Human Molecular Genetics, 2000, Vol. 9, No. 5 713-723
© 2000 Oxford University Press
Allele diversity and germline mutation at the insulin minisatellite
Department of Genetics, University of Leicester, University Road, Leicester LE1 7RH, UK
Received 13 December 1999; Revised and Accepted 20 January 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Previous analysis of germline mutation at highly unstable GC-rich minisatellites with continuous allele size distributions revealed similar meiotic recombinational mechanisms operating at all loci investigated. The insulin minisatellite has been studied intensively due to its associations with diabetes, polycystic ovary syndrome, obesity and birth size. Its bimodal allele size distribution in Caucasians suggests a much lower mutation rate and possible differences in the mutation process compared with highly unstable minisatellites. Mutation at the insulin minisatellite therefore was studied both indirectly from allele diversity surveys and directly by recovering de novo mutants from sperm DNA. Structural analysis of variant repeat distributions in 876 alleles identified 189 different alleles, almost all of which could be assigned to one of three very distinct lineages. Variation within a lineage was minor and due mainly to the gain or loss of one or a few repeat units. These events most probably arise by mitotic replication slippage at a frequency of perhaps 10–3 per gamete. Sperm DNA analysis revealed a second class of mutation occurring at a frequency of ~2 x 10–5 that involved highly complex intra- and inter-allelic rearrangements very similar to those seen at unstable minisatellites. These complex rearrangements were not seen in somatic DNA and are probably meiotic in origin. Minisatellite homozygosity did not reduce the frequency of these mutants in sperm. The insulin minisatellite therefore appears to evolve by two distinct processes: one involving slippage-like events and the second resulting in complex recombinational turnover of allele structure.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Minisatellites include some of the most polymorphic loci in the human genome and show variation both in the number of tandem repeats and in the interspersion patterns of variant repeats within alleles. Analysis of variant repeat distribution, usually by minisatellite variant repeat mapping by PCR (MVR-PCR) (1), allows allele structures to be compared from which indirect inferences can be drawn about the types of mutational mechanism that might generate population variability. Mutation processes can also be explored directly by characterizing de novo mutants detected either in families or in sperm.
Minisatellite mutation has been previously analysed directly in sperm at the highly variable human loci MS205, MS32, B6.7 and CEB1 (2–5). These GC-rich loci all have high mutation rates to new length alleles in the male germline (0.4–13% per sperm) and show extreme population variability, with continuous allele size distributions and heterozygosities of >98%. Similar mutation processes operate at all of these loci, resulting either in intra-allelic rearrangements or in conversion-like transfers of repeats between alleles. Both types of rearrangement often are complex. Conversions frequently show polarity towards one end of the repeat array, with repeat transfers usually occurring in-register between alleles, suggesting an allele pairing function in DNA flanking the unstable end of the repeat array. The recombinational nature and germline specificity of these mutations, together with the discovery that at least two of these loci are also very active in meiotic crossover (6; J. Buard and A.J. Jeffreys, unpublished data), strongly suggest that almost all mutations at these minisatellites arise at meiosis by aberrant processing of recombination initiation complexes. It is not known whether this recombinational instability is driven by allele heterozygosity. Mismatches in length and repeat array sequence between interacting alleles may serve to trigger the abortion of recombination initiation complexes and instead result in complex rearrangements without crossover. The rarity of homozygotes at unstable minisatellites has prevented this question from being addressed. In contrast to germline instability at hypervariable loci, somatic mutations occur far less frequently and involve simple intra-allelic deletions and duplications presumably arising through polymerase slippage or unequal sister chromatid exchange (7).
Although unstable minisatellites mutate by a similar recombinational process, it is unclear whether the same mechanism operates at all human minisatellites, including the majority of loci which display much lower allele length variabilities [typically 70–80% heterozygosity, implying germline mutation rates of perhaps 3–5 x 10–5 per gamete (8,9)]. These low variability loci often show discontinuous allele length distributions presumably resulting from genetic drift enhanced by low mutation rates, and possibly coupled with mutation processes exhibiting an inherent bias towards specific allele lengths and even selection for certain allele sizes. In Caucasian populations, bimodal allele size distributions are seen at the insulin minisatellite (10) and at D2S44 (11), whereas D19S20 (12) and MS51 (D11S97) (13) display trimodal size distributions (unpublished data). MVR-PCR at D2S44 (11), D19S20 and MS51 (unpublished data) demonstrated that each allele size class was derived largely from a different ancestral lineage. Population diversity analysis of the Harvey-ras1 (HRAS1) minisatellite revealed evidence for repeat transfers between alleles, suggesting that conversions do occur at this relatively stable locus, though at an unknown frequency (14).
The insulin GC-rich minisatellite has been the subject of intensive analysis due to its association with type 1 diabetes (15), type 2 diabetes (16), polycystic ovary syndrome (17), adult obesity (18) and infant birth size (19). These associations may result from the influence of the minisatellite on transcriptional regulation of the insulin gene (INS) and the gene for insulin-like growth factor II (IGF2) (20). It has a bimodal size distribution in Caucasians with small class I alleles (28–44 repeats) and large class III alleles (138–159 repeats) at frequencies of ~70 and ~30%, respectively (10). Class II alleles of intermediate size are rare in Caucasian populations. This bimodal size distribution, combined with strong linkage disequilibrium surrounding the minisatellite (21), suggests that mutation processes operating at this locus may be fundamentally different from those observed at more unstable minisatellites. To gain further insights into allele diversity and germline mutation processes at the insulin minisatellite, we have developed an MVR-PCR system capable of mapping the majority (>90%) of repeat variants at this locus and have used this system to characterize allele variability in Caucasians and the structural basis of mutation in sperm.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Allele mapping by MVR-PCR
Variant repeat distributions within insulin minisatellite alleles have been typed previously by sequence analysis of relatively few alleles (for a review see ref. 22). Eleven variant repeats named a–j based on the 14 bp consensus, ACAGG- GGTGTGGGG, were identified in Caucasian alleles (23,24). We developed an MVR-PCR system using seven MVR-specific primers to detect the six most common variant repeats (a–f) plus the h-type repeat due to its putative role as a transcriptional enhancer (25) (Table 1). To maximize PCR specificity, the register of the repeat unit was redefined, resulting in the gain of a 5' repeat and loss of a 3' repeat when compared with previous allele structures. There are inconsistencies in the literature over definitions of variant repeat unit sequences (23,25,26). We therefore used the original repeat unit definition as a guide to variant identification (see Table 2 in ref. 23) and named repeats with upper case letters to distinguish them from previous definitions (Table 1). Alleles initially were typed using the INS-MD primer which detects both D and F repeat variants, and not with INS-MF which detects only F repeats. Further analysis of 36 alleles with the discriminating INS-MF MVR primer demonstrated that all 166 variant repeats detected in these alleles by INS-MD had the sequence of F-type repeats.
