Downregulated mitochondrial transcription factor A enhances mycoplasma infection to promote the metastasis of hepatocellular carcinoma

Abstract

Mycoplasma is widespread in various hosts and may cause various diseases in animals. Interestingly, the occurrence of mycoplasma infection was observed in many tumor types. However, the mechanism regulating its infection is far from clear. We unexpectedly found that the knockdown of mitochondrial transcription factor A (TFAM) remarkably enhanced mycoplasma infection in hepatocellular carcinoma (HCC) cells. More importantly, we found that mycoplasma infection facilitated by TFAM knockdown significantly promoted HCC cell metastasis. Mycoplasma infection was further found to be positively correlated with poor prognosis in patients with HCC. Mechanistically, the decreased TFAM expression upregulated the transcription factor Sp1 to increase the expression level of Annexin A2 (ANXA2), which was reported to interact with membrane protein of mycoplasma. Moreover, we found that mycoplasma infection enhanced by the TFAM downregulation promoted HCC migration and invasion by activating the nuclear factor‐κB signaling pathway. The downregulation of TFAM enhanced mycoplasma infection in HCC cells and promoted HCC cell metastasis. Our study contributes to the understanding of the pathological role of mycoplasma infection and provides supporting evidence that targeting TFAM could be a potential strategy for the treatment of HCC with mycoplasma infection.

First, we found that the downregulation of TFAM facilitated mycoplasma infection in HCC cells. Second, mycoplasma infection promoted the metastasis of HCC cells and was significantly correlated with the prognosis of HCC patients. Third, the decreased TFAM expression upregulated the transcription factor Sp1 to increase ANXA2 expression and then promoted mycoplasma infection.

Abbreviations

ANXA2

annexin A2

CCU

color change unit

CTC

circulating tumor cell

FMMU

Fourth Military Medical University

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

IHC

immunohistochemistry

NF‐κB

nuclear factor‐κB

p‐p65

phosphorylated p65 subunit of NF‐κB

qPCR

quantitative PCR

ROS

reactive oxygen species

RT

room temperature

Sp1

specificity protein 1

TFAM

mitochondrial transcription factor A

WGA

wheat germ agglutinin

1. INTRODUCTION

Infection with certain viruses, bacteria, and parasites is responsible for approximately 16.1% of new cancer cases worldwide. ¹ For instance, the major risk factors for HCC are chronic hepatitis B virus and HCV infection, and they are associated with malignant progression. ² However, the role of other infectious microorganisms in human HCC has not been extensively studied.

Mycoplasma, a common contaminant in cultured mammalian cells, is now being seen as a key mediator in modifying cancer susceptibility and tumor progression. ³ , ⁴ Mycoplasma infection commonly exists in numerous tumor tissues, including gastric cancer, breast cancer, lung cancer, and prostate cancer, indicating that mycoplasma infection is intimately associated to carcinogenesis. ⁵ , ⁶ , ⁷ Moreover, this evidence also indicates that mycoplasma in cultured tumor cells might directly originate from infected tumor tissues and not only from the culture environment. Mycoplasma infection has recently been proven to affect the host's metabolic pathways ⁸ and considerably facilitate the development of gastric cancer. ⁹ However, the factors mediating mycoplasma infection and the biological effects of mycoplasma infection in HCC remain unclear.

Mitochondria are catabolic organelles and the primary generator of cellular ROS and ATP, which play crucial parts in innate immunological reactions to infection, cellular damage, and stress. ¹⁰ , ¹¹ Mitochondrial transcription factor A, a protein that binds to mtDNA, is crucial for maintaining the genome. ¹² , ¹³ Recently, we found that in metastatic HCC tissues, TFAM was dramatically downregulated and was correlated with overall survival and time to tumor recurrence in HCC patients. ¹⁴ Moreover, TFAM has been linked to the development of malignant behavior by influencing the invasion gene signatures of melanoma cells and metabolic profiles in colon tumor cells. ¹⁵ Nonetheless, it is unclear whether the dysregulated expression of TFAM is related to mycoplasma infection and thus affects tumor progression.

