< 3rd term

DNA forms and RNA structure

Last update on the 25th of September, 2017

Here we observe three DNA forms and tRNA, calculate torsion angles and analyze interactions of bases and their pairs. Initial assay was made with command-line 3DNA set of programs. Obtained files made possible visualization with Jmol and take snapshots. Further analysis was made with LibreOffice Calc spreadsheet software. Cytosine was drawn in MarvinSketch.

List of downloads
File Link
A-DNA gatc-a.pdb
B-DNA gatc-b.pdb
Z-DNA gatc-z.pdb
tRNA compared to DNA by torsion angles 1qrt_torsion.ods
tRNA and DNA torsion angles deformed_dna_trna.ods

DNA forms

gatc-a.pdb, gatc-b.pdb, gatc-z.pdb

With 3DNA fiber program sample models of A-, B- and Z-DNA were generated in PDB format. In all three forms major and minor grooves were noticed taking into account depth and width of grooves. Models are shown in figure 1.

Fig. 1. Sample A-, B- and Z-forms of DNA.
A- and B-forms are 5-times repeated "GATC" sequence, Z-form is 10 times repeated "GC" sequence. Major and minor grooves are signed.

Then, in B-form deoxycytidine 16:A (in PDB notion) was highlighted and inspected which atoms are turned to the major groove and which — to the minor groove. A 3D-model is presented in figure 2.

Fig. 2. Deoxycytidine 16:A in B-DNA.

In total, atoms c16.n1, c16.c2, c16.o2 are turned to the minor groove and atoms c16.c4, c16.n4, c16.c5, c16.c6 are turned to the major groove. The orientation of c16.n3 cannot be unequivocally determined (see the representation of colored base, fig. 3).

Fig. 3. Cytosine base.
Red atoms are turned to the major groove, blue - to the minor groove.

With Jmol measurement tools main spiral properties of DNAs were spotted (table 1). It turned out that in A-form major groove is narrower then minor one, though it is still deeper. The orientation of deoxycytidine atoms remains the same.

Table 1. Main spiral properties of DNA forms.
In major and minor grooves width lines phosphates used to measure distance are given in PDB notation.
Prop|Form A B Z
Spiral type Right Right Left
Step, Å 28.03 33.75 43. 5
Bases per coil 11 10 12
Major groove width, Å 7.98
5:A.P — 30:B.P
4:A.P — 31:B.P		 17.91
8:A.P — 29:B.P
9:A.P — 28:B.P		 15.17
14:A.P — 27:B.P
13:A.P — 28:B.P
Minor groove width, Å 16.97
31:B.P — 13:A.P
30:B.P — 14:A.P		 11.69
34:B.P — 11:A.P
35:B.P — 10:A.P		 9.87
38:B.P — 7:A.P
37:B.P — 8:A.P

Torsion angles

1qrt_torsion.ods, deformed_dna_trna.ods

Objects of study in this section are tRNA from 1QRT PDB entry and DNA from 1KSX entry. The PDB files were converted to the old format with remediator program and analyzed with find_pair (-t) filename stdout | analyze pipeline. Several files were generated. First, the information from *.out files was used.

Fig. 4. Torsion angles of nucleotides.

The first property of interest is which form of DNA tRNA is similar to. That was determined with torsion angles study (see fig. 4). To do that, the values of torsion angles were retrieved from *.out files (all DNA models gatc-* were also put through the mentioned pipeline). As the pipeline calculates angles for only those nucleotides that are engaged in base pairs, the comparison seems probable. Retrieved values were put into spreadsheet software and were handled with following analysis: absolute differences of DNA and RNA angle values were taken, averages and standard deviations were calculated; the lower average means greater similarity of tRNA to a particular DNA form, the lower standard deviation means bigger consistence of particular angle value in tRNA (as DNA values are the same except Z-DNA). It should be meant that greater standard deviation means lower applicability of this test or wrong order of Z-DNA angles. First issue is solved as it is the only reasonable test for this task, second issue is solved with observation that tRNA angles do not alter (sign too) as Z-form ones. Results are given in the table 2; all raw data is stored in 1qrt_torsion.ods.

Table 2. tRNA similarity to DNA forms based on torsion angles.
Numbers are absolute differences between tRNA and DNA angle values reduced to averages and standard deviations (stdev).
B* - no altering angle values as in Z-form.
Angle Strand Criterion A B Z Verdict
α 1 average 56,86 67,96 106,69 A
stdev 67,88 55,74 78,16 A
2 average 48,23 61,01 99,40 A
stdev 60,37 50,06 54,69 A
β 1 average 155,24 146,28 133,40 B*
stdev 157,74 131,03 144,04 B
2 average 154,44 140,79 146,10 B
stdev 155,43 133,52 140,58 B
γ 1 average 45,44 51,99 132,93 A
stdev 61,38 58,32 102,42 A
2 average 35,59 43,07 135,17 A
stdev 48,99 46,06 114,62 A
δ 1 average 6,22 58,47 31,86 A
stdev 8,68 9,01 20,92 A
2 average 6,24 58,09 30,31 A
stdev 5,75 5,88 20,95 A
ε 1 average 39,69 39,83 60,67 A|B
stdev 75,91 73,47 57,44 A|B
2 average 31,81 34,68 61,41 A
stdev 63,98 60,56 41,75 A
ζ 1 average 24,83 94,06 87,62 A
stdev 42,07 45,92 81,87 A
2 average 35,39 103,34 75,33 A
stdev 57,90 65,44 74,64 A
χ 1 average 44,83 82,32 136,75 A
stdev 99,33 67,67 109,49 A
2 average 57,30 84,87 128,70 A
stdev 107,89 74,12 108,47 A

Despite huge values, it seems possible to say that tRNA is more similar to A-DNA than to other forms. The second point of interest are average values of torsion angles in sample tRNA and DNA. They were calculated in deformed_dna_trna.ods spreadsheet. Also, the most deformed nucleotides were determined with rank test. First, the absolute differences of each base angle and average were calculated. Then, each base for each angle were ranked from the biggest to the smallest difference. Ranks were summed for each base and, finally, sums were ranked from the lowest to the biggest. Since then, the base with "1" rank is the most deformed. Such test were made for all 4 strands (see table 3).