|
Allele variability at the insulin minisatellite
As part of a parallel study to investigate the effects of insulin minisatellite allele structure on susceptibility to type 1 diabetes (unpublished data), we typed >50 000 repeats in 876 alleles from the parents of 219 Caucasian UK-based type 1 diabetic affected sib pair families. A full list of MVR-PCR allele codes is available at: http://www.le.ac.uk/genetics/ajj/insulin . Class I alleles are associated with predisposition to type 1 diabetes, whereas class III alleles are associated with protection against type 1 diabetes (10). This choice of sample population therefore resulted in some enrichment for class I alleles, with 704 (80%) and 164 (19%) alleles of class I and class III sizes, respectively. It is unlikely that this pre-selection resulted in a qualitative difference between alleles typed in these families compared with the general population since all parental alleles whether transmitted or not transmitted to affected offspring were analysed, and the relative risk to sibs of type 1 diabetic patients conferred by the insulin minisatellite is low (
s = 1.25) (27). MVR-PCR analysis revealed 189 different alleles over 39 size classes. Allele nomenclature reflects allele group, repeat number and a further discriminator (e.g. allele R42.1 is the first allele of 42 repeats in length identified in the group of rare alleles collectively called R). Only two alleles (R42.1 and R188.1) bore no structural similarity to any other allele (data not shown). R42.1 was also unusual in being the only allele of class I size which was linked to the class III-associated absence of an HphI site 23 bp upstream of the insulin gene translation start site (28). MVR codes of the remaining 99.8% of alleles could be assigned readily by eye to one of three lineages: class I, class IIIA and class IIIB (Fig. 1A). The three class II allele codes detected in this survey all aligned with class I or IIIB alleles, indicating that class II alleles do not comprise a distinct group, at least in Caucasians. Each lineage had a very different distribution of variant repeats, preventing alleles from different lineages from being meaningfully aligned. In contrast, variation within a lineage was minor, allowing allele structures to be readily aligned. By converting variant sites within each lineage of aligned alleles into a multi-state matrix and analysing variants by multi-dimensional scaling (data not shown), class I and class IIIA lineages could be divided further into two groups: class IC/ID and class IIIAi/IIIAii (Fig. 1A). These subgroups were also clearly detectable by visual inspection of MVR codes.
|
Allele diversity revealed by MVR-PCR was used to estimate germline mutation rates from Ewens’ distribution (29) (Table 2). Overall diversity suggested a mutation rate of ~10–3 per gamete, though the assumptions of an infinite allele model, mutation drift equilibrium and absence of selection are all dubious for this locus. Class III alleles tended to be more diverse than class I, though further subdivision into class IC, ID, IIIA and IIIB alleles showed no consistent relationship between allele length and estimated mutation rate, which varied substantially between allele classes (Table 2).
|
Patterns of variation in the insulin minisatellite
Patterns of variation among aligned alleles in a given lineage give clues about mutation processes (Fig. 1). Most variation was due to small deletions or duplications of repeats, mainly within uninterrupted arrays of 3–11 consecutive A-type repeats. For the smaller An arrays (3–5 repeats), most changes involved the gain or loss of a single A repeat. There were also less frequent instances of switches of repeat type between alleles without any apparent change in repeat copy number. Some alleles contained higher order repeat motifs, e.g. the ABABAB motif present at the 3' end of many class I alleles, suggesting that repeat turnover can occur at levels larger than a single repeat.
Although most differences between aligned alleles could be explained by small and simple rearrangements, there was also evidence for more complex turnover. Allele IIIB128.1 shows a deletion relative to other class IIIB alleles plus a 20 repeat motif at the site of deletion not present in any other allele typed (Fig. 1A). Similarly, the class II-sized allele IC60.1 aligned with class IC alleles but showed a complex triplication at the 3' end of the array, doubling allele size (Fig. 1B). The largest allele detected, R213.1, was identical to many class IIIB alleles at the 5' end but matched the class IIIA allele IIIA149.4 at the 3' end (Fig. 1B), with complex repeat duplications separating these regions of MVR identity. Upstream of the minisatellite, this allele is linked to a tyrosine hydroxylase microsatellite variant commonly associated with class IIIA haplotypes (21; unpublished data), suggesting that R213.1 was produced by an inter-allelic conversion-like event between a class IIIA and IIIB allele accompanied by a complex rearrangement at the conversion junction, as frequently occurs during mutation at unstable minisatellites. Further evidence for inter-allelic transfer comes from allele R44.1 which is similar to the class IC allele IC38.3 except for the insertion of an F-type repeat 14 repeats from the 3' end of the array (Fig. 1B). Most class IIIA alleles also contain an F-type repeat at this position, suggesting that R44.1 was produced by an in-register gene conversion-like transfer from an IIIA to an IC allele. However, major differences in repeat composition between lineages, e.g. E-type repeats present in class III but not class I alleles and H-type repeats restricted to IIIA alleles, suggest that such inter-allelic transfers can only occur at very low frequency.
