Clinical information of the subjects
This study included 148 (77.08%) healthy control (HC) and 44 (22.92%) latent infection (LI) group, and clinical information of two groups were summarized in Table 1.
Table 1
The clinical information between healthy control and latent infection group (x ± s).
Indicators
|
healthy control
|
latent infection
|
Number/n
|
148
|
44
|
Age(median±SD, range)
|
37.65 ± 8.84(18-60)
|
42.34 ± 9.11(22-57)
|
Male
|
78 (52.70%)
|
29 (65.91%)
|
Female
|
70 (47.30%)
|
15 (34.09%)
|
γ Interferon / basic level N
|
0.10 ± 0.34
|
0.27 ± 0.75
|
γ Interferon / actual release level(T-N)
|
-0.02 ± 0.22
|
10.46 ± 14.30
|
γ Interferon / positive actual level(M-N)
|
40.63 ± 7.12
|
36.86 ± 12.51
|
Leukocyte count
|
6.81 ± 1.58
|
6.91 ± 1.65
|
Lymphocyte ratio (%)
|
31.85 ± 6.47
|
30.45 ± 6.91
|
Monocyte ratio (%)
|
7.44 ± 2.04
|
7.69 ± 1.80
|
Analysis of differential metabolite distribution
The distribution of serum metabolites in healthy control and latent infection group were analyzed by LC-MS. Before the analysis by SMICA-P13.0 software, the data was normalized to ensure more intuitive and reliable results. Principal Component Analysis (PCA) showed that there was no significant difference between healthy control and latent infected groups (Fig. 1A). Partial Least Squares Discriminant Analysis (OPLS-DA) could distinguish the latent infected group from the healthy control group, suggesting that there were differences in metabolite components between the two groups (Fig. 1B). Through metabolomics technology, a series of serum metabolic markers of drug-resistant and drug-sensitive groups were screened in the early stage. Through PCA and OPLS-DA analysis (Fig. 1A and 1B), drug-resistant and drug-sensitive groups could be significantly distinguished from healthy control and latent patients groups, suggesting that drug treatment may significantly affect the composition of serum metabolites.
Screening differential metabolites in different groups
The serum metabolites of patients of health control, latent infection, drug sensitivity, and drug resistance group were statistically tested by one-way ANOVA test. The results showed that there were 565 significantly different metabolites in the four groups (Table 2).
Table 2
Screening results of differential metabolites among groups (partial results)
Metabolite
|
Mean
|
p
|
Heal
|
Lat
|
Res
|
Sen
|
Betaine
|
3.94
|
4.03
|
6.61
|
6.91
|
1.38E-25
|
4-Hydroxybenzaldehyde
|
-1.39
|
-1.57
|
-0.66
|
-0.23
|
2.79E-14
|
(R)-(+)-2-Pyrrolidone-5-carboxylic acid
|
2.89
|
3.07
|
-0.66
|
-0.23
|
3.31E-25
|
PHENACYLAMINE
|
0.14
|
0.00
|
-0.66
|
-0.23
|
6.29E-12
|
Hypoxanthine
|
0.30
|
0.11
|
0.51
|
0.59
|
0.00378
|
4-formyl Indole
|
-1.80
|
-1.94
|
-0.66
|
-0.23
|
1.10E-19
|
Coumarin
