Timothy (Phleum pretense L.) is
extraordinary roughage because it is rich in fiber promoting performance of
racing horses and helping dairy cows maintain high milk production. As the only
cultivated variety registered in 1990 in China, Phleum pratense L. cv.
Minshan (Minshan) is highly adapted
to the climate of Minshan County, Gansu
Province (Cao 2003).
The specific cold and moist climate in
turn limits its
seed production and breeding. In the last twenty years, the production and
promotion of Minshan has been
hindered due to variety degeneration (Du 2003). Variety improvement and new
cultivars of high quality and yield are eagerly demanded to meet the needs of
the emerging racing horse industry (Wang et
al. 2018a).
Information
carried on chloroplast genome (cp genome) has been widely applied to gene
mapping, variety identification, plant barcode sequences screening, population
genetics, gene diversity study and molecular assisted breeding (Parks et al. 2009). Codon is the key link
connecting nucleic acids and proteins and plays a vital role in biological genetic information transmission. Among 20 amino acids
forming proteins in organisms, except for methionine and tryptophan which are
encoded by unique codons, other 18 amino acids correspond to 2–6 synonymous
codons. Synonymous codons are used differently in organisms even in different
genes in one genome and different parts of one gene, which is called synonymous
codon usage bias (SCUB) (Li et al. 2012). SCUB is an important feature of organism evolution
and exists in numerous living organisms (Sau et al. 2006; Parks et al.
2009; Chen et al. 2014; Li et al. 2019; Zhou et al. 2019). SCUB analyses enable the scientific community to
increase target gene expression genetically, make the exogenous genes more
efficient and stable as well as variety improvement (Li et al. 2019).
The cp genome of Minshan was
assembled and reported using Illumina pair-end sequencing data (Cui et al. 2019). To comprehensively
understand the architectures of Minshan cp genomes and provide useful
information for molecular assisted breeding, synonymous codon usage of Minshan
was studied in this study.
Materials
and Methods
Sequence
data
Healthy fresh leaves were sampled
from Lanzhou Scientific Observation and Experiment Field Station of the
Ministry of Agriculture for Ecological System in the Loess Plateau Area
(36°01′N, 103°45′E, altitude 1700 m), Gansu, China, on July 22th
2019. The complete cp genome sequencing of Minshan was determined (Genbank
accession number: MN551180) based on Illumina NovaSeq platform at Benagen Tech
Solution Co., Ltd. (Wuhan, China). GeSeq was employed to annotate the assembled
genome (Tillich et al. 2017). After
filtering the repeated sequences and the sequences length less than 300 bp, 49
sequences with the start codon of ATG, TTG, CTG, ATT, ATC, GTG and ATA, also
the end codon of TGA, TAG, and TAA, were used to carry on the subsequent
analysis.
Relative synonymous codon usage
A great deal of codon usage indices were acquired via the program Codon W
(version 1.3, https://sourceforge.net/projects/codonw/), including the relative synonymous codon usage (RSCU) value , the codon
adaptation index (CAI), the effective number of codons (ENC), the nucleotides G
and C content of all 49 coding sequences (GC), the frequency of G + C at the
third position of synonymous codons (GC3s), and the silent base
compositions (A3s, T3s, G3s, and C3s)
(Li et al. 2019). The G+C content at
the first, second, third positions of codons (GC1, GC2,
GC3) and the average G+C content of the first and second positions
(GC12) were calculated by the online CUSP function from EMBOSS (http://imed.med.ucm.es/EMBOSS/) (Wang et al. 2018b).
Identification of the optimal codon
According to RSCU value of each codon, the highest frequency synonymous
codons with the largest RSCU value were identified (Li et al. 2019). Using ENC analysis as preference standard, the 49
sequences of Minshan were ordered from high to low, and the highest 5%
sequences and the lowest 5% sequences were taken to form the high and low
expression gene group, separately. ΔRSCU was subtracted the RSCU value of
each codons in the low expressed gene group from the high expressed gene group.
The codons with ΔRSCU value larger than 0.08 were recognized as high
expressed. The optimal codons were identified as the ones which were high
frequency and high expressed (Wang et al.
2019).
Correspondence analysis
Correspondence analysis is a vital multivariate tool to explore codon
usage change trends (Choudhury et al.
2017). The corresponding analysis of genes and codon bias was carried out by
Codon W based on the RSCU values. RSCU values of 49 coding sequences in Minshan
were spread into a 40-dimensional vector space. The axes were related to the
influencing factors on SCUB and the data of different axes were obtained
according to codon base bias and genes. The correlation analysis among GC, ENC,
CAI, G3s, GC3, Axis 1, Axis 2, Axis 3 and Axis 4 was
accomplished through SPSS 16.0 based on the Spearman’s rank correlation method
(P < 0.05 or P < 0.01). The graphs were depicted using EXCEL 2016.