Detection of de novo mutants
The low germline mutation rate at the insulin minisatellite estimated from the allele diversity survey precluded mutation detection either in families or by small pool PCR analysis of sperm DNA (3). Sperm mutants were therefore recovered by restriction digestion of sperm DNA with HinfI which cleaves outside the repeat array, followed by electrophoresis to recover size fractions enriched in mutants and depleted in progenitor molecules; mutants were then recovered from each of these fractions by single molecule PCR (7) (see Materials and Methods). Given the small size of the minisatellite repeat, we focused on mutants derived from the smaller class I alleles rather than class III alleles. Mutants gaining or losing three or more repeats could be recovered quantitatively, whereas ±1 repeat changes were almost completely lost in the progenitor allele fractions. Mutants selected to contain between 1 and 110 repeats were analysed in sperm DNA from three non-diabetic Caucasian donors who between them contained the three most common alleles seen in Caucasians. Donor 1 was a ID/IIIA heterozygote (ID40.2/IIIA152.2), donor 2 a ID/IIIB heterozygote (ID42.4/IIIB145.1) and donor 3 a ID/ID homozygote with both alleles identical in size and MVR code (ID44.1/ID44.1) (Fig. 1A). Mutants were also analysed in blood DNA of donor 1. Examples of size-validated mutants detected by this approach are shown in Figure 2. The structures of all mutants were determined subsequently by MVR-PCR. A full list of mutant allele MVR codes is available at: http://www.le.ac.uk/genetics/ajj/insulin . In class I/III heterozygotes, the majority of mutants were derived from the class I allele, with only 8% derived by deletion of the class III allele (data not shown).
|
Mutation rates of class I alleles were determined for changes of at least three repeats from surveys of 1.3–2.6 x 106 progenitor DNA molecules per sperm or blood sample (Table 3). Deletions occurred at a similar frequency of 4–8 x 10–6 in sperm and blood. In contrast, expansions were far more common in sperm (8–24 x 10–6 in sperm versus 1 x 10–6 in blood). Overall sperm mutation rates were similar in all three men tested, including the class I homozygote. Most sperm mutants were similar in size to the class I progenitor, with 86% of mutants involving length changes of 20 repeats or less. There were, however, some very large mutations. For example, a 15 repeat sperm mutant detected in donor 1 arose by simple deletion of >90% of the 152 repeat class III progenitor allele, whereas an expansion of the 44 repeat progenitor in donor 3 more than doubled allele length to 109 repeats. Class I and class III allele sizes therefore can be interchanged by a single mutational step. Major expansions were investigated further in sperm from the class I homozygous donor by screening 15 x 106 progenitor molecules for expansions of 40–200 repeats; of the additional mutants detected, none had gained more than 65 repeats, suggesting that there is a ceiling to allele expansion.
|
Mutation processes in the soma
Almost all mutants in blood arose from a simple intra-allelic deletion or duplication of a contiguous block of repeats located apparently at random in the progenitor allele (Fig. 3). The nomenclature of mutants reflects the donor and tissue of origin, repeat number and a further discriminator (e.g. mutant B1-38.2 is the second mutant of 38 repeats identified from blood DNA of donor 1). One mutant (B1-38.2) was seen three times, suggesting somatic mosaicism for this mutant. Only one mutant (B1-75.1) was more complex, with a triplication plus deletion within one triplicated unit. None of the class I blood mutants recovered from the I/IIIA heterozygote showed any evidence for gaining repeats from the class III allele.
|
Deletion processes in the germline
Deletion mutants in sperm were similar to those seen in blood, arising primarily by simple intra-allelic deletion. All sperm samples also showed evidence of mutational mosaicism (Table 3, Fig. 4), indicating that at least some of these deletions arose pre-meiotically. This was particularly noticeable for deletions of two repeats which were relatively common in both blood and sperm but whose frequency could not be estimated readily given partial loss of such mutants in progenitor allele fractions. Curiously, all three sperm samples were mosaic for the same two-repeat deletion in a CAC motif 12–14 repeats from the beginning of the allele, suggesting that this motif may serve as a mitotic deletion hotspot. In two of the men, about half of these deletions were associated with an additional rearrangement, namely an identical single repeat switch (AFACA
AFAFA) close to the 3' end of the deletion, raising the possibility that both classes of mutant (two-repeat deletion with or without distal repeat switch) may have originated from the same mitotic mutation event. One possible explanation is that these mutants arose by unequal crossover between sister chromatids misaligned by two repeats. Branch migration of the initial Holliday junction could then create heteroduplex DNA including a C/F mismatched repeat exactly at the site of where half of the mutants show a CF switch. If this heteroduplex site escapes mismatch repair before DNA replication, then these two repeat types will segregate into daughter germline stem cells which could then proliferate to give rise to the observed mosaicism for two different mutants of common origin. However, the reciprocal two-repeat gain mutant(s) predicted from sister chromatid exchange was detected only once (S1-42.1; data not shown), though such small gains are difficult to detect. An alternative explanation is therefore a two repeat slip within a single allele which could also create the necessary precursor heteroduplex for generating two mutant types.
|
Expansion processes in the germline
Only a minority of sperm length-gain mutants (36%) showed simple and perfect intra-allelic duplications or triplications similar to those seen in blood (data not shown). Most duplication units were relatively small (2–10 repeats), with the largest duplication seen involving 21 repeats. Three expansions (a duplication in donor 2, and a duplication and triplication in donor 3) showed evidence of mosaicism, suggesting a pre-meiotic component to gain mutation. All 45 remaining mutants showed more complex rearrangements (Fig. 5). All of these complex expansions were different, consistent with a meiotic origin, and were on average larger than the products of simple duplication. For the I/III heterozygous sperm donors, ~55% of these expansions appeared to involve rearrangements restricted to a single allele, though the possibility of transfers of short repeat segments between alleles could not be excluded. The largest expansion seen, in mutant S3-109.1, arose by an imperfect duplication of almost the entire length of the allele together with a reduplication of the 3' end of the allele. In some cases, these apparently intra-allelic rearrangements can be very complex, e.g. mutant S3-93.1 where the centre of the allele has been profoundly remodelled in a single but multi-step mutation event.
|
The remaining 45% of complex expansions showed evidence, in some cases definitive, for repeat transfer between alleles (Fig. 5). For example, mutant S1-51.1 shows the transfer of an EAAA motif from the class III to the class I allele which is devoid of E-type repeats. The transfer is in-register between alleles paired at their 3' ends, and appears to have been accompanied by an additional complex duplication of the junction between donor and recipient alleles generating two EAAAC motifs at the site of gene conversion. In some cases, e.g. mutant S2-51.1, the inserted segment does not contain repeat types unique to the donor allele but nevertheless does align perfectly in-register with the donor, again strongly suggesting in-register transfer from class III to class I alleles. With the exception of mutant S2-48.5, all such putative inter-allelic conversions involved alleles paired at the 3' end nearest the insulin gene.
