Mitochondrial portraits of human populations using median networks.

Analysis of variation in the hypervariable region of mitochondrial DNA (mtDNA) has emerged as an important tool for studying human evolution and migration. However, attempts to reconstruct optimal intraspecific mtDNA phylogenies frequently fail because parallel mutation events partly obscure the true evolutionary pathways. This makes it inadvisable to present a single phylogenetic tree at the expense of neglecting equally acceptable ones. As an alternative, we propose a novel network approach for portraying mtDNA relationships. For small sample sizes (< approximately 50), an unmodified median network contains all most parsimonious trees, displays graphically the full information content of the sequence data, and can easily be generated by hand. For larger sample sizes, we reduce the complexity of the network by identifying parallelisms. This reduction procedure is guided by a compatibility argument and an additional source of phylogenetic information: the frequencies of the mitochondrial haplotypes. As a spin-off, our approach can also assist in identifying sequencing errors, which manifest themselves in implausible network substructures. We illustrate the advantages of our approach with several examples from existing data sets.

Characterization of cells with high alkaline phosphatase activity derived from human bone and marrow: preliminary assessment of their osteogenicity.

Confluent cellular layers are reproducibly obtained (from 21 of 24 specimens) by outgrowth from composite pieces of human trabecular bone and marrow. The cells resemble fibroblasts in terms of morphology, esterase profile, and production of collagen type 1. However, the cells displayed some osteoblastlike features. Both the primary outgrowths and passaged cultures had high alkaline phosphatase activities (37 nmols min-1 X microgram DNA-1) in the range displayed by embryonic osteoblastlike cells. The cellular alkaline phosphatase activity, which showed similarity to the bone isoenzyme on kinetic criteria, was stimulated by 1,25-dihydroxyvitamin D3 but decreased by PTH (1-34). In addition, the cell preparations were shown to increase osteocalcin (bone Gla protein) production in response to 1,25-dihydroxyvitamin D3. The osteogenic potential of the bone and marrow-derived cells has been assessed in an in vivo diffusion chamber assay in which congenitally athymic (nude) mice were used as hosts. None of the 25 chambers examined showed evidence of osteogenesis, although the cells remained viable and fibroblastlike. The alkaline phosphatase activities decreased to less than 1% of the original, high in vitro values. The findings question the hypothesis that bone and marrow-derived cells are osteoblasts or osteoblastlike cells, rather than a mixture of cell lines of the bone and marrow stromal system.

Ancient DNA suggests a recent expansion of European cattle from a diverse wild progenitor species.

A total of 11 Bos primigenius and Bos taurus bones from archaeological sites between 500 and 12000 years old were examined for the presence of DNA. It was possible to amplify and sequence mitochondrial control region DNA extracted from seven of the 11 samples, including two Pleistocene B. primigenius samples. We compared the results with published data by constructing phylogenetic networks. The two B. primigenius samples clustered with the extant B. taurus samples in the networks. The similarity between B. primigenius and modern taurine cattle confirms that these should be considered members of a single species. The sequences obtained from the B. taurus specimens were either identical to the reference sequence for modern European cattle or closely related to it. They included two sequences not previously documented. The network analysis of the ancient data highlights the intermediary nature of the B. primigenius sequences between modern European and African B. taurus and the proximity of the ancient DNA B. taurus sequences to modern European B. taurus. Further analysis of the extant data in the light of the ancient DNA results suggests that a degree of Pleistocene diversity survives in the extant European Bos population that is mainly derived from a more recent population expansion.

Distribution of laminin, fibronectin, and interstitial collagen type III in soft tissue tumours.

The distributions of laminin, fibronectin, and interstitial collagen type III have been investigated in a series of 60 soft tissue tumours by immunochemistry. Positive laminin staining was seen in sites predicted by the distribution of ultrastructurally visible basal lamina. Pericellular laminin was present in all benign tumours of Schwann cell and smooth muscle origin examined, in the two malignant Schwannomas examined, and in six of 13 leiomyosarcomas. It was also evident around nests of cells in an alveolar soft part sarcoma and around malignant endothelial cells in an angiosarcoma. In fibroblastic and fibrohistiocytic tumours it was found only in blood vessel walls. The results of laminin staining led to revision of the original histopathological diagnosis in seven of the 60 cases studied. Fibronectin was abundant in the stroma of most neoplasms, both benign and malignant. It was also found in a distribution parallel to that of laminin. In some tumours this was clearly distinguishable from the distribution of interstitial collagen. Intracellular fibronectin was shown consistently only in mast cell granules. Its demonstration in synovial cells, fibroblasts, and histiocytes was more variable. Interstitial collagen type II had the most irregular distribution of the three proteins. It was as plentiful in tumours of smooth muscle origin as in tumours of fibroblastic origin, but was scanty in fibrous histiocytomas. Its distribution appeared similar to that of laminin and fibronectin in leiomyomas, but differed from these two proteins in Schwann cell tumours and other neoplasms. In one leiomyosarcoma fibronectin, laminin, and type III collagen appeared to be lost concomitantly from tumour cell peripheries.