ENC‑plot analysis
ENC displays the degree of codons deviated from random selection (Li et al. 2019). ENC value ranges from 20
to 61 with the boundary value of 35 and the ENC value less than 35 means strong
codon preference, otherwise, weak codon preference takes place. There is
extreme preference when the ENC value is 20. In contrast, there is no preference
with the ENC value of 61, indicating random selection of codons (Wang et al. 2018b). ENC-plot mapping analysis
is used to explore the dominating reason affecting the SCUB. The ENC plot of
ENC versus GC3s was drawn by EXCEL 2016. The ENC formula of expected
curve is as follows:
%20352-358_files/image002.png){kind=small}
The genes would be distributed alongside or next to the expected curve
when SCUB is merely involved by mutation, however if the points of genes are far away from the expected curve, it
illustrates SCUB is primarily affected by natural selection other than mutation
pressure (Wang et al. 2018b).
Neutrality plot analysis
GC
content of the cp genome is highly conserved, while under the condition of
various evolution pressures, different bases preferences would happen, and
synonymous codon mutation usually occurs at the third position. In the neutral
graph, the GC12 value of each gene is used as vertical coordinate,
and the corresponding GC3 value is used as horizontal axis (Wei et al. 2014). If the point distributes
alongside or nearby the diagonal line which means that GC12 is equal
to GC3, it implies that other evolution pressure excluding mutation
pressure is weak. Oppositely, if GC12 and GC3 correlates
weakly, the regression coefficient is approximately to 0, it shows that the base
composition of the 3 positions are significantly different, revealing that
natural selection is the dominating factor affecting SCUB (He et al. 2013).
Statistical analysis
Correlation analysis was conducted by SPSS 16.0 employing the correlation
method of Spearman’s rank (P < 0.05 or P < 0.01). The graphs were depicted using EXCEL 2016.
Results
The codon usage pattern of Minshan
Nucleotide A and
T were abundant in Minshan cp genome. The average GC,
GC1, GC2 and GC3 content of Minshan cp genome
were 38, 47, 39 and 30%. The frequencies of A3s,
T3s, G3s, C3s, and GC3s were 43, 46, 17, 17 and 27%, respectively. The
length of 49 amino acids was between 101 and 1473 with the average of 338. The
%20352-358_files/image004.jpg)
Fig. 1: Contributions
of 40 axes from a correspondence analysis
%20352-358_files/image006.jpg)
Fig. 2: Correspondence
analysis of synonymous codon usage towards the codons in Minshan cp genome
%20352-358_files/image008.jpg)
Fig. 3: Correspondence
analysis of synonymous codon usage towards the coding genes in Minshan cp genome
value of ENC was between 39.91 and 60.65 with an average of 49.44, implying weak preference of synonymous usage. All CAI values of these 49 sequences
were less than 0.35, which additionally demonstrated the weak preference of
synonymous usage (Table 1).
According to the RSCU values of 61
codons in Minshan cp genome, 18 high frequency
synonymous codons with the largest RSCU value, were observed (Table 2). 26
codons were identified as the high expressed codons (Table 3). 10 codons
including AGA, TGT, TTT, TAT, TTA, CAA, CAT, GCT, GAA and GTT were located in
the intersection set of high frequency codons and high expressed codons and
they were identified as the optimal codons, of which, 6 ended with T and 4
ended with A, implying the synonymous codons is biased in using A and T ended
codons in Minshan cp genome.
Correspondence analysis of SCUB
Synonymous codons and the cluster of 49 coding genes in Minshan were
characterized by different color points in 40 dimensional axes and went through
the correspondence analysis (Fig. 1, 2 and 3). The first four axes comprised
33.35% of the whole variation and Axis 1 and Axis 2 were the two chief
contributors to SCUB in Minshan, which accounted for 9.18 and 8.85% of the
total variation, respectively (Fig. 1). The codons ending with G or C
distributed dispersedly and away from Axis 1 and Axis 2 while A or T ended
codons were closer to Axis 1 (Fig. 2), indicating the nucleotide
constitution for mutation pressure may associate with SCUB. Different gene
types showed different distribution patterns (Fig. 3). The genes of photosystem
I and photosystem II distributed in the third and fourth quadrant, and rubisCO
large subunit distributed next to Axis 1. For the genes of Cytochrome b/f
complex, the petA and petD distributed in the second and third
quadrant, petB was close to Axis 1. Moreover, rpoB and ropC1 points
were also close to Axis 1. However, ATP synthase, NADH dehydrogenase, Ribosomal
proteins (LSU) and Ribosomal proteins (SSU) distributed discretely, indicating
that other SCUB influencing factors such as natural selection may work.