The positioning of sperm rearrangements was analysed further by locating breakpoints along mutant alleles (Fig. 6). Breakpoints for simple duplications and complex rearrangements were distributed similarly along alleles, with little evidence for clustering other than avoidance of the 5' end of the repeat array which shows constancy over virtually all minisatellite alleles (Fig. 1). This 5' avoidance was less evident for deletion mutations.
|
Finally, sperm mutants recovered from the class I homozygous man showed a range of expansions similar to those seen in the class I/III heterozygotes, with 56% of length-gain mutants showing complex or very complex rearrangements. Although it is impossible to detect inter-allelic transfers in a homozygote, the frequency and spectrum of mutation in this man suggest that homozygosity has had no obvious influence on either the sperm mutation rate or the process.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Previous analyses of insulin minisatellite variability, including studies on association with diabetes, polycystic ovary syndrome, adult obesity and birth size, have focused almost exclusively on allele length variation and neighbouring single nucleotide polymorphisms (17–19,22). As a result, little was known about the level of structural and lineage diversity at this locus. We therefore developed a multi-state MVR system to determine allele structure. The level of variability in Caucasians so revealed is considerably higher than previously detected by allele length analysis due in part to size homoplasy (189 different alleles were defined by MVR code, compared with 39 alleles defined by size) and in part to the difficulty in resolving class III alleles of similar lengths without MVR analysis.
In Caucasians, virtually all alleles fall into three distinct lineages of alignable alleles which correspond to the bimodal allele length distribution and further divide the larger class III alleles into two very distinct lineages, IIIA and IIIB. Thus, whereas the insulin minisatellite shows substantial polymorphism, the underlying lineage diversity in Caucasians is severely restricted. A similar restricted set of lineages in Caucasians has been seen at mini- satellite MS205, compared with greater lineage diversity in particular in African populations, consistent with a recent African origin for modern humans (30). Almost no information exists on insulin minisatellite diversity in non-Caucasians, though the lack of similarity of the previously published allele HI-4 of African-American origin (31) to any Caucasian allele suggests that other populations may show different or additional allele lineages. More extensive MVR surveys are needed to determine whether restricted lineage diversity is specific to Caucasians, reflecting a bottleneck in the founding of Caucasian populations, or is common to non-African populations as predicted by the Out-of-Africa hypothesis, or instead is universal in all populations and was established prior to the diversification of all modern humans.
The ability to align allele structures allows, in principle, the subdivision of allele lineages and the construction of allele phylogenies. Visual inspection and multi-dimensional scaling analysis both indicated some degree of structuring within insulin minisatellite lineages. For example, 15 of the 126 class IIIA alleles share a similar additional 10–11 repeat motif (class IIIAii) (Fig. 1A), suggesting that these alleles form a single monophyletic sublineage. However, many of the more subtle character states frequently show evidence of homoplasy (e.g. class IC, ID and IIIA alleles all terminate in a polymorphism for two or three B-type repeats) and, in the absence of an explicit mutation model that can define the relative rates and reversibility of different types of mutation, allele alignments cannot be used to derive meaningful phylogenies.
We used allele diversity at the insulin minisatellite to estimate the germline mutation rate at ~9 x 10–4 per gamete, though this estimate varies substantially between lineages, from 1.9 x 10–4 for class ID alleles to 63 x 10–4 for class IIIA alleles (Table 2). Such estimates need to be treated with caution given the likelihood of selection acting at this locus plus the potential for departure from mutation-drift equilibrium, in particular if there has been a major bottleneck in Caucasians. Parallel changes, in particular in the short class I alleles, will also lead to underestimation of mutation rates through invalidation of the infinite allele model. Parallel changes do indeed occur, as shown by some sperm mutants having structures identical to alleles detected in the Caucasian survey (mutant S1-38.2 is identical to allele ID38.3, as are S2-40.1/ID40.1 and S3-42.1/ID42.1; data not shown). Nevertheless, class I mutation rates determined in sperm (0.14–0.3 x 10–4) (Table 3) are much lower than estimated from population data. This discrepancy is almost certainly due to the difficulty of detecting changes of 1–2 repeats that are likely to dominate mutation. Indeed, two-repeat changes were detected in all sperm samples tested, and in one case were present at a frequency possibly as high as 10–4 per sperm as a result of mutational mosaicism. It is also possible that mutation in the female germline may be a major contributor to allele diversity and that such mutants may be detectable in families by careful allele length analysis, in particular of class III alleles.
The major mode of germline mutation at the insulin minisatellite revealed from the allele diversity study involves insertions and deletions of 1–2 repeats, occurring preferentially within homogeneous stretches of A-type repeats. This stepwise mode of mutation will minimize allele length diversification and help to maintain the bimodal allele length distribution seen at the insulin minisatellite. This mutation process is reminiscent of microsatellite instability being facilitated by homogeneous repeat arrays (32–34), and suggests that polymerase slippage may play an important role in mutation at the insulin minisatellite. Polymerase slippage could possibly be promoted by the formation of unusual DNA conformations such as hairpin G-quartet structures, which in vitro form most readily within arrays of A-type repeats (35). Although such small events are likely to arise during the mitotic stages of germ cell development, a meiotic component to such instability cannot be excluded, in particular as intra-allelic meiotic mutation at the highly unstable minisatellite CEB1 has been shown to cluster within homogeneous repeat regions (5).
The allele diversity study also revealed evidence for a second much less frequent mode of mutation involving complex rearrangements, such as imperfect duplications, reshuffling of repeats and the transfer of information between alleles by a gene conversion-like process. The existence of such a low frequency mutation pathway was proved directly by mutation analysis of class I alleles in sperm. Given the fact that at least some of these mutation events are recombinational in nature, and, further, that none of these mutations show mosaicism in sperm, these complex rearrangements are most likely to be the result of aberrant meiotic recombination within the repeat array. A meiotic origin is supported further by the absence of highly complex mutations in the somatic (blood) DNA analysed. This complex mutation pathway is strikingly similar to that observed at highly unstable minisatellites with much higher mutation rates (2–5), and suggests that meiotic repeat instability is not restricted to the most unstable loci but may instead be a general feature of GC-rich minisatellites. There is growing evidence that minisatellite instability is modulated by DNA flanking the repeat array (6,36,37) that, in one case at least, provides a meiotic recombination hotspot that appears to drive repeat instability at one end of the array (6,37). Although polarity of mutation in the insulin minisatellite is weak, the in-register transfers seen between alleles paired at their 3' ends do point to a possible pairing function in DNA downstream of the minisatellite towards the insulin gene itself and therefore the possible existence of an, albeit weak, recombinational promoter of minisatellite mutation in this region.
