Research Article

Prognostic and Immunological Roles of MMP-9 in Pan-Cancer

Yudan Zeng ,1 Mengqian Gao [D,1 Dongtao Lin (D,1 Guoxia Du,1 and Yongming Cai (2,3,4

1School of Public Health, Guangdong Pharmaceutical University, Guangzhou, China

2College of Medical Information Engineering, Guangdong Pharmaceutical University, Guangzhou, China

3Guangdong Provincial TCM Precision Medicine Big Data Engineering Technology Research Center, Guangzhou, China

4Key Specialty of Clinical Pharmacy, The First Affiliated Hospital of Guangdong Pharmaceutical University, China

Correspondence should be addressed to Yongming Cai; cym@gdpu.edu.cn

Received 12 September 2021; Revised 12 November 2021; Accepted 13 December 2021; Published 7 February 2022

Academic Editor: Wan-Ming Hu

Copyright @ 2022 Yudan Zeng et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. Matrix metalloproteinase-9 (MMP-9) can degrade the extracellular matrix and participate in tumor progression. The relationship between MMP-9 and immune cells has been reported in various malignant tumors. However, there is a lack of comprehensive pan-cancer studies on the relationship between MMP-9 and cancer prognosis and immune infiltration. Method. We used data from TCGA and GTEx databases to comprehensively analyze the differential expression of MMP-9 in normal and cancerous tissues. Survival analysis was performed to understand the prognostic role of MMP-9 in different tumors. We then analyzed the expression of MMP-9 across different tumors and at different clinical stages. Based on the results, we assessed the correlation between MMP-9 expression and immune-associated genes and immunocytes. Finally, we calculated the tumor mutation burden (TMB) of 33 cancer types and analyzed the correlation between MMP-9 and TMB, DNA microsatellite instability, and DNA repair genes. Results. MMP-9 significantly affected the prognosis and metastasis of various cancers. It was associated based on overall survival, disease-specific survival in five tumors, progression-free interval in seven tumors, and clinical stage in eight tumors, as well as with prognosis and metastasis in adrenocortical carcinoma and kidney renal clear cell carcinoma. It was also coexpressed with immune-related genes and DNA repair genes. The expression of MMP- 9 was positively correlated with the markers of T cells, tumor-associated macrophages, Th1 cells, and T cell exhaustion. Furthermore, MMP-9 expression was highly correlated with macrophage M0 in 28 tumors. In addition, its expression was associated with TMB in eight cancer types and DNA microsatellite instability in six cancer types. Conclusion. MMP-9 is related to immune infiltration in pan-cancer and can be used as a biomarker related to cancer prognosis and metastasis. Our findings provide prognostic molecular markers and new ideas for immunotherapy.

1. Introduction

Matrix metalloproteinase-9 (MMP-9) [1] is a significant matrix metalloproteinase that is involved in many biological processes by degrading the extracellular matrix. MMP-9 plays an important role in the onset, progression, and metas- tasis of gastric [2], lung [3], colon [4], and breast cancers [5]. Metastasis is a major cause of mortality in patients with can- cer. MMP-9 promotes metastasis and angiogenesis through decomposition of the extracellular matrix [6, 7]. Infiltration of immune cells can also affect cancer metastasis and prog- nosis. Recently, many studies [8-10] evaluated the potential of MMP-9 as a biomarker for the prognosis of various can-

cers, including cervical [11, 12], ovarian [13, 14], pancreatic [15], and breast cancers [16].

In recent years, the incidence of cancer and its morbidity and mortality have shown an increasing trend. Cancer is a major cause of death worldwide and is second only to car- diovascular disease. The WHO estimates that malignant tumors will become the main cause of global mortality after 2030 [17]. The tumor microenvironment (TME) influences tumor growth and development. Tumor-associated macro- phages (TAMs) are macrophages that infiltrate the tumor tissue and most immune cells in the TME. Tekin et al. [18] found that macrophages release MMP-9 in pancreatic can- cer. TAMs can support the proliferation, invasion, and

FIGURE 1: Flow chart of this article.

TCGA data from UCSC Xena

Mutation data

RNA-sequence data

Clinical data

MMP-9 expression

Relationship between MMP-9 and TMB, MSI in pan-cancer

Association between MMP-9 and immune in pan-cancer

Clinicopathology analysis

Survival analysis

Immune genes co-expression analysis

Macrophage correlation analysis

8 cancer types

11 cancer types

2 cancer type

DLBC as a control group

GSEA

Immune cell infiltration

DNA repair genes co-expression analysis

Immune cell markers co-expression analysis

metastasis of tumor cells. Therefore, the development of antitumor drugs that can target macrophage polarization is urgently required. Immunotherapy is highly suitable for patients with cancer because of its excellent efficacy. How- ever, not all patients can benefit from immunotherapy and research has shown that tumor mutation burden (TMB) and DNA microsatellite instability (MSI) can be used as pre- dictive markers for immunotherapy efficacy. TMB [19] has a good predictive value for immunotherapy in a variety of tumors. In addition, MSI [20] has been regarded as an important molecular marker for the prognosis and adjuvant treatment of colorectal cancer and other solid tumors. In view of the complexity of tumor progression, pan-cancer analysis has been widely used in cancer research and consid- erable progress has been made in understanding various tumor features, including cancer susceptibility variation, oncogenic pathway cooccurrence and mutual exclusion, and biological regulation network disorder [21-23].

MMP-9 has been found to be closely related to immu- nity and tumor progression; however, most studies have focused on single cancers. Here, we systematically studied MMP-9 expression and its correlation with prognosis and metastasis in 33 cancer types to help us fully understand the role of MMP-9 in tumors. A flowchart of the study is shown in Figure 1. We also analyzed the relationship between MMP-9 expression and immune cell infiltration.

2. Materials and Methods

2.1. Data Acquisition. Gene expression profiles, mutation data, and clinical information of 33 cancers in TCGA database were downloaded from UCSC Xena [24] (http://

xena.ucsc.edu/).The disease-specific survival (DSS) and progression-free interval (PFI) data were downloaded from TCGA Pan-Cancer (PANCAN) of UCSC Xena. After excluding cases with missing survival time data, 11,057 sam- ples were included in the study.

2.2. Gene Expression Analysis. We used “wilcox.test” to ana- lyze the differential expression of MMP-9 in normal and tumor tissue samples, as well as the differential expression of MMP-9 in different cancer types in TCGA database, and drew a box diagram.

In view of the small number of normal tissue samples in TCGA database, we included data from the GTEx (geno- type-tissue expression) database [25] using the “Match TCGA normal and GTEx data” option in the GEPIA2 database [26] (http://gepia2.cancer-pku.cn/#analysis) for the differential analysis to ensure more reliable results.

2.3. Survival and Clinical Analysis. The expression of MMP- 9 was extracted from the gene expression profile data, and the samples were divided into high- and low-expression groups according to the median MMP-9 expression. We used the Kaplan-Meier method to analyze the survival information and “survival” [27] and “survminer” to draw the survival curve. We also performed COX analysis of the survival data, and the R package “forestplot” was used to visualize the results.

A boxplot using tumor stage as a variable was graphed to observe the differences in MMP-9 expression at different clin- ical stages and analyze the relationship between the expression level of MMP-9 and tumor metastasis in different cancers. This was carried out using the R package “limma” [28].

MMP9 expression

10

0

5

ACC

-

BLCA


BRCA

CESC

CHOL


COAD

DLBC

ESCA


GBM

0

HNSC

KICH

KIRC

KIRP

-

I

LAML

LGG

LIHC


LUAD

LUSC

MESO

OV

PAAD

PCPG


PRAD

READ

SARC

0-

SKCM

STAD


TGCT

THCA

THYM

-

UCEC


UCS

UVM

Type

Tumor

Normal

(a)

FIGURE 2: MMP-9 expression levels in different tumor types in various databases. (a) Expression level of MMP-9 in different tumors of TCGA database; MMP-9 expression in tumor samples is significantly higher than normal in many cancer types. The P values are

Transcripts per million (TPM)

3000

6000

9000

12000

15000

18000

0

T (n=77)

N (n= 128)

ACC

T (n=404)

BLCA

N (n = 28)

T (n = 1085)

BRCA

N (n= 291)

T (n= 306)

CESC

N (n=13)

T (n=36)

CHOL

N (n= 9)

T (n=275)

COAD

N (n=349)

T (n=47)

DLBC

N (n = 337)

T (n=182)

ESCA

N (n= 286)

T (n= 163)

GBM

N (n=207)

T (n=519)

HNSC

N (n = 44)

T (n= 66)

KICH

N (n= 53)

T (n= 523)

KIRC

N (n= 100)

T (n=286)

KIRP

N (n=60)

T (n= 173)

LAML

N (n=70)

(b)

T (n= 518)

LGG

N (n= 207)

T (n= 369)

LIHC

N (n=160)

T (n= 483)

LUAD

N (n = 347)

T (n=486)

LUSC

N (n = 338)

T (n= 87)

MESO

T (n= 426)

N (n=88)

OV

T (n=179)

N (n= 171)

PAAD

T (n=182)

N (n= 3)

PCPG

T (n= 492)

N (n= 152)

PRAD

T (n=92)

N (n= 318)

READ

T (n=262)

N (n=2)

SARC

T (n= 461)

N (n= 558)

SKCM

T (n=408)

N (n= 211)

STAD

T (n= 137)

N (n= 165)

TGCT

T (n= 512)

N (n = 337)

THCA

T (n= 118)

N (n= 339)

THYM

T (n=174)

N (n= 91)

UCEC

T (n= 57)

N (n=78)

UCS

T (n= 79)

UVM

indicated as *P < 0.05, ** P <0.01, and *** P < 0.001. (b) Expression level of MMP-9 in different tumors of data matching TCGA normal

and GTEx data by GEPIA2 database; marked red cancer means that MMP-9 is highly expressed in tumor tissues and marked green

cancer represents that MMP-9 is highly expressed in normal tissues.

