HOME
BACKGROUND
The hereditary background of an organism is well documented in the DNA of its genes and in the proteins that it produces. We would expect siblings to have greater similarity in their DNA and proteins than two unrelated individuals of the same species. Based on the principle of common descent, we should also expect two species, judged to be closely related by other criteria, to have a greater proportion of their DNA and proteins in common than more distantly related species.
RESEARCH
The Animals and The Gene
The research group collected a wide variety of animal specimens from Island Beach State Park, New Jersey. We also used specimens supplied by the American Museum of Natural History. (See an early version of our database.) Our goal was to identify evolutionary relationships among them by examining a gene they all share. The gene we studied is the 18s gene, which codes for rRNA of the small subunit of cellular ribosomes. Much of this gene is conserved among animal species. However, the nucleotides that do differ from species to species can be used to determine evolutionary relatedness.
The 18s gene is too large to read in one chunk, so we set out to read three segments of the gene. The segments are defined by the primers that border them. Our goal was to read a "1F/5R" segment, approximately 940 bases long; an "S2/S3" segment, 550 bases long; and an "A2/9R" segment which encompassed 680 bases. If all had gone well, we would end up with more than 90% of the bases of each 18s gene we read. Inevitably, we had some problems. We ended up with reliable data for only the "S2/S3" section, which contains enough code that we were able to continue with our phylogenetic endeavor.
Augmenting the Gene
Our first step was to grind up our specimens and extract the DNA. We were able to get DNA out of most of our 45 original specimens.
Our next step was to use the Polymerase Chain Reaction (PCR) to amplify each section of the gene that we hoped to look at. Each section was handled separately; thus a single specimen would undergo three PCRs. PCR creates double-stranded copies of the double-stranded original DNA, but only for the section bordered by the primers. For an excellent animation of how PCR works, visit the Dolan DNA Learning Center. After the PCRs were completed, we cleaned each sample so that the extraneous reactants were removed and we were left with a sample of DNA. Each sample contained a species-specific collection of identical copies of a particular 18S gene fragment. Not all PCRs worked. Only those that did continued on to sequencing.
Reading the Gene
Next, we sequenced each sample. In the interest of reading both sides of the DNA strands, samples were divided again. For example, the tube of "S2/S3" copies of the beetle DNA had to be divided into two sequencing tubes. One would amplify one side of the DNA fragment. The other tube would be used to amplify the complementary strand. Interestingly, the basic method of sequencing is very similar to PCR. The successful sequences were run on a special electrophoresis gel. As the DNA fragments run through the gel, a laser reads off the DNA sequence - which is a letter by letter code of the fragment. Our goal was to have the sequences for each "S2/S3" fragment of each species and further, to have the complementary code.
Not all the sequences worked, nor did the gel get a good read of all of the DNA we loaded. The gel was connected to a computer that translated the information into graphs (a program called Chromas) which had nucleotide letters associated with the peaks on the graphs. (The Dolan DNA Learning Center has a good animation of sequencing and the gel.)
In the most successful cases, we ended up with the code for the "S2/S3" segment in each direction (the code of one strand and the code of the complementary strand). We used an internet site called Clustal to run comparisons on the letter codes of the two strands. The results guided us in editing the original Chromas files that came off the sequencing gel. We ended with a final version for each fragment that we were confident had been double-checked. The data "cleaning" was probably the most tedious and potentially subjective part of the project. In some cases we did not have both strands, and in these cases we had to rely on the information we had.
Comparing Genes
The last two steps were the most rewarding. First, we did a an internet BLAST search (hosted and run by the National Center for Biotechnology Information) on the final "S2/S3" sequence for each animal. BLAST returns its "best guess" as to the identity of our specimens. In some cases, the database had record of our organism and so was able to recognize the code. In a couple of cases, however, our specimen's code has never been published, and so it was not in the database; instead the BLAST returned a related species..
We returned to Clustal a second time. This time we entered the final sequences for all our specimens, side by side. Clustal returned a phylogenetic tree based on the similarities and differences among the "S2/S3" codes. Our slime mold specimen was set as an outgroup as it was the only non-animal among our collection. The resulting "tree" can be accessed below.
CLICK ON THE PICTURE TO SEE OUR TREE
Notice that parts of the tree do not make sense! This is due to several factors. One, our final group of specimens was not very large and not very diverse. Furthermore, the 18s gene is a very well conserved gene, and we were working with only a third of it at that. One way to elicit a more probable tree is to include more specimens. To fill out our tree we collected published 18s sequences from GenBank and repeated the Clustal analysis. The fuller tree can be accessed below.
CLICK ON THE PICTURE TO SEE OUR TREE
The Last Word
The intent of this project was to develop a phylogeny for a variety of organisms using current techniques in biotechnology. In addition the project incorporated many aspects of field ecology. Specimens were collected and identified, then their DNA was analyzsed. As you can see from the trees above, our project did not produce the phylogenetic trees that we anticipated. For example, the arthropods failed to yield a monophyletic group. When we included DNA sequences from Genbank, our tree took on a more predictable form. This certainty demonstrated the concept of using as large of a collection of taxa in your study as possible. However, even with the additional taxa the tree contained irregularities, such as the Silverside which one would expect to see grouped with the other vertebrates.
In terms of our experience, this was an amazing project and one that many of us felt could be done in the high school setting. Even if some of the equipment is prohibitive, linking with resources in the local community could provide the assistance necessary.