Finally, the recombinational nature of minisatellite mutation raises the possibility that these complex turnover events are triggered in heterozygotes either by a difference between allele lengths or by heteroduplex DNA forming on strand invasion between alleles (38,39). The existence of insulin minisatellite alleles with a significant population frequency allowed sperm instability to be determined directly in a man homozygous for alleles of identical length and structure. Ironically, this man showed the highest sperm mutation rate of the three individuals tested, with no significant change in the spectrum and complexity of mutation. It therefore appears that allele length imbalance does not promote mutation and that complex repeat turnover can arise even between alleles that show no repeat differences when aligned at their 3' ends.
In conclusion, MVR-PCR analysis at the insulin minisatellite has provided new insights into lineages present in Caucasians, and has the potential to define further the relationship between minisatellite variation and diabetes risk (unpublished data). It has also demonstrated two distinct modes of germline instability, the major one involving simple stepwise mutations reminiscent of microsatellite instability, and the minor one (which nevertheless plays an important role in the generation of allele diversity) involving complex recombinational mechanisms as seen for other more unstable GC-rich minisatellites.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
PCR primers
Primers flanking the minisatellite were:
INS-1296, 5'-CTG CTG AGG ACT TGC TGC TTG-3';
MINS-A, 5'-TGC CGC CAC CCC CAG ATC A-3';
MINS-B, 5'-GGC CAG ACC TGT CCC TGC T-3'; and
MINS-C, 5'-GGG GCA AAT GTC TCC AGG AGA-3'.
Allele-specific primers were:
INS-23+, 5'-CAG AAG GAC AGT GAT CTG GGT-3'; and
INS-23–, 5'-CAG AAG GAC AGT GAT CTG GGA-3'.
MVR-PCR primers were as follows, with the 5' TAG extension indicated in lower case:
INS-MA, 5'-tca tgc gtc cat ggt ccg gaA CCC CTG TCC CCA C-3';
INS-MB, 5'-tca tgc gtc cat ggt ccg gaA CCC CTG TCC CCA GG-3';
INS-MC, 5'-tca tgc gtc cat ggt ccg gaA CCC CTG TCC CCA G-3';
INS-MD, 5'-tca tgc gtc cat ggt ccg gaA CCC CTG TCC CCG G-3';
INS-ME, 5'-tca tgc gtc cat ggt ccg gaA CCC CTA TCC CCA C-3';
INS-MF, 5'-tca tgc gtc cat ggt ccg gaA CCC CTG TCC CCG GG-3';
INS-MH, 5'-tca tgc gtc cat ggt ccg gaA CCC CTG TGC CCA C-3'.
The reverse MVR-PCR primer used was INS-MER, 5'-ctg ctg agg act tgc tgc ttg CAG GGG TGT GGG GAT-3', where the 5' INS-1296 sequence is indicated in lower case.
DNA source
DNA from the population of type 1 diabetic-affected sib pair families was derived from Epstein–Barr virus-transformed cell lines from the British Diabetic Association-Warren Repository (Porton Down, UK). Clinical criteria for diabetics were as defined (40).
Amplification of insulin minisatellite alleles
Allele sizes were determined initially by amplification of 20 ng of genomic DNA in 10 µl reactions using the buffer described previously (41) supplemented with 12 mM Tris base, 1 µg/ml carrier herring sperm DNA, 0.4 µM of each flanking primer, 0.07 U/µl Taq polymerase (Advanced Biotechnologies, Leatherland, UK) and 0.007 U/µl Pfu polymerase (Stratagene, Amsterdam, The Netherlands). Class I alleles were amplified using INS-1296 plus INS-23+ at 96°C for 40 s, 61°C for 30 s and 70°C for 3 min for 20 cycles on an MJ Tetrad thermal cycler (MJ Research, Waltham, USA). Class III alleles were amplified using INS-1296 plus INS-23– at 96°C for 40 s, 61°C for 30 s and 70°C for 3 min for 22 cycles. Samples were electrophoresed through a 40 cm long 1% SeaKem LE (FMC Bioproducts, Rockland, MD) agarose gel in 1x TBE buffer (89 mM Tris–borate pH 8.3, 2 mM EDTA) at 3 V/cm for 20 h and detected by Southern blot hybridization using a 32P-labelled probe generated by PCR amplification of a class I allele.
Alleles were separated prior to MVR-PCR typing. Alleles from –23 HphI heterozygotes were separated by allele-specific PCR as described above. Amplified class I and class III alleles were diluted 500- and 100-fold, respectively, in dilution buffer (5 mM Tris–HCl pH 7.5, 5 µg/ml herring sperm DNA) prior to MVR-PCR typing. Alleles from –23 HphI homozygotes were separated by amplification of alleles with primers INS-1296 plus INS-23+ or INS-23– as above for 32 cycles. Amplified alleles were electrophoresed through a 40 cm long 1% agarose gel, visualized by ethidium bromide staining using a Dark Reader (Clare Chemical Research, Ross on Wye, UK) to prevent UV damage and excised from the gel. Class I alleles were released from the gel by adding 50 µl of dilution buffer and freezing–thawing–vortexing three times. Each supernatant was diluted 1000-fold in dilution buffer. Class III alleles were gel purified using the Qiaex II gel purification kit (Qiagen, Dorking, UK) according to the manufacturer’s instructions and diluted 100-fold in dilution buffer.
