The Discovery of m-RNA

The Discovery of m-RNA
and important spin-offs:
the source of rRNA, transience of mRNA, basics of molecular taxonomy, and evidence for a genetic basis of biological clocks

Before these discoveries could be made, a method or "tool" had to be invented that worked. In theory, what was being considered was doing a sort of reverse melting of double-stranded nucleic acids. The brew would consist of melted single-stranded DNA (ssDNA) and RNA from that same species. IF the RNA was indeed a transcript of the DNA, THEN there would be complementary stretches on the DNA to which the RNA should anneal. A stretch of DNA/RNA double helix would be called "hybrid," as shown:

The major hurdle was finding a "control" type of RNA that was very close to being pure, which could then be used to test the method. As 85% of E. coli RNA is ribosomal RNA (rRNA), it was easy to purify some ribosomes and then isolate rRNA from them. However, the question arose as to whether or not rRNA would have different origin from all other RNA's. Maybe it was not a direct genetic transcript! But this method-in-principle would answer that question.

Is Ribosomal-RNA a Genetic Transcript?

So in order to test the method, rRNA was used:

However, in order to "see" or "observe" if any hybrid was formed, radio-active tracers had to be used. (See a short lesson on isotope tagging appended to Individual Quiz #1.) Notice that the DNA is tritiated and the rRNA has radio-phosphorous incorporated into its backbone. So, in general, these were the reagents used in the very first experiments. Notice that the fourth line consists of ribosomes taken from an extremely distantly related bacterium (E. coli and Bacillus sp. are less related that we humans are to our local tree). This we can call "heterologous" rRNA, which would not likely have much simularity with the E. coli rRNA. (Or would it? Maybe ribosome functions are so narrowly defined that the RNA base sequence is highly conserved! But we shall soon find out!)

So the following reagents were made with the listed degrees of radioactivity ("specific radioactivity" = cpm/µg)

Jumping ahead a little over a year to when it was discovered that ssDNA adhered tenaciously to nitrocellulose, the researchers "stuck" the ssDNA to small circles of nitrocellulose filters ("ncFilters"). These various filters could then be soaked in various solutions of RNA's and/or enzymes with the following results. Again, all this tediousness was necessitated to test the validity of the "tool" under the most stringent conditions. Because it was expected that large amounts of DNA would be present but only minute traces of RNA bound to the DNA, the specific radioactivity of the DNA could be low and that of the RNA must be very high (to be "bright" enough to see even small amounts).

Thus if any hybrid at all was formed, both isotopes would be found there, if only the top two "reagents" (above) were used in the annealing process. But with the use of these extremely large molecules, there was expected to be a great deal of coincidental trapping of RNA in the tangled mass in non-complementary regions. Thus as a final step, the hybrid was treated with RNase, which can digest single-stranded RNA but not RNA that is in a helix with DNA. Thus a "clean" experiment was devised:

When the above experiment was done, the following results were observed. A dozen or more "DNA-filters" were placed in a solution of the P-32 rRNA, and one by one they removed at timed intervals. As more time elapsed the RNA molecules had more chance of meeting up with a complementary stretch of DNA until a time came when all the sites were occupied (saturated). The treated filters were then immersed in RNase to destroy any non-specifically bound RNA (RNA in hybrid is in a double helix and is not susceptible to RNase). The filters were then counted in a scintillation spectrometer that could discern the differences in the decays of the weak tritiums and the very strong P-32's. Notice that the tritium counts (gray flecks) formed a horizontal line up at the 100,000 cpm level. That indicated that all the filters contained about 100 µg of DNA. The P-32 counts however increased during the first minutes until they sought to saturate all the complementary sites on the DNA. They plateaued at 900,000 cpm, which indicated 0.3µg of rRNA. So 0.3 µg RNA / 100 µg DNA = 0.3 percent hybrid.

How many copies of rDNA are there in E. coli?