Table 3. Average values of torsion angles in sample nucleic acids.
Given nucleic acids are presented in PDB by IDs 1KSX (DNA) and 1QRT (tRNA). "1" and "2" denote strands of nucleic acids. Values in brackets in "D" line stand for PDB denotion of the most deformed nucleotides in each strand concerning torsion angles by rank test.
Angle|Strand DNA1 DNA2 tRNA1 tRNA2
α -20,52 -20,51 -40,59 -44,78
β 31,41 31,40 22,39 12,07
γ 20,76 20,74 53,30 51,66
δ 140,19 140,19 85,31 85,18
ε -136,21 -136,22 -118,15 -125,43
ζ -88,04 -88,03 -63,15 -53,57
χ -116,43 -116,44 -121,76 -116,62
D 27T (12:K) 28T (12:O) 10G (52:B) 7G (65:B)

DNA average values are more consistent between strands than tRNA ones. This can be explained as DNA usually takes regular double-strand structure, wereas tRNA double-strand structure is variable which leads to unique abilities of its tertiary structure. There is no reason in showing the most deformed nucleotides as it would be llittle use of comparing them to "normal" nucleotides (which also must be determined).

tRNA stems and base pairs

Fig. 5. tRNA stems and respective base pairs.
Red — acceptor stem, yellow — T-stem, limegreen — D-stem, cyan — anticodon stem, grey — others.

Stems are those elements of RNA secondary structure, which look similar to DNA in terms of connection: consequative nucleotides are connected with hydrogen bonds to the other strand. In given tRNA 4 stems were identified (fig. 5). As tRNA plays a highly conserved role, it stems are named due to their functional or sequence features.

Hydrogen bonds in stems are not the only type of connection between nucleotydes which sustain tRNA secondary and tertiary structures. Those for particular tRNA are given in figure 6.

Fig. 6. Non-stem base interactions in tRNA extracted from *.out file.
12   (0.007) ....>B:..54_:[..U]U-**--A[..A]:..58_:B<.... (0.003)     |
13   (0.010) ....>B:..55_:[..U]U-**+-G[..G]:..18_:B<.... (0.010)     x
25   (0.004) ....>B:..13_:[..A]A-**+-A[..A]:..45_:B<.... (0.006)     |
26   (0.004) ....>B:..14_:[..A]A-**--A[..A]:..21_:B<.... (0.008)     |
27   (0.006) ....>B:..15_:[..G]G-**+-C[..C]:..48_:B<.... (0.011)     x
28   (0.014) ....>B:..19_:[..G]G-----C[..C]:..56_:B<.... (0.005)     +
29   (0.003) ....>B:..46_:[..U]U-**+-U[..U]:..47_:B<.... (0.007)     +

tRNA is rich with non-Watson-Crick base pairs. All of them are shown in figure 7A.

Fig. 7A. Non-Watson-Crick base pairs in tRNA extracted from *.out file.
12   (0.007) ....>B:..54_:[..U]U-**--A[..A]:..58_:B<.... (0.003)     |
13   (0.010) ....>B:..55_:[..U]U-**+-G[..G]:..18_:B<.... (0.010)     x
14   (0.009) ....>B:..37_:[..A]A-**--U[..U]:..33_:B<.... (0.004)     |
15   (0.007) ....>B:..38_:[..U]U-**--U[..U]:..32_:B<.... (0.005)     |
21   (0.005) ....>B:..44_:[..C]C-**--A[..A]:..26_:B<.... (0.005)     |
25   (0.004) ....>B:..13_:[..A]A-**+-A[..A]:..45_:B<.... (0.006)     |
26   (0.004) ....>B:..14_:[..A]A-**--A[..A]:..21_:B<.... (0.008)     |
27   (0.006) ....>B:..15_:[..G]G-**+-C[..C]:..48_:B<.... (0.011)     x
29   (0.003) ....>B:..46_:[..U]U-**+-U[..U]:..47_:B<.... (0.007)     +

Three non-canonical base pairs (12, 14 and 27) are comprised of "canonical" bases. However, interaction are non-Watson-Crick, which is shown in fig. 7B.

Fig. 7B. Non-canonical base pairs with common sets of bases.

Putative stacking interactions

Stacking is also the interaction which builds nucleic acid structure. 3DNA allows to quantify square of overlap area of two consequative base pairs. The most and the least stacked pairs were found, images were obtained with ex_str and stack2img programs. Also, the position of pairs was observed in Jmol. Figure 8 presents the composite picture.

Fig. 8. 3DNA and Jmol visualization of sample stacked pairs in tRNA.
A, B: the most stacked pairs (12,26 Å2) G43≡C27 and C44=A26; C, D: the least stacked (0,00 Å2) pairs G19≡C56 and U46=U47.

As it is seen, the most stacked pairs are indeed placed close and overlap hugely. The least stacked pairs are placed relatively far from each other and not parallel.