MVR-PCR mapping of separated alleles
MVR-PCR was performed in 7 µl reactions using 1 µl (~0.1 pg) of purified allele DNA in the modified buffer described above with 0.035 U/µl Taq polymerase and 0.0035 U/µl Pfu polymerase plus one MVR primer together with 0.25 µM INS-1296 and TAG primers. MVR primer concentrations used were: INS-MA, 8 nM; INS-MB, 10 nM; INS-MC, 25 nM; INS-MD, 15 nM; INS-ME, 50 nM; INS-MF, 10 nM; and INS-MH, 3 nM. INS-MF was used to type only a subset of alleles. PCRs were cycled at 96°C for 40 s, 58°C for 30 s and 70°C for 2 min for eight cycles followed by 96°C for 40 s, 65°C for 30 s and 70°C for 2 min for 12 cycles. Amplified DNA was electrophoresed through a 40 cm long 1.5% LE agarose gel at 3 V/cm for 18 h and detected by Southern blot hybridization. The MVR system is very sensitive to input DNA concentration, so a test MVR of each allele using only INS-MA was performed initially and input DNA subsequently diluted according to signal strength and quality.
Determination of complete structures of class III alleles
MVR-PCR of class III alleles, as described above, could only accurately type the first ~100 repeats in the array. The remainder of the allele was typed by creating deletion amplicons covering the 3' end of the array. Reverse MVR was performed in 10 µl reactions with 0.4 µM primers INS-23– and INS-MER for one cycle at 96°C for 40 s, 60°C for 30 s and 70°C for 2 min followed by 26 cycles at 96°C for 40 s, 65°C for 30 s and 70°C for 2 min. INS-MER is a composite primer with the 3' sequence specific to E-type repeats and the 5' sequence identical to INS-1296. PCR generated a population of amplicons starting with the INS-1296 primer sequence and extending from each E-type repeat to the 3' flanking site. Amplicons were separated by electrophoresis through a 1% LE agarose gel in the presence of ethidium bromide, and the DNA of 1–2 appropriate amplicons (depending on allele size) was gel purified, diluted and MVR mapped as before. Full allele codes were assembled from overlapping codes generated from the whole allele and each deletion amplicon.
Detection of mutants
Mutants were recovered by a modification of procedures described (7). Blood and sperm DNAs were prepared as described previously under conditions designed to minimize the risk of contamination (3,41). A 70 µg aliquot of each DNA was digested with HinfI, which cleaves 255 bp 5' and 144 bp 3' to the insulin minisatellite, and electrophoresed through a 40 cm long 1% LE agarose gel in 1x TBE buffer at 4 V/cm for 15 h in the presence of ethidium bromide. Thirty gel slices spanning 0.5–2 kb and covering the position of the class I progenitor allele (1.0 kb) were excised on a Dark Reader, and DNA from each slice was recovered by electroelution. Fraction size ranges were determined by agarose gel electrophoresis of an aliquot of each fraction. The purity of each fraction was estimated by PCR amplification of 1/40 of each fraction in the modified PCR buffer described above, using 0.2 µM flanking primers MINS-A and MINS-B, at 96°C for 20 s, 58°C for 30 s and 70°C for 3 min for 17 cycles, followed by Southern blot hybridization of PCR products and comparison of hybridization intensity with corresponding PCR products from a dilution series of unfractionated genomic DNA. The progenitor allele was limited to 3–5 adjacent fractions, and other fractions were depleted at least 100-fold in progenitor allele. The number of amplifiable progenitor molecules was determined by pooling progenitor fractions and subjecting limiting dilutions of this pool to Poisson analysis (90 PCRs each containing 0.5–1.5 molecules per PCR, amplified as above for 28 cycles), and gave yields of 50–60% after gel fractionation. Mutants in non-progenitor fractions were detected by PCR amplification, as above for 24 cycles, of multiple aliquots of each fraction containing at most 200 progenitor molecules per PCR. Mutants were purified by electrophoresis of PCR products through a 40 cm long 1% agarose gel; 3–5 gel slices spanning the expected location of the mutant were excised, DNA recovered by freeze–thaw, re-amplified as above for 24 cycles using the nested flanking primers INS-1296 and MINS-C and amplified mutants purified by electrophoresis, with visualization with ethidium bromide and excision from the gel. Purified mutants were MVR mapped as above.
The use of multiple size fractions allowed genuine mutants to be distinguished from PCR artefacts. These fractions typically covered a four-repeat range for fractions close to the progenitor, and were up to 15 repeats wide for the largest fractions. The survey of one blood and three sperm DNAs yielded 158 mutants plus 37 artefacts (20 gains and 17 deletions) of hybridization intensity similar to authentic mutants; these artefacts were distributed evenly across all four samples. In almost all cases, the artefacts deviated substantially in size from that predicted from the fraction, in most cases lying 5–35 repeats outside the permissible size window. All artefacts resulted from a simple, perfect duplication or deletion of a contiguous block of repeats in one allele, as shown by MVR mapping (data not shown), and presumably arose by a slippage event during the first cycle of PCR. PCR artefacts are therefore structurally distinct from the majority of sperm mutants showing complex rearrangements.
|
ACKNOWLEDGEMENTS |
|---|
We are grateful to John Todd (Cambridge Institute for Medical Research, Cambridge, UK) and Sarah Nutland for kindly supplying DNA from type 1 diabetic families, to Yuri Dubrova for multi-dimensional scaling analysis, to semen donors, and to Jérôme Buard and colleagues for helpful discussions. Sample collection was supported by the British Diabetic Association and Wellcome Trust. The work of A.J.J. was supported in part by an International Research Scholars Award from the Howard Hughes Medical Institute and in part by grants from the Wellcome Trust, Medical Research Council and Royal Society.
|
FOOTNOTES |
|---|
+ To whom correspondence should be addressed. Tel: +44 116 252 3435; Fax: +44 116 252 3378; Email: ajj@le.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Jeffreys, A.J., MacLeod, A., Tamaki, K., Neil, D.L. and Monckton, D.G. (1991) Minisatellite repeat coding as a digital approach to DNA typing. Nature, 354, 204–209.[Medline]
2 May, C.A., Jeffreys, A.J. and Armour, J.A.L. (1996) Mutation rate heterogeneity and the generation of allele diversity at the human minisatellite MS205 (D16S309). Hum. Mol. Genet., 5, 1823–1833.
3 Jeffreys, A.J., Tamaki, K., MacLeod, A., Monckton, D.G., Neil, D.L. and Armour, J.A.L. (1994) Complex gene conversion events in germline mutation at human minisatellites. Nature Genet., 6, 136–145.[Web of Science][Medline]
4 Tamaki, K., May, C.A., Dubrova, Y.E. and Jeffreys, A.J. (1999) Extremely complex repeat shuffling during germline mutation at human minisatellite B6.7. Hum. Mol. Genet., 8, 879–888.