The E. coli genome is about 3.5 x 109 daltons or
about 5.25 x 106 bp.
Prokaryotic rRNA consists of equal amounts of two sizes:
1 x 106 daltons and 0.5 x 106 daltons.

Since %Hybrid = 0.3%, then
0.003 x 3.5 x 109 daltons = 10.5 x 106 daltons of rDNA

10.5 x 106 daltons ÷ 1.5 x 106 = 7 copies of each type of rDNA

Another necessary test of the method was that of the stoichiometry of the reaction. IF an equal amount of identical - but non-radioactive - rRNA were added to the "brew", half of the complementary sites would anneal with "cold" RNA and half with "hot" RNA. Thus a final count of the P-32 radioactivity would be half that seen in the foregoing experiment. This is what is called a "competition" experiment: does the cold RNA compete for the same sites as does the "hot" RNA? If they do, then they must be identical. Well, our "reagents" were devised to be identical - both were rRNA from E. coli. Again, this is a check on the method's validity, before applying it to systems of unknown degrees of similarity.

The data collected from this 1-to-1 competition experiment supported the model we have had in mind. Notice that the level of radioactivity is only half that in the first experiment.

This following experiment breaks new ground into the realm of molecular taxonomy: if a tremendous excess ("XS") of "cold" rRNA is added from a heterologous source (another bacterial species or from your lab partner!) we can see if there is any similarity at all between that rRNA and that of the E. coli rRNA. If there is similarity, then the "cold" RNA's will compete for the E. coli rRNA sites, and lowering the amount of radioactivity counted. But if they are very dissimilar, then no competition will result, and the results will look like those of experiment #1.

Well, what do you say? The rRNA from your heterologous source was indeed totally dissimilar to that found in the E. coli rRNA - no competition! Thus, among other conclusions: rRNA sequences are not highly conserved throughout nature.

OH, OH! Mary Lou got into the lab!

After the above experiments were done, she thought she had a bright idea: that the main experiment ought to be repeated to show reproducibility. So she grabbed two cultures of E. coli that were growing in media containing tritiated thymidine. One was an old culture from the day before and the other was one just started three hours before and then abandoned. From each she isolated their tritiated DNA, and used them for repeats of Experiment #1. Here are her puzzling results:

When Cheryl and her brother saw those results, they were sure that they were seeing some sort of short-term proliferation of the rRNA genes. In other words, they were thinking that when a gene is needed a lot, the cell might somehow make more copies for awhile (something like tandem alleles). Then once the growth rate subsided, the excess copies of the gene were discarded. At any rate, if the cell started with one copy, it would increase the numbers of copies by integral numbers - something akin to quantum jumps as one should not have half-genes. When they looked at Mary Lou's data, they seemed to think that the quantum jump was something like 2 copies becoming 3 copies.

By Mary Lou was not finished with her brighter colleagues yet. She with the help of Prof. Bengston devised ways to make tritiated DNA's from cultures growing at various speeds (the device used was a "chemostat"). Look at the data that was derived! What is showed was that there was a smooth rise in the amount of DNA devoted to coding for rRNA as the growth rate increased. (Note that the horizontal axis shows generations per hour. When it is zero, the culture is dormant, and when it is "1" then the culture is doubling every 60 minutes; and when it is "3" it is doubling every 20 minutes, which is E. coli's maximum rate.)

This was extremely puzzling - seemingly no integral jumps in copies of genes per genome. How could this be?

A few months later, Jacob (of Jacob, Monod, Brenner and Cuzin of Lactose Operon fame), thought he had the answer as he hypothesized how the E. coli chromosome must replicate. Here is his "replicon" model. Note that the closer a gene might be situation to "oriC", the more copies of it there are in a rapidly growing cell (bottom picture).

