comparing melting temperature values from Homodimer vs Two State melting hybridization

No replies
vtmilam
Offline
Joined: 09/06/2014

Dear Michael:
First of all, thanks for providing this web-based thermodynamic calculator! My group has used it for several years to calculate delta G and Tm for many perfectly complementary sequences in which intrastrand loops are negligible. However, we have been moving on to more complicated aptamer sequences that are likely to self-fold. So, I have questions regarding the appropriate approach to calculate “duplex” melt temperatures for two identical aptamer sequences (forming an AA “homodimer”).

The aptamer sequence, A, is as follows:
5' - TTTTTTTTTT CCCGTCTTCC AGACAAGAGT GCAGGG -3'
I used the following conditions for the“Homodimer Simulations”:
[A_0] = 0.001 M
Na+ = 150 mM; Mg++ = 0

The results indicated a Tm (conc) = 58.6 deg C and delta G=7.8 kcal/mol

My questions:
1. Am I right in presuming that the Tm value is for homodimer formation between two identical aptamer sequences? If so, the Tm values seems unusually high (especially when compared to the third set of simulations shown near the end of this message).
2. Is it implied that the delta G value is a negative value, not a positive value (meaning hybridization is thermodynamically favorable below Tm?)
3. Also, what is temperature value associated with delta G = 9.1 kcal/mol?
4. Since the hybridized structure cannot be viewed in this simulation, is there any assumption about the alignment of sequences in the duplex (e.g. only “blunt-ended” duplexes are considered, even if mismatches occur?)
5. Are G-T bonds also assumed as possible in these particular simulations? (Note: occasionally, I’ve seen infrequent instrastrand G-T bonds predicted in your QuickFold for other sequences). Also, G quadruplexes are not considered in your simulations, right?

Just for fun, I then wanted to compare results from your other simulation tools with the Homodimer Simulations.
I used the “Hybridization of two different strands” with the following conditions:
Sequence 1 = 5' - TTTTTTTTTT CCCGTCTTCC AGACAAGAGT GCAGGG -3'
Sequence 2 = 5' - TTTTTTTTTT CCCGTCTTCC AGACAAGAGT GCAGGG -3'
[A_0] = 0.001 M
[B_0] = 0.001 M
Na+ = 150 mM; Mg++ = 0

The results indicated that Tm (conc) = 58.2 deg C and delta G=7.8 kcal/mol -- so very similar results for these first two sets of simulations.

Lastly, I wanted to see the actual secondary structures predicted by UNAFOLD.
I used the “Two State melting hybridization” simulations with the following conditions
Sequence 1 = 5' - TTTTTTTTTT CCCGTCTTCC AGACAAGAGT GCAGGG -3'
Sequence 2 = 5' - TTTTTTTTTT CCCGTCTTCC AGACAAGAGT GCAGGG -3'
DNA at 22 deg C, Na+ = 150 mM; Mg++ = 0
Strand concentration = 0.001 M

The results indicated that Tm (conc) = 26.2 deg C and delta G= -5 kcal/mol and the heterodimer shows one internal loop sandwiched between two 4-base long hybridization segments. No G-T bonding is predicted.
My question is as follows:
1. Based on the predicted duplex structure, the relatively low Tm makes sense but why is this melting temperature value so much lower compared to the first two sets of simulations?

Ok, I appreciate your answer to my lengthy set of questions above!