5 Buard, J., Bourdet, A., Yardley, J., Dubrova, Y.E. and Jeffreys, A.J. (1998) Influences of array size and homogeneity on minisatellite mutation. EMBO J., 17, 3495–3502.[Web of Science][Medline]
6 Jeffreys, A.J., Neil, D.L. and Neumann, R. (1998) Repeat instability at human minisatellites arising from meiotic recombination. EMBO J., 17, 4147–4157.[Web of Science][Medline]
7 Jeffreys, A.J. and Neumann, R. (1997) Somatic mutation processes at a human minisatellite. Hum. Mol. Genet., 6, 129–136.
8 Nakamura, Y., Leppert, M., O’Connell, P., Wolff, R., Holm, T., Culver, M., Martin, C., Fujimoto, E., Hoff, M., Kumlin, E. and White, R. (1987) Variable number of tandem repeat (VNTR) markers for human gene mapping. Science, 235, 1616–1622.
9 Armour, J.A.L., Povey, S., Jeremiah, S. and Jeffreys, A.J. (1990) Systematic cloning of human minisatellites from ordered array charomid libraries. Genomics, 8, 501–512.[Web of Science][Medline]
10 Bell, G.I., Horita, S. and Karam, J.H. (1984) A polymorphic locus near the human insulin gene is associated with insulin-dependent diabetes mellitus. Diabetes, 33, 176–183.[Abstract]
11 Holmlund, G. and Lindblom, B. (1998) Different ancestor alleles: a reason for the bimodal fragment size distribution in the minisatellite D2S44 (YNH24). Eur. J. Hum. Genet., 6, 597–602.[Web of Science][Medline]
12 Nakamura, Y., Cowen, J., Krapcho, K., O’Connell, P., Leppert, M., Lathrop, G.M., Lalouel, J.M. and White, R. (1988) Isolation and mapping of a polymorphic DNA sequence pJCZ3.1 on chromosome 19 [D19S20]. Nucleic Acids Res., 16, 1229.
13 Armour, J.A.L., Wong, Z., Wilson, V., Royle, N.J. and Jeffreys, A.J. (1989) Sequences flanking the repeat arrays of human minisatellites: association with tandem and dispersed repeat elements. Nucleic Acids Res., 17, 4925–4935.
14 Ding, S.F., Larson, G.P., Foldenauer, K., Zhang, G.X. and Krontiris, T.G. (1999) Distinct mutation patterns of breast cancer-associated alleles of the HRAS1 minisatellite locus. Hum. Mol. Genet., 8, 515–521.
15 Bennett, S.T., Wilson, A.J., Esposito, L., Bouzekri, N., Undlien, D.E., Cucca, F., Nistico, L., Buzzetti, R., the IMDIAB Group, Bosi, E. et al. (1997) Insulin VNTR allele-specific effect in type 1 diabetes depends on identity of untransmitted paternal allele. Nature Genet., 17, 350–352.[Web of Science][Medline]
16 Ong, K.K.L., Phillips, D.I., Fall, C., Poulton, J., Bennett, S.T., Golding, J., Todd, J.A. and Dunger, D.B. (1999) The insulin gene VNTR, type 2 diabetes and birth weight. Nature Genet., 21, 262–263.[Web of Science][Medline]
17 Waterworth, D.M., Bennett, S.T., Gharani, N., McCarthy, M.I., Hague, S., Batty, S., Conway, G.S., White, D., Todd, J.A., Franks, S. and Williamson, R. (1997) Linkage and association of insulin gene VNTR regulatory polymorphism with polycystic ovary syndrome. Lancet, 349, 986–990.[Web of Science][Medline]
18 O’Dell, S.D., Bujac, S.R., Miller, G.J. and Day, I.N. (1999) Associations of IGF2 ApaI RFLP and INS VNTR class I allele size with obesity. Eur. J. Hum. Genet., 7, 821–827.[Web of Science][Medline]
19 Dunger, D.B., Ong, K.K.L., Huxtable, S.J., Sherriff, A., Woods, K.A., Ahmed, M.L., Golding, J., Pembrey, M.E., Ring, S., the ALSPAC Study Team et al. (1998) Association of the INS VNTR with size at birth. Nature Genet., 19, 98–100.[Web of Science][Medline]
20 Paquette, J., Giannoukakis, N., Polychronakos, C., Vafiadis, P. and Deal, C. (1998) The INS 5' variable number of tandem repeats is associated with IGF2 expression in humans. J. Biol. Chem., 273, 14158–14164.
21 Bennett, S.T., Lucassen, A.M., Gough, S.C.L., Powell, E.E., Undlien, D.E., Pritchard, L.E., Merriman, M.E., Kawaguchi, Y., Dronsfield, M.J., Pociot, F. et al. (1995) Susceptibility to human type 1 diabetes at IDDM2 is determined by tandem repeat variation at the insulin gene minisatellite locus. Nature Genet., 9, 284–292.[Web of Science][Medline]
22 Bennett, S.T. and Todd, J.A. (1996) Human type 1 diabetes and the insulin gene: principles of mapping polygenes. Annu. Rev. Genet., 30, 343–370.[Web of Science][Medline]
23 Bell, G.I., Selby, M.J. and Rutter, W.J. (1982) The highly polymorphic region near the human insulin gene is composed of simple tandemly repeating sequences. Nature, 295, 31–35.[Medline]
24 Owerbach, D. and Aagaard, L. (1984) Analysis of a 1963-bp polymorphic region flanking the human insulin gene. Gene, 32, 475–479.[Web of Science][Medline]
25 Kennedy, G.C., German, M.S. and Rutter, W.J. (1995) The minisatellite in the diabetes susceptibility locus IDDM2 regulates insulin transcription. Nature Genet., 9, 293–298.[Web of Science][Medline]
26 Owerbach, D. and Gabbay, K.H. (1996) The search for IDDM susceptibility genes: the next generation. Diabetes, 45, 544–551.[Abstract]
27 Todd, J.A. and Farrall, M. (1996) Panning for gold: genome-wide scanning for linkage in type 1 diabetes. Hum. Mol. Genet., 5, 1443–1448.[Abstract]
28 Ullrich, A., Dull, T.J., Gray, A., Brosius, J. and Sures, I. (1980) Genetic variation in the human insulin gene. Science, 209, 612–615.
