Assessing the underlying pattern of human germline mutations: lessons from the factor IX gene.

Germline mutations cause or predispose to most disease. Hemophilia B is a useful model for studying the underlying pattern of recent germline mutations in humans because the observed pattern of mutation in factor IX more closely reflects the underlying pattern of mutation than the observed pattern for many other genes. In addition, it is possible to identify and correct for biases inherent in ascertaining only those mutations that cause hemophilia. Aspects of the pattern of germline mutation in the factor IX gene are becoming clear: 1) in the United States, two-thirds of mutations causing mild disease arose from three founders whereas almost all the mutations resulting in either moderate or severe disease arose independently, generally within the past 150 years; 2) direct estimates of the rates of mutation in humans indicate that transitions are more frequent than transversions, which in turn are more frequent than deletions and insertions; 3) transitions at CpG are elevated approximately 24-fold relative to transitions at non-CpG dinucleotides; 4) transversions at CpG are elevated approximately eightfold relative to transversions at non-CpG dinucleotides; 5) the sum total of the dinucleotide mutation rates produces a bias against G and C bases that would be sufficient to maintain the G+C content of the factor IX gene at its evolutionarily conserved level of 40%; and 6) the pattern of mutation is similar for Caucasians residing in the United States and for Asians residing in Asia. Two ideas emerge from this and from an analysis of the pattern of recent deleterious mutations compared with ancient neutral mutations that have been fixed during evolution into the factor IX gene. First, the bulk of germline mutations are likely to arise from endogenous processes rather than environmental mutagens. Second, the factor IX protein is composed mostly of two classes of amino acids: critical residues in which all single-base missense changes will disrupt protein function, and "spacer" residues in which the precise nature of the residue is unimportant but the peptide bond is necessary to keep the critical residues in register. More work is necessary to assess the veracity and generality of these ideas.     

Missense mutations and evolutionary conservation of amino acids: evidence that many of the amino acids in factor IX function as "spacer" elements.

We report 31 point mutations in the factor IX gene and explore the relationship between the level of evolutionary conservation of an amino acid and the probability of a mutation causing hemophilia B. From our total sample of 125 hemophiliacs and from those reported by others, we identify 95 independent missense mutations, 94 of which occur at amino acids that are evolutionarily conserved in the available mammalian factor IX sequences. The likelihood of a missense mutation causing hemophilia B depends on whether the residue is also conserved in the factor IX-related proteases: factor VII, factor X, and protein C. Most of the possible missense mutations in generically conserved residues (i.e., those conserved in factor IX and in all the related proteases) should cause disease. In contrast, missense mutations in factor IX-specific residues (i.e., those conserved in human, cow, dog, and mouse factor IX but not in the related proteases) are sixfold less likely to cause disease. Missense mutations at nonconserved residues are 33-fold less likely to cause disease. At least three models are compatible with these observations. A comparison of sequence alignments from four and nine species of factor IX and an examination of the missense mutations occurring at CpG residues suggest a model in which most residues fall on opposite ends of a spectrum. In about 40% of residues, virtually any missense mutation in a minority of the residues will cause disease, while virtually no missense mutations will cause disease in most of the remaining residues. Thus, many of the residues in factor IX are spacers; that is, the main chains are presumably necessary to keep other amino acid interactions in register, but the nature of the side chain is unimportant.     

The rates and patterns of deletions in the human factor IX gene.

Deletions are commonly observed in genes with either segments of highly homologous sequences or excessive gene length. However, in the factor IX gene and in most genes, deletions (of > or = 21 bp) are uncommon. We have analyzed DNA from 290 families with hemophilia B (203 independent mutations) and have found 12 deletions > 20 bp. Eleven of these are > 2 kb (range > 3-163 kb), and one is 1.1 kb. The junctions of the four deletions that are completely contained within the factor IX gene have been determined. A novel mutation occurred in patient HB128: the data suggest that a 26.8-kb deletion occurred between two segments of alternating purines and pyrimidines and that a 2.3-kb sense strand segment derived from the deleted region was inserted. For our sample of 203 independent mutations, we estimate the "baseline" rates of deletional mutation per base pair per generation as a function of size. The rate for large (> 2 kb) deletions is exceedingly low. For every mutational event in which a given base is at the junction of a large deletion, there are an estimated 58 microdeletions (< 20 bp) and 985 single-base substitutions at that base. Analysis of the nine reported deletion junctions in the factor IX gene literature reveals that (i) five are associated with inversions, orphan sequences, or sense strand insertions; (ii) four are simple deletions that display an excess of short direct repeats at their junctions; (iii) there is no dramatic clustering of junctions within the gene; and (iv) with the exception of alternating purines and pyrimidines, deletion junctions are not preferentially associated with repetitive DNA.     