2.4. Immunological Correlation Analysis. We used the “Gene” module of TIMER [29] (https://cistrome.shinyapps .io/timer/) to explore the correlation between MMP-9 expression and abundance of immune infiltrates in adreno-

cortical carcinoma (ACC), kidney renal clear cell carcinoma

(KIRC), and lymphoid neoplasm diffuse large B-cell lym- phoma (DLBC). In addition, we employed the “Immune- Gene” module in the TIMER2.0 database [30](http://timer

.comp-genomics.org/) to explore the association between MMP-9 expression and macrophage immune infiltration. The R package “CIBERSORT” [31] was used to evaluate the infiltration of immune cells in all samples. Coexpression analysis of MMP-9 and immune cells was performed using Spearman’s correlation. In addition, we calculated the corre- lation coefficient between various immune markers and MMP-9 using “limma.”

p valueHazard ratio
ACC<0.0011.713(1.329-2.207)
BLCA0.0481.076(1.001-1.156)
BRCA0.1560.936(0.854-1.026)
CESC0.7570.978(0.847-1.128)
CHOL0.4180.897(0.690-1.167)
COAD0.8800.990(0.870-1.127)
DLBC0.1350.763(0.534-1.088)
ESCA0.4610.938(0.792-1.111)
GBM0.0311.123(1.011-1.248)
HNSC0.8821.006(0.926-1.094)
KICH0.0521.737(0.996-3.029)
KIRC<0.0011.194(1.100-1.297)
KIRP0.6100.948(0.774-1.162)
LAML0.3050.902(0.741-1.098)
LGG<0.0011.249(1.118-1.396)
LIHC0.0111.140(1.031-1.261)
LUAD0.3091.048(0.957-1.148)
LUSC0.6021.024(0.936-1.120)
MESO0.4731.041(0.932-1.163)
OV0.3590.962(0.886-1.045)
PAAD0.0851.121(0.984-1.277)
PCPG0.9890.997(0.620-1.601)
PRAD0.6880.895(0.521-1.539)
READ0.8980.981(0.731-1.317)
SARC0.2781.043(0.967-1.125)
SKCM0.0090.919(0.862-0.979)
STAD0.7200.979(0.874-1.097)
TGCT0.0931.921(0.898-4.110)
THCA0.2121.219(0.893-1.662)
THYM0.8020.937(0.566-1.552)
UCEC0.0360.874(0.771-0.992)
UCS0.6171.045(0.879-1.243)
UVM<0.0012.019(1.535-2.656)

0.50

1.0

2.0

4.0

Hazard ratio

FIGURE 3: Continued.

(a)

1.00

Cancer: ACC

Overall survival

0.75

+

0.50

0.25

p = 0.003

+

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

39

36

24

18

10

7

4

2

1

1

1

1

1

Low

40

39

34

26

20

17

12

9

7

6

3

1

1

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

☒ High

☒ Low

(b)

FIGURE 3: Continued.

1.00

Cancer: BLCA

Overall survival

0.75

0.50

0.25

p = 0.027

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Time (years)

MMP9 levels

High

203

145

68

38

31

18

11

7

6

5

4

2

2

2

0

0

Low

203

144

71

49

36

29

16

14

7

4

2

1

1

1

0

0

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Time (years)

MMP9 levels

☒ High

Low

(c)

1.00

Cancer: DLBC

Overall survival

0.75

#

0.50

0.25

p = 0.017

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Time (years)

MMP9 levels

High

23

22

19

11

10

7

6

6

6

6

4

3

3

2

2

2

2

1

0

0

0

Low

24

16

11

7

3

2

2

2

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Time (years)

MMP9 levels

+ High ☒

☒ + Low

(d)

Cancer: KIRC

FIGURE 3: Correlation between MMP-9 and overall survival for various cancer types of TCGA database. (a) Multivariate Cox regression analysis to identify prognosis in 33 cancer types. (b-e) Kaplan-Meier survival curves comparing the high and low expression levels of MMP-9 in different types of cancer. The high expression of MMP-9 was related to the low overall survival rate (b) in ACC (P= 0.003), (c) in BLCA (P=0.027), (d) in KIRC (P=0.001), and (e) in LIHC (P=0.009). The low expression of MMP-9 was related to the low overall survival rate of (f) in DLBC (P = 0.017).

1.00

Overall survival

0.75

0.50

0.25

p = 0.001

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

265

218

176

143

106

68

37

23

18

15

7

2

1

Low

266

222

184

148

112

82

62

39

23

16

6

1

0

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

Low

(e)

Cancer: LIHC

1.00

Overall survival

0.75

0.50

#

0.25

#

p = 0.009

0.00

0

1

2

3

4

5

6

7

8

9

10

Time (years)

MMP9 levels

High

184

123

65

41

30

18

16

5

3

2

0

Low

184

141

78

52

36

25

13

4

3

2

1

1

2

3

4

5

6

7

8

9

10

Time (years)

MMP9 levels

+ High

-+ Low

(f)

2.5. Mutation Analysis. TMB refers to the number of somatic mutations that occur after germline mutations are removed from the tumor genome. We used PERL scripts to calculate the TMB of each sample. The MSI values were derived from TCGA database. We then analyzed the correlation between

MMP-9 and TMB and MSI and designed a radar map using the R package “fmsb.”

2.6. Gene Set Enrichment Analysis (GSEA). We used GSEA to group and classify the genes according to multiple

FIGURE 4: Continued.

Cancer: ACC

1.00

Disease-specific survival

0.75

+

0.50

0.25

p = 0.003

+

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

38

35

24

18

10

7

4

2

1

1

1

1

1

Low

39

38

33

25

19

17

12

9

7

6

3

1

1

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

Low

(a)

Cancer: KIRC

1.00

Disease-specific survival

0.75

0.50

0.25

p = 0.002

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

294

244

201

166

124

82

48

29

21

18

8

2

1

Low

295

246

196

163

131

102

81

52

29

20

10

3

1

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

☒ Low

(b)

FIGURE 4: Continued.

Cancer: DLBC

1.00

+

Disease-specific survival

0.75

#

+

0.50

0.25

p = 0.010

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Time (years)

MMP9 levels

High

24 23

20

11

10

7

6

6

6

6

4

3

3

2

2

2

2

1

0

0

0

Low

24

15

10

7

3

2

2

2

2

1

0

0

0

0

0

0

0

0

0

0

0

0

0

1

2

3

4

5

6

7

8

9

10

11 12

13

14

15 1

£ 17

18

1

20

Time (years)

MMP9 levels

+ High ☒

-+ Low ☒

(c)

Cancer: UCEC

1.00

+

+

Disease-specific survival

0.75

0.50

0.25

p = 0.018

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

Time (years)

MMP9 levels

High

28 3223

2351 5169 591

11

86

60

35

21

9

6

4

2

2

2

2

2

1

1

1

0

0

Low

28224

516.

115

79

59

45

26

17

11

3

2

1

0

0

0

0

0

0

0

0

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

A

18

1

20

Time (years)

MMP9 levels

-+ High ☒

☒ Low

(d)

Cancer: SKCM

1.00

Disease-specific survival

0.75

0.50

0.25

p = 0.029

0.00

0

1

2

3

4

5

6

7

8

9

10

I

12

13

14 4

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

Time (years)

MMP9 levels

High

225 203 16613

$7 119

90

73

64

59

47

43

38

32

26

22

14

13

11

11

6

4

3

3

3

3

3

2

2

2

1

0

Low

226 6 186 122

88

72

60

46

41

32

28

25

20

16

13

11

9

9

9

8

6

6

4

4

3

3

2

2

2

2

2

2

1

0

1

2

3

4

5

6

7

8

9

10

1

12

13

1 14

15

16 17 18 19 2

0 21 2

23

24

6 27

26

8 29

30

Time (years)

MMP9 levels

High

Low

(e)

.