In our current study, we found that TFAM knockdown facilitated mycoplasma infection in HCC cells, which in turn significantly promoted the metastasis of HCC cells. The underlying molecular mechanism was further explored. Our study contributes to the understanding of the pathological role of mycoplasma infection and provides supporting evidence that targeting TFAM could be a potential strategy for the treatment of HCC with mycoplasma infection.

2. MATERIALS AND METHODS

2.1. Cell culture and patient sample collection

The human HCC cells SNU‐739 and SNU‐368 were acquired from the Korean Cell Line Bank; HLE, Hep‐3B, BEL‐7402, and MHCC‐97 L were kept in our laboratory. The cells were cultured in RPMI‐1640 (Gibco) or DMEM (Gibco) medium supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin solution. Cells were typically kept at 37°C in a humidified environment with 5% CO₂. BAY 11‐7082 (a NF‐κB inhibitor) and XJB‐5‐131 (a mitochondria‐targeted ROS scavenger) came from the MedChemExpress (Cat. No. HY‐13453; HY‐129460). The FMMU Center for DNA Typing used short tandem repeat DNA testing to verify the authenticity of all HCC cell lines. In addition, 100 tissue samples from HCC patients were gathered at the Xijing Hospital affiliated with the FMMU in Xi'an, China (Table 1). All participants supplied written informed permission and the study was authorized by the ethics committee of the FMMU (KY20213188‐1).

TABLE 1.

Clinical characteristics of patients with hepatocellular carcinoma (HCC)

Variable	No. of cases
Total	100
Age, years
<54	52
≥54	48
Gender
Female	85
Male	15
HBsAg
Negative	10
Positive	90
AFP, ng/ml
<200	58
≥200	42
HCC size, cm
<5	30
≥5	70
Tumor number
Single	81
Multiple	19
Differentiation
I + II	30
III + IV	70
TNM
I + II	79
III + IV	21
PVTT
Negative	87
Positive	13
Adjuvant therapy
No	72
TACE	28
Survival
Alive	58
Dead	42
Relapse
No	34
Yes	66

Abbreviations: AFP, alpha‐fetoprotein; HBsAg, hepatitis B surface antigen; PVTT, portal vein tumor thrombus; TACE, transarterial chemoembolization.

2.2. Immunofluorescence assay

Hepatocellular carcinoma cells were fixed with 4% paraformaldehyde for 40 min, treated with 0.5% Triton X‐100 for 15 min, and blocked with 2% BSA at RT for 35 min. The samples were first added with primary Abs (Table 2) and then left at 4°C overnight. On the second day, the corresponding secondary Ab was added for 1 h at RT. Finally, the samples were stained with DAPI at RT for 15 min. Images were taken using a laser‐scanning confocal microscope (Olympus FV3000). ImageJ software was used to measure the fluorescence intensity and area in the region of dsDNA and p37 on cell membranes, and TFAM in cytoplasm (NIH).

TABLE 2.

Primary Abs used for western blotting (WB), immunohistochemistry (IHC), and immunofluorescence (IF)

Antibody	Company (Cat. No.)	Working concentration dilutions
TFAM	Abcam (ab47517)	WB 1:1000, IF 1:500
dsDNA	Abcam (ab27156)	IF 1:500
Custom‐made rabbit polyclonal anti‐P37	ABclonal	IHC 1:1000, WB 1:1000
Tubulin	TDYBIO (TDY054F)	WB 1:3000
ANXA2	ABclonal (A11235)	IHC 1:200, WB 1:1000
NF‐κB p65	Abcam (ab7970)	WB 1:1000
NF‐κB p65 (phospho S536)	Abcam (ab76302)	WB 1:1000
Sp1	Abcam (ab157123)	WB 1:1000

Abbreviations: ANXA2, annexin A2; NF‐κB, nuclear factor‐κB; Sp1, specificity protein 1; TFAM, mitochondrial transcription factor A.