Germline origins in the human F9 gene: frequent G:C→A:T mosaicism and increased mutations with advanced maternal age

The factor IX gene (F9) is an advantageous system for analyzing recent spontaneous germline mutation in humans. Herein, the male:female ratio of mutation ("r") in F9 have been estimated by Bayesian analysis from 59 germline origin families. The overall "r" in F9 was estimated at 3.75. The "r"s varied with the type of mutation. The "r"s ranged from 6.65 and 6.10 for transitions at CpG and A:T to G:C transitions at non-CpG dinucleotides, respectively, to 0.57 and 0.42 for microdeletions/microinsertions and large deletions (>1 kb), respectively. The "r" for the two subtypes of non-CpG transitions differed (6.10 for A:T to G:C vs 0.80 for G:C to A:T). Somatic mosaicism was detected in 11% of the 45 origin individuals for whom the causative mutation was visualized directly by genomic sequencing of leukocyte DNA (estimated sensitivity of approximately one part in 20). Four of the five defined somatic mosaics had G:C to A:T transitions at non-CpG dinucleotides, hinting that this mutation subtype may occur commonly early in embryogenesis. The age at conception was analyzed for 41 US Caucasian families in which the age of the origin parent and the year of conception for the first carrier/hemophiliac were available. No evidence for a paternal age effect was seen. However, an advanced maternal age effect was observed (P=0.03) and was particularly prominent for transversions (average of the 79th percentile when maternal age was normalized for the year of conception). This suggests that an increased maternal age results in a higher rate of transmitted mutation, whereas the increased number of mitotic replications associated with advanced paternal age has little, if any, effect on the rate of transmitted mutation.     

ATM-heterozygous germline mutations contribute to breast cancer-susceptibility

Approximately 0.5%-1% of the general population has been estimated to be heterozygous for a germline mutation in the ATM gene. Mutations in the ATM gene are responsible for the autosomal recessive disorder ataxia-telangiectasia (A-T) (MIM 208900). The finding that ATM-heterozygotes have an increased relative risk for breast cancer was supported by some studies but not confirmed by others. In view of this discrepancy, we examined the frequency of ATM germline mutations in a selected group of Dutch patients with breast cancer. We have analyzed ATM germline mutations in normal blood lymphocytes, using the protein-truncation test followed by genomic-sequence analysis. A high percentage of ATM germline mutations was demonstrated among patients with sporadic breast cancer. The 82 patients included in this study had developed breast cancer at age <45 and had survived >/=5 years (mean 15 years), and in 33 (40%) of the patients a contralateral breast tumor had been diagnosed. Among these patients we identified seven (8.5%) ATM germline mutations, of which five are distinct. One splice-site mutation (IVS10-6T-->G) was detected three times in our series. Four heterozygous carriers were patients with bilateral breast cancer. Our results indicate that the mutations identified in this study are "A-T disease-causing" mutations that might be associated with an increased risk of breast cancer in heterozygotes. We conclude that ATM heterozygotes have an approximately ninefold-increased risk of developing a type of breast cancer characterized by frequent bilateral occurrence, early age at onset, and long-term survival. The specific characteristics of our population of patients may explain why such a high frequency was not found in other series.     