+

+

+

+

+

+

+

+

+

0.50

1.0

2.0

4.0

Hazard ratio

(f)

p valueHazard ratio
ACC<0.0011.690(1.297-2.203)
BLCA0.0341.099(1.007-1.198)
BRCA0.4680.962(0.865-1.069)
CESC0.3200.921(0.783-1.084)
CHOL0.2970.883(0.699-1.115)
COAD0.8590.986(0.846-1.150)
DLBC0.0920.624(0.360-1.081)
ESCA0.7640.971(0.804-1.173)
GBM0.0341.131(1.010-1.267)
HNSC0.0910.929(0.854-1.012)
KICH0.0421.746(1.020-2.988)
KIRC<0.0011.232(1.125-1.349)
KIRP0.9030.986(0.786-1.237)
LGG<0.0011.273(1.124-1.441)
LIHC0.1721.089(0.964-1.231)
LUAD0.7081.020(0.920-1.131)
LUSC0.4710.955(0.843-1.082)
MESO0.7750.979(0.846-1.132)
OV0.2990.953(0.870-1.044)
PAAD0.0781.137(0.986-1.312)
PCPG0.9901.003(0.597-1.686)
PRAD0.4831.310(0.616-2.786)
READ0.7640.942(0.638-1.391)
SARC0.6151.022(0.939-1.113)
SKCM0.0150.918(0.857-0.983)
STAD0.9261.006(0.878-1.153)
TGCT0.1921.683(0.770-3.679)
THCA0.3920.839(0.562-1.254)
THYM0.5730.799(0.366-1.744)
UCEC0.0340.849(0.730-0.988)
UCS0.6191.046(0.877-1.246)
UVM<0.0011.986(1.492-2.643)

FIGURE 4: Correlation between MMP-9 and DSS for various cancer types of TCGA database. (a-e) Kaplan-Meier survival curves comparing the high and low expression levels of MMP-9 in different types of cancer. The high expression of MMP-9 was related to the low DSS (a) in ACC (P= 0.003) and (b) in KIRC (P=0.018). The low expression of MMP-9 was related to the low DSS (c) in DLBC (P=0.010), (d) in UCEC (P=0.018), and (e) in SKCM (P=0.02) and (f) multivariate Cox regression analysis to identify prognosis in 33 cancer types.

FIGURE 5: Continued.

Cancer: ACC

Progression-free interval

1.00

0.75

+

0.50

0.25

p = 0.002

+

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

39

24

13

8

5

4

2

1

1

1

1

1

1

Low

40

32

28

19

15

14

10

7

5

4

2

1

1

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

☒ Low

(a)

Cancer: UVM

Progression-free interval

1.00

0.75

0.50

0.25

p = 0.009

0.00

0

1

2

3

4

5

6

Time (years)

MMP9 levels

High

39

24

13

Low

6

15

1

40

31

24

0

0

3

2

1

0

1

2

3

4

5

6

Time (years)

MMP9 levels

☒ High

☒ Low

(b)

FIGURE 5: Continued.

Cancer: KIRC

Progression-free interval

1.00

0.75

0.50

0.25

p = 0.001

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

300

218

169

140

98

62

34

22

14

10

5

0

0

Low

301

232

178

149

120

90

64

39

21

12

5

1

0

0

1

2

3

4

5

6

7

8

9

10

11

12

Time (years)

MMP9 levels

High

☒ Low

(c)

Cancer: THCA

1.00

1

Progression-free interval

+++

0.75

H

H

+

0.50

0.25

p = 0.025

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Time (years)

MMP9 levels

High

284

228

158

106

72

51

39

33

24

16

13

10

4

3

2

0

Low

284

256

168

107

76

54

41

27

19

15

9

9

4

3

1

0

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

Time (years)

MMP9 levels

++ High ☒

☒ Low

(d)

FIGURE 5: Continued.

Progression-free interval

1.00

Cancer: GBM

0.75

0.50

0.25

p = 0.021

0.00

0

1

2

3

4

Time (years)

MMP9 levels

High

83

14

3

0

0

Low

83

25

9

4

1

0

1

2

3

4

Time (years)

MMP9 levels

+ High ☒

☒ Low

(e)

1.00

Cancer: DLBC

Progression-free interval

0.75

0.50

0.25

p = 0.004

0.00

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

1

17

18

19

20

Time (years)

MMP9 levels

High

24 22 20 11 10 7

Low

6

6

24

1 12

9

6

2

2

2

2

6

6

4

2

2 0

2

1

1

1

1

1

0

0

0

0

0

0

0

0

0

0

0

2

0

0

0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Time (years)

MMP9 levels

-+ High ☒

☒ Low

(f)

(g)
p valueHazard ratio
ACC<0.0011.536(1.210-1.949)
BLCA0.1121.061(0.986-1.140)
BRCA0.8840.994(0.916-1.078)
CESC0.0500.867(0.752-1.000)
CHOL0.1500.848(0.677-1.062)
COAD0.8370.988(0.882-1.108)
DLBC0.0020.670(0.521-0.860)
ESCA0.6750.971(0.845-1.116)
GBM0.0071.156(1.041-1.285)
HNSC0.2750.962(0.897-1.031)
KICH0.0301.520(1.040-2.220)
KIRC<0.0011.194(1.106-1.288)
KIRP0.8461.016(0.868-1.188)
LGG<0.0011.271(1.144-1.411)
LIHC0.7291.015(0.933-1.104)
LUAD0.4571.029(0.954-1.112)
LUSC0.9941.000(0.907-1.102)
MESO0.5191.045(0.914-1.196)
OV0.0800.938(0.872-1.008)
PAAD0.1831.085(0.963-1.222)
PCPG0.0471.264(1.003-1.594)
PRAD0.0951.153(0.976-1.362)
READ0.7401.043(0.814-1.336)
SARC0.0171.077(1.014-1.144)
SKCM0.1340.959(0.908-1.013)
STAD0.1790.925(0.826-1.036)
TGCT0.5261.061(0.884-1.273)
THCA0.0301.182(1.016-1.375)
THYM0.8530.971(0.713-1.323)
UCEC0.0780.914(0.828-1.010)
UCS0.3771.078(0.912-1.274)
UVM<0.0011.721(1.328-2.230)
FIGURE 5: Correlation between MMP-9 and PFI for various cancer types of TCGA database. (a-g) Kaplan-Meier survival curves comparing the high and low expression levels of MMP-9 in different types of cancer. The high expression of MMP-9 was related to the low PFI (a) in ACC (P=0.002), (b) in UVM (P=0.009), (c) in KIRC (P=0.001), (d) in THCA (P=0.025), and (e) in GBM (P=0.021). The low expression of MMP-9 was related to the low PFI (f) in DLBC (P=0.004) and (g) in CESC (P=0.031) and (h) multivariate Cox regression analysis to identify prognosis in 33 cancer types.

Progression-free interval

1.00

Cancer: CESC

0.75

H

#

0.50

0.25

p = 0.031

0.00

0

1

2

3

4

5

6

7

8

9

10

11

L

13 3

14

15

16

17

18

1

19

20

Time (years)

MMP9 levels

High

9

Low

154 112

66

45

34

24

20

15

12

8

6

6

4

4

2

2

2

0

0

0

155 107

55

37

23

15

14

11

9

9

8

7

5

1

0

0

0

0

0

0

0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Time (years)

MMP9 levels

High

Low

0.50

0.71

1.0

1.41

2.0

Hazard ratio

(h)

Cancer: ACC

12

0.072

0.0098

0.78

0.073

9

0.8

MMP9 expression

0.81

6

3

0

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage I

Stage III

Stage II

Stage IV

(a)

Cancer: BLCA

0.79

20

0.00086

0.0024

0.16

0.22

MMP9 expression

15

0.63

10

5

0

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage I

Stage III

Stage II

Stage IV

(b)

FIGURE 6: Continued.

FIGURE 6: Continued.

Cancer: BRCA

Cancer: THCA

0.092

0.44

0.7

12

0.00094

20

0.00035

0.0008

0.52

0.16

0.035

0.47

MMP9 expression

15

0.5

MMP9 expression

8

0.00091

10

4

5

0

0

Stage I

Stage II

Stage III

Stage IV

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage

Stage

Stage I

Stage III

Stage I

Stage III

Stage II

Stage IV

Stage II

Stage IV

(c)

(d)

Cancer: KIRC

0.29

15

0.1

0.0052

0.041

7.5e-05

MMP9 expression

10

0.8

5

0

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage I

Stage III

Stage II

Stage IV

(e)

Cancer: KIRP

0.21

0.014

0.04

0.17

10

0.82

MMP9 expression

0.053

5

0

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage I

Stage III

Stage II

Stage IV

(f)

Cancer: SKCM

0.21

0.58

7.9e-06

15

0.24

0.85

MMP9 expression

0.00079

10

5

0

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage I

Stage III

Stage II

Stage IV

(g)

FIGURE 6: Continued.

FIGURE 6: MMP-9 expression in different tumor stages of TCGA database (a) in ACC, (b) in BLCA, (c) in BRCA, (d) in THCA, (e) in KIRC, (f) in KIRP, (g) in SKCM, and (h) in ESCA.

Cancer: ESCA

0.17

0.041

0.16

0.79

10

0.017

MMP9 expression

0.00066

5

0

T

T

Stage I

Stage II

Stage III

Stage IV

Stage

Stage

Stage I

Stage III

Stage II

Stage IV

(h)

functional gene sets, such as the GO gene set. We used the package “clusterProfiler” [32] of R (ver. 3.6.3) to analyze the GO enrichment of MMP-9 in ACC, KIRC, and DLBC.

2.7. Other Analyses. We extracted the expression of common immune checkpoint genes and DNA repair genes of 33 tumors and used Spearman correlation coefficients to evalu- ate their correlation with MMP-9 expression.