Correlation analysis
To fully explore the SCUB affected by evolution
pressure of mutation or natural selection, correlation among GC, ENC, CAI, GC3s,
GC3, Axis 1, Axis 2, Axis 3, and Axis 4 were calculated (Table 4).
Axis 1 showed significant positive correlation with ENC, GC3s and GC3
(r = 0.415**, P < 0.01; r =
0.556**, P < 0.01; r = 0.599**, P < 0.01) while Axis 2 exhibited
significant positive correlation with ENC (r = 0.496**, P < 0.01) but significant negative Table 1: Codon usage indices of 49 coding genes
in Minshan cp genome
|
Gene |
GC |
GC3 |
GC3s |
CAI |
ENC |
Gene |
GC |
GC3 |
GC3s |
CAI |
ENC |
|
atpA |
0.41 |
0.29 |
0.28 |
0.18 |
48.20 |
psaB |
0.41 |
0.32 |
0.28 |
0.17 |
47.90 |
|
atpB |
0.42 |
0.30 |
0.29 |
0.20 |
46.20 |
psbA |
0.42 |
0.33 |
0.29 |
0.31 |
39.91 |
|
atpE |
0.41 |
0.34 |
0.31 |
0.18 |
57.47 |
psbB |
0.43 |
0.28 |
0.25 |
0.20 |
47.04 |
|
atpF |
0.38 |
0.31 |
0.28 |
0.15 |
49.63 |
psbC |
0.44 |
0.34 |
0.30 |
0.19 |
49.17 |
|
atpI |
0.39 |
0.32 |
0.29 |
0.16 |
48.40 |
psbD |
0.44 |
0.34 |
0.30 |
0.24 |
49.69 |
|
ccsA |
0.35 |
0.28 |
0.23 |
0.14 |
47.33 |
rbcL |
0.44 |
0.30 |
0.27 |
0.29 |
48.10 |
|
cemA |
0.34 |
0.30 |
0.27 |
0.16 |
54.91 |
rpl14 |
0.39 |
0.23 |
0.22 |
0.18 |
48.68 |
|
clpP |
0.42 |
0.36 |
0.30 |
0.18 |
52.22 |
rpl16 |
0.45 |
0.27 |
0.22 |
0.13 |
44.40 |
|
infA |
0.40 |
0.39 |
0.37 |
0.19 |
61.00 |
rpl20 |
0.36 |
0.31 |
0.27 |
0.12 |
52.92 |
|
matK |
0.33 |
0.28 |
0.25 |
0.17 |
48.09 |
rpl22 |
0.36 |
0.33 |
0.30 |
0.17 |
49.89 |
|
ndhA |
0.34 |
0.33 |
0.20 |
0.13 |
44.95 |
rpoA |
0.36 |
0.27 |
0.26 |
0.14 |
50.83 |
|
ndhB |
0.38 |
0.33 |
0.30 |
0.16 |
46.76 |
rpoB |
0.40 |
0.31 |
0.29 |
0.15 |
50.32 |
|
ndhC |
0.41 |
0.36 |
0.30 |
0.17 |
59.32 |
rpoC1 |
0.40 |
0.31 |
0.29 |
0.16 |
51.21 |
|
ndhD |
0.36 |
0.30 |
0.26 |
0.13 |
47.86 |
rpoC2 |
0.39 |
0.31 |
0.30 |
0.16 |
52.51 |
|
ndhE |
0.33 |
0.28 |
0.25 |
0.13 |
57.52 |
rps11 |
0.44 |
0.24 |
0.22 |
0.18 |
43.39 |
|
ndhF |
0.34 |
0.26 |
0.21 |
0.15 |
46.66 |
rps14 |
0.39 |
0.31 |
0.28 |
0.14 |
49.68 |
|
ndhG |
0.35 |
0.25 |
0.21 |
0.13 |
48.32 |
rps18 |
0.32 |
0.26 |
0.24 |
0.15 |
46.59 |
|
ndhH |
0.38 |
0.29 |
0.24 |
0.15 |
49.59 |
rps2 |
0.37 |
0.31 |
0.27 |
0.17 |
47.85 |
|
ndhI |
0.35 |
0.27 |
0.25 |
0.17 |
44.98 |
rps3 |
0.34 |
0.27 |
0.26 |
0.20 |
48.62 |
|
ndhJ |
0.39 |
0.30 |
0.26 |
0.16 |
50.46 |
rps4 |
0.37 |
0.25 |
0.23 |
0.16 |
48.07 |
|
ndhK |
0.38 |
0.31 |
0.27 |
0.16 |
52.06 |
rps7 |
0.40 |
0.24 |
0.21 |
0.16 |
48.31 |
|
petA |
0.41 |
0.32 |
0.31 |
0.17 |
49.09 |
rps8 |
0.36 |
0.26 |
0.23 |
0.11 |
44.14 |
|
petB |
0.40 |
0.30 |
0.23 |
0.20 |
43.25 |
ycf3 |
0.41 |
0.48 |
0.44 |
0.14 |
60.65 |
|
petD |
0.40 |
0.31 |
0.28 |
0.16 |