|
-2.54
|
-2.82
|
-0.66
|
-0.23
|
1.59E-24
|
L-Lysine
|
-2.29
|
-2.65
|
-0.66
|
-0.23
|
1.12E-23
|
2-Hydroxycinnamic acid
|
0.14
|
0.06
|
-0.66
|
-0.23
|
3.75E-13
|
Val Gly
|
-1.62
|
-2.02
|
-0.66
|
-0.23
|
2.55E-17
|
Theophylline
|
-2.76
|
-2.57
|
-0.66
|
-0.23
|
1.29E-18
|
butamben
|
1.33
|
1.21
|
-0.66
|
-0.23
|
5.09E-26
|
p-CHLOROPHENYLALANINE
|
0.78
|
0.87
|
-0.66
|
-0.23
|
2.12E-25
|
Acetylcarnitine
|
0.55
|
0.57
|
3.68
|
3.74
|
4.50E-24
|
DL-Tryptophan
|
3.05
|
3.04
|
-0.66
|
-0.23
|
2.79E-25
|
Pro Leu
|
-1.78
|
-1.87
|
-0.66
|
-0.23
|
7.77E-19
|
Inosine
|
-3.95
|
-6.22
|
-0.66
|
-0.23
|
1.03E-24
|
Phe Phe
|
1.32
|
1.01
|
-0.66
|
-0.23
|
4.11E-23
|
Leu Leu Phe
|
-1.24
|
-1.36
|
-0.66
|
-0.23
|
2.80E-11
|
Palmitoyl-L-carnitine
|
-1.45
|
-1.59
|
-0.66
|
-0.23
|
1.27E-14
|
11β-PGF2α Ethanolamide
|
-0.11
|
-0.21
|
-0.66
|
-0.23
|
5.84E-09
|
PC(16:0/0:0)[U] / PC(16:0/0:0)[rac]
|
6.46
|
6.38
|
-0.66
|
-0.23
|
1.13E-20
|
1-heptadecanoyl-sn-glycero-3-phosphocholine
|
-0.14
|
-0.36
|
-0.66
|
-0.23
|
9.57E-05
|
Tyr Arg Leu Ile Val
|
-1.25
|
-3.35
|
-0.66
|
-0.23
|
6.60E-19
|
w/o MS2:δ-Valerolactam
|
-0.15
|
0.29
|
-0.66
|
-0.23
|
0.00847
|
w/o MS2:Hydroxyhydroquinone
|
-1.98
|
-1.09
|
-0.66
|
-0.23
|
6.42E-15
|
w/o MS2:CYCLOCREATINE
|
-2.69
|
-3.05
|
-0.66
|
-0.23
|
1.45E-24
|
w/o MS2:2-Aminopropiophenone
|
-2.32
|
-2.29
|
-0.66
|
-0.23
|
2.91E-23
|
w/o MS2:Pyroglutamic acid
|
-1.69
|
-1.69
|
-0.29
|
0.51
|
3.22E-18
|
w/o MS2:N-HYDROXYMETHYLNICOTINAMIDE
|
-2.07
|
-2.03
|
-0.66
|
-0.23
|
3.76E-20
|
w/o MS2:Mechlorethamine
|
-2.30
|
-2.49
|
-0.66
|
-0.23
|
1.62E-24
|
w/o MS2:1-Benzylimidazole
|
-2.06
|
-2.34
|
-0.66
|
-0.23
|
1.50E-23
|
w/o MS2:3-[Bis(2-hydroxyethyl)amino]propanenitrile
|
-2.57
|
-2.53
|
-0.66
|
-0.23
|
5.18E-24
|
w/o MS2:Indoleacetaldehyde
|
-4.20
|
-5.67
|
-0.66
|
-0.23
|
1.24E-24
|
w/o MS2:3-thio-Pheneacrylic Acid methyl ester
|
1.61
|
1.67
|
-0.66
|
-0.23
|
5.11E-26
|
w/o MS2:Cotinine
|
-2.49
|
-2.23
|
2.55
|
2.85
|
2.29E-21
|
w/o MS2:3-Methylethcathinone
|
-4.17
|
-3.67
|
-0.66
|
-0.23
|
5.97E-25
|
w/o MS2:Thiabendazole
|
-0.84
|
-0.83
|
-0.66
|
-0.23
|
5.37E-10
|
w/o MS2:Kynurenine
|
-2.26
|
-2.66
|
-0.66
|
-0.23
|
3.17E-24
|
w/o MS2:PROPOXUR
|
-4.33
|
-6.90
|
-0.66
|
-0.23
|
1.42E-24
|
w/o MS2:Tuckolide; Decarestrictine D
|
-2.31
|
-2.78
|
-0.66
|
-0.23
|
2.19E-23
|
w/o MS2:5-Hydroxy-4-[3-(2-hydroxy-2-propanyl)-2-oxiranyl]-1-methyl-7-oxabicyclo[4.1.0]hept-3-en-2-one
|
-1.45
|
-1.69
|
-0.66
|
-0.23
|
1.33E-14
|
w/o MS2:Melatonin
|
-2.66
|
-3.27
|
-0.66
|
-0.23
|
3.70E-26
|
w/o MS2:3,4,5-Trimethoxycinnamic acid
|
-2.35
|
-2.94
|
-0.66
|
-0.23
|
2.13E-22
|
w/o MS2:Ribothymidine
|
-0.40
|
-0.38
|
-0.66
|
-0.23
|
0.00216
|
w/o MS2:Phe Ile
|
-3.01
|
-3.80
|
-0.66
|
1.04
|
8.36E-25
|
w/o MS2:Phe Asp
|
-2.82
