Zp parameter A/vs B DNA

[amadeus2]

Dear Xiang-Jun,

I am currently faced with short single strand sequences obtained by ab initio optimizations and I wish to understand if they are A or B  forms. I have read your illuminating NAR03 paper, where the "Zp" criterion to distinguish A from B DNA is discussed. I used the analyze program, eg:

Code: [Select]

analyze -torsion=caa.tor caa.pdb
and obtained

Code: [Select]

[...]
      Dp: perpendicular distance of the 3' P atom to the glycosydic bond
            [as per the MolProbity paper of Richardson et al. (2010)]

              base       v0      v1      v2      v3      v4     tm       P    Puckering    Zp      Dp
   1 A:...1_:[.DA]A   -11.2    28.7   -34.0    28.5   -11.0    34.0   179.8    C2'-endo    2.29    2.01
   2 A:...2_:[.DA]A   -40.8    45.6   -33.4    10.7    18.6    45.3   137.5    C1'-exo     2.51    2.44
   3 A:...3_:[.DC]C   -44.6    43.3   -25.7     0.4    27.4    45.3   124.5    C1'-exo     ---     ---

Now, according to your NAR2003 paper, this should be an A DNA form. To check if I was correct, I worked out the bdl024.pdb B DNA example of the /examples/analyze_rebuild  x3dna folder and, to my surprise, by using the above command I got:

Code: [Select]


[...]
              base       v0      v1      v2      v3      v4     tm       P    Puckering    Zp      Dp
   1 A:...1_:[..C]C   -20.3    33.1   -33.1    22.1    -1.2    34.5   163.5    C2'-endo    1.60    1.87
   2 A:...2_:[..G]G   -23.9    36.6   -34.9    22.3     0.8    37.1   160.0    C2'-endo    1.46    1.59
   3 A:...3_:[..C]C   -24.8     5.6    13.9   -28.7    33.8    33.3    65.3    C4'-exo     2.36    3.05
   4 A:...4_:[..G]G   -25.7    37.1   -33.9    20.1     3.4    37.1   156.2    C2'-endo    2.19    2.01
   5 A:...5_:[..A]A   -20.9    31.2   -29.3    17.7     2.0    31.6   158.0    C2'-endo    1.92    2.02
   6 A:...6_:[..A]A   -22.6    30.8   -27.2    14.8     4.8    30.6   152.6    C2'-endo    1.87    2.02
   7 A:...7_:[..T]T   -34.2    33.1   -20.1     0.8    20.9    34.8   125.2    C1'-exo     2.03    2.27
   8 A:...8_:[..T]T   -35.8    35.4   -22.0     1.8    21.1    36.7   126.9    C1'-exo     2.09    2.31
   9 A:...9_:[..C]C   -21.5    31.7   -29.8    18.0     2.0    32.2   157.9    C2'-endo    1.17    1.63
  10 A:..10_:[..G]G   -36.1    45.1   -36.0    16.8    11.4    43.6   145.6    C2'-endo    1.41    1.25
  11 A:..11_:[..C]C   -21.3    35.1   -35.2    23.4    -1.5    36.6   163.8    C2'-endo    1.11    1.67
[...]

where almost all the "Zp s" are higher than 1.5 Angstroem, even if bdl024.pdb  is a clear example of B DNA.  I have read here http://x3dna.org/highlights/sugar-pucker-correlates-with-phosphorus-base-distance and here: http://forum.x3dna.org/general-discussions/a-dna-definition/ and I realized I am possibly missing something.

Now the questions:
1. Is the "Zp" parameter of NAR03, the same as the one returned by by the analyze -t=myfile.tor myfile.pdb command?  If not, how can I get the correct "Zp"?
2. How can I obtain an output like the one  in http://forum.x3dna.org/general-discussions/a-dna-definition/, where A and B steps are tentatively assigned by 3DNA?

Code: [Select]

[...]
    step       Xp      Yp      Zp     XpH     YpH     ZpH    Form
   1 GC/GC   -4.37    8.77   -1.05   -4.95    8.83   -0.22     B
   2 CG/CG   -2.56    8.63    0.13   -2.38    8.49    1.55     B
   3 GC/GC   -3.74    9.12   -0.62   -4.46    9.06   -1.34     B
[...[

I am sorry if these are old aquestions, but I wasn't able to find answers in the 3DNA site nor in the forum.

As a newbie I would appreciate your help.

Amedeo

[mauricio esguerra]

Hi,

You can just try:

Code: [Select]

find_pair bdl084.pdb stdout | analyze stdin
And look at the automatically generated output file bdl084.out

The Zp and Zph values will be there.

For your structure just type:

Code: [Select]

find_pair caa.pdb stdout | analyze stdin
And this will produce the caa.out file with your results for the average of the projections on the Z-axis in both reference frames.


Hope this helps,

Mauricio Esguerra

[amadeus2]

Thank you very much for yor kind reply,

I was already aware of find_pair -s myfile.pdb stdout | analyze stdin --note the -s option--  and of course I had already tried (I first had to find a way to make my ab initio outputs pdb-compilant, and I also had to write an "analyze" code of my own just to compute  bp_step parameters to test everything was correct in my ab initio generated pdbs), but unfortunately the above procedure (at variance with double strand cases) does not furnish the wished information for my single strands.

