Analysis of 8000-year-old brain tissue from the Windover site • 67 



A B C D 



Figure 8. Identification of human mtDNA sequences in DNA troni 8UUU-year-old adult male brain. Total DNA 

 isolated and purified as described in text. An aliquot was digested with EcoRI and electrophoresed on a 1 % agarose gel 

 along with an undigested aliquot and samples of authentic human KB cell mtDNA treated similarly. After staining and 

 photography, gel was blotted and hybridized to a human mtDNA-specific probe (Chang and Clayton 1985), washed and 

 autoradiographed. Central lanes containing enzymes, buffers, and carrier tRNA used in extraction, and enzyme 

 digestion showed no hybridization to probe. Lane A, undigested KB cell mtDNA; lane B, EcoRI-digested KB cell 

 mtDNA; lane C. undigested ancient DNA; lane D. EcoRl-digested ancient DNA. OC, L. and EcoRl refer to positions 

 of open- (nicked) circular, linear, and 8050-base pair (bp) EcoRI (D-loop-containing) fragment of KB cell mtDNA, 

 respectively (Anderson etal., 1981; Houcket al., 1979). Left: Ethidium-stainedgel. /?/g/ir.Autoradiograph of blotted 

 gel. Lanes A and B were exposed for 3 days, lanes C and D for 7 days. 



Resistance to enzyme digestion is an intrinsic property of the 

 old DNA because a bacterial plasmid DNA mixed with this 

 DNA sample did digest to completion under the same condi- 

 tions. The inability to completely digest the DNA may be due 

 to base modification leading to a loss of restriction endo- 

 nuclease site recognition. 



The old DNA lacked supercoiled, covalently closed cir- 

 cles (Figure 8, right). The reason for the absence of super- 

 coiled molecules has not been investigated further at present, 

 but many spontaneous processes can lead to single-strand 

 nicks in DNA, converting covalently closed molecules to 

 open-circular forms (Vinograd et al. 1965). Multiple single- 

 stranded nicks or damage resulting in a double-stranded scis- 

 sion would lead to linear, full-length molecules, as was seen. 



High molecular weight and a surprisingly high fraction of 

 intact open-circular mtDNA was observed in these DNA 

 samples (Figure 8). Both of these observations may be due to 

 DNA damage caused by depurination leading to intcrstrand- 

 crosslinking of DNA (Goftln et al. 1984). Crosslinking 

 would raise the apparent double-stranded molecular weight 

 of the DNA, greatly increase the lifetime of circular DNA 

 forms, interfere with restriction digestions, and perhaps also 

 interfere with DNA hybridization experiments by preventing 

 strand separation. This type of damage has been shown to be 

 enhanced in aqueous solution (Goffin et al. 1984). 



The presence and condition of nuclear DNA was also in- 

 vestigated. Restriction endonuclease-digested DNA was 



Zagreb Paleopathology Symp. J9H8 



probed with radiolabeled nuclear DNA and RNA sequences 

 present as multiple copies in the human genome (an Alu 

 sequence, a large and small ribosomal RNA) (Houck et al. 

 1979; Long and Dawid 1980). None of these probes 

 hybridized to restriction fragments of a defined size (data not 

 shown). If nuclear DNA was damaged in a similar manner as 

 mtDNA and largely resistant to restriction endonuclease di- 

 gestion, discrete restriction fragment bands of multicopy or 

 single copy genes would not occur in hybridization experi- 

 ments. In contrast, mtDNA occurs as a small 16 kb circular 

 molecule and migrates as a discrete species without depend- 

 ing upon recognition of specific undamaged sequences by 

 restriction endonuclease; therefore, it can be detected in hy- 

 bridization experiments. Alkaline cleavage of the 8000-year- 

 old DNA also reveals a significant amount of DNA damage 

 (data not shown). It is estimated that these alkali-sensitive 

 sites, many of which may represent apurinic nucleotides, are 

 present at the 1% level. 



An attempt was made to clone and compare DNA se- 

 quences from several brains with nucleotide sequences of 

 genes or other DNA sequences already known. Thus far, old 

 brain nucleotide sequences have not corresponded to any 

 known DNA sequence. A small library of DNA fragments 

 was constructed using a partial Alu I digest and an M13 

 cloning vector. Approximately 1000 clones containing small 

 inserts (50-1000 bp) were isolated and 90 were screened for 

 homology to human mtDNA, human Alu rep)eat sequences. 