48.83 |
ycf4 |
0.41 |
0.34 |
0.30 |
0.17 |
47.57 |
|
psaA |
0.43 |
0.34 |
0.30 |
0.19 |
51.85 |
Average |
0.39 |
0.30 |
0.27 |
0.17 |
49.44 |
Table 2: Codon usage in Minshan cp genome
|
Amino acid |
Codon |
Number |
RSCU |
Amino acid |
Codon |
Number |
RSCU |
|
Ala (A) |
GCT |
460 |
1.75 |
Asn (N) |
AAT |
485 |
1.49 |
|
GCC |
146 |
0.56 |
AAC |
165 |
0.51 |
||
|
GCA |
323 |
1.23 |
Pro (P) |
CCT |
274 |
1.53 |
|
|
GCG |
120 |
0.46 |
CCC |
165 |
0.92 |
||
|
Cys (C) |
TGT |
128 |
1.48 |
CCA |
192 |
1.07 |
|
|
TGC |
45 |
0.52 |
CCG |
86 |
0.48 |
||
|
Asp (D) |
GAT |
485 |
1.53 |
Gln (Q) |
CAA |
442 |
1.52 |
|
GAC |
150 |
0.47 |
CAG |
140 |
0.48 |
||
|
Glu (E) |
GAA |
654 |
1.45 |
Arg (R) |
CGT |
227 |
1.39 |
|
GAG |
245 |
0.55 |
CGC |
94 |
0.58 |
||
|
Phe (F) |
TTT |
634 |
1.35 |
CGA |
208 |
1.28 |
|
|
TTC |
307 |
0.65 |
CGG |
65 |
0.4 |
||
|
Gly (G) |
GGT |
388 |
1.27 |
AGA |
275 |
1.69 |
|
|
GGC |
131 |
0.43 |
AGG |
108 |
0.66 |
||
|
GGA |
486 |
1.59 |
Ser (S) |
TCT |
330 |
1.66 |
|
|
GGG |
221 |
0.72 |
TCC |
223 |
1.12 |
||
|
His (H) |
CAT |
273 |
1.46 |
TCA |
201 |
1.01 |
|
|
CAC |
102 |
0.54 |
TCG |
101 |
0.51 |
||
|
Ile (I) |
ATT |
696 |
1.51 |
AGT |
255 |
1.28 |
|
|
ATC |
253 |
0.55 |
AGC |
85 |
0.43 |
||
|
ATA |
438 |
0.95 |
Thr (T) |
ACT |
382 |
1.74 |
|
|
Lys (K) |
AAA |
605 |
1.48 |
ACC |
162 |
0.74 |
|
|
AAG |
212 |
0.52 |
ACA |
241 |
1.1 |
||
|
Leu (L) |
TTA |
638 |
2.11 |
ACG |
93 |
0.42 |
|
|
TTG |
333 |
1.1 |
Val (V) |
GTT |
364 |
1.52 |
|
|
CTT |
389 |
1.29 |
GTC |
115 |
0.48 |
||
|
CTC |
119 |
0.39 |
GTA |
356 |
1.49 |
||
|
CTA |
251 |
0.83 |
GTG |
120 |
0.5 |
||
|
CTG |
86 |
0.28 |
Tyr (Y) |
TAT |
473 |
1.55 |
|
|
Met (M) |
ATG |
386 |
1 |
TAC |
139 |
0.45 |
|
|
Trp (W) |
TGG |
311 |
1 |
The highest frequency used
synonymous codons (the largest RSCU value) are in bold
RSCU, relative synonymous codon usage
correlation with CAI (r =-0.424**, P
< 0.01) and Axis 3 negatively correlated with GC and CAI significantly
(r = -0.410**, P < 0.01; r =
-0.362*, P < 0.05), moreover Axis
4 showed no significant correlation with other indices, suggesting Axis 1 and
Axis 3 were the major contributors for codon nucleotide constitution variation,
and Axis 1, Axis 2 and Axis 3 all contributed to SCUB. GC3s
correlated with GC3, ENC and Axis 1 significantly (r = 0.896**, P < 0.01; r = 0.614**, P < 0.01; r = 0.556**, P < 0.01), implying that the codon
nucleotide base constitution for the pressure of mutation may affect SCUB. CAI
positively correlated with GC (r = 0.515**, P
< 0.01) and negatively correlated with Axis 2 and Axis 3 (r = 0.424**, P < 0.01; r = 0.362*, P < 0.01), indicating natural
selection may play a considerable role in SCUB.