3. Results

3.1. mRNA Expression Levels of MMP-9 in Different Types of Human Cancers. To determine the differences in the expres- sion levels of MMP-9 in various human cancers, we exam- ined the MMP-9 expression levels using the RNA-seq data of multiple malignancies from TCGA database. The differ- ential expression of MMP-9 between tumor and adjacent normal tissues across tumor types is shown in Figure 2(a). Except for tumors without normal tissue data, MMP-9 expression was significantly higher in tumor samples than in normal samples.

Owing to the insufficiency of normal tissue data in TCGA database, we included data from the GTEx database to supplement TCGA data for the differential analysis (Figure 2(b)). MMP-9 was highly expressed in the tissues of bladder urothelial carcinoma (BLCA), breast invasive car- cinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), colon adenocarci- noma (COAD), esophageal carcinoma (ESCA), glioblastoma multiforme (GBM), head and neck squamous cell carcinoma (HNSC), KIRC, kidney renal papillary cell carcinoma (KIRP), liver hepatocellular carcinoma (LIHC), lung adeno- carcinoma (LUAD), lung squamous cell carcinoma (LUSC),

ovarian serous cystadenocarcinoma (OV), pancreatic adeno- carcinoma, rectum adenocarcinoma, skin cutaneous mela- noma (SKCM), stomach adenocarcinoma, testicular germ cell tumors, uterine corpus endometrial carcinoma (UCEC), and uterine carcinoma compared with normal tissues. Inter- estingly, the expression of MMP-9 was higher in the normal tissues of thymoma than in tumor tissues.

3.2. Association between MMP-9 Expression and Cancer Prognosis. Next, we investigated whether the expression level of MMP-9 is associated with patient prognosis. Using uni- variate survival analysis, we found a significant correlation between prognosis and MMP-9 expression in many cancer types, including uterine, kidney, skin, brain, liver, and blad- der cancers. Additionally, we used the Kaplan-Meier method to plot the survival curves and found that ACC (P= 0.003), BLCA (P= 0.027), KIRC (P=0.001), and LIHC (P= 0.009) patients with high MMP-9 levels had a poor prognosis (Figures 3(b)-3(e)). However, DLBC patients with high MMP-9 expression had a better prognosis (P=0.017) (Figure 3(f)).

Considering the possibility that there may also be non- tumor-related factors leading to death during the follow-up period, we analyzed the relationship between gene expression and DSS. Notably, MMP-9 expression significantly affected the prognosis in five cancer types (Figures 4(a)-4(e)), includ- ing ACC (P=0.003), KIRC (P=0.002), DLBC (P=0.010), UCEC (P = 0.018), and SKCM (P = 0.029). These results sug- gest that high MMP-9 expression is an independent risk factor for poor prognosis in ACC and KIRC.

To further examine the prognostic potential of MMP-9 in different cancers, we evaluated the PFI of the 33 cancer types. Higher MMP-9 expression levels were associated with

FIGURE 7: Continued.

Purity

B Cell

CD8+ T Cell

CD4+ T Cell

Macrophage

Neutrophil

Dendritic cell

MMP9 expression level (log2 TPM)

10

. cor= - 0.067

partial.cor = 0.238

partial.cor = 0.102

partial.cor = 0.035.

partial.cor = 0.187

partial.cor = 0.266.

partial,cor .= 0,322

.p = 5.71e-01

PF 4.22e-02

.p = 3,90e-01

P =7.69€-01

P = 1.13e-01

P=2.31e,02

P .= 5,51e-03

5

ACC

0

0.2

0.4

0.6

0.8

1.0

0.11

0.12

0.13

0.20

0.25

0.30

0.35 0.07

0.09

0.11

0.13

0.15

0.08

0.12

0.16

0.12

0.14

0.16

0.18

0.49

0.50

0.51

0.52

0.53

Infiltration level

(a)

Purity

B Cell

CD8+ T Cell

CD4+ T Cell

Macrophage

Neutrophil

Dendritic cell

MMP9 expression level (log2 TPM)

10.0

-0,146

partial.cor .= 0.139 .

partial.çor = 0.073

partial.cor = 0.211.

partial,cor = 0.178

partial.cor = 0.211

Partial cor,= 0.25.

p = 1.63e-03

p = 2.77e-03

p = 1.26e-01

p = 5.03e-06

P .= 1.522-04

p = 5.11e-06

D = 6.24e-08

7.5

KIRC

5.0

2.5 -

0.0

0.25

0.50

0.75

1.00 0.0

0.1

0.2

0.3

0.4

0.5

0.0

0.2

0.4

0.6

0.0

0.1

0.2

0.3

0.4

0.5 0.0

0.1

0.2

0.3

0.4

0.0

0.1

0.2

0.3

0.0

0.4

0.8

1.2

Infiltration level

(b)

Purity

B Cell

CD8+ T Cell

CD4+ T Cell

Macrophage

Neutrophil

Dendritic cell

MMP9 expression level (log2 TPM)

Cor = - 0,31

partial,cor =- 0,362

parțial.cor =- 0.374

P=1.40e-01

P .= ‘9.49e-02

partial.cor = - 0.1

partial.cor = - 0,197

partial.cor = 0.233,

partial.cor, = 0.408

.P =“4.6e-02.

.P= 6,67e-01

: P’= 3.91e-01

:p = 3:09e-01”

P = 6.67e-02

12

8

DLBC

4

0.25

0.50

0.75

1.00.00

0.05

0.10

0.15

0.20

0.10

0.15

0.20

0.25

0.0

0.1

0.2

0.3

0.4

0.5 0.00

0.05

0.10

0.15

0.1

0.2

0.3

0.3

0.4

0.5

0.6

0.7

Infiltration level

(c)

FIGURE 7: (a-c) Correlation analysis between MMP-9 expression and six kinds of infiltrating immune cells by TIMER database (a) in ACC, (b) in KIRC, and (c) in DLBC and (d) correlation analysis between MMP-9 expression and immune infiltration of macrophage by TIMER 2.0 database.

Macrophage_EPIC

Macrophage_TIMER

Macrophage_XCELL

Macrophage M0_CIBERSORT

Macrophage MO_CIBERSORT-ABS

_Macrophage M1_CIBERSORT

_Macrophage M1_CIBERSORT-ABS

Macrophage M1_QUANTISEQ

Macrophage M1_XCELL

Macrophage M2_CIBERSORT

Macrophage M2_CIBERSORT-ABS

Macrophage M2_QUANTISEQ

Macrophage M2_XCELL

Macrophage M2_TIDE

_Macrophage/Monocyte_MCPCOUNTER

Partial_Cor

ACC (n = 79)

1

BLCA (n = 408)

BRCA (n = 1100)

BRCA-Basal (n = 191)

BRCA-Her2 (n=82)

☒ ☒ ☒

☒ ☒

BRCA-LumA (n = 568)

BRCA-LumB (n = 219)

CESC (n = 306)

CHOL (n=36)

COAD (n = 458)

DLBC (n = 48)

ESCA (n = 185)

GBM (n = 153)

HNSC (n= 522)

HNSC-HPV- (n=422)

HNSC-HPV+ (n=98)

KICH (n = 66)

KIRC (n = 533)

KIRP (n = 290)

LGG (n = 516)

LIHC (n = 371)

0

LUAD (n = 515)

LUSC (n = 501)

MESO (n = 87) ☒

OV (n = 303) ☒

PAAD (n = 179) ☒

PCPG (n = 181)

PRAD (n = 498)

READ (n = 166)

SARC (n = 260)

SKCM (n = 471)

SKCM-Metastasis (n = 368)

SKCM-Primary (n = 103) ☒

☒ ☒

STAD (n = 415)

TGCT (n= 150)

THCA (n=509)

THYM (n = 120)

UCEC (n = 545)

UCS (n = 57)

☒ ☒

UVM (n= 80)

-1

p>0.05

☒ p … 0.05

(d)

shorter PFI in ACC (P=0.002), uveal melanoma (UVM) (P=0.009), KIRC (P=0.001), thyroid carcinoma (THCA) (P=0.025), and GBM (P=0.021) and longer PFI in DLBC (P=0.004) and CESC (P=0.031) (Figures 5(a)-5(g)).

These results indicate that high MMP-9 expression might be a risk factor for poor prognosis in ACC, BLCA, KIRC, LIHC, UVM, THCA, and GBM, while low MMP-9 expression might be a risk factor for poor prognosis in DLBC, UCEC, SKCM, and CESC.

3.3. Relationship between MMP-9 Expression and the Clinical Stage. Next, we analyzed the expression of MMP-9 in rela- tion to the tumor stage in the 33 cancer types and found that it was closely related to the clinical stage in eight tumors (Figures 6(a)-6(h)). MMP-9 was differentially expresses according to the clinical stage and was specifically positively correlated with the tumor stage in ACC, BLCA, and KIRC, in which MMP-9 expression increased with tumor progres- sion. These results suggest that MMP-9 expression has the potential to influence cancer prognosis by affecting lymph node metastasis. These results suggest that MMP-9 is involved in promoting cancer progression or metastasis.