|
-3.58
|
-0.66
|
-0.23
|
1.14E-24
|
w/o MS2:1-Methyladenosine
|
-2.37
|
-2.78
|
-0.66
|
-0.23
|
3.16E-25
|
w/o MS2:AG-17
|
-1.65
|
-3.18
|
-0.66
|
-0.23
|
1.71E-23
|
w/o MS2:(±)-Octanoylcarnitine
|
-2.73
|
-1.95
|
-0.66
|
-0.23
|
1.20E-20
|
w/o MS2:4-(1-Acetyloxypropen-2-yl-)-2-methoxyphenylisobutyrat; 4-(1-Acetoxy-2-propen-1-yl)-2-methoxyphenyl 2-methylpropanoate
|
-3.43
|
-4.72
|
-0.66
|
-0.23
|
2.20E-25
|
w/o MS2:Ser Lys Ser
|
-0.38
|
-0.54
|
-0.66
|
-0.23
|
3.23E-06
|
w/o MS2:PGH2
|
-3.93
|
-6.90
|
-0.66
|
-0.23
|
3.52E-24
|
w/o MS2:Trimethylolpropane trimethacrylate
|
-1.68
|
0.00
|
-0.66
|
-0.23
|
1.39E-10
|
w/o MS2:Lisuride
|
-4.18
|
-5.54
|
-0.66
|
-0.23
|
1.41E-24
|
w/o MS2:N-Oleoyl Glycine
|
-1.60
|
-1.46
|
0.84
|
-0.23
|
9.58E-18
|
w/o MS2:Ala Met Lys
|
-2.63
|
-6.90
|
-0.66
|
-0.23
|
2.14E-23
|
w/o MS2:Spiromesifen
|
-3.80
|
-5.71
|
-0.66
|
-0.23
|
1.01E-24
|
w/o MS2:13,14-dihydro-19(R)-hydroxyPGE1
|
-1.00
|
-1.90
|
-0.66
|
-0.23
|
4.89E-15
|
w/o MS2:16,16-dimethyl-6-keto Prostaglandin E1
|
-4.05
|
-6.90
|
-0.66
|
-0.23
|
2.59E-24
|
w/o MS2:Pro Arg Ile
|
-0.46
|
-1.20
|
-0.66
|
-0.23
|
3.97E-08
|
w/o MS2:Asn Phe Ile
|
-2.50
|
-6.90
|
-0.66
|
-0.23
|
2.14E-23
|
w/o MS2:Arg Gly Tyr
|
-3.70
|
-4.83
|
-0.66
|
-0.23
|
1.27E-24
|
Unknown
|
1.35
|
0.50
|
-0.66
|
-0.23
|
2.76E-19
|
w/o MS2:Myriocin
|
-1.01
|
-1.78
|
-0.66
|
-0.23
|
1.25E-14
|
w/o MS2:His Lys Met
|
-4.37
|
-6.08
|
-0.66
|
-0.23
|
1.46E-24
|
w/o MS2:Ala Asn Val Asp
|
-3.38
|
-4.23
|
-0.66
|
-0.19
|
1.17E-24
|
w/o MS2:Phe Leu Arg
|
-0.58
|
-1.07
|
-0.66
|
-0.23
|
1.44E-08
|
w/o MS2:Arg Met Met
|
-3.93
|
-5.10
|
-0.66
|
-0.23
|
7.15E-24
|
w/o MS2:HC Toxin
|
-2.40
|
-6.90
|
-0.66
|
-0.23
|
2.14E-23
|
w/o MS2:PE(16:0/0:0)
|
-1.33
|
-1.75
|
-0.66
|
-0.23
|
1.64E-16
|
w/o MS2:Pristimerin
|
-0.73
|
-0.63
|
-0.66
|
-0.23
|
2.85E-10
|
w/o MS2:Buprenorphine
|
-0.22
|
-0.43
|
-0.66
|
-0.23
|
0.00079
|
w/o MS2:(3E)-7-Hydroxy-3,7-dimethyl-3-octen-1-yl 6-O-(6-deoxy-?-L-mannopyranosyl)-?-D-glucopyranoside
|
-2.51
|
-6.90
|
-0.66
|
-0.23
|
2.14E-23
|
w/o MS2:PG(18:1(9Z)/0:0)
|
-3.30
|
-5.33
|
-0.66
|
-0.23
|
7.09E-25
|
w/o MS2:Arg Thr Asp Arg
|
-2.93
|
-3.16
|
-0.66
|
-0.23
|
1.38E-24
|
w/o MS2:Phe Glu Ser Phe Gly
|
-4.37
|
-4.19
|
-0.66
|
-0.23
|
1.83E-25
|
w/o MS2:Leu Leu Asp Leu Leu
|
3.21
|
3.38
|
-0.66
|
-0.23
|
2.35E-26
|
w/o MS2:Arjunglucoside II; (2S,3R,4S,5S,6R)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl (4aS,6aS,6bR,9R,10R,11R,12aR)-10,11-dihydroxy-9-(hydroxymethyl)-2,2,6a,6b,9,12a-hexamethyl-1,3,4,5,6,6a,6b,7,8,8a,9,10,11,12,12a,12b,13,14b-octadecahydropicene-4a(2H)-carboxylate
|
-3.52
|
-4.46
|
-0.66
|
-0.23
|
2.35E-23
|
Heal means healthy control, Lat means latent infection, Res means resistance, and Sen means sensitivity group.