2.3. Western blot analysis

Protein lysates were extracted using RIPA lysis buffer (Beyotime), and the BCA method was used to calculate the concentrations of the lysates. Total protein (80 μg) was subjected to SDS‐PAGE on a 10% polyacrylamide gel and then transported to a methylcellulose membrane (Millipore). The appropriate primary Abs listed in Table 2 and HRP secondary Abs were used to probe membranes (Maixin Biotech). An enhanced chemiluminescence detection kit was used for detection (Millipore).

2.4. Reverse transcription‐qPCR

Total RNA was extracted from HCC cells using RNAiso Plus (Cat. No. 9109; Takara). cDNA was produced from the total RNA (500 ng) using the PrimeScript RT reagent kit (Cat. No. RR047A; Takara). We carried out RT‐qPCR on a CFX 96 real‐time system (Bio‐Rad Laboratories) using the SYBR Premix Ex Taq II kit (Cat. No. RR820A; Takara). The primers utilized are listed in Table 3. The 2^−△△Ct method was used to calculate the relative expression of the target genes. GAPDH was used as an internal control.

TABLE 3.

Sequence of primers

Gene	Forward primer	Reverse primer
2.1 Primers used in quantitative PCR analysis
ANXA2	TCTACTGTTCACGAAATCCTGTG	AGTATAGGCTTTGACAGACCCAT
p37	ACAGGAGTAGTCAAGCAAGAGG	ACACAGTAGAGTCTTGAATTGGAGT
TFAM	ATGGCGTTTCTCCGAAGCAT	TCCGCCCTATAAGCATCTTGA
Sp1	CCTGGATGAGGCACTTCTGT	GCCTGGGCTTCAAGGATT
2.2 siRNA
siANXA2‐1	GCAGGAAAUUAACAGAGUCUA	UAGACUCUGUUAAUUUCCUGC
siANXA2‐2	CCAGAAAGUAUUUGAUAGGUA	UACCUAUCAAAUACUUUCUGG
siSP1‐1	CAGAUACCAGACCUCUUCU	AGAAGAGGUCUGGUAUCUG
siSP1‐2	UGAUCUGCCUCUCAACUGCCC	UACCGGGAAACUGGAGCACU
sip65‐1	GATTGAGGAGAAACGTAAA	CTAACTCCTCTTTGCATTT
sip65‐2	GCGACAAGGUGCAGAAAGA	CGCUGUUCCACGUCUUUCU
2.3 Primers used in gene cloning
TFAM	TCCGCTCGAGATGGCGTTTCTCCGAAGC	ATGGGGTACCGAACACTCCTCAGCACCATATTTTC

Abbreviations: ANXA2, annexin A2; Sp1, specificity protein 1; TFAM, mitochondrial transcription factor A.

2.5. Mycoplasma infection

Mycoplasma hyorhinis (ATCC 17981; Yise) was grown at 37°C for 72 h in modified Hayflick medium with 20% heat‐inactivated FBS as previously reported. ¹⁶ Mycoplasma titers were quantified as CCU per milliliter according to a previous report. When HCC cells have confluence of roughly 80%, 10⁵ CCU/ml of Mycoplasma hyorhinis was added to the cells after serum starvation for 24 h. ¹⁷ The HCC cells were then cocultured with mycoplasmas for 48 h before being harvested.

