58
A. Jayaraman et al.
experimental results of Sakakibara et al. [87], where they saw a higher T m for
(POG) 10 versus (POG) 5 . Next, we investigate the effect of charged residues on the
stability of the CLP triple helix. Figure 7b shows our computational melting curves
for (PKG) 4 (POG) 4 (DOG) 4 , (PKG) 4 (POG) 6 (DOG) 4 , (POG) 12 and (POG) 14 (denoted
as sequences I, II, III and IV, respectively, for brevity). We observe that the introduction of charged amino acids while keeping CLP length constant destabilizes the triple
helix as seen by the reduction in T m for III to I and IV to II. The destabilizing effect
of charged residues is a result of the increased electrostatic repulsion between CLP
strands since like charges are located at the same relative position along the triple
helix, thus, repelling each other. These results mirror the observations of Condon
and Jayaraman where they observed that I had a lower melting temperature than III.
Furthermore, we also observe a slight increase in T m when the length of the central
(POG) block of I is increased by two repeat units to yield II. The additional stabilization results from the length effect as shown in Fig. 7a since a larger number of
(POG) units increases the overall enthalpic gain for the triple helix to be hybridized
(e.g., U hyb ) due to a larger number of stabilizing h-bonds.
Next, we use our CLP model to test the hypothesis of how altering the balance
of charges while keeping sequence length constant would affect the stability of the
triple helix. To do so, we started with the sequence II and varied the number of (PKG)
and (DOG) repeat units to obtain (PKG) 3 (POG) 7 (DOG) 4 and (PKG) 4 (POG) 7 (DOG) 3
(denoted as sequences V and VI, respectively). We hypothesize that V and VI should
have higher melting temperatures than the original sequence, II, due to reduced
electrostatic repulsion between CLP strands. Figure 7c shows that the T m of the triple
helix is unchanged when the balance of charges is varied while holding sequence
length constant. It is important to note that the overall charge imbalance per sequence
for V and VI is either − 1 or + 1, respectively, and as a result, these findings may
only be applicable to CLP systems containing small charge imbalances. Therefore,
future work will be focused on exploring the effects of larger charge imbalances on
the melting of the triple helix.
To investigate the effect of changes in CLP amino sequence on the dimensions of
the CLP triple helix, we calculate the end-to-end distances and diameters for a series
of charged and neutral CLP sequences. We observe that the end-to-end distances of I
and III are identical while the end-to-end distances of II and IV are identical (Fig. 8a).
I and III (or II and IV) have the same CLP length and therefore it makes sense that
they would also have the same average end-to-end distance. For III (i.e., (POG) 12 ),
the end-to-end distance agrees with experimental results obtained via small angle Xray scattering experiments which report the length of the (POG) 12 triple helix to be
8.9 nm [107]. We also see that charged sequences, such as I and II, have distributions
of diameters (measured at the ends of the helix) that are shifted to larger values as
compared to neutral sequences such as III and IV (Fig. 8b). The larger diameters for
charged sequences are a result of increased electrostatic repulsion involving charged
residues that are within proximity of one another. Another impressive aspect of our
CLP model is that our computational diameters for III agree with X-ray diffraction
data for a similar sequence, (POG) 4 (POA)(POG) 5 , which report the diameter of the
triple helix to be 1 nm with a resolution of 0.19 nm [108]. Clearly, our CLP CG
A. Jayaraman et al.
experimental results of Sakakibara et al. [87], where they saw a higher T m for
(POG) 10 versus (POG) 5 . Next, we investigate the effect of charged residues on the
stability of the CLP triple helix. Figure 7b shows our computational melting curves
for (PKG) 4 (POG) 4 (DOG) 4 , (PKG) 4 (POG) 6 (DOG) 4 , (POG) 12 and (POG) 14 (denoted
as sequences I, II, III and IV, respectively, for brevity). We observe that the introduction of charged amino acids while keeping CLP length constant destabilizes the triple
helix as seen by the reduction in T m for III to I and IV to II. The destabilizing effect
of charged residues is a result of the increased electrostatic repulsion between CLP
strands since like charges are located at the same relative position along the triple
helix, thus, repelling each other. These results mirror the observations of Condon
and Jayaraman where they observed that I had a lower melting temperature than III.
Furthermore, we also observe a slight increase in T m when the length of the central
(POG) block of I is increased by two repeat units to yield II. The additional stabilization results from the length effect as shown in Fig. 7a since a larger number of
(POG) units increases the overall enthalpic gain for the triple helix to be hybridized
(e.g., U hyb ) due to a larger number of stabilizing h-bonds.
Next, we use our CLP model to test the hypothesis of how altering the balance
of charges while keeping sequence length constant would affect the stability of the
triple helix. To do so, we started with the sequence II and varied the number of (PKG)
and (DOG) repeat units to obtain (PKG) 3 (POG) 7 (DOG) 4 and (PKG) 4 (POG) 7 (DOG) 3
(denoted as sequences V and VI, respectively). We hypothesize that V and VI should
have higher melting temperatures than the original sequence, II, due to reduced
electrostatic repulsion between CLP strands. Figure 7c shows that the T m of the triple
helix is unchanged when the balance of charges is varied while holding sequence
length constant. It is important to note that the overall charge imbalance per sequence
for V and VI is either − 1 or + 1, respectively, and as a result, these findings may
only be applicable to CLP systems containing small charge imbalances. Therefore,
future work will be focused on exploring the effects of larger charge imbalances on
the melting of the triple helix.
To investigate the effect of changes in CLP amino sequence on the dimensions of
the CLP triple helix, we calculate the end-to-end distances and diameters for a series
of charged and neutral CLP sequences. We observe that the end-to-end distances of I
and III are identical while the end-to-end distances of II and IV are identical (Fig. 8a).
I and III (or II and IV) have the same CLP length and therefore it makes sense that
they would also have the same average end-to-end distance. For III (i.e., (POG) 12 ),
the end-to-end distance agrees with experimental results obtained via small angle Xray scattering experiments which report the length of the (POG) 12 triple helix to be
8.9 nm [107]. We also see that charged sequences, such as I and II, have distributions
of diameters (measured at the ends of the helix) that are shifted to larger values as
compared to neutral sequences such as III and IV (Fig. 8b). The larger diameters for
charged sequences are a result of increased electrostatic repulsion involving charged
residues that are within proximity of one another. Another impressive aspect of our
CLP model is that our computational diameters for III agree with X-ray diffraction
data for a similar sequence, (POG) 4 (POA)(POG) 5 , which report the diameter of the
triple helix to be 1 nm with a resolution of 0.19 nm [108]. Clearly, our CLP CG