Table 3: The codon statistics within high and low expressed genes
and ΔRSCU value for each codon in Minshan cp genome
|
Amino acid |
Codon |
High expressed gene |
Low expressed gene |
ΔRSCU |
Amino acid |
Codon |
High expressed gene |
Low expressed gene |
ΔRSCU |
||||
|
Frequency |
RSCU |
Frequency |
RSCU |
Frequency |
RSCU |
Frequency |
RSCU |
||||||
|
Ala (A) |
GCT* |
11 |
1.63 |
0 |
0.80 |
0.83 |
Asn (N) |
AAT |
8 |
1.33 |
9 |
1.50 |
-0.17 |
|
GCC |
4 |
0.59 |
5 |
2.00 |
-1.41 |
AAC* |
4 |
0.67 |
3 |
0.50 |
0.17 |
||
|
GCA |
8 |
1.19 |
3 |
1.20 |
-0.01 |
Pro (P) |
CCT |
3 |
0.80 |
5 |
1.43 |
-0.63 |
|
|
GCG* |
4 |
0.59 |
0 |
0.00 |
0.59 |
CCC* |
5 |
1.33 |
1 |
0.29 |
1.04 |
||
|
Cys (C) |
TGT* |
2 |
1.33 |
3 |
1.20 |
0.13 |
CCA |
3 |
0.80 |
5 |
1.43 |
-0.63 |
|
|
TGC |
1 |
0.67 |
2 |
0.80 |
-0.13 |
CCG* |
4 |
1.07 |
3 |
0.86 |
0.21 |
||
|
Asp (D) |
GAT |
5 |
1.43 |
5 |
1.43 |
0.00 |
Gln (Q) |
CAA* |
6 |
1.50 |
7 |
1.40 |
0.10 |
|
GAC |
2 |
0.57 |
2 |
0.57 |
0.00 |
CAG |
2 |
0.50 |
3 |
0.60 |
-0.10 |
||
|
Glu (E) |
GAA* |
15 |
1.67 |
11 |
1.29 |
0.38 |
Arg (R) |
CGT* |
6 |
1.24 |
5 |
1.07 |
0.17 |
|
GAG |
3 |
0.33 |
6 |
0.71 |
-0.38 |
CGC |
2 |
0.41 |
5 |
1.07 |
-0.66 |
||
|
Phe (F) |
TTT* |
5 |
2.00 |
9 |
0.90 |
1.10 |
CGA* |
9 |
1.86 |
5 |
1.07 |
0.79 |
|
|
TTC |
0 |
0.00 |
11 |
1.10 |
-1.10 |
CGG |
0 |
0.00 |
5 |
1.07 |
-1.07 |
||
|
Gly (G) |
GGT |
7 |
1.40 |
9 |
1.50 |
-0.10 |
AGA* |
11 |
2.28 |
4 |
0.86 |
1.42 |
|
|
GGC* |
4 |
0.80 |
2 |
0.33 |
0.47 |
AGG |
1 |
0.21 |
4 |
0.86 |
-0.65 |
||
|
GGA |
7 |
1.40 |
9 |
1.50 |
-0.10 |
Ser (S) |
TCT |
0 |
0.00 |
8 |
1.92 |
-1.92 |
|
|
GGG |
2 |
0.40 |
4 |
0.67 |
-0.27 |
TCC* |
6 |
2.57 |
5 |
1.20 |
1.37 |
||
|
His (H) |
CAT* |
2 |
2.00 |
0 |
0.00 |
2.00 |
TCA |
1 |
0.43 |
5 |
1.20 |
-0.77 |
|
|
CAC |
0 |
0.00 |
2 |
2.00 |
-2.00 |
TCG |
0 |
0.00 |
2 |
0.48 |
-0.48 |
||
|
Ile (I) |
ATT |
11 |
1.18 |
22 |
1.69 |
-0.51 |
AGT* |
6 |
2.57 |
3 |
0.72 |
1.85 |
|
|
ATC* |
5 |
0.54 |
6 |
0.46 |
0.08 |
AGC |
1 |
0.43 |
2 |
0.48 |
-0.05 |
||
|
ATA* |
12 |
1.29 |
11 |
0.85 |
0.44 |
Thr (T) |
ACT |
3 |
1.00 |
5 |
1.54 |
-0.54 |
|
|
Lys (K) |
AAA* |
13 |
1.63 |
10 |
0.95 |
0.68 |
ACC |
1 |
0.33 |
4 |
1.23 |
-0.90 |
|
|
AAG |
3 |
0.38 |
11 |
1.05 |
-0.67 |
ACA* |
4 |
1.33 |
3 |
0.92 |
0.41 |
||
|
Leu (L) |
TTA* |
6 |
1.57 |
4 |
0.71 |
0.86 |
ACG* |
4 |
1.33 |
1 |
0.31 |
1.02 |
|
|
TTG |
3 |
0.78 |
8 |
1.41 |
-0.63 |
Val (V) |
GTT* |
6 |
1.60 |
4 |
1.00 |
0.60 |
|
|
CTT |
6 |
1.57 |
13 |
2.29 |
-0.72 |
GTC |
1 |
0.27 |
3 |
0.75 |
-0.48 |
||
|
CTC* |
2 |
0.52 |
1 |
0.18 |
0.34 |
GTA* |
7 |
1.87 |
6 |
1.50 |
0.37 |
||
|
CTA |
2 |
0.52 |
6 |
1.06 |
-0.54 |
GTG |
1 |
0.27 |
3 |
0.75 |
-0.48 |
||
|
CTG* |
4 |
1.04 |
2 |
0.35 |
0.69 |
Tyr (Y) |
TAT* |
5 |
2.00 |
6 |
1.00 |
1.00 |
|
|
Met (M) |
ATG |
7 |
1.00 |
10 |
1.00 |
0.00 |
TAC |
0 |
0.00 |
6 |
1.00 |
-1.00 |
|
|
Trp (W) |
TGG |
7 |
1.00 |
9 |
1.00 |
0.00 |
|||||||
RSCU, relative synonymous codon usage
* indicates the high expression codons (ΔRSCU>0.08)
Table 4: Correlation coefficients of the indices influencing codon bias in Minshan
cp genome
|
Indices |
GC |
ENC |
CAI |
GC3s |