Accumulation of Spontaneous Mutations in the Ciliate Tetrahymena thermophila

Knowledge of the rate and fitness effects of mutations is essential for understanding the process of evolution. Mutations are inherently difficult to study because they are rare and are frequently eliminated by natural selection. In the ciliate Tetrahymena thermophila, mutations can accumulate in the germline genome without being exposed to selection. We have conducted a mutation accumulation (MA) experiment in this species. Assuming that all mutations are deleterious and have the same effect, we estimate that the deleterious mutation rate per haploid germline genome per generation is U = 0.0047 (95% credible interval: 0.0015, 0.0125), and that germline mutations decrease fitness by s = 11% when expressed in a homozygous state (95% CI: 4.4%, 27%). We also estimate that deleterious mutations are partially recessive on average (h = 0.26; 95% CI: –0.022, 0.62) and that the rate of lethal mutations is <10% of the deleterious mutation rate. Comparisons between the observed evolutionary responses in the germline and somatic genomes and the results from individual-based simulations of MA suggest that the two genomes have similar mutational parameters. These are the first estimates of the deleterious mutation rate and fitness effects from the eukaryotic supergroup Chromalveolata and are within the range of those of other eukaryotes.     

A high incidence of clustered microsatellite mutations revealed by parent-offspring analysis in the African freshwater snail, <Emphasis Type="BoldItalic">Bulinus forskalii</Emphasis> (Gastropoda,Pulmonata)

Genotyping of 11 microsatellites in 432 offspring from 28 families of the hermaphroditic, freshwater snail Bulinus forskalii detected 10 de novo mutant alleles. This gave an estimated mutation rate of 1.1 × 10^–3 per locus per gamete per generation. There was a trend towards repeat length expansion and, unlike most studies, multi-step mutations predominated, suggesting that the microsatellite mutation process does not conform to a strict stepwise mutation model. Interestingly, the ten mutant alleles appear to have arisen from only six independent germline mutation events within the microsatellite array, with seven of them residing in three mutational clusters. Our results extend observations of clustered microsatellite mutations to another taxonomic group and type of mating system, self-fertile gastropods, and provide compelling evidence of premeiotic germline mutations, a phenomenon that could greatly impact upon our understanding of mutation dynamics but which has received little attention.     

Recurrent nonsense mutations at arginine residues cause severe hemophilia B in unrelated hemophiliacs

Summary Direct sequencing of the regions of the factor IX gene of likely functional significance was performed in four patients with severe hemophilia B. In two of the individuals, a transition at the dinucleotide CpG caused a nonsense mutation at arginine 333. In the other two individuals, a transition at CpG caused a nonsense mutation at arginine 29. Since these patients are all unrelated, as shown by differing alleles of the TaqI polymorphism in intron four or extensive nonoverlapping pedigrees, the mutations arose independently. In addition, the origin of one arginine 333 mutation in one family has been traced to the germline of the maternal grandfather. The frequent occurrence of mutations at arginine codons that contain the sequence CGN can be explained by the dramatic elevation of transitions at CpG. As a result, approximately one in four individuals with hemophilia B is expected to have a mutation at arginine and nonsense mutations at one of six arginine residues should be common causes of severe hemophilia.     

Structure and quantum chemical analysis of NAD+-dependent isocitrate dehydrogenase: hydride transfer and co-factor specificity

Mutations in human coagulation factor IX cause an X-linked bleeding disorder Hemophilia B, which can be classified as severe, moderately severe and mild based on the plasma levels of factor IX among affected individuals with respect to normal factor IX activity assayed in the patients' plasma (<1%, 2-5%, 6-30%, respectively). Recently, we identified hemophilia B to be a disease with mutations showing clinical variation and speculated that this phenotypic heterogeneity might be a replacement-specific property. Here, we have analyzed the differences in sequence and structural properties among identical mutations with varying phenotypes (IMVPs) by comparing with mutations with uniform phenotypes (MUPs), with recurring reports in Haemophilia B mutation database. Classification of mutations into IMVPs and MUPs has been done based on rigorous systematic evaluation of the clotting activity each mutation is associated with. IMVPs (n = 51) occur in less conserved mutant sites with more tolerated substitutions compared to MUPs (n = 100). A preponderance of CpG site mutations and Arg as the mutated residue in IMVPs compared to Cys in MUPs was observed. Hence, a CpG site substitution at less conserved Arg site might have an increased propensity of expressing variable phenotypes. The changes in intrinsic properties associated with the mutation are less drastic for IMVPs than for MUPs, though no significant differences were observed in structural properties. Based on this study and available literature we speculate that modifier genes at other loci, epigenetic interactions and environment may serve individually or cumulatively to bring about the clinical variation implicating hemophilia B to be deviation from classical Mendelian disorder with complete penetrance. We demonstrate that phenotypic heterogeneity appears to be site-specific also owing to the lesser conservation of the mutant site.     