2.6. Cell migration and invasion assays

Invasion and migration experiments in Transwell chambers and wound healing assay were undertaken to evaluate cells' capacity for metastasis. The wound healing experiment was executed following the routine workflow. ¹⁸ Cells in dishes were pretreated with the proliferation inhibitor mitomycin C (0.5 μM) for 2 h, which effectively inhibited HCC cell proliferation without significantly increasing cytotoxicity. Wound healing was observed and photographs were taken 48 h after scratching. The migration area was then measured using ImageJ software. For the Transwell assay, HCC cells were added in the upper chamber (invasion assay, upper chambers were pretreated with 10 mg/ml growth factor‐reduced Matrigel). Medium containing 20% FBS was added to the bottom well of the Transwell chambers (Cat. No. CLS3470‐48EA; Corning Inc.) and covered with Matrigel (Cat. no. 356235; Corning Inc.). After 24 h of incubation (invasion assay for 48 h), cells that migrated to the lower surface of the filter membrane were fixed, stained with crystal violet, and then photographed under a microscope. Migration and invasion were defined by counting the cells in three microscopic fields per well, and the degree of migration and invasion was represented as the average number of cells per microscopic field.

2.7. Knockdown and overexpression of target genes

Lentiviruses for TFAM knockdown were created as previously reported. ¹⁴ Lentiviral vectors for TFAM knockdown were constructed by GenePharma Co., Ltd. (cat. no. D01001). Lentivirus was collected by cotransfecting 293 T cells with lentiviral expression construct and packaging plasmids. Hepatocellular carcinoma cells (5 × 10⁴/well) were seeded and incubated overnight. Lentivirus was then added into cells (10 MOI) with serum‐free medium and incubated overnight. Successfully transfected cell clones were screened using puromycin (5 μg/ml). The siRNA‐transfected cells were grown in 6‐well plates until cell attachment was complete and transfected with siRNAs and siRNA‐Mate (Tsingke Biotechnology). Table 3 lists the target sequences of siRNA. siRNAs were transfected with Lipofectamine 2000 reagent (Cat. No. 11668019; Invitrogen) as directed.

The TFAM expression plasmid was established as previously described ¹⁹ and maintained in our laboratory. Briefly, the coding sequence of TFAM was amplified from cDNA derived from SW480 cells using the primers listed in Table 3 and cloned into the pcDNA™3.1(+) vector. Then the vector was transfected into HCC cells using Lipofectamine 2000 according to the manufacturer's instructions. The monoclonal cell strain screened by G418 with TFAM overexpression was used for further examination. For restoration of TFAM in shTFAM HCC cells, the above‐mentioned WT TFAM expression plasmid was introduced in HCC cells with stable TFAM knockdown using the Lipofectamine 2000. The monoclonal cell strain with obviously elevated TFAM expression was selected.

2.8. Immunohistochemical analysis

Paraffin‐embedded HCC tissue sections were dewaxed in xylene and then rehydrated in absolute ethanol, 95% ethanol, 90% ethanol, 85% ethanol, and 75% ethanol. Antigen retrieval was carried out by boiling tissue sections in citrate buffer (Sigma‐Aldrich) for 3 min then blocking with a blocking solution (Maixin Biotechnology) for 25 min at RT. Primary Abs (Table 2) were added to the tissue and left at 4°C overnight. On the second day, the IHC kit (Maixin Biotechnology) was used for detection according to instructions. The IHC staining was observed under a microscope. Two pathology specialists independently quantified staining intensity and percent p37, TFAM, and ANXA2 positivity in a double‐blinded manner.

2.9. Mycoplasma detection

Mycoplasma was detected using the MycoAlert PLUS Mycoplasma Detection Kit (Lonza). Briefly, the HCC cells were cultured for 3 days without medium change. The cell supernatant was centrifuged at 200 g for 5 min to remove any remaining cells. Supernatant (100 μl) was transferred to a white microtiter plate (96‐well), then 100 μl MycoAlert PLUS reagent was added. Luminescence values were measured using a FLUOstar Omega microplate reader (BMG Labtech), resulting in reading A. Another 100 μl MycoAlert PLUS substrate was added to the sample. The luminescence signal was measured, resulting in reading B. If B/A > 1.2, the presence of mycoplasma infection is indicated.