3.4. Correlation between MMP-9 Expression and Immune Cell Infiltration. Many studies have shown that MMP-9 is related to immune cells [33, 34]. Therefore, we evaluated the correlation between MMP-9 and immune cell infiltration in 33 tumors. Through survival analysis and clinical correla- tion analysis, we found that MMP-9 was related to poor prognosis and metastasis in ACC and KIRC. DLBC was used as the control group. The correlation between the expression level of MMP-9 and six types of infiltrating immune cells in ACC, KIRC, and DLBC is shown in Figures 7(a)-7(c). The expression of MMP-9 was positively correlated with the infiltration of B cells, CD8+ cells, CD4+ cells, and macro- phages in ACC and KIRC, while it was mostly negatively correlated in DLBC. In addition, our results indicated a marked correlation between MMP-9 expression and the macrophage M0 in 28 cancer types (Table 1). MMP-9 was positively correlated with the macrophage M1 in four tumors (Figures 8(a)-8(d)). The levels of infiltrating macro- phage M2 were positively correlated with MMP-9 expres- sion in HNSC, CESC, and COAD (Figures 8(e)-8(g)) and negatively correlated in SKCM, LIHC, and THCA (Figures 8(h)-8(j)). In addition, TIMER2.0 analysis showed that MMP-9 had a strong positive correlation with macro- phages (Figure 7(d)). These results showed that high MMP-9 expression was positively correlated with immune cell infiltration.

3.5. Correlation between the MMP-9 Expression Level and Immune Cell Markers. The TME [35] can affect survival and tumor metastasis. We performed immune cell marker gene coexpression analyses in ACC, KIRC, and DLBC and found that the expression of MMP-9 was mainly positively correlated with the expression levels of most marker sets of T cells, TAMs, M2 macrophages, Th1 cells, and T cell exhaustion, especially in ACC (Table 2), while no such cor- relation was observed in DLBC.

TABLE 1: Correlation analysis between MMP-9 and macrophage MO of TCGA database (the P values are indicated as *P < 0.05, ** P < 0.01, and *** P < 0.001).
Cancer typeCorP value
ACC0.59
BLCA0.43***
BRCA0.58***
CESC0.29***
COAD0.36
DLBC0.56
ESCA0.29***
GBM0.61
HNSC0.32
KICH0.67***
KIRC0.52
KIRP0.61
LGG0.49***
LIHC0.24
LUAD0.23***
LUSC0.29***
MESO0.42
PAAD0.32***
PCPG0.61***
PRAD0.62***
SARC0.64***
SKCM0.30***
STAD0.26***
TGCT0.46***
THCA0.23***
THYM0.33***
UCEC0.29***
UCS0.72***

3.6. Coexpression of DNA Repair Genes with MMP-9 and GSEA. To better understand the potential mechanism of MMP-9 expression in cancers, we analyzed its expression in ACC, KIRC, and DLBC using GSEA. The results showed that MMP-9 was mainly enriched in immune-related path- ways in KIRC, such as immune response regulating cell sur- face receptor signaling and regulation of immune effector process (Figure 9(c)), and in pathways related to gene silenc- ing and RNA modification in ACC and DLBC (Figures 9(a) and 9(b)). We further used RNA sequence data from TCGA database to evaluate the correlation between MMP-9 and five DNA repair genes and found that MMP-9 was associ- ated with multiple DNA repair genes in various tumors (Figure 9(d)). More specifically, MMP-9 was moderately positively correlated with MSH2 in ACC and negatively cor- related with EPCAM and PMS2 in KIRC. In addition,

FIGURE 8: Continued.

Cancer: CESC

Cancer: LGG

10.0

R = 0.27, p= 3.7e=06

R = 0.31, p = 6.3e-09:

6

7.5

MMP9

MMP9

4

5.0

2

2.5

0.0

0

0.00

0.05

0.10

0.15

0.00

0.05

0.10

Macrophages M1

Macrophages M1

(a)

(b)

Cancer: LUAD

Cancer: OV

R =0.31, p = 5.6e-13

8

R =0.3, p = 2.6e-08

7.5

6

MMP9

5.0

MMP9

4

2.5

2

0.0

0.00

0.05

0.10

0.15

0.00

0.05

0.10

0.15

Macrophages M1

Macrophages M1

(c)

(d)

Cancer: HNSC

Cancer: CESC

10.0

10.0

R=0.19;p=2.3e:05

R = 0.21, p =0.00037

7.5

7.5

MMP9

MMP9

5.0

5.0

2.5

2.5

0.0

0.0

0.1

0.2

0.0

0.1

0.2

0.3

Macrophages M2

Macrophages M2

(e)

(f)

FIGURE 8: Correlation between MMP-9 gene expression and infiltrating levels of macrophage M1 and macrophage M2 of TCGA database in pan-cancer. MMP-9 was positively correlated with macrophage M1 (a) in CESC, (b) in LGG, (c) in LUAD, and (d) in OV. MMP-9 was positively correlated with macrophage M2 (e) in HNSC, (f) in CESC, and (g) in COAD. MMP-9 was negatively correlated with macrophage M2 (h) in SKCM, (i) in LIHC, and (j) in THCA.

Cancer: COAD

Cancer: SKCM

12

8

R = 0.18, p = 0.00014

R =- 0.29, p =2.6e-09 .

9

6

MMP9

MMP9

6

4

3

2

0.0

0.1

0.2

0

Macrophages M2

0.0

0.1

0.2

0.3

0.4

Macrophages M2

(g)

(h)

Cancer: LIHC

Cancer: THCA

8

R = - 0.31, p = 4.9e-08

R =- 0.26, p = 2.6e-07

7.5

6

MMP9

5.0

MMP9

4

2.5

2

0.0

0

0.0

0.1

0.2

0.3

0.4

0.5

0.1

0.2

0.3

0.4

Macrophages M2

Macrophages M2

(i)

(j)

MMP-9 showed a significant correlation with DNA repair genes in LGG and LIHC.

3.7. Correlation between the MMP-9 Expression Level and TMB, MSI, and Immune Checkpoint Genes. TMB and MSI are important for immunotherapy response. Here, we calcu- lated the TMB of each tumor sample and analyzed the cor- relation between MMP-9 and TMB in 33 tumors. MMP-9 was positively correlated with TMB in six tumors, including

ACC, BRCA, COAD, brain lower grade glioma (LGG), OV, and UCEC, and negatively correlated with HNSC and LUSC (Figure 10(a)). Next, we analyzed the correlation between MSI and MMP-9 levels. MSI was positively correlated with MMP-9 in COAD and sarcoma, whereas it was negatively correlated in four tumors (Figure 10(b)). In addition, most immune checkpoint genes were coexpressed with MMP-9, especially PDCD1 and CTL4, which are the targets of immune checkpoint inhibitors.

TABLE 2: Correlation analysis between MMP-9 and related genes and markers of immune cells of TCGA database (the P values are indicated as *P < 0.05, ** P < 0.01, and *** P < 0.001).
DescriptionGene markersACCKIRCDLBC
CorP valueCorP valueCorP value
CD8+ T cellCD8A0.336**0.1110.0980.506
CD8B0.363***0.0950.0210.886
T cell (general)CD3D0.405***0.192***0.1300.379
CD3E0.371***0.201***0.0980.510
CD20.342**0.182***0.1230.403
B cellCD19-0.0400.7230.306***0.0810.583
CD79A0.0500.6640.316***0.0660.654
MonocyteCD860.292**0.248***0.1010.493
CSF1R0.2200.0510.255***0.2310.114
TAMCCL20.0710.537-0.0710.1020.2370.105
CD680.2540.0240.255***0.3430.017
IL-100.545***0.302***0.3600.012
M1 macrophageNOS20.408***-0.0630.1480.1760.233
IRF50.1690.1370.0620.1530.0490.739
PTGS20.505***0.228***0.1510.305
M2 macrophageCD1630.400***0.305***0.1760.232
VSIG40.350**0.342***0.1200.418
MS4A4A0.375***0.302***0.3400.018
NeutrophilsCEACAM8 (CD66b)0.1870.0980.0040.9250.1500.310
ITGAM (CD11b)0.290**0.202***0.4870.000
Natural killer cellKIR2DL10.1010.376-0.0290.5030.0470.753
KIR2DL30.0340.765-0.0600.1660.0630.671
KIR2DL40.318**0.0850.1060.473
KIR3DL10.1420.211-0.1110.1410.338
KIR3DL2-0.2020.074-0.0140.7540.1030.485
KIR3DL30.1590.1610.0260.5410.0430.771
KIR2DS40.1450.2030.0140.7550.0650.658
Dendritic cellHLA-DPB10.1610.1570.147***0.1960.182
HLA-DQB10.1170.3060.0370.3910.1280.384
HLA-DRA0.1370.2270.143***0.1200.418
HLA-DPA10.0750.5100.118**0.1310.376
NRP1 (BDCA-4)0.2390.0340.0440.3050.1390.348
CD1C (BDCA-1)0.0110.9240.0900.0090.952
ITGAX (CD11c)0.340**0.271***0.5360.000
Th1TBX210.434***0.0530.2240.0760.610
STAT40.463***0.178***0.0670.653
STAT10.301**0.0620.1540.0960.517
IFNG (TNF-y)0.397***0.0980.1370.354
TNF (TNF-a)0.0700.5410.0690.1100.2040.165
TABLE 2: Continued.
DescriptionGene markersACCKIRCDLBC
CorP valueCorP valueCorP value
Th2GATA30.0260.8200.0440.3080.1340.363
STAT6-0.2060.069-0.0540.2160.3650.011
STAT5A0.1980.0800.193***0.1170.429
IL-130.0350.7620.0440.3110.0990.504
TfhBCL60.0870.4450.191***0.1000.498
IL-210.0001.0000.165***0.0610.680
Th17STAT30.2000.0780.0870.3710.009
IL-17A0.0001.0000.0650.1350.0230.879
TregFOXP30.1620.1530.385***0.1960.181
CCR80.0150.8990.250***0.1810.217
STAT5B0.0170.880-0.193***0.2290.118
T cell exhaustionTGFB10.522***0.411***0.0900.542
PDCD1 (PD-1)0.399***0.141**0.0070.961
CTLA40.392***0.155***0.1770.230
LAG30.412***0.169***0.0260.862
HAVCR2 (TAM-3)0.299**0.0400.3580.0540.715
GZMB0.551***0.140**0.1630.269