Mutation rates in humans. II. Sporadic mutation-specific rates and rate of detrimental human mutations inferred from hemophilia B

We estimated the rates per base per generation of specific types of mutations, using our direct estimate of the overall mutation rate for hemophilia B and information on the mutations present in the United Kingdom's population as well as those reported year by year in the hemophilia B world database. These rates are as follows: transitions at CpG sites 9.7x10-8, other transitions 7.3x10-9, transversions at CpG sites 5.4x10-9, other transversions 6.9x10-9, and small deletions/insertions causing frameshifts 3.2x10-10. By taking into account the ratio of male to female mutation rates, the above figures were converted into rates appropriate for autosomal DNA-namely, 1.3x10-7, 9.9x10-9, 7.3x10-9, 9.4x10-9, 6.5x10-10, where the latter is the rate for all small deletion/insertion events. Mutation rates were also independently estimated from the sequence divergence observed in randomly chosen sequences from the human and chimpanzee X and Y chromosomes. These estimates were highly compatible with those obtained from hemophilia B and showed higher mutation rates in the male, but they showed no evidence for a significant excess of transitions at CpG sites in the spectrum of Y-sequence divergence relative to that of X-chromosome divergence. Our data suggest an overall mutation rate of 2.14x10-8 per base per generation, or 128 mutations per human zygote. Since the effective target for hemophilia B mutations is only 1.05% of the factor IX gene, the rate of detrimental mutations, per human zygote, suggested by the hemophilia data is approximately 1.3.     

A Novel Mutation of the Fibrillin Gene Causing Ectopia Lentis

Ectopia lentis (EL), a dominantly inherited connective tissue disorder, has been genetically linked to the fibrillin gene on chromosome 15 (FBN1) in earlier studies. Here, we report the first EL mutation in the FBN1 gene confirming that EL is caused by mutations of this gene. So far, several mutations in the FBN1 gene have been reported in patients with Marfan syndrome (MFS). EL and MFS are clinically related but distinct conditions with typical manifestations in the ocular and skeletal systems, the fundamental difference between them being the absence of cardiovascular involvement in EL. We report a point mutation, cosegregating with the disease in the described family, that displays EL over four generations. The mutation changes a conserved glutamic acid residue in an EGF-like motif, which is the major structural component of the fibrillin and is repeated throughout the polypeptide. In vitro mutagenetic studies have demonstrated the necessity of an analogous glutamic acid residue for calcium binding in an EGF-like repeat of human factor IX. This provides a possible explanation for the role of this mutation in the disease pathogenesis.     

Whole genome sequencing for quantifying germline mutation frequency in humans and model species: cautious optimism

Factors affecting the type and frequency of germline mutations in animals are of significant interest from health and toxicology perspectives. However, studies in this field have been limited by the use of markers with low detection power or uncertain relevance to phenotype. Whole genome sequencing (WGS) is now a potential option to directly determine germline mutation type and frequency in family groups at all loci simultaneously. Medical studies have already capitalized on WGS to identify novel mutations in human families for clinical purposes, such as identifying candidate genes contributing to inherited conditions. However, WGS has not yet been used in any studies of vertebrates that aim to quantify changes in germline mutation frequency as a result of environmental factors. WGS is a promising tool for detecting mutation induction, but it is currently limited by several technical challenges. Perhaps the most pressing issue is sequencing error rates that are currently high in comparison to the intergenerational mutation frequency. Different platforms and depths of coverage currently result in a range of 10-10(3) false positives for every true mutation. In addition, the cost of WGS is still relatively high, particularly when comparing mutation frequencies among treatment groups with even moderate sample sizes. Despite these challenges, WGS offers the potential for unprecedented insight into germline mutation processes. Refinement of available tools and emergence of new technologies may be able to provide the improved accuracy and reduced costs necessary to make WGS viable in germline mutation studies in the very near future. To streamline studies, researchers may use multiple family triads per treatment group and sequence a targeted (reduced) portion of each genome with high (20-40 ×) depth of coverage. We are optimistic about the application of WGS for quantifying germline mutations, but caution researchers regarding the resource-intensive nature of the work using existing technology.     