2.10. Detection of MitoROS

Mitochondrial ROS was detected using the fluorescent probe MitoSOX (M36008; Invitrogen) following the manufacturer's instructions. Briefly, HCC cells were incubated with the working solution for 10 min at 37°C in darkness. Images were taken using a laser‐scanning confocal microscope (Olympus FV3000). ImageJ software was used to measure the fluorescence intensity.

2.11. Statistical analysis

We used GraphPad Prism 8.3.1 software to analyze the data. The data are presented as mean ± SD. Values of p < 0.05 were statistically significant. Pearson's correlation coefficient of normally distributed variables and Spearman's correlation coefficient of nonparametric testing were used to evaluate the association between continuous variables. Differences between two groups were inspected using two‐sided Student's t‐test. Kaplan–Meier survival curves were drawn and compared by the log–rank test.

3. RESULTS

3.1. Knockdown of TFAM enhanced mycoplasma infection in HCC cells

Previously, mtDNA stress induced by TFAM loss has been reported to prime the antiviral innate immune response. ²⁰ To further investigate functional role of mtDNA stress in tumor cells, HCC cells with TFAM knockdown were stained with anti‐dsDNA Ab. Unexpectedly, much more dsDNA was distributed around the HCC cells with TFAM knockdown when compared to control cells (Figures 1A and S1). The similar pattern in HCC cells was further verified by staining with DAPI, a commonly used DNA‐specific probe (Figure 1B). To further characterize the location of these dsDNA molecules, WGA, a probe for the plasma membrane, was used. As shown in Figures 1C,D, the majority of dsDNA was colocalized with the cell membrane, a feature of mycoplasma infection as previously reported. ²¹ , ²² To determine whether the dsDNA originates from mycoplasma, the main immunogen p37 of mycoplasma was indirectly stained with a fluorescent Ab. As expected, p37 was colocalized with dsDNA and significantly increased in HCC cells with TFAM knockdown (Figures 1E,F and S1), implying that TFAM knockdown significantly enhanced mycoplasma infection.

FIGURE 1.

Mitochondrial transcription factor A (TFAM) knockdown enhanced mycoplasma infection in hepatocellular carcinoma (HCC) cells. (A, B) Representative confocal microscopy images of immunofluorescence staining with anti‐TFAM, anti‐dsDNA Abs, and DAPI in HCC cells with stable TFAM knockdown as indicated. Blue, DAPI; green, TFAM, wheat germ agglutinin (WGA), or p37; red, dsDNA. shCtrl, control shRNA; shTFAM, shRNA against TFAM. Scale bar, 40 μm. (C, D) Representative confocal microscopy images of immunofluorescence staining with WGA (a cell membrane dye), anti‐dsDNA Ab, and DAPI in HCC cells with stable TFAM knockdown. (E, F) Representative confocal microscopy images of immunofluorescence staining with anti‐p37, anti‐dsDNA Abs, and DAPI in HCC cells with stable TFAM knockdown

3.2. Restoration of TFAM expression reversed enhanced mycoplasma infection in HCC cells

To further verify the causal link between expression of TFAM and mycoplasma infection, WT TFAM was restored in HCC cells with TFAM knockdown (Figure 2A,B). The results showed that the fluorescence intensity of dsDNA in the TFAM restoration group was much lower than that in the TFAM knockdown group (Figure 2C). The dsDNA staining area was also significantly smaller than that in the TFAM knockdown group (Figure 2D). Furthermore, compared with the TFAM knockdown group, the expression of p37 was decreased after restoration of TFAM expression (Figure 2E).

FIGURE 2.

