To determine which of the three models of replication applied to E. coli cells, Matthew Meselson and Franklin Stahl needed a way to distinguish old and new DNA. They accomplished this by using two isotopes of nitrogen, ¹⁴N (the common form) and ¹⁵N (a rare, heavy form). Meselson and Stahl grew a culture of E. coli in a medium that contained ¹⁵N as the sole nitrogen source; after many generations, all the E. coli cells had ¹⁵N incorporated into all the purine and pyrimidine bases of their DNA (see Figure 8.8). Meselson and Stahl took a sample of these bacteria, switched the rest of the bacteria to a medium that contained only ¹⁴N, and then took additional samples of bacteria over the next few cellular generations. In each sample, the bacterial DNA that was synthesized before the change in medium contained ¹⁵N and was relatively heavy, whereas any DNA synthesized after the switch contained ¹⁴N and was relatively light.

Meselson and Stahl distinguished between the heavy ¹⁵N-laden DNA and the light ¹⁴N-containing DNA with the use of equilibrium density gradient centrifugation (Figure 9.2). In this technique, a centrifuge tube is filled with a heavy salt solution and a substance of unknown density—in this case, DNA fragments. The tube is then spun in a centrifuge at high speeds. After several days of spinning, a gradient of density develops within the tube, with high-density material at the bottom and low-density material at the top. The density of the DNA fragments matches that of the salt: light molecules rise and heavy molecules sink.

Meselson and Stahl found that DNA from bacteria grown only on medium containing ¹⁵N produced a single band at the position expected of DNA containing only ¹⁵N (Figure 9.3a). DNA from bacteria transferred to the medium with ¹⁴N and allowed one round of replication also produced a single band, but at a position intermediate between that expected of DNA containing only ¹⁵N and that expected of DNA containing only ¹⁴N (Figure 9.3b). This result is inconsistent with the conservative replication model, which predicts one heavy band (the original DNA molecules) and one light band (the new DNA molecules). A single band of intermediate density is predicted by both the semiconservative and the dispersive models.

To distinguish between these two models, Meselson and Stahl grew the bacteria in medium containing ¹⁴N for a second generation. After a second round of replication in medium with ¹⁴N, two bands of equal intensity appeared, one in the intermediate position and the other at the position expected of DNA containing only ¹⁴N (Figure 9.3c). All samples taken after additional rounds of replication produced two bands, and the band representing light DNA became progressively stronger (Figure 9.3d). Meselson and Stahl’s results were exactly as expected for semiconservative replication and were incompatible with those predicated for both conservative and dispersive replication. TRY PROBLEM 18

CONCEPTS