4. Discussion

MMP-9 can degrade the extracellular matrix components and promote tumor invasion and metastasis. The high expression of MMP-9 is closely related to the development, invasion, and metastasis in many cancers. Here, we found that MMP-9 promotes cancer development and progression in some cancers, suggesting that MMP-9 expression can be used to predict metastasis, especially in kidney cancer. In addition, correlation analysis showed that the expression of MMP-9 was correlated with different levels of immune infil- tration and immunological markers. Finally, we evaluated the relationship between MMP-9 expression and TMB and MSI. The results showed that MMP-9 may be used as a bio- marker for pan-cancer prognosis.

In this study, we obtained the expression levels of MMP- 9 and the prognosis and relevant indices of 33 cancer types from TCGA database. Differential expression of MMP-9 in cancer and normal tissues was observed in all cancers, with MMP-9 being overexpressed in tumor tissue across cancer types. This suggested that dysregulated or excessive MMP- 9 could cause tumorigenesis. As for the survival analysis, higher expression levels of MMP-9 were correlated with poorer prognosis in patients with ACC, BLCA, KIRC, and LIHC. In contrast, high levels of MMP-9 were favorable for the prognosis of lymphoma. The results indicated that MMP-9 promotes bladder and cervical cancer invasion and metastasis. MMP-9 is a potential prognostic biomarker for various cancers, including lung, ovarian, pancreatic, and breast cancers [11, 16]. However, in our study, analysis based on three survival indicators showed that high MMP-

9 expression was associated with poor prognosis in ACC and KIRC. The correlation between MMP-9 and renal can- cers has not been reported in previous studies. In addition, our analysis of OS, DSS, and PFI showed that high expres- sion of MMP-9 is a protective factor in DLBC; however, this has not been observed in previous studies. MMP-9 promotes metastasis via ECM decomposition [36]. The expression of MMP-9 was related to the clinical stage in eight tumors, sug- gesting that MMP-9 may be involved in tumor metastasis. In addition, MMP-9 increased with the progression of cancer in three types of urological tumors. These results suggest that MMP-9 may be used as an indicator of prognosis and metastasis in pan-cancer.

Furthermore, we found that MMP-9 expression was cor- related with immune infiltration levels in multiple cancer types, especially ACC and KIRC. It was positively correlated with the infiltration of B cells, CD8+ cells, CD4+ cells, and macrophages in ACC and KIRC, while it was mostly nega- tively correlated in DLBC. This suggests that MMP-9 may lead to poor prognosis by participating in tumor immune infiltration. Moreover, MMP-9 expression levels were mainly positively correlated with immune cell markers. Notably, in ACC, MMP-9 was moderately correlated with four Th1 marker genes (TBX21, STAT4, STAT1, and IFNG), suggesting that it may be involved in Th1 differentiation. Th1 cells induce the activation of macrophages, NK cells, B cells, and CD8+ T cells [37]. Concurrently, we also found that MMP-9 was moderately correlated with the immune markers of CD8+ T cells (CD8A and CD8B) and T cells (CD3D, CD3E, and CD2). These results suggest that MMP-9 may promote cell-mediated inflammatory

FIGURE 9: Continued.

0.8

Running enrichment score

0.6

0.4

0.2

DLBC

0.0

15

Ranked list metric

10

5

0

-5

-10

-15

10000

20000

30000

40000

50000

Rank in ordered dataset

GO_GENE_SILENCING

GO_GENE_SILENCING_BY_RNA

GO_MRNA_BINDING

GO_NEGATIVE_REGULATION_OF_BLOOD_VESSEL_ENDOTHELIAL_CELL_MIGRATION

GO_NEGATIVE_REGULATION_OF_CELLULAR_AMIDE_METABOLIC_PROCESS

(a)

FIGURE 9: Continued.

0.6

Running enrichment score

0.4

0.2

ACC

0.0

15

Ranked list metric

10

5

0

-5

-10

-15

10000

20000

30000

40000

50000

Rank in ordered dataset

GO_CELLULAR_RESPONSE_TO_BIOTIC_STIMULUS

GO_DEFENSE_RESPONSE_TO_GRAM_POSITIVE_BACTERIUM

GO_KERATINOCYTE_DIFFERENTIATION

GO_MODIFICATION_OF_MORPHOLOGY_OR_PHYSIOLOGY_OF_OTHER_ORGANISM

GO_RNA_3_END_PROCESSING

(b)

0.5

Running enrichment score

0.4

0.3

0.2

KIRC

0.1

0.0

10

Ranked list metric

5

0

-5

-10

10000

20000

30000

40000

50000

Rank in ordered dataset

GO_EPIDERMIS_DEVELOPMENT

GO_IMMUNE_RESPONSE_REGULATING_CELL_SURFACE_RECEPTOR_SIGNALING_PATHWAY

GO_NEGATIVE_REGULATION_OF_HYDROLASE_ACTIVITY

GO_POSITIVE_REGULATION_OF_CYTOKINE_PRODUCTION

GO_REGULATION_OF_IMMUNE_EFFECTOR_PROCESS

(c)

Coexpression across cancer types

FIGURE 9: Pathway analysis of MMP-9 in different cancers and DNA repair gene coexpression analysis with MMP-9 of TCGA database. (a) GO functional annotation of MMP-9 in DLBC, (b) GO functional annotation of MMP-9 in ACC, (c) GO functional annotation of MMP-9 in KIRC, and (d) DNA repair gene coexpression analysis with MMP-9. Each small rectangular module represents the coexpression of DNA repair genes and MMP-9 in cancer, where the upper left corner is the P value, where *P< 0.05, ** P < 0.01, and *** P < 0.001, and the lower right corner is the correlation coefficient.

EPCAM

0.9

P value

MSH2

0

MSH6

PMS2

0.4

MLH1

Cor

ACC

BLCA

BRCA

CESC

CHOL

COAD

DLBC

ESCA

GBM

HNSC

KICH

KIRC

KIRP

LAML

LGG

LIHC

LUAD

LUSC

MESO

OV

PAAD

PCPG

PRAD

READ

SARC

SKCM

STAD

TGCT

THCA

THYM

UCEC

UCS

UVM

-0.4

Cancer typest

(d)

FIGURE 10: Continued.

BLCA ACC **

UVM

BRCA ***

UCS

CESC

0.4

UCEC **

CHOL

THYM

0.2

COAD*

THCA

0

DLBC

TGCT

0.2

ESCA

S TAD

0.4

GBM

SKCM

HNSC ***

SARC

KICH

READ

KIRC

PRAD

KIRP

PCPG

LAML

PAAD

LGG ***

OV*

LIHC

LUAD LUSC ***

MESO

(a)

BLCA

ACC

UVM

BRCA

UCS

CESC

0.3

UCEC

CHOL

THYM

0.15

COAD ***

THCA

0

DLBC

TGCT **

-0.15

ESCA*

STAD

0.3

GBM

SKCM **

HNSC

SARC*

KICH

READ

KIRC **

PRAD

KIRP

PCPG

LAML

PAAD

LGG

OV

LIHC

LUAD LUSC

MESO

(b)

Coexpression across cancer types

FIGURE 10: Correlation between MMP-9 gene expression and TMB and MSI and coexpression between MMP-9 and immunological checkpoint genes of TCGA database in pan-cancer. (a) Correlation between MMP-9 and TMB in 33 cancer types. (b) Correlation between MMP-9 and MSI in 33 cancer types. (c) Coexpression of MMP-9 and immunological checkpoint genes, *P < 0.05, ** P < 0.01, and *** P < 0.001.