To further screen the differential metabolites among different groups, variable importance in the project (VIP) was used to screen the key variables in the grouping based on OPLS-DA analysis. As shown in Fig. 2, the VIP scores of pos_879 (Cotinine) compound was the largest, which was a significant difference metabolite in the four groups (Table 2), and was up-regulated in the drug sensitive group and down-regulated in the latent infection group. Pos_6484 compounds increased in the healthy control group, followed by the latent infection, drug sensitivity, and drug resistance group. pos_ 8430(unknown)、pos_ 2688(unknown)、pos_ 1509(Phencyclidine)、pos_2594(Ranitidine)、pos_4682(unknown)、pos_297(unknown)、pos_874(unknown)、pos_10662(unknown)、pos_1154(C10H16N4O; PlaSMA ID-740)、pos_ 3932 (NCGC00384560-01_C16H24O9_beta-D-Glucopyranoside, 3-hydroxy-2-(4-hydroxy-3-methoxyphenyl)propyl)、pos_3048(13E-Docosenamide)、pos_3416(unknown)、pos_3563 (unknown) and other 13 compounds were up-regulated in drug sensitive and drug resistant group and down-regulated in latent infection and healthy control group. They can be used as potential markers to distinguish drug sensitive patients, including pos_3048 (13e docosenamide) was a significantly different metabolite in four groups (Table 2).
Cluster analysis of serum metabolic biomarkers
The cluster analysis of the differential metabolites of patients in groups of healthy control, latent infection, drug resistance patients, drug sensitivity group (Fig. 3). We found that in the healthy control and latent infection groups, pos_6484(PC(16:0/0:0)[U] / PC(16:0/0:0)[rac]), pos_1123(DL-Tryptophan), pos_8385(Leu Leu Asp Leu Leu)were up-regulated, and pos_2176((±)-Octanoylcarnitine), pos_2011(Phe Ile), pos_906(Theophylline), pos_879(Cotinine), pos_4727(Ala Asn Val Asp)0, pos_8368(Phe Glu Ser Phe Gly), pos_2488(4-(1-Acetyloxypropen-2-yl-)-2-methoxyphenylisobutyrat, 4-(1-Acetoxy-2-propen-1-yl)-2-methoxyphenyl 2-methylpropanoate), pos_5241(Arg Met Met), pos_3272(Ala Met Lys), pos_5257(HC Toxin), pos_4158(Asn Phe Ile), pos_6142((3E)-7-Hydroxy-3,7-dimethyl-3-octen-1-yl 6-O-(6-deoxy-?-L-mannopyranosyl)-?-D-glucopyranoside), pos_2962(PGH2), pos_3878(16,16-dimethyl-6-keto Prostaglandin E1), pos_1154(PROPOXUR), pos_4682(His Lys Met), pos_1800(Inosine), pos_772(Indoleacetaldehyde), pos_7613(Arg Thr Asp Arg), pos_1010(3-Methylethcathinone), pos_4201(Arg Gly Tyr), pos_9608(Arjunglucoside II; (2S,3R,4S,5S,6R)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl (4aS,6aS,6bR,9R,10R,11R,12aR)-10,11-dihydroxy-9-(hydroxymethyl)-2,2,6a,6b,9,12a-hexamethyl-1,3,4,5,6,6a,6b,7,8,8a,9,10,11,12,12a,12b,13,14b-octadecahydropicene-4a(2H)-carboxylate), pos_6824(PG(18:1(9Z)/0:0)), pos_3043(Lisuride), pos_3744(Spiromesifen)were down-regulated.