29 Ewens, W.J. (1972) The sampling theory of selectively neutral alleles. Theor. Pop. Biol., 3, 87–112.[Web of Science][Medline]
30 Armour, J.A.L., Anttinen, T., May, C.A., Vega, E.E., Sajantila, A., Kidd, J.R., Kidd, K.K., Bertranpetit, J., Paabo, S. and Jeffreys, A.J. (1996) Minisatellite diversity supports a recent African origin for modern humans. Nature Genet., 13, 154–160.[Web of Science][Medline]
31 Rotwein, P., Yokoyama, S., Didier, D.K. and Chirgwin, J.M. (1986) Genetic analysis of the hypervariable region flanking the human insulin gene. Am. J. Hum. Genet., 39, 291–299.[Web of Science][Medline]
32 Weber, J.L. (1990) Informativeness of human (dC-dA)n(dG-dT)n polymorphisms. Genomics, 7, 524–530.[Web of Science][Medline]
33 Chung, M.Y., Ranum, L.P.W., Duvick, L.A., Servadio, A., Zoghbi, H.Y. and Orr, H.T. (1993) Evidence for a mechanism predisposing to intergenerational CAG repeat instability in spinocerebellar ataxia type I. Nature Genet., 5, 254–258.[Web of Science][Medline]
34 Garza, J.C., Slatkin, M. and Freimer, N.B. (1995) Microsatellite allele frequencies in humans and chimpanzees, with implications for constraints on allele size. Mol. Biol. Evol., 12, 594–603.[Abstract]
35 Catasti, P., Chen, X., Moyzis, R.K., Bradbury, E.M. and Gupta, G. (1996) Structure–function correlations of the insulin-linked polymorphic region. J. Mol. Biol., 264, 534–545.[Web of Science][Medline]
36 Monckton, D.G., Neumann, R., Guram, T., Fretwell, N., Tamaki, K., MacLeod, A. and Jeffreys, A.J. (1994) Minisatellite mutation rate variation associated with a flanking DNA sequence polymorphism. Nature Genet., 8, 162–170.[Web of Science][Medline]
37 Jeffreys, A.J., Murray, J. and Neumann, R. (1998) High-resolution mapping of crossovers in human sperm defines a minisatellite-associated recombination hotspot. Mol. Cell, 2, 267–273.[Web of Science][Medline]
38 Nassif, N., Penney, J., Pal, S., Engels, W.R. and Gloor, G.B. (1994) Efficient copying of nonhomologous sequences from ectopic sites via P-element-induced gap repair. Mol. Cell. Biol., 14, 1613–1625.
39 Paques, F., Leung, W.Y. and Haber, J.E. (1998) Expansions and contractions in a tandem repeat induced by double-strand break repair. Mol. Cell. Biol., 18, 2045–2054.
40 Bain, S.C., Prins, J.B., Hearne, C.M., Rodrigues, N.R., Rowe, B.R., Pritchard, L.E., Ritchie, R.J., Hall, J.R.S., Undlien, D.E., Ronningen, K.S. et al. (1992) Insulin gene region-encoded susceptibility to type 1 diabetes is not restricted to HLA-DR4-positive individuals. Nature Genet., 2, 212–215.[Web of Science][Medline]
41 Jeffreys, A.J., Neumann, R. and Wilson, V. (1990) Repeat unit sequence variation in minisatellites: a novel source of DNA polymorphism for studying variation and mutation by single molecule analysis. Cell, 60, 473–485.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Mead, T. E. F. Webb, T. A. Campbell, J. Beck, J. M. Linehan, S. Rutherfoord, S. Joiner, J. D. F. Wadsworth, J. Heckmann, S. Wroe, et al. Inherited prion disease with 5-OPRI: Phenotype modification by repeat length and codon 129 Neurology, August 21, 2007; 69(8): 730 - 738. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Awata, E. Kawasaki, H. Ikegami, T. Kobayashi, T. Maruyama, K. Nakanishi, A. Shimada, H. Iizuka, S. Kurihara, M. Osaki, et al. Insulin Gene/IDDM2 Locus in Japanese Type 1 Diabetes: Contribution of Class I Alleles and Influence of Class I Subdivision in Susceptibility to Type 1 Diabetes J. Clin. Endocrinol. Metab., May 1, 2007; 92(5): 1791 - 1795. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kralovicova, T. R. Gaunt, S. Rodriguez, P. J. Wood, I. N.M. Day, and I. Vorechovsky Variants in the Human Insulin Gene That Affect Pre-mRNA Splicing: Is -23HphI a Functional Single Nucleotide Polymorphism at IDDM2? Diabetes, January 1, 2006; 55(1): 260 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Jeffreys and R. Neumann Factors influencing recombination frequency and distribution in a human meiotic crossover hotspot Hum. Mol. Genet., August 1, 2005; 14(15): 2277 - 2287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D.H. Stead, M. E. Hurles, and A. J. Jeffreys Global Haplotype Diversity in the Human Insulin Gene Region Genome Res., September 1, 2003; 13(9): 2101 - 2111. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Vafiadis, H. Ounissi-Benkalha, M. Palumbo, R. Grabs, M. Rousseau, C. G. Goodyer, and C. Polychronakos Class III Alleles of the Variable Number of Tandem Repeat Insulin Polymorphism Associated with Silencing of Thymic Insulin Predispose to Type 1 Diabetes J. Clin. Endocrinol. Metab., August 1, 2001; 86(8): 3705 - 3710. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D.H. Stead, J. Buard, J. A. Todd, and A. J. Jeffreys Influence of allele lineage on the role of the insulin minisatellite in susceptibility to type 1 diabetes Hum. Mol. Genet., December 1, 2000; 9(20): 2929 - 2935. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. G.M. v. Overveld, R. J.F.L. Lemmers, G. Deidda, L. Sandkuijl, G. W. Padberg, R. R. Frants, and S. M. van der Maarel Interchromosomal repeat array interactions between chromosomes 4 and 10: a model for subtelomeric plasticity Hum. Mol. Genet., November 1, 2000; 9(19): 2879 - 2884. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||