentage of cells with gaps or splits, and the number of nonexchange SCs were examined by Pearson correlation analysis. Two-way ANOVA was used to compare the mean number of MLH1 foci in the population of cells with and without gaps/splits.
Results Variation in recombination frequency Pachytene-stage cells from 11 individuals were analyzed. A representative pachytene spermatocyte is shown in Fig. 1. In all, 1100 pachytene-stage spermatocytes were analyzed, with an overall mean of 48.0±4.7 MLH1 (range: 21 to 65) foci per cell. This corresponds to a genetic length of 2400 cM, which is similar to the 2590 cM obtained from linkage data (Kong et al. 2002). The numbers of MLH1 foci per cell were compared among the 11 individuals, and significant variability both within and among individuals was observed (ANOVA, P