Restoration of mitochondrial transcription factor A (TFAM) expression reversed the increased mycoplasma infection facilitated by TFAM knockdown in hepatocellular carcinoma (HCC) cells. (A, B) Representative confocal microscopy images of immunofluorescence staining with anti‐TFAM and anti‐dsDNA Abs, and DAPI in shTFAM HCC cells with treatment as indicated. Blue, DAPI; green, TFAM; red, dsDNA. EV, empty vector; shCtrl, control shRNA; shTFAM, shRNA against TFAM; TFAM, expression vector encoding TFAM. Scale bar, 40 μm. (C, D) Statistical analysis of the fluorescence intensity and area of dsDNA on cell membranes in (A) and (B). n ≥ 60 for each group. Statistical differences were determined using a t‐test. *p < 0.05; **p < 0.01. (E) Western blot analysis of the expression of TFAM and p37 in HCC cells with treatment as indicated. Tubulin was used as the internal loading

3.3. Overexpression of TFAM dramatically attenuated mycoplasma infection in HCC cells

Next, we sought to address whether TFAM overexpression attenuates mycoplasma infection in HCC cells. Mitochondrial transcription factor A was then overexpressed in two mycoplasma‐free HCC cell lines, HLE and Hep‐3B, which originally expressed low levels of TFAM (Figure 3A,B). ¹⁴ We found that the fluorescence intensity of dsDNA in HCC cells with TFAM overexpression was much lower than that in control cells after coculture with mycoplasmas for 48 h (Figures 3C and S2). Moreover, the staining area of dsDNA on cell membranes was significantly smaller in HCC cells with TFAM overexpression (Figure 3D). Consistently, p37 expression was decreased in TFAM‐overexpressing cells (Figure 3E), indicating that TFAM overexpression attenuated mycoplasma infection in HCC cells.

FIGURE 3.

Mitochondrial transcription factor A (TFAM) overexpression dramatically attenuated mycoplasma infection in hepatocellular carcinoma (HCC) cells. (A, B) Representative confocal microscopy images of immunofluorescence staining with anti‐TFAM, anti‐dsDNA Abs, and DAPI in HCC cells with treatment as indicated. Blue, DAPI; green, TFAM; red, dsDNA. Scale bar, 40 μm. EV, empty vector; TFAM, expression vector encoding TFAM. (C, D) Statistical analysis for fluorescence intensity and area of dsDNA on cell membranes in (A) and (B). n ≥ 60 for each group. Statistical difference was determined by t‐test. *p < 0.05; **p < 0.01. (E) Western blot analysis for the expression of TFAM and p37 in HCC cells with treatment as indicated. Tubulin was used as internal loading. The cells mentioned above were all infected with Mycoplasma.

3.4. Mitochondrial transcription factor A inhibited mycoplasma infection of HCC cells by downregulating ANXA2 expression

To further explore the underlying mechanisms in HCC cells, we analyzed the expression of molecules reported to interact with p37 ⁹ , ²² , ²³ , ²⁴ , ²⁵ in TFAM‐knockdown cells using RNA sequencing data. ¹⁴ As shown in Figure 4A, ANXA2 was the most upregulated gene at the mRNA level in HCC cells with TFAM knockdown. The upregulation of ANXA2 expression was further validated by western blotting in HCC cells with TFAM knockdown upon mycoplasma infection (Figures 4B and S3). Moreover, knockdown of ANXA2 in HCC cells with TFAM knockdown significantly reduced p37 expression and mycoplasma infection (Figures 4C,D), which is consistent with a previous report that ANXA2 is a host receptor that mediates mycoplasma infection. ⁹ Immunohistochemical staining analysis also showed that the expression level of ANXA2 was negatively correlated with the expression level of TFAM in human HCC tissue (Figure 4E,F). We also confirmed that Sp1, the transcription factor of ANXA2, was also upregulated in HCC cells with TFAM knockdown (Figure 4G–I). Furthermore, the mitochondrial ROS (a classical Sp1 inducer) was then measured in HCC cells and was found to be also increased in TFAM knockdown cells (Figure 4J,K). Moreover, the levels of TFAM knockdown‐induced Sp1 and ANXA2 were decreased when the mitochondrial ROS scavenger XJB‐5‐131 were added (Figure 4L). These results suggested that decreased TFAM promoted the expression of Sp1 by elevating mitochondrial ROS production.

