66 • W.W. Hauswirth, CD. Dickel, G.H. Doran, P.J. Laipis, and D.N. Dickel 



B 



\ 



Figure 1. A. Scanning electron micro- 

 graph of cerebral cortex. Remains of 

 neuropil made up of intermingling tu- 

 bular processes can be seen. Occasion- 

 al structures representing neuronal re- 

 mains can be identified at lower left. 

 Bar = 5 ^.m. B. Electron micrograph of 

 pyramidal neuron of cerebral cortex. 

 Cell shape can be inferred from in- 

 cluded lipofuscin granules (Luna 

 1968). These electron dense granules 

 show characteristic peripheral vacuole 

 (inset). Remainder of cell shows finely 

 granular, moderately electron dense 

 material with a clear region in center 

 (also seen by light microscopy), which 

 may represent nucleus (Luna 1968). 

 Two circular profiles at upper right rep- 

 resent remains of neuritic processes. 

 (xSOOO, inset x 22,000). 



Scanning electron microscopy revealed a background of 

 processes and presumptive neurons observable as accumula- 

 tions of granular structures with an outer membranelike 

 covering (Figure 7 A). Transmission electron microscopy in- 

 dicated the processes have a pattern reminiscent of myeli- 

 nated structures; however, myelin lamellae were not identi- 

 fied. No organelles associated with neurons or their 

 processes were present. The most striking finding was an 

 accumulation of electron dense bodies which corresponded 

 to the yellow, granular pigment seen by light microscopy 

 (Figure 7B). These granules resemble lipofuscin pigment 

 (Tauboldetal. 1975; Adams and Lee 1982:234-237). Consi- 

 stent with this interpretation was the finding of more pigment 

 in the older brain (female) than in the younger brain (male). 



ISOLATION AND DEMONSTRATION OF HUMAN DNA 



Nucleic acids were extracted and purified from 15g of rela- 

 tively peat-free cortex by solubilizing, chloroform-phenol 

 extracting, and centrifuging in a CsCl-ethidium bromide 

 density gradient. Material banding at a density of 1 .55 g/cc 

 was collected and identified as DNA by DNase sensitivity 

 and RNase resistance. High molecular weight DNA of 8-20 

 kilobases was clearly present in an ethidium bromide stained 

 gel of this DNA (Figure 8. left). 



To determine whether this DNA was of human origin, a gel 

 was blotted and hybridized to a probe specific for human 

 mitochondrial DNA (miDNA) (Chang and Clayton 1985), 

 The probe hybridized to appropriate sized species in un- 

 digested brain DNA demonstrating that human mtDNA was 

 present (Figure 8, right). To confirm the presence of human 

 DNA a dot blot of 80()()-ycar-old DNA was probed with an 

 Alu repeat sequence (Figure 8, right). The Alu sequence 



hybridized to old human DNA but not to a peat sample from 

 the same level. The experiment was repeated several times 

 using DNA samples from different old brains with similar 

 results. 



The total yield of DNA was about 1 ^.g/g tissue, or 1% of 

 that normally isolated from fresh tissue. Also, the amount of 

 mtDNA present in the old DNA sample appears low relative 

 to total isolated DNA. A comparison of hybridization sug- 

 gests that about 0.05% of the total old DNA was mtDNA; 

 DNA isolated from fresh brain tissue yields 0.5%- 1% 

 mtDNA. Quantitation of Alu sequences on dot blots allowed 

 an independent estimate of the fraction of human DNA se- 

 quences in the old DNA samples (data not shown). We esti- 

 mated that Alu sequences were present at 1% of the level of 

 that from an equivalent amount of human placental DNA. 

 The low yield of human mtDNA sequences could have sever- 

 al potential causes: preferential loss of mitochondrial se- 

 quences may occur during extraction; preferential degrada- 

 tion of mitochondrial sequences may occur during 8000 

 years or during the immediate postmortem period; and signif- 

 icant amounts of nucleic acids from the surrounding plant 

 material may be present in the old-brain cortex sample. If the 

 latter situation is the case, the apparent fraction of any specif- 

 ic human sequences would be diluted by plant DNA se- 

 quences. Although the surrounding peat does contain about 

 the same amount (on a per weight basis) of DNA as brain 

 tissue, this DNA does not hybridize to the human mtDNA 

 probe (Figure 9). 



When the mtDNA was digested with Eco R I , the expected 

 (Anderson et al. 1981) 8kb fragment which should appear 

 after hybridization with the probe was not present (Figure 8, 

 right). However, partial conversion of open-circular to linear 

 molecules did take place, as would be expected if only a 

 fraction of the Eco Rl recognition sequences were present. 



Ztign'h PaleopalhoUffty Symp. 198fi 