TNFRSF9

1

CD44

CD86

☒ ☒

☒ ☒

☒ ☒

CD274

TIGIT

TNFSF15

TNFRSF18

CD40

TNFRSF4

P value

VSIR

TNFRSF25

CD27

TNFRSF8

☒ ☒

TNFSF9

CD70

☒ ☒

☒ ☒

BTNL2

TNFSF18

HHLA2

PDCD1LG2

IDO1

VTCN1

TMIGD2

ICOSLG

0

IDO2

TNFSF14

☒ ☒

CD160

LGALS9

0.7

PDCD1

CD80

KIR3DL1

CD276

ADORA2A

HAVCR2

CD200R1

CD28

CD48

CTLA4

Cor

CD40LG

ICOS

LAG3

CD244

TNFSF4

LAIR1

NRP1

TNFRSF14

CD200

BTLA

ACC

0.4

BLCA

BRCA

CESC

CHOL

COAD

DLBC

ESCA

GBM

HNSC

KICH

KIRC

KIRP

LAML

LGG

LIHC

LUAD

LUSC

MESO

OV

PAAD

PCPG

PRAD

READ

SARC

SKCM

STAD

TGCT

THCA

THYM

UCEC

UCS

UVM

Cancer types

(c)

responses by participating in Th1 differentiation and T cell activation. Th1 cells regulate macrophage function at multi- ple levels. In addition, MMP-9 was associated with macro- phage immune marker genes. More specifically, MMP-9 expression was positively correlated with IL-10 (a TAM marker), which is often associated with tumor immune eva- sion. The markers of M2 macrophages were moderately cor- related with MMP-9 expression in tumors, suggesting that MMP-9 may be involved in the differentiation of macro- phages. Most importantly, in ACC, MMP-9 expression was strongly correlated with most markers of T cell exhaustion, including TGFB1, PDCD1, CTLA4, LAG3, and GZMB. T cell exhaustion is one of the main causes of immune dys- function that leads to a poor prognosis [38]. This suggests that MMP-9 may be the cause of poor prognosis in patients with ACC. At the same time, T cell exhaustion is also one of the reasons for poor immunotherapy response. In contrast, TAMs are important cellular components of the TME [39]

and imbalance of M1/M2 plays a key role in tumor progres- sion, immune escape, and drug resistance [40]. Therefore, the development of antineoplastic drugs that target macro- phage polarization is important. Tekin et al. [18] found that M0 macrophages secrete MMP-9 in the early stages of pan- creatic cancer development, which promotes tumor progres- sion. This is consistent with the findings of our study. In addition, we found that MMP-9 was highly positively corre- lated with M0 macrophage levels in 27 types of tumors. Although research has shown that M2 macrophages can alter miR-149-5p to increase the expression of MMP-9 in liver cancer [41], in our study, MMP-9 and M2 macrophages were negatively correlated in LIHC. These results indicated that MMP-9 is involved in the recruitment and activation of immune cells and that MMP-9 inhibition may be another approach for tumor immunotherapy based on macrophages.

In this study, MMP-9 expression was associated with TMB in eight cancer types and with MSI in six cancer types.

In ACC, MMP-9 was highly correlated with the markers of T cell exhaustion, which can be reversed by PD-1 inhibitors. A recent study [42] identified TMB as a marker for evaluat- ing the therapeutic effect of PD-1 inhibitors. Therefore, we analyzed the relationship between MMP-9 expression and TMB expression. Our results also showed that MMP-9 has a significant positive correlation with TMB in a variety of cancers. This suggests that in these cancers, patients with high MMP-9 expression may be more suitable for immuno- suppressive therapy. Furthermore, MSI plays an important role in the diagnosis, prognosis, and treatment of multiple tumors, especially colon cancer [43]. Our results showed that MSI is positively correlated with MMP-9 in COAD. In brief, patients with high MMP-9 expression may be more suitable for immunotherapy.

Immune checkpoints are closely related to tumor immune escape. Hence, we analyzed the relationship between the expression of MMP-9 and certain common immune checkpoint genes. The results showed that MMP- 9 was significantly associated with immune checkpoints in most tumor types. This may be related to the poor prognosis of some tumors in the survival analysis. Another study [44] indicated that inhibition of MMP-2/MMP-9 improves the efficacy of PD-1 or CTLA4 blockade in the treatment of pri- mary and metastatic tumors.

Monferran et al. [45] reported that the DNA repair pro- tein Ku interacts with MMP-9 at the cell membrane of highly invasive hematopoietic cells. Our results also showed that MMP-9 was correlated with various DNA repair genes. These findings may help in understanding the role of MMP-9 in gene expression and gene repair. The GSEA results also suggested that MMP-9 participates in immune regulation. This is consistent with the results of our previ- ous analysis. This suggests that MMP-9 is a potential target for immunotherapy.

Although we comprehensively analyzed MMP-9 expres- sion in 33 tumors, many deficiencies exist in our study. First, our data source was relatively single and simple as we used mainly TCGA database data. Second, our findings require further validation in the clinical setting. Third, although we found that the expression of MMP-9 is related to immune cell infiltration and survival, we could not prove its causal relationship, and hence, its prognostic value needs to be fur- ther studied.

In conclusion, MMP-9 can be used as a pan-cancer prog- nostic biomarker involving immune infiltration, especially in kidney cancer. These findings may contribute to clinical decision-making and cancer immunotherapy.

Abbreviations

MMP-9: Matrix metalloproteinase-9

TMB: Tumor mutation burden

DSS: Disease-specific survival

PFI: Progression-free interval

TME: The tumor microenvironment

MSI: DNA microsatellite instability

ACC: Adrenocortical carcinoma

BLCA: Bladder urothelial carcinoma

BRCA:Breast invasive carcinoma
CESC:Cervical squamous cell carcinoma and endocervi- cal adenocarcinoma
CHOL:Cholangiocarcinoma
COAD:Colon adenocarcinoma
DLBC:Lymphoid neoplasm diffuse large B cell lymphoma
ESCA:Esophageal carcinoma
GBM:Glioblastoma multiforme
HNSC:Head and neck squamous cell carcinoma
KICH:Kidney chromophobe
KIRC:Kidney renal clear cell carcinoma
KIRP:Kidney renal papillary cell carcinoma
LAML:Acute myeloid leukemia
LGG:Brain lower grade glioma
LIHC:Liver hepatocellular carcinoma
LUAD:Lung adenocarcinoma
LUSC:Lung squamous cell carcinoma
MESO:Mesothelioma
OV:Ovarian serous cystadenocarcinoma
PAAD:Pancreatic adenocarcinoma
PCPG:Pheochromocytoma and paraganglioma
PRAD:Prostate adenocarcinoma
READ:Rectum adenocarcinoma
SARC:Sarcoma
SKCM:Skin cutaneous melanoma
STAD:Stomach adenocarcinoma
TGCT:Testicular germ cell tumors
THCA:Thyroid carcinoma
THYM:Thymoma
UCEC:Uterine corpus endometrial carcinoma
UCS:Uterine carcinosarcoma
UVM:Uveal melanoma.

Data Availability

The datasets obtained from UCSC Xena (http://xena.ucsc .edu/), partial analysis by GEPIA2 database (http://gepia2 .cancer-pku.cn/#analysis) TIMER (https://cistrome.shinyapps .io/timer/), and TIMER2.0 database (http://timer.comp- genomics.org/).

Conflicts of Interest

The authors declare that there are no potential conflicts of interest.

Acknowledgments

This study was supported by the National Key Clinical Specialty Construction Project (clinical pharmacy) and High Level Clinical Key Specialty (clinical pharmacy) in Guang- dong province. This work was supported by the project of the Chinese Ministry of Education (no. 2017A11001) and Research on Prediction Trend of Population Infected with COVID-19 Based on Big Data (2020KZDZX1126).

References

[1] T. Klein and R. Bischoff, “Physiology and pathophysiology of matrix metalloproteases,” Amino Acids, vol. 41, no. 2, pp. 271-290, 2011.

[2] L. Liu, Y. Ye, and X. Zhu, “MMP-9 secreted by tumor associ- ated macrophages promoted gastric cancer metastasis through a PI3K/AKT/Snail pathway,” Biomedicine & Pharmacother- apy, vol. 117, 2019.

[3] H. Zhang, B. Zhao, Z. G. Zhai, J. D. Zheng, Y. K. Wang, and Y. Y. Zhao, “Expression and clinical significance of MMP-9 and P53 in lung cancer,” European Review for Medical and Pharmacological Sciences, vol. 25, no. 3, pp. 1358-1365, 2021.

[4] P. Ruan, Y. Gu, H. Zhou et al., “Expression and clinical signif- icance of CD74 and MMP-9 in colon adenocarcinomas,” Jour- nal of BU ON .: official journal of the Balkan Union of Oncology, vol. 25, no. 2, pp. 927-932, 2020.

[5] Z. Ran, L. Hou, H. Guo, K. Wang, and X. Li, “Expression of VEGF, COX-2 and MMP-9 in breast cancer and their relation- ship with ultrasound findings,” International Journal of Clini- cal and Experimental Pathology, vol. 11, no. 9, pp. 4264-4269, 2018.

[6] C. Gialeli, A. D. Theocharis, and N. K. Karamanos, “Roles of matrix metalloproteinases in cancer progression and their pharmacological targeting,” The FEBS Journal, vol. 278, no. 1, pp. 16-27, 2011.

[7] C. Mehner, A. Hockla, E. Miller, S. Ran, D. C. Radisky, and E. S. Radisky, “Tumor cell-produced matrix metalloproteinase 9 (MMP-9) drives malignant progression and metastasis of basal-like triple negative breast cancer,” Oncotarget, vol. 5, no. 9, pp. 2736-2749, 2014.

[8] H. Huang, “Matrix metalloproteinase-9 (MMP-9) as a cancer biomarker and MMP-9 biosensors: recent advances,” Sensors, vol. 18, no. 10, p. 3249, 2018.