Metabolic Pathway Analysis
KEGG (Kyoto Encyclopedia of genes and genes) annotation and enrichment analysis were performed on the differential metabolites among four groups with p adjust < 0.05. The results are shown in Fig. 4 and Table 3.
The significantly enriched pathway included the Oxytocin signaling pathway, Platelet activation, Retrograde endocannabinoid signaling, Serotonergic synapse, and Caffeine metabolism (Fig. 4A). And the significant compounds were Prostaglandin H2 and Theophylline. Simultaneously, the annotation of level 2 KEGG pathway included Amino acid metabolism, Biosynthesis of other secondary metabolites, Nucleotide metabolism, Endocrine system, Immune system, Lipid metabolism, and Nervous system (Fig. 4B). And the significant compounds were Indole-3-acetaldehyde, Theophylline, Inosine, and Prostaglandin H2 (Table 3). Among these, Prostaglandin H2 and Theophylline were showed by enriched and annotated pathways at the same time.
Combined with the above results, Theophylline (pos_906) and Inosine (pos_1800) in drug resistance patients, the drug sensitivity group was higher than that in the healthy control and latent infection groups (Fig. 3).
Table 3
KEGG pathway enrichment results
ID
|
Description
|
Count
|
p value
|
p adjust
|
Enrichment Fold
|
Compound Sig
|
ko00232
|
Caffeine metabolism
|
1
|
0.01
|
0.03
|
74.77
|
Theophylline;
|
ko04611
|
Platelet activation
|
1
|
0.01
|
0.03
|
104.68
|
Prostaglandin H2;
|
ko04723
|
Retrograde endocannabinoid signaling
|
1
|
0.01
|
0.03
|
82.64
|
Prostaglandin H2;
|
ko04921
|
Oxytocin signaling pathway
|
1
|
0.01
|
0.03
|
120.79
|
Prostaglandin H2;
|
ko04726
|
Serotonergic synapse
|
1
|
0.03
|
0.04
|
37.39
|
Prostaglandin H2;
|
ko00380
|
Tryptophan metabolism
|
1
|
0.05
|
0.06
|
19.39
|
Indole-3-acetaldehyde;
|
ko00590
|
Arachidonic acid metabolism
|
1
|
0.05
|
0.06
|
20.94
|
Prostaglandin H2;
|
ko00230
|
Purine metabolism
|
1
|
0.06
|
0.06
|
16.70
|
Inosine
|
Relationship between serum markers and disease progression
Through OPLS-DA analysis, one-way ANOVA analysis, cluster analysis, and pathway analysis of differential metabolites, four potential serum markers were screened. The relative contents of these four serum markers (Inosine, Prostaglandin E1, Theophylline, and Cotinine 1) in the four groups were compared. The results showed that the relative contents of Inosine (Fig. 5A) and Prostaglandin E1 (Fig. 5C) were the lowest in the latent infection group and the highest in the drug sensitive group, which suggested that these two metabolites are expected to be potential biomarkers for the diagnosis of latent infection. In addition, the relative contents of Theophylline (Fig. 5B) and Cotinine 1 (Fig. 5D) increased in healthy control, latent infections, drug resistance, and drug sensitivity group, suggesting that these two metabolites can be used as potential markers to monitor disease progression and provide important ideas for timely control of disease deterioration.