Dinucleotides - Isomerization Properties

    All of the 80 K ATDs show multiple peaks indicating that the dinucleotides have different conformations. Each peak has been identified as the stacked, H-bonded, or open form. However, as the temperature inside the drift cell increases, the ATD peaks begin to merge, eventually forming one symmetric peak between 200 and 400 K (shown at left).
   This merging is a sign that the different conformers are starting to isomerize. Another way to look at it is shown above at right. For each dinucleotide, at least two conformers are present that must be separated by some barrier with energy E_a. At low temperatures (and hence low internal energy), each conformer remains in its respective well and they can be separated in the drift cell (appearing as separate peaks in the ATD). As the temperature is raised, the internal energy increases and eventually surpasses E_a, allowing the two conformers to interconvert. At this point the separate peaks in the ATD begin to merge since the ion spends time in the drift cell in each conformation. At high enough temperatures, the two conformers can interconvert fast enough that they yield a single, time-averaged peak in the ATD. The question is: Can we get an estimate of the height of the isomerization barrier?
    The shape of the ATD will depend on how fast the conformers isomerize as they drift through the cell. The slower the interconversion process, the more distinct the peaks become in the ATD. A theoretical model for ATDs, taking into consideration that the ion may be reacting to form another species, has been published by I.R. Gatland [Gatland, I.R. Case Studies in Atomic Physics 1974, 4, 369-437]. We have modified this program to fit our situation of one conformation isomerizing into a different conformation. In the modeling, the only unknown variables are the rate constants for isomerization. That is, k for the stacked open conversion and k for the open stacked conversion, etc.
    The experimental ATDs are then fit with this model by adjusting the various rate constants until the shape of the theoretical ATD matches that of the experimental ATD. An example of the fits is shown above.
    Since the ATDs are fit over a series of temperatures, the rate constants k are known as a function of temperature. Thus, a simple Arrhenius plot will yield the barrier height, E_a.
    For the dinucleotides with two conformers present, the ATD fits are relatively straightforward: stacked open, open stacked, stacked H-bonded, H-bonded stacked. An example of an Arrhenius plot for a two-conformer system is shown above at left. A summary of the barrier heights for all the two-conformer systems is shown in the table at right.
    For dAC, dCC, dGC, and dTC, the problem of fitting the ATDs is more complex. Three conformers are observed for each system and the theoretical model was only developed to handle two reacting species. Additionally, it is difficult to tell which conformer isomerizes into what. For example, does the isomerization follow stacked H-bonded open or does the stacked form isomerize directly into an open form?
    For dAC, dGC, and dTC, the isomerization of all three conformers takes place between 100 and 200 K, making it difficult to determine the order of isomerization. For dCC, however, two peaks are still present in the 300 K ATDs (below).
    The cross section obtained from the shortest-time peak in the 300 K ATD agrees very well with the theoretical value of the stacked conformer (132 ± 1 Ų vs. 130 ± 2 Ų). Therefore, this peak most likely represents only the stacked form and is not a mixture of conformers. The cross section from the longest-time peak in the 300 K ATD (146 ± 1Ų), on the other hand, falls between the theoretical values of the H-bonded form (142 ± 2 Ų), and the open form (152 ± 2 Ų). Therefore, this peak is most likely a combination of the H-bonded and open forms.
    The ratio of the two peaks in the 300 K ATD is 1:3. In the 80 K ATD, the ratio of the three peaks is 1:2:1. Thus, only the H-bonded and open forms are isomerizing at low temperatures and the stacked form does not get involved in the isomerization process until ~400 K. A possible reaction coordinate diagram describing the isomerization of dCC is shown above. Because the H-bonded and open conformers isomerize between 100 and 200 K, the barrier between these two forms must be small. The stacked form, however, does not begin to isomerize until 400-500 K, suggesting a larger barrier. The stacked form most likely isomerizes into the H-bonded form or at least goes through an H-bonded intermediate.
    Assuming this scenario, the two barrier heights can be determined by modeling the dCC ATDs at two different temperature ranges. Between 80 and 200 K, the experimental ATDs were fit using only the rate constants for the H-bonded open and open H-bonded conversions (and ignoring the fastest-time peak in the ATD which corresponds to the stacked form). Between 400 and 500 K, the experimental ATDs were fit using only the rate constants for the stacked H-bonded and H-bonded stacked conversions. The resulting Arrhenius plots are shown at right.