T lymphocyte alteration in CHB patients
Compared with the healthy controls, the percentages of CD4+T and CD8+T cells in CHB patients were lower (P<0.01). The percentages of Th17, Treg, ratio of Th17/Treg and Tfh cells in CHB patients were significantly higher than the healthy controls (P<0.01). But there was no statistically significant difference in ratio of CD4+T /CD8+T between the two groups (P>0.05) (Table1 and Fig. 1).
Table 1 Difference statistics of T lymphocyte subsets (n=50, Mean ±SD)
T lymphocytes
|
|
CHB (%)
|
Health (%)
|
Value P
|
CD4+T
|
|
31.17±6.59
|
36.58±4.76
|
0.007
|
CD8+T
|
|
24.18±7.54
|
32.07±6.27
|
0.001
|
CD4+T /CD8+T
|
|
1.37±0.35
|
1.18±0.30
|
0.092
|
Th17
|
|
11.07±4.83
|
0.49±0.24
|
0.000
|
Treg
|
|
9.76±2.07
|
4.67±3.08
|
0.000
|
Th17/Treg
|
|
1.18±0.57
|
0.20±0.22
|
0.000
|
Tfh
|
|
11.22±3.18
|
8.67±1.82
|
0.001
|
The pathway proteins alteration in CHB patients
Compared with the healthy controls, the serum levels of HMGB1, PI3K, PDK1, and Akt in CHB patients were significantly higher (P<0.01). No difference in PTEN was found in CHB patients and healthy controls (P>0.05) (Table 2 and Fig. 2).
Table 2 Statistic of HMGB1-PTEN pathway protein level (n=50, Mean ±SD)
HMGB1-PTEN pathway
|
CHB
|
Health
|
value P
|
HMGB1
|
0.7053±0.0040
|
0.6699±0.0106
|
0.000
|
PTEN
|
0.2348± 0.0823
|
0.2759±0.0022
|
0.376
|
PI3K
|
0.3287±0.0024
|
0.2795±0.0054
|
0.000
|
PDK1
|
0.5058±0.0003
|
0.4535±0.0016
|
0.000
|
Akt
|
1.3899±0.0043
|
0.7762±0.0049
|
0.000
|
HMGB1-PTEN pathway protein network
The protein interaction network diagram was constructed through the STRING database, and then the relationship between HMGB1-PTEN pathway was analyzed. As shown in Fig. 3, we found that Akt was the core factor, which connect the HMGB1 with other proteins (PTEN, PI3K, PDK1). The connection involved varied aspects such as literature mining, experiments, databases, co-expression, gene neighborhood, gene fusion, co-occurrence, etc. The data showed that the combined score of evidence suggesting a functional link among PI3K, PDK1 and Akt were above 0.8. The combined scores among PTEN, PI3K and Akt were above 0.9. The results suggested that HMGB1, PTEN, PI3K, PDK1 and Akt had close relationship.
The GeneMANIA database was used to further analyze the functional expression correlation between HMGB1, PTEN, PI3K, PDK1, and Akt. As shown in Fig. 4, HMGB1, PTEN, PI3K, PDK1, and Akt mainly interacted through physical interactions (67.64%), co-expression (13.50%), co-localization (6.17%), pathway-mediated (4.35%) and other aspects. Depending on Akt, these proteins (HMGB1, PTEN, PI3K, and PDK1) were linked. The activated relationship among PI3K, PDK1and Akt, and the negative relationship among PTEN, PI3K and Akt have been found. HMGB1 and PDK1are co-expressed, suggesting that there may be a positive regulatory relationship between HMGB1 and PI3K, PDK1 and Akt
The correlation analysis between HMGB1-PTEN pathway and immune cell
In order to analyze the relationship between HMGB1-PTEN pathway and immune cell, we performed Pearson/Spearman test (Table 3). HMGB1 was negatively correlated with CD8+T cells(P<0.05) and positively correlated with Th17, Treg and Tfh cells(P<0.01). PTEN was negatively correlated with Treg cells(P<0.05), while no significant correlations were found with CD4+T, CD8+T, Th17, Tfh cells. PI3K, PDK1 and Akt were all negatively correlated with CD4+T and CD8+T cells, and positively correlated with Th17, Treg and Tfh cells (P<0.01).
Table 3 Correlation analysis of HMGB1-PTEN pathway and T lymphocytes
Pathway
|
CD4+T
|
CD8+T
|
Th17
|
Treg
|
Tfh
|
HMGB1
|
-0.276/0.055
|
-0.364/0.010
|
0.736/0.00
|
0.611/0.00
|
0.393/0.005
|
PTEN
|
0.261/0.071
|
0.052/0.721
|
0.105/0.471
|
-0.283/0.049
|
0.125/0.391
|
PI3K
|
-0.365/0.010
|
-0.444/0.001
|
0.766/0.00
|
0.688/0.00
|
0.377/0.008
|
PDK1
|
-0.379/0.007
|
-0.451/0.001
|
0.759/0.00
|
0.701/0.00
|
0.378/0.007
|
Akt
|
-0.380/0.007
|
-0.455/0.001
|
0.759/0.00
|
0.698/0.00
|
0.385/0.006
|
The numbers to the left and right sides of each slash line represent the correlation coefficient and P value, respectively.
To further identify the important parameters that HMGB1-PTEN pathway axis can affect, stepwise regression analysis was conducted. The multiple regression equation of HMGB1-PTEN pathway and T lymphocytes was constructed in CHB. The adjusted R2 of all proteins were good. The results showed that it was highly consistent in five proteins, mainly linearly correlated with CD4+T, CD8+T, and ratio of CD4+T/CD8+T (Table 4). Similarly, in the Healthy controls, the regression results showed that the HMGB1-PTEN pathway axis could significantly affect the T lymphocyte and was highly consistent, which were significantly correlated with CD4+T, CD8+T, and ratio of CD4+T/CD8+T. There was a linear correlation between HMGB1-PTEN pathway protein and the number of immune cells, and did not change with pathophysiological status.
Table 4 Multiple stepwise regression analysis of HMGB1-PTEN pathway and T lymphocytes in CHB
Pathway
|
CD4+T
|
CD8+T
|
CD4+T/CD8+T
|
Th17
|
Treg
|
Th17/Treg
|
Tfh
|
HMGB1
|
-7.973/0.997**
|
16.105/0.997**
|
22.261/0.997**
|
—
|
—
|
—
|
—
|
PTEN
|
6.006/0.926**
|
—
|
—
|
—
|
—
|
3.597/0.926**
|
—
|
PI3K
|
-8.233/0.997**
|
16.469/0.997**
|
22.814/0.997**
|
—
|
—
|
—
|
—
|
PDK1
|
-8.215/0.997**
|
16.371/0.997**
|
22.624/0.997**
|
—
|
—
|
—
|
—
|
Akt
|
-8.153/0.997**
|
16.254/0.997**
|
22.437/0.997**
|
—
|
—
|
—
|
—
|
The numbers to the left and right sides of each slash line represent the regression coefficient and coefficient of determination, respectively. *, **: p=0.05,0.01, respectively.