Fibronectin and type III collagen in epithelial neoplasms of gastrointestinal tract and salivary gland.

The distribution of fibronectin (FN) and collagen type III (IIIC) have been compared in a series of epithelial neoplasms from the gastrointestinal tract and salivary gland. The difficulty of distinguishing between FN of epithelial and fibroblastic origin is emphasised and evidence is presented for the validity of this distinction. In carcinomas FN was sometimes, but not invariably, lost from epithelial cell surfaces. Both FN and IIIC were increased in reactive connective tissue stroma. It is concluded that loss of cell surface FN is unlikely to be a useful diagnostic marker for malignancy, but that the occurrence of this phenomenon in vivo as in vitro indicates that it is biologically significant.

Length heteroplasmy in the first hypervariable segment of the human mtDNA control region.

The first hypervariable segment of the human mtDNA control region contains a homopolymeric tract of cytosines between nt 16184 and 16193, interrupted at position 16189 by a thymine, according to the Cambridge reference sequence. A variant commonly found in population screening is a T-to-C transition at nt 16189, resulting in an uninterrupted homopolymeric tract. Direct sequencing of individuals with this variant produces a characteristic blurred sequence in nucleotides beyond the tract. Sequencing clones from these individuals revealed that this is caused by high levels of length heteroplasmy in the homopolymeric tract and low levels of length heteroplasmy in the four adenines following the tract. We have developed a rapid method involving densitometry of sequencing gels to quantify the relative proportions of different length variants present in an individual. We have used this to study the proportions of length variants in individuals from three twin pairs and two maternal lineages. While unrelated individuals usually have different proportions of length variants, all maternally related individuals studied have the same proportions, even if they are only distantly related. It is not obvious how identical heteroplasmic profiles are maintained in maternally related individuals, but some possible mechanisms are suggested.

Salt-soluble elastin from lathyritic chicks.

The isolation of a salt-soluble homogeneous elastin from the aortas of lathyritic chicks by chromatography on DEAE-cellulose and salt precipitation is described. These new techniques, as well as some previously published by other workers, were evaluated with the help of antiserum raised in sheep against insoluble chick elastin. The purified elastin was very basic and behaved in a predictable manner in coacervation studies. The protein migrated in sodium dodecyl sulphate-polyacrylamide gels as a single band moving slightly faster than pyruvate kinase (mol.wt. 57000).

Lethal osteogenesis imperfecta and a collagen gene deletion. Length polymorphism provides an alternative explanation.

A 300 base pair deletion near the 3'-end of the gene encoding Type II (cartilage) collagen has been implicated in the pathogenesis of perinatal lethal osteogenesis imperfecta. We have found similar deletions occurring at a high frequency in normal Asian Indian and West Indian populations generated by a length polymorphism just beyond the 3'-end of the gene. We suggest that this polymorphism provides an alternative explanation of the original results.

Segregation analysis of the structural genes of the major fibrillar collagens provides further evidence of molecular heterogeneity in type II Ehlers-Danlos syndrome.

Type II Ehlers-Danlos syndrome (EDS) is one of a group of disorders characterized by striking abnormalities of the soft connective tissues. The major fibrillar collagens (types I and III) found in these tissues have important stress-bearing functions and abnormal collagen could therefore account for the clinical features of this condition. We have used a number of restriction site dimorphisms, tightly linked to the structural genes of type I collagen (COL1A1 COL1A2) and type III collagen (COL3A1), to investigate the segregation of corresponding alleles in three pedigrees in which type II EDS was clearly inherited as a dominant trait. Discordant segregation of all three collagen genes was seen in a large pedigree that included 17 affected individuals with the typical phenotype of type II EDS. Thus mutations in neither type I nor type III collagen genes were responsible for the disease in this family. In a second small pedigree discordant segregation of the disease with both type I collagen loci was observed while the concordant segregation seen at COL3A1 could easily have arisen by chance (P = 0.5). The third pedigree was uninformative at all three collagen loci because of inability to discriminate between the parental alleles. These results suggest that there may be molecular heterogeneity of type II EDS since abnormalities of type I collagen have been described in other individuals phenotypically similar to those in our study.

Tandem competitive polymerase chain reaction (TC-PCR): a method for determining ratios of RNA and DNA templates.

A sensitive and accurate method for determining the ratios of RNA and DNA templates by polymerase chain reaction (PCR) is presented. A common competitor containing tandemly arranged internal standards differing from the target template by the presence of different restriction enzyme sites is coamplified with the target templates under identical conditions. Products from each template and internal standard are identified by the band pattern after digestion with the restriction enzyme. As the amount of the common competitor is kept constant for all target templates, the ratio of PCR products from the templates reflects their ratio in the reaction mix before amplification. The method was used to study the relative abundance of mRNA for the pro-alpha1 and pro-alpha2 chains of type I collagen and for estimating disturbances of normal ratio in the inherited bone disorder, osteogenesis imperfecta.

Variable levels of a heteroplasmic point mutation in individual hair roots.