Is "Zp" implemented for single strands? And now I am very curious, what is the difference between the two "Zp" , the one stored in myfile.outs and the one located in myfile.tor  obtained using the -t=myfile.tor option of analyze?

Sorry if my first question was badly asked and if I was unclear.

[mauricio esguerra]

Hi,

That being said, it looks like your input file might be a single strand instead of a double strand. If such is the case the definition of the Zp value for a single-base to single-base STEP might be slightly different as it most likely is not an average, which is the case for the BASE-PAIR step Zp values.

In the output file using the -t option as you have done one can additionally see the following information:

Code: [Select]

   Zp: z-coordinate of the 3' phosphorus atom (P) expressed in the
            standard base reference frame; it's POSITIVE when P is on
            the +z-axis side (base in anti conformation); NEGATIVE if
            P is on the -z-axis side (base in syn conformation)

I am also interested in seeing Xiang-Jun's reply as the single-stranded Zp parameter is a new implementation in 3DNA.

I am guessing that it might be "risky" to call something A-DNA or B-DNA outside of the context of a double-strand.


Cheers,

M.

[xiangjun]

Thank you, Amedeo, for posting such swell-thought-out question regarding Zp vs A/B-DNA classification. Thank you, Mauricio, for your quick response which answers Amedeo's question per the 3DNA NAR03 paper and the paper titled "A-form conformational motifs in ligand-bound DNA structures".

Amedeo's question prompted me to think deeper on the Zp parameter for single-stranded vs double-stranded DNA/RNA structures. The Zp value calculated with the analyze -torsion treats the input structure as single-stranded by expressing the 3'-phosphorus on the reference frame of the preceding base, while the default Zp is for a double-stranded structure as defined in the above two papers. The 1.5 Å cut-off refers to the double-stranded Zp, which is an average of the z-coordinates of the two 'middle' phosphorus atoms in the middle step frame.

The single-stranded Zp was introduced mostly as a more robust substitute than the Richardson Dp parameter to characterize C2'-endo vs C3'-endo sugar puckering. Note that here the cut-off is 2.9 Å. See my post "Sugar pucker correlates with phosphorus-base distance".

To the best of my knowledge, Zp is the most discriminative parameter to classify A- vs B-DNA (duplex). It also helps to check other parameters, such as slide, and x-displacement etc.

HTH,

Xiang-Jun

[xiangjun]

While I was posting my previous reply, you guys have come up with another round of exchange on the topic. Very impressive...THANK YOU!

The single-stranded Zp was not intended as a classifier for A- vs B-DNA as noted in my previous reply. I did see the confusion here, and I will write a post with more info on this topic shortly. In the meantime, please continue to report your suggestions, findings, etc.

Xiang-Jun

[Jeff 3D]

I don't think you would want to classify a single strand as A or B, only a double-strand.

[xiangjun]

I don't think you would want to classify a single strand as A or B, only a double-strand.

Thanks for chiming in. Could you please eleborate on your point?

As a follow-up to my previous reply, I've written a post titled "Single- and double-stranded Zp" to clarify the naming confusion.

Xiang-Jun

[amadeus2]

I wish to thank Xiang-Jun for his excellent Single- and double-stranded Zp post which clarifies everything. I also quote his question for Jeff 3D. It should be possible, at leat in principle,  to classify a single strand as A, B, AB or whatever, although strong irregularities are to be expected. Indeed  I have found an example in the literature where even the ss-5'-AAA-3' trimer in solution has been characterized as B-DNA by using  2D-NMR and molecular dynamics simulations. I have not found crystallographic structures of ss-oligonucleotides, but I am not a crystallographer, hence a sligthly off-topic question: « are you aware of examples of "pure" ss-DNA crystals, i.e. without proteins, ligands, etc.? » 

Thank you all

Amedeo



 

[xiangjun]

Hi Amedeo,

It should be possible, at leat in principle,  to classify a single strand as A, B, AB or whatever, although strong irregularities are to be expected.

I agree with you on this point. It is logical to imagine that each of the two single-strand dinucleotides in an A-form double-helical step would have A-like conformation; similar argument holds for the B-form.

Indeed  I have found an example in the literature where even the ss-5'-AAA-3' trimer in solution has been characterized as B-DNA by using  2D-NMR and molecular dynamics simulations.

Could you provide the reference?

I have not found crystallographic structures of ss-oligonucleotides, but I am not a crystallographer, hence a sligthly off-topic question: « are you aware of examples of "pure" ss-DNA crystals, i.e. without proteins, ligands, etc.? »

I do not have an example at hand. However, I believe a search of the NDB should help.

Xiang-Jun

[amadeus2]

First of all thank you for the suggestion of NAD, I tried PDB database without success.

As to the citation I cannot find it anymore,  but I remember that it was a work by Professor Ignacio Tinoco,  Jr.  Nevertheless, unless you are already aware of, I hope you will find:

1. Achraya et al JACS (2003) vol. 125 p. 9948
2. Isaksoon et al Biochemistry (2004) vol. 43 pag. 15996

useful as well.

Amedeo

[xiangjun]

Thanks for the two references.

Xiang-Jun