Where did I come from?

Mitochondrial DNA and Human Evolution

Allan Wilson, working with colleague Mark Stoneking and research student Rebecca Cann, studied the patterns of human genetic variation to answer the questions of when and where did humans arise. They applied methods which had previously been used to study other animals. From this, they obtained the first pictures of how large groups of humans are related. These results provided evidence supporting one hypothesis regarding the origin of modern humans and enabled them to estimate when that event occurred.

Use the maternal inheritance of mitochondrial DNA to estimate the genealogy of humans through the female line:

Human Groups Sampled
People from 5 geographic regions were represented in the study.

20 Africans, representing the sub-Saharan region
This was one of the most controversial samples because most of the individuals were African Americans. Cann and colleagues justified their choice by arguing that the individuals sampled could probably trace their maternal ancestry to Africa. They thought that any non-African ancestry in the individuals would have come from men, and so not be inherited via mitochondria. Only one sample came from the !Kung group of aboriginal South Africans.

AfricanAmericanWoman.png    Kung_motherandchild.png

34 Asians
People from China, Vietnam, Laos, the Philippines, Indonesia and Tonga.

Children_in_Middle_Vietnam.png   Dr_Gao_Yaojie.png

46 Caucasians
People from Europe, North Africa and the Middle East

Egyptian_Girl.png   StephanieJimenez.png

21 aboriginal Australians

Australian_woman.png   Australian_girl.png

26 aboriginal New Guineans

NGwoman1.png   olderNGwoman.png

Tissues Used
This study was based on DNA extracted from human placentas, obtained from hospitals, and from cells grown in tissue culture. The majority (145 of 147) of the samples were placentas. These are large organs, weighing about 500 grams each, and we do not know if the techniques for assessing genetic variation required such a large amount of material. It may have been that it would have been difficult to get more than a small blood sample from each donor, and that amount might have been insufficient.

Genetic Variation
This study did not determine the DNA sequence of a region of the mitochondrial DNA. That approach for assessing genetic variation was developed later. Instead, a method known as restriction fragment length polymorphism (RFLP) was used to "sample" the mtDNA genome in several places, using 12 different restriction enzymes. Cann, Stoneking and Wilson estimated that they sampled about 9% of the 16,569 bases in the human mtDNA genome. Here is an overview of how the RFLP method works.

The RFLP method identified 467 binding sites, of which 195 were variable, that is they occurred in only some individuals. The data then was a list of which of these variable sites were present or absent in each individual.

Estimating the Tree
Genealogical or phylogenetic trees can be estimated from observed genetic variation because genetic change accumulates over time. There are different methods for estimating such trees. Some methods are based on average similarity or difference between pairs of sequences. The method used by Cann, Stoneking and Wilson is a logical method called maximum parsimony. In this method many different possible trees are tested. For each tree, they ask "If this is the right tree, then how many evolutionary changes are required to explain the observed genetic variation?" The simplest explanation is considered the best. This means that the tree which requires the fewest number of evolutionary changes is the best estimate of evolutionary history. So, Cann and colleagues searched for the tree which best explained the variation in RFLP binding sites among the individuals sampled.

The following diagram shows that in the left hand tree only one change is needed to explain the pattern of presence and absence of binding sites, but in the right hand tree two changes are needed. The parsimony method would choose the left had tree because it provides a simpler explanation of evolutionary history.


This study is concerned with estimating the genealogy of humans back to their most recent common ancestor. To do so, we need to be able to identify the oldest point on the tree, theroot. The approach taken by Cann and colleagues was to put the root at the mid-point of the tree. This is the point halfway between the two most distant tips.


In using the mid-point as the root, you are assuming that the rate of evolution is the same in all of the lineages descending from the root.

This account is based on:
Cann, R. L., M. Stoneking, and A. C. Wilson. 1987. Mitochondrial DNA and human evolution. Nature 325:31-36.