During direct sequencing of the first hypervariable segment of the human mitochondrial control region, we identified one individual with a heteroplasmic point mutation at nt 16,256. We used primer extension to analyze the proportions of each mitochondrial haplotype in peripheral blood, buccal cells, and single hair roots from this individual and from eight members of his maternal lineage. Significant levels of heteroplasmy were found in only three individuals, and, in these cases, the proportions of each haplotype were similar in both blood and buccal cells. From the changes in mitochondrial haplotypes within mother-offspring pairs, we calculated that the most likely size of a mitochondrial bottleneck during development was 1-27 segregating units. However, highly variable levels of heteroplasmy were found in single hair roots, even among roots from the same individual. We analyzed a large number of hair roots from one individual and found that the proportion of one haplotype was within a range of 9% to > 99% in different roots. Roots originating from within a small patch of skin had haplotype proportions as variable as those from different areas of skin.

Isolation of salt-soluble elastin from ligamentum nuchae of copper-deficient calf.

This paper describes the isolation and amino acid analysis of un-cross-linked elastin obtained by neutral salt extraction from the ligamentum nuchae of a calf fed from birth to 9 months on a diet low in copper.

Site 73 in hypervariable region II of the human mitochondrial genome and the origin of European populations.

The majority of published human mitochondrial DNA sequence data are confined to hypervariable region I in the control region. By contrast, this paper focuses on a nucleotide site in hypervariable region II. Unlike most non-European populations whose mtDNA sequences have been studied in the literature, the British 'white Caucasian' population has a high level of variation at site 73 (following the site numbering by Anderson et al. 1981). This variation appears to have its origin largely in a mutation from guanine to adenine at that site with an estimated minimum age between 15,000 and 25,000 years. The data of Piercy et al. (1993) suggest that roughly half of the British 'white Caucasian' mitochondrial gene pool is descended from a common maternal ancestor who carried this mutation at site 73. This site also plays a central role in distinguishing the five major European mtDNA clusters identified in Richards et al. (1996). We suggest that the lineages carrying an A at site 73, together with some other lineages, may have their origins in a small founder population which expanded after the last glacial maximum about 20,000 years ago. We conclude that, in addition to region I sequences, site 73 is worth determining in studies of Caucasian populations.

Heteroplasmic point mutations in the human mtDNA control region.

As part of an investigation of the fixation mechanisms of mtDNA mutations in humans, we sequenced the first hypervariable segment of the control region in 180 twin pairs and found evidence of site heteroplasmy in 4 pairs. Significant levels of two mitochondrial haplotypes differing by a single point mutation were found in two MZ pairs, and within each pair, both members had similar levels of heteroplasmy. Two DZ pairs were found in which the predominant mitochondrial haplotype differed within the pair. We measured proportions of mitochondrial haplotypes within two twin pairs and their maternal relatives, using primer extension. In both maternal lineages, most family members were heteroplasmic, and the proportions of each genotype varied widely in different individuals. We used the changes in haplotype proportions within mother-offspring pairs to calculate the size range of potential bottlenecks in mitochondrial numbers occurring during development of the offspring. In most individuals, the most likely effective bottleneck sizes ranged from 3 to 20 segregating units, though in two individuals a small bottleneck was very unlikely and there was no upper limit on its possible size. We also used the data from this study, together with unpublished data from other populations, to estimate the frequency of site heteroplasmy in normal human populations. From this, we calculated that the rate of mutation and fixation in the first hypervariable segment of the human mtDNA control region is between 1.2 x 10(-6) and 2.7 x 10(-5) per site per generation. This range is in good agreement with published estimates calculated by other methods.

A new missense mutation of fibrillin in a patient with Marfan syndrome.

A patient with Marfan syndrome was shown to be heterozygous for a G to A transition at nucleotide 3952 of the FBNI gene. This would result in a cysteine to tyrosine substitution at amino acid 1223 in the fibrillin protein.

Mutation screening by a combination of biotin-SSCP and direct sequencing.

We have developed a mutation detection strategy that combines single strand conformational polymorphism (SSCP) analysis of one strand of a double-stranded amplification product with direct sequencing of the other. Using this strategy, which we find economical of both time and resources, we have identified a G to A transition, which substitutes a serine for glycine residue at position 862 in the major helix of the alpha 1 chain of Type I collagen. We use this mutation, which causes a lethal form of osteogenesis imperfecta, to illustrate the technique.

Phylogeography of mitochondrial DNA in western Europe.

For most of the past century, prehistorians have had to rely on the fossil and archaeological records in order to reconstruct the past. In the last few decades, this evidence has been substantially supplemented from classical human genetics. More recently, phylogenetic analyses of DNA sequences that incorporate geographical information have provided a high-resolution tool for the investigation of prehistoric demographic events, such as founder effects and population expansions. These events can be dated using a molecular clock when the mutation rate and founder haplotypes are known. We have previously applied such methods to sequence data from the mitochondrial DNA control region, to suggest that most extant mitochondrial sequences in western Europe have a local ancestry in the Early Upper Palaeolithic, with a smaller proportion arriving from the Near East in the Neolithic. Here, we describe a cladistic notation for mitochondrial variation and expand upon our earlier analysis to present a more detailed portrait of the European mitochondrial record.