GC3 |
Axis 1 |
Axis 2 |
Axis 3 |
Axis 4 |
|
GC |
1 |
||||||||
|
ENC |
-0.017 |
1 |
|||||||
|
CAI |
0.515** |
-0.227 |
1 |
||||||
|
GC3s |
0.356* |
0.614** |
0.176 |
1 |
|||||
|
GC3 |
0.334* |
0.565** |
0.127 |
0.896** |
1 |
||||
|
Axis 1 |
0.025 |
0.415** |
0.148 |
0.556** |
0.599** |
1 |
|||
|
Axis 2 |
0.086 |
0.496** |
-0.424** |
0.199 |
0.178 |
0.007 |
1 |
||
|
Axis 3 |
-0.410** |
-0.066 |
-0.362* |
-0.03 |
-0.061 |
-0.006 |
0.001 |
1 |
|
|
Axis 4 |
0.055 |
0.251 |
-0.064 |
0.247 |
0.28 |
0.011 |
0.008 |
-0.009 |
1 |
**correlation is significant at the 0.01 level.
*correlation is
significant at the 0.05 level.
ENC plot analysis
Most points of the total 49 genes in Minshan cp genome distributed
discretely (Fig. 4). The points of clpP,
ndhE, rp114, rps18, rps4 and ycf located on and the points of ndhH, ndhI, ndhK and ccsA were close to the expected curve, indicating mutation pressure
was the major factor affecting their SCUB. Meanwhile, the rest of genes could
be divided into two groups, the points of infA,
ndhJ and clpP located above the expected curve and other genes were below
the expected curve, both of which were apart from the expected curve implying
natural selection affected their SCUB momentously in the terms of warranting
the most effective use of codons.
Neutrality plot analysis
%20352-358_files/image010.jpg)
Fig. 4: ENC-plot
analysis of Minshan cp genome. ENC, effective number of codons. GC3s,
the frequencies of nucleotide G + C at the third position of synonymous codons.
The curve shows the expected relationship between ENC values and GC3s under
random codon usage assumption
%20352-358_files/image012.jpg)
Fig. 5:
Neutrality plot analysis of Minshan cp genome. GC12,
the average frequencies of nucleotide G + C at the first and second positions
of synonymous codons. GC3, the frequencies of nucleotide G + C at
the third position of synonymous codons. The curve shows that GC12
is equal to GC3
It is an effective way to study the degree of mutation pressure against
natural selection in SCUB in cp genome employing the method of neutrality plot
analysis. The point of infA was
diagonally distributed (Fig. 5), suggesting no significant difference existed
among GC1, GC2 and GC3. Besides, GC3 correlated
negatively with GC12 in other 48 coding sequences of Minshan cp
genome, and the correlation was very little (r = -0.1133). The results showed
that natural selection influenced the SCUB for 48 coding sequence in Minshan
expect for infA which was mainly
affected by the pressure of mutation.
Discussion
Minshan is the only timothy cultivar in China, and severe variety degradation
has restricted its promotion and production. Research on SCUB in Minshan cp
genome could help reveal its biological architectures, gene evolution and
assist molecular breeding in further study. SCUB affects the speed and
efficiency of mRNA translation and the folding characteristics of polypeptide
chain (Brule and Grayhack 2017; Hu et al. 2019). SCUB is quite different according to different species, tissues and
genes (Qiu et al. 2011; He et al. 2013; Chakraborty et al. 2017; Paulet et al.