As the author of this web-site found, the seven separate rRNA genes are found scattered on both sides of oriC in the "early" half of the chromosome. Later this author also demonstrated that by placing the lac-operon at various positions around the circular map of E. coli, the amount of ß-galactosidase was commensurate with the distance from oriC. And later yet, this author showed that the placement of the various genes for making the many different amino acids were placed on the chromosome such that those that made the most plentiful amino acids (met, leu, etc.) were early on the map, and those that made rare amino acids (his) were very late.

Evidence of the Existence of a Biological Clock at the Genetic Level

Development requires two things (among others, of course):

  1. That some genes turn on at different times, and
  2. That the mRNA transcripts only last a short time so that the expression of earlier functions does not carry over into later periods.

However, in the simple viral setting, clock mechanisms may not be present. There are several different scenarios possible:

  1. Once the viral gene is within the host cell, it begins to be expressed and stays "on" the whole time thereafter. This could imply that -
    • a clock could depend upon the order in which genes enter or are revealed (unwrapped) in the cell.
  2. Genes are turned on at different times and then stay "on" thereafter until the host cell dies. Any finding supporting this implies that a mechanism exists for deciding which gene is turned on when.

  3. Genes are turned on at different times, and then their expression is turned off shortly thereafter. This will necessitate not only the timing director of #2, but also a transient mRNA that is short-lived.

Interestingly, the experimental history shows that #3 was explored first:

This was called a "pulse/chase" experiment because at a specified time after infecting E. coli with T4, a "pulse" of P-32 phosphate was added so that any [macromolecular] mRNA made thereafter would incorporate P-32, and then in another 2 or three minutes a thousand-fold excess non-radioactive (ordinary) P-31 phosphate would be added as a "chaser" to the growth medium. This huge excess ("XS") of P-31 would "quench" the uptake of P-32 by the extreme dilution. Thus, after the "chase" was added, all subsequently synthesized RNA would not be radioactive. The only radioactive RNA would be that which was made during the time between the pulse and the chase. How long the RNA remained radioactive would indicate the life-span of the RNA, and would thus answer the which pathway in the above graph was the true one - that of persistent mRNA or a rapidly decaying, transient mRNA.

When this experiment was actually run, rapid decay always occurred. (The only long-lived mRNA known is that coding for hemoglobin in mammalian red blood cells.) The macromolecular P-32 started disappearing immediately after the addition of the chaser. Again, for emphasis, this mRNA is very transient and disappears within a couple of minutes.

The next problem attacked was concerned whether the genes expressed during a 2 minute pulse/chase early in the viral infection period (the "latent" period) were the same or different from the genes expressed late in the infection cycle.

Solving the problem requires the "competition experiment" talked about earlier. P-32 RNA's from an early pulse/chase period were mixed with a huge excess of P-31 RNA's isolated much later in the infection cycle. The results showed "no competition", which meant that the early mRNA's were different from the late mRNA's because they didn't compete for the same sites. A reciprocal supportive experiment in which early P-31 was used to compete against "late" P-32 mRNA's, and again there was no competition. Thus as the infection cycle continues different genes are turned on and then off, and the respective mRNA's come and go. Repetitions of this experiment show that the same genes turn on at their appointed times "just like clockwork."

Thus while trying to develop the method for looking to see if there were early and late genes in bacterial viruses, a number of major findings were made - many not pertaining to viruses, but of extremely fundamental importance to ALL cells:

  1. rRNA is indeed a transcription product. (Later tRNA's were also shown to be transcriptional products.)
  2. That while rRNA's and tRNA's have long half-lives, mRNA's are highly transient.
  3. specific RNA's (and small fragments of ssDNA) can be used in taxonomic studies in competition studies.
  4. discovered that the E. coli chromosome replicates from a fixed point (oriC) and goes bi-directionally.
  5. that placement of genes on the E.coli chromosome can correlate with the amount of "gene-product" (protein) that the cell normally needs
  6. that hybridization of RNA to DNA can serve as an accurate tool for looking for the genetic basis of biological clocks (think development) because there are early and late mRNA's.

I want to go to the TOP OF PAGE or ESCAPE! or go back to the Virology Home Page!