An Excess of G over C Nucleotides in Mutagenesis of Human Genetic Diseases

Strand asymmetries in DNA evolution, including indel and single nucleotide substitutions, were reported in prokaryotes. Recently, an excess of G>A over C>T substitutions in hemophilia B patients was recognized in our molecular diagnostic practices. Further analysis demonstrated biased point mutations between sense and antisense strands when unique changes in factor IX were counted. Similar mutation spectra of factor IX and the HGMD prompted us to speculate that the excess of G>A over C>T may be present in genes other than factor IX. Data from nine genes (each has ≥100 missense mutations) retrieved from HGMD, international factor IX database, and Dr. Sommer’s lab database in the City of Hope National Medical Center, Duarte, CA, USA were analyzed for their point mutation spectra. Similar to factor IX, all genes selected in this study have biased G>A over C>T unique mutations when nonsense mutations were excluded. The biased missense point mutations were recently convincingly documented by the statistic data of categorized missense mutation in HGMD. The consistence of the genetic observation and the genomic data from HGMD strongly indicate that biased point mutations, possibly a phenotypic selection, are more widespread than previously thought. The biased mutations have immediate clinical impact in molecular diagnostics.     

Why does the human factor IX gene have a G + C content of 40%?

The factor IX gene has a G + C content of approximately 40% in all mammalian species examined. In human factor IX, C----T and G----A transitions at the dinucleotide CpG are elevated at least 24-fold relative to other transitions. Can the G + C content be explained solely by this hot spot of mutation? Using our mathematical model, we show that the elevation of mutation at CpG cannot alone lower the G + C content below 45%. To search for other hot spots of mutation that might contribute to the reduction of G + C content, we assessed the relative rates of base substitution in our sample of 160 families with hemophilia B. Seventeen independent single-base substitutions are reported herein for a total of 96 independent point mutations in our sample. The following conclusions emerge from the analysis of our data and, where appropriate, the data of others: (1) Transversions at CpG are elevated an estimated 7.7-fold relative to other transversions. (2) The mutation rates at non-CpG dinucleotides are remarkably uniform; none of the observed rates are either more than twofold above the median for transitions or more than threefold above the median for transversions. (3) The pattern of recent mutation is compatible with the pattern during mammalian evolution that has maintained the G + C content of the factor IX gene at approximately 40%.     

Quantification of the paternal allele bias for new germline mutations in the retinoblastoma gene

New germline mutations in the human retinoblastoma gene are known to arise preferentially on paternally derived chromosomes, but the magnitude of that bias has not been measured. We evaluated 49 cases with a new germline mutation and found that in 40 cases (82%) the mutation arose on the paternally derived allele. We also evaluated 48 cases likely to have a somatic initial mutation; in this group the initial mutation arose on paternal or maternal chromosomes with approximately equal frequency. There was no statistically significant difference in the average age of fathers of children with new paternal germline mutations from the average age of fathers of children with new maternal germline mutations or somatic initial mutations. Combining the data with that from previous reports from other groups, the proportion of new germline mutations arising on a paternally derived allele is 85% (based on 72 cases; 95% confidence interval = 76–93%). This number can be useful in the genetic counseling of some families with retinoblastoma. Received: 18 December 1996 / Accepted: 30 April 1997     

The type of somatic mutation at APC in familial adenomatous polyposis is determined by the site of the germline mutation: a new facet to Knudson's 'two-hit' hypothesis.