2017; Zhang et al. 2018; Cai et al. 2019; Hu et al. 2019). Among numerous affecting
factors, mutation pressure and natural selection are of great importance
(Prabha et al. 2017). SCUB is the
result of long-term competition between the nucleotide constitution for
mutation pressure and natural selection. Research on SCUB in Minshan also could
conduce to find out the main influencing factor in its evolution and advance
the understanding of the balance between them (Sharp and Li 1987; Olejniczak
and Uhlenbeck 2006; Zalucki et al.
2007; Wang et al. 2018b).
In this study, nucleotide composition in Minshan cp genome was abundant in A or T, showing A
or T bias. 10 optimal synonymous codons were identified and of which, 6 ended
with A and 4 ended with T. Moreover, the correspondence analysis reflected that A or T ended codons were closer
to Axis 1 than others. The nucleotide preference in Minshan may be related to the relative
evolutionary conservation of cp genome (Hu et
al. 2019). Nucleotide A or T preference in SCUB was in line with earlier
studies in Oncidium gower ramsey
(Chen et al. 2011), seed plants (Meng
et al. 2008), Lonicera japonica (He et al.
2017), solanum (Zhang et al. 2018)
and Elaeagnus angustifolia (Wang et al. 2019), indicating that the codon bias may correlate to the base composition for mutation
pressure.
ENC is an important index to reflect
the preference degree of unequal use of synonymous codons (Gupta et al. 2004). In Minshan, ENC values ranged from 39.91 to
60.65 with the average of 49.44, implying the synonymous codon usage bias was
weak. On ENC plot, only the points of clpP, ndhE, rp114, rps18, rps4 and ycf were located on and the
points of ndhH, ndhI, ndhK and ccsA were close to the expected curve,
others presented a discrete distribution. Additionally, except for infA which diagonally distributed in neutrality plot, GC3 and GC12 correlated weakly in
other 48 genes in Minshan cp genome.
All of the results suggest SCUB in Minshan is predominantly influenced by
natural selection and different genes have different evolutionary pressure (Mukhopadhyay
et al. 2008; Li et al. 2019; Wang et al.
2019).
Conclusion
The research is the first one which systematically analyzes codon usage
pattern in Minshan cp genome and
comprehensively explores the influencing factors on SCUB. Weak preference of
synonymous usage in Minshan exists, and nucleotide constitution, mutation
stress and natural selection all have an effect on SCUB, of which natural
selection is the major contributor. It still needs further study to clarify
whether natural selection has effect on the evolution of functional genes in
Minshan cp genome. The results exhibit the architectures of Minshan cp genomes
and afford useful information for codon modification and molecular assisted
breeding for further study in the future.
Acknowledgements
This work was supported by Quality evaluation and pollution-free
production standard system construction of Minshan Timothy (GSZYTC-ZCJC-18027),
the Key Laboratory of Superior Forage
Germplasm in the Qinghai-Tibetan Plateau (2020-ZJ-Y12), and the National
Science Foundation of China (31700338).
References
Brule CE, EJ Grayhack
(2017). Synonymous codons: choose wisely for expression. Trends Genet 33:283‒297
Cai MS, AC Cheng, MS
Wang, LC Zhao, DK Zhu, QH Luo, F Liu, XY Chen (2019). Characterization of
synonymous codon usage bias in the duck plague virus UL35 gene. Intervirology 52:266‒278
Cao ZZ (2003). The cultivation
and production of Phleum pretense in
China. Grassl Chin 6:72‒74
Chakraborty S, D Nag,
TH Mazumder, A Uddin (2017). Codon usage pattern and prediction of gene
expression level in Bungarus species.
Gene 604:48‒60
Chen X, XN Cai, QZ
Chen, HX Zhou, Y Cai, A Ben (2011). Factors affecting synonymous codon usage
bias in chloroplast genome of Oncidium Gower
Ramsey. Evol Bioinform 7:271‒278
Chen Y, YZ Shi, HJ
Deng, T Gu, J Xu, JX Ou, ZG Jiang, YR Jiao, T Zou, C Wang (2014).
Characterization of the porcine epidemic diarrhea virus codon usage bias. Infect Genet Evol 28:95‒100
Choudhury MN, A
Uddin, S Chakraborty (2017). Codon usage bias and its influencing factors for
Y-linked genes in human. Comput Biol Chem
69:77‒86
Cui GX, Y Lu, XX Wei,
XL Wang, CM Wang, YQ Gao, HR Duan (2019). Characterization of the complete chloroplast
genome of Phleum pretense L. cv.
Minshan. Mitochondr DNA 4:4180‒4181
Du WH (2003).
Research advance in nutritive value, cultivation and utilization of Timothy. Grassl Turf 4:7‒11
Gupta SK, TK
Bhattacharyya, TC Ghosh (2004). Synonymous codon usage in Lactococcus lactis: mutational bias versus translational selection.