[9] R. Roy, J. Yang, and M. A. Moses, “Matrix metalloproteinases as novel biomarkers and potential therapeutic targets in human cancer,” Journal of Clinical Oncology, vol. 27, no. 31, pp. 5287-5297, 2009.

[10] L. Chen, J. Zhang, Y. He, and X. Y. Ding, “Matrix metalloproteinase-9 expression of GCTSC in peripheral tissue and central tissue of GCTB,” Journal of Cellular Biochemistry, vol. 119, no. 7, pp. 5805-5812, 2018.

[11] M. Zajkowska, M. Zbucka-Krętowska, I. Sidorkiewicz et al., “Human plasma levels of vascular endothelial growth factor, matrix metalloproteinase 9, and tissue inhibitor of matrix metalloproteinase 1 and their applicability as tumor markers in diagnoses of cervical cancer based on ROC analysis,” Cancer Control, vol. 25, no. 1, article 107327481878935, 2018.

[12] M. Zajkowska, M. Zbucka-Krętowska, I. Sidorkiewicz et al., “Plasma levels and diagnostic utility of macrophage-colony stimulating factor, matrix metalloproteinase-9 and tissue inhibitor of metalloproteinase-1 as tumor markers in cervical cancer patients,” Tumour Biology, vol. 40, no. 7, article 1010428318790363, 2018.

[13] A. T. Reiner, S. Tan, C. Agreiter et al., “EV-associated MMP9 in high-grade serous ovarian cancer is preferentially localized to annexin V-binding EVs,” Disease Markers, vol. 2017, Article ID 9653194, 2017.

[14] C. M. Coticchia, A. S. Curatolo, D. Zurakowski et al., “Urinary MMP-2 and MMP-9 predict the presence of ovarian cancer in women with normal CA125 levels,” Gynecologic Oncology, vol. 123, no. 2, pp. 295-300, 2011.

[15] B. Grünwald, J. Vandooren, E. Locatelli et al., “Matrix metalloproteinase-9 (MMP-9) as an activator of nanosystems for targeted drug delivery in pancreatic cancer,” Journal of Controlled Release, vol. 239, no. 239, pp. 39-48, 2016.

[16] S. G. Dofara, S. L. Chang, and C. Diorio, “Gene polymor- phisms and circulating levels of MMP-2 and MMP-9: a review of their role in breast cancer risk,” Anticancer Research, vol. 40, no. 7, pp. 3619-3631, 2020.

[17] C. Mattiuzzi and G. Lippi, “Current cancer epidemiology,” Journal of Epidemiology and Global Health, vol. 9, no. 4, pp. 217-222, 2019.

[18] C. Tekin, H. L. Aberson, C. Waasdorp et al., “Macrophage- secreted MMP9 induces mesenchymal transition in pancreatic cancer cells via PAR1 activation,” Cellular Oncology, vol. 43, no. 6, pp. 1161-1174, 2020.

[19] T. A. Chan, M. Yarchoan, E. Jaffee et al., “Development of tumor mutation burden as an immunotherapy biomarker: utility for the oncology clinic,” Annals of Oncology, vol. 30, no. 1, pp. 44-56, 2019.

[20] L. Chang, M. Chang, H. M. Chang, and F. Chang, “Microsatel- lite instability: a predictive biomarker for cancer immunother- apy,” Applied Immunohistochemistry & Molecular Morphology, vol. 26, no. 2, pp. e15-e21, 2018.

[21] X. Cui, X. Zhang, M. Liu et al., “A pan-cancer analysis of the oncogenic role of staphylococcal nuclease domain- containing protein 1 (SND1) in human tumors,” Genomics, vol. 112, no. 6, pp. 3958-3967, 2020.

[22] J. Hu, J. Xu, X. Feng, Y. Li, F. Hua, and G. Xu, “Differential expression of the TLR4 gene in pan-cancer and its related mechanism,” Frontiers in Cell and Developmental Biology, vol. 9, article 700661, 2021.

[23] Y. Miao, J. Wang, Q. Li et al., “Prognostic value and immuno- logical role of PDCD1 gene in pan-cancer,” International Immunopharmacology, vol. 89, no. Part B, article 107080, 2020.

[24] M. J. Goldman, B. Craft, M. Hastie et al., “Visualizing and interpreting cancer genomics data via the Xena platform,” Nature Biotechnology, vol. 38, no. 6, pp. 675-678, 2020.

[25] GTEx Consortium, “The genotype-tissue expression (GTEx) project,” Nature Genetics, vol. 45, no. 6, pp. 580-585, 2013.

[26] Z. Tang, B. Kang, C. Li, T. Chen, and Z. Zhang, “GEPIA2: an enhanced web server for large-scale expression profiling and interactive analysis,” Nucleic Acids Research, vol. 47, no. W1, pp. W556-W560, 2019.

[27] T. M. Therneau and P. M. Grambsch, Modeling Survival Data: Extending the Cox Model, Springer, New York, 2000.

[28] M. E. Ritchie, B. Phipson, D. Wu et al., “limma powers differ- ential expression analyses for RNA-sequencing and microar- ray studies,” Nucleic Acids Research, vol. 43, no. 7, article e47, 2015.

[29] T. Li, J. Fan, B. Wang et al., “TIMER: a web server for compre- hensive analysis of tumor-infiltrating immune cells,” Cancer Research, vol. 77, no. 21, pp. e108-e110, 2017.

[30] T. Li, J. Fu, Z. Zeng et al., “TIMER2.0 for analysis of tumor- infiltrating immune cells,” Nucleic Acids Research, vol. 48, no. W1, pp. W509-W514, 2020.

[31] A. M. Newman, C. L. Liu, M. R. Green et al., “Robust enumer- ation of cell subsets from tissue expression profiles,” Nature Methods, vol. 12, no. 5, pp. 453-457, 2015.

[32] G. Yu, L. G. Wang, Y. Han, and Q. Y. He, “clusterProfiler: an R package for comparing biological themes among gene clus- ters,” OMICS, vol. 16, no. 5, pp. 284-287, 2012.

[33] V. C. Ardi, T. A. Kupriyanova, E. I. Deryugina, and J. P. Quig- ley, “Human neutrophils uniquely release TIMP-free MMP-9 to provide a potent catalytic stimulator of angiogenesis,” Pro- ceedings of the National Academy of Sciences of the United States of America, vol. 104, no. 51, pp. 20262-20267, 2007.

[34] E. I. Deryugina, E. Zajac, A. Juncker-Jensen, T. A. Kupriya- nova, L. Welter, and J. P. Quigley, “Tissue-infiltrating neutro- phils constitute the major in vivo source of angiogenesis- inducing MMP-9 in the tumor microenvironment,” Neoplasia, vol. 16, no. 10, pp. 771-788, 2014.

[35] B. Arneth, “Tumor Microenvironment,” Medicina, vol. 56, no. 1, p. 15, 2019.

[36] S. Mondal, N. Adhikari, S. Banerjee, S. A. Amin, and T. Jha, “Matrix metalloproteinase-9 (MMP-9) and its inhibitors in cancer: A minireview,” European Journal of Medicinal Chemis- try, vol. 194, article 112260, 2020.

[37] M. Kanamori, H. Nakatsukasa, M. Ito, S. Chikuma, and A. Yoshimura, “Reprogramming of Th1 cells into regulatory T cells through rewiring of the metabolic status,” International Immunology, vol. 30, no. 8, pp. 357-373, 2018.

[38] M. Ando, M. Ito, T. Srirat, T. Kondo, and A. Yoshimura, “Memory T cell, exhaustion, and tumor immunity,” Immuno- logical Medicine, vol. 43, no. 1, pp. 1-9, 2020.

[39] A. J. Petty and Y. Yang, “Tumor-associated macrophages: implications in cancer immunotherapy,” Immunotherapy, vol. 9, no. 3, pp. 289-302, 2017.

[40] C. Yunna, H. Mengru, W. Lei, and C. Weidong, “Macrophage M1/M2 polarization,” European Journal of Pharmacology, vol. 877, article 173090, 2020.

[41] G. Liu, L. Yin, X. Ouyang, K. Zeng, Y. Xiao, and Y. Li, “M2 macrophages promote HCC cells invasion and migration via miR-149-5p/MMP9 signaling,” Journal of Cancer, vol. 11, no. 5, pp. 1277-1287, 2020.

[42] D. Sha, Z. Jin, J. Budczies, K. Kluck, A. Stenzinger, and F. A. Sinicrope, “Tumor mutational burden as a predictive bio- marker in solid tumors,” Cancer Discovery, vol. 10, no. 12, pp. 1808-1825, 2020.

[43] G. L. de Angelis, L. Bottarelli, C. Azzoni et al., “Microsatellite instability in colorectal cancer,” Acta Biomedica, vol. 89, no. 9-S, pp. 97-101, 2018.

[44] Y. Ye, X. Kuang, Z. Xie et al., “Small-molecule MMP2/MMP9 inhibitor SB-3CT modulates tumor immune surveillance by regulating PD-L1,” Genome Medicine, vol. 12, no. 1, p. 83, 2020.

[45] S. Monferran, J. Paupert, S. Dauvillier, B. Salles, and C. Muller, “The membrane form of the DNA repair protein Ku interacts at the cell surface with metalloproteinase,” The EMBO Journal, vol. 23, no. 19, pp. 3758-3768, 2004.