APC is often cited as a prime example of a tumor suppressor gene. Truncating germline and somatic mutations (or, infrequently, allelic loss) occur in tumors in FAP (familial adenomatous polyposis). Most sporadic colorectal cancers also have two APC mutations. Clues from attenuated polyposis, missense germline variants with mild disease and the somatic mutation cluster region (codons 1,250-1,450) indicate, however, that APC mutations might not result in simple loss of protein function. We have found that FAP patients with germline APC mutations within a small region (codons 1,194-1,392 at most) mainly show allelic loss in their colorectal adenomas, in contrast to other FAP patients, whose 'second hits' tend to occur by truncating mutations in the mutation cluster region. Our results indicate that different APC mutations provide cells with different selective advantages, with mutations close to codon 1,300 providing the greatest advantage. Allelic loss is selected strongly in cells with one mutation near codon 1,300. A different germline-somatic APC mutation association exists in FAP desmoids. APC is not, therefore, a classical tumor suppressor. Our findings also indicate a new mechanism for disease severity: if a broader spectrum of mutations is selected in tumors, the somatic mutation rate is effectively higher and more tumors grow.     

Haemophilia B: database of point mutations and short additions and deletions--eighth edition.

The eighth edition of the haemophilia B database (http://www.umds.ac. uk/molgen/haemBdatabase.htm ) lists in an easily accessible form all known factor IX mutations due to small changes (base substitutions and short additions and/or deletions of <30 bp) identified in haemophilia B patients. The 1713 patient entries are ordered by the nucleotide number of their mutation. Where known, details are given on: factor IX activity, factor IX antigen in circulation, presence of inhibitor and origin of mutation. References to published mutations are given and the laboratories generating the data are indicated.     

Ionizing radiation and genetic risks. XI. The doubling dose estimates from the mid-1950s to the present and the conceptual change to the use of human data on spontaneous mutation rates and mouse data on induced mutation rates for doubling dose calculations

This paper provides an overview of the concept of doubling dose, changes in the database employed for calculating it over the past 30 years and recent advances in this area. The doubling dose is estimated as a ratio of the average rates of spontaneous and induced mutations in a defined set of genes. The reciprocal of the doubling dose is the relative mutation risk per unit dose and is one of the quantities used in estimating genetic risks of radiation exposures. Most of the doubling dose estimates used thus far have been based on mouse data on spontaneous and induced rates of mutations. Initially restricted to mutations in defined genes (with particular focus on the seven genes at which induced recessive mutations were studied in different laboratories), the doubling dose concept was subsequently expanded to include other endpoints of genetic damage. At least during the past 20 years, the magnitude of the doubling dose has remained unchanged at approximately 1 Gy for chronic low LET radiation exposures.One of the assumptions underlying the use of the doubling dose based on mouse data for predicting genetic risks in humans, namely, that the spontaneous rates of mutations in mouse and human genes are similar, is incorrect; this is because of the fact that, unlike in the mouse, the mutation rate in humans differs between the two sexes (being higher in males than in females) and increases with paternal age. Further, an additional source of uncertainty in spontaneous mutation rate estimates in mice has been uncovered. This is related to the non-inclusion of mutations which arise as germinal mosaics and which result in clusters of identical mutations in the following generation. In view of these reasons, it is suggested that a prudent way forward is to revert to the use of human data on spontaneous mutation rates and mouse data on induced mutation rates for doubling dose calculations as was first done in the 1972 BEIR report of the US National Academy of Sciences. The advantages of this procedure are the following: (i) estimates of spontaneous mutation rates in humans, which are usually presented as sex-averaged rates, automatically include sex differences and paternal age-effects; (ii) since human geneticists count all mutations that arise anew irrespective of whether they are part of a cluster or not, had clusters occurred, they would have been included in mutation rate calculations and (iii) one stays close to the aim of risk estimation, namely, estimation of the risk of genetic diseases in humans.On the basis of detailed analyses of the pertinent data, it is now estimated that the average spontaneous mutation rate of human genes (n=135 genes) is: (2.95+/-0.64)x10(-6) per gene and the average induced mutation rate of mouse genes (n=34) is: (0.36+/-0.10)x10(-5) per gene per Gy for chronic low LET radiation. The resultant doubling dose is (0.82+/-0.29) Gy. The standard error of the doubling dose estimate incorporates sampling variability across loci for estimates of spontaneous and induced mutation rates as well as variability in induced mutation rates in individual mouse experiments on radiation-induced mutations. We suggest the use of a rounded doubling dose value of 1 Gy for estimating genetic risks of radiation. Although this value is the same as that used previously, its conceptual basis is different and the present estimate is based on more extensive data than has so far been the case.