J Biomol Str Dyn 21:527‒536
He L, J Qian, X Li, Z
Sun, X Xu, S Chen (2017). Complete chloroplast genome of medicinal plant Lonicera japonica: genome rearrangement,
intron gain and loss, and implications for phylogenetic studies. Molecules 22:249–261
He S, SJ Zhang, LS
Xiu, ZB Xing (2013). Codon usage analysis in squalene synthase gene. Genom Appl Biol 32:232‒239
Hu XY, YQ Xu, YZ Han,
SH Du (2019). Codon usage bias analysis of the chloroplast genome of Ziziphus jujube var spinosa. J For Environ
39:621‒628
Li GL, ZL Pan, SC
Gao, YY He, QY Xia, J Yan, HP Yao (2019). Analysis of synonymous codon usage of
chloroplast genome in Porphyra
umbilicalis. Genes Genomics
41:1173‒1181
Li ML, ZY Zhao, JH Chen, BY Wang, Z Li, J Li, MS Cai (2012).
Characterization of synonymous codon usage bias in the pseudorabies virus US1 gene. Virol Sin 27:303‒315
Meng Z, L Wei, L Xia (2008).
Patterns of synonymous codon usage bias in chloroplast genomes of seed plants. For Stud Chin 10:235‒242
Mukhopadhyay P, S Basak, TC Ghosh (2008). Differential
selective constraints shaping codon usage pattern of housekeeping and
tissue-specific homologous genes of rice and Arabidopsis. DNA Res
15:347‒356
Olejniczak M, OC
Uhlenbeck (2006). tRNA residues that have coevolved with their anticodon to
ensure uniform and accurate codon recognition. Biochimie 88:943‒950
Parks M, R Cronn, A
Liston (2009). Increasing phylogenetic resolution at low taxonomic levels using
massively parallel sequencing of chloroplast genomes. BMC Biol 7; Article
84
Paulet D, A David, E
Rivals (2017). Ribo-seq enlightens codon usage bias. DNA Res 24:303‒310
Prabha R, DP Singh, S
Sinha, K Ahmad, A Rai (2017). Genome-wide comparative analysis of codon usage
bias and codon context patterns among cyanobacterial genomes. Mar Genom 32:31‒39
Qiu S, K Zeng, T
Slotte, S Wright, D Charlesworth (2011). Reduced efficacy of natural selection
on codon usage bias in selfing Arabidopsis
and Capsella species. Genome Biol Evol 3:868‒880
Sau K, SK Gupta, S
Sau, SC Mandal, TC Ghosh (2006). Factors influencing synonymous codon and amino
acid usage biases in Mimivirus. Biosystems
85:107‒113
Sharp PM, WH Li (1987).
The rate of synonymous substitution in enterobacterial genes in inversely
related to codon usage bias. Mol Biol Evol
4:222‒230
Tillich M, P Lehwark,
T Pellizzer, ES Ulbricht-Jones, A Fischer, R Bock S Greiner (2017). GeSeq-versatile
and accurate annotation of organelle genomes. Nucl Acids Res 45: 6‒11
Wang HX, ZD Cao, JC
Xiang, P Fu (2018a). Present situation of grass industry and countermeasures of
transformational development in Dingxi. Pratac
Sci 7:1811‒1817
Wang J, TY Wang, LY
Wang, JG Zhang, YF Zeng (2019). Assembling and Analysis of the whole
chloroplast genome sequence of Elaeagnus
angustifolia and its codon usage bias. Acta
Bot Sin 9:1559‒1572
Wang L, H Xing, Y
Yuan, X Wang, M Saeed, J Tao, W Feng, G Zhang, X Song, X Sun (2018b).
Genome-wide analysis of codon usage bias in four sequenced cotton species. PLoS One 13; Article e0194372
Wei L, J He, X Jia, Q
Qi, ZS Liang, H Zheng, Y Ping, SY Liu, JC Sun (2014). Analysis of codon usage
bias of mitochondrial genome in Bombyx
mori and its relation to evolution. BMC
Evol Biol 14:262-273
Zalucki YM, PM Power,
MP Jennings (2007). Selection for efficient translation initiation biases codon
usage at second amino acid position in secretory proteins. Nucl Acids Res 35:5748‒5754
Zhang R, L Zhang, W
Wang, Z Zhang (2018). Differences in codon usage bias between photosynthesis-related genes and
genetic system-related genes of chloroplast genomes in cultivated and wild solanum species. Intl J Mol Sci
19:3124–3148
Zhou CL, LY Peng, X Wang, JL Chen, L Wang, H Chen, ZX Lai,
SC Liu (2019). Codon bias and evolution analysis of AtGAI in Amaranthus tricolor
L. J Chin Agric Univ 24:10‒22