FIGURE 4.

Mitochondrial transcription factor A (TFAM) inhibited mycoplasma infection of hepatocellular carcinoma (HCC) cells by downregulating annexin A2 (ANXA2) expression. (A) Expression heatmap of RNA sequencing in TFAM knockdown and control SNU‐368 cells (n = 3 per group). (B) Western blot analysis for the expression of TFAM, ANXA2, specificity protein 1 (Sp1), and p37 in HCC cells with TFAM knockdown. Tubulin was used as internal loading. M. Hyo, Mycoplasma hyorhinis. (C) Western blot analysis for the expression of ANXA2 and p37 in HCC cells 48 h after ANXA2 knockdown. Tubulin was used as internal loading. (D) Quantitative PCR (qPCR) analysis for the mRNA expression of ANXA2 and p37 in HCC cells with ANXA2 knockdown. Statistical difference was determined by t‐test. *p < 0.05; **p < 0.01. (E) Representative immunohistochemical staining of TFAM and ANXA2 in human HCC tissues. Scale bar, 50 μm. (F) Spearman's correlation analysis between the TFAM and ANXA2 expression level based on the immunohistochemical (IHC) score. (G) qPCR analysis for the mRNA expression of TFAM and Sp1 in HCC cells with TFAM knockdown. Statistical difference was determined by t‐test. *p < 0.05; **p < 0.01. (H) Western blot analysis for the expression of Sp1 in HCC cells with TFAM knockdown. Tubulin was used as internal loading. (I) Western blot analysis for the expression of ANXA2 and Sp1 in HCC cells with Sp1 knockdown. Tubulin was used as internal loading. (J, K) Representative confocal microscopy images and statistical analysis of staining with MitoSOX (5 μM) in HCC cells. Scale bar, 40 μm. (L) Western blot analysis for the expression of Sp1 and ANXA2 in HCC cells treated with mitochondrial ROS scavenger XJB‐5‐131 (10 μM). Ctrl, control; EPCAM, epithelial cellular adhesion molecule; LRP6, low‐density lipoprotein receptor‐related protein 6; TLR4, Toll‐like receptor 4

3.5. Mycoplasma infection is negatively associated with TFAM expression in HCC tissues and positively correlated with poor prognosis in HCC patients

We further investigated the relationship between TFAM expression and mycoplasma infection in human HCC tissue samples by IHC (Figure 5A). Overall, 76% of HCC tissues were positive for p37, indicating that mycoplasma infection is relatively common in HCC tissues (Figure 5B). In addition, we found that the expression of p37 was negatively associated with TFAM expression in human HCC tissues (Figure 5C). Furthermore, HCC patients positive for p37 had significantly poorer overall survival and disease‐free survival than those negative for p37 (Figure 5D,E), suggesting that mycoplasma infection might be an unfavorable prognostic factor in HCC patients.

FIGURE 5.

Mycoplasma infection is negatively associated with mitochondrial transcription factor A (TFAM) expression in hepatocellular carcinoma (HCC) tissues and positively associated with poor prognosis in HCC patients. (A) Representative immunohistochemical (IHC) staining of TFAM and p37 in human HCC tissues. Scale bar, 50 μm. (B) Percentage of p37‐positive HCC tissues. (C) Spearman's correlation analysis between the TFAM and p37 expression level based on the IHC score. (D, E) Kaplan–Meier analysis of overall survival and disease‐free survival curves in HCC patients by the p37 expression (n = 100). Statistical significance was determined using the log–rank test. *p < 0.05; **p < 0.01

3.6. Mycoplasma infection facilitated metastasis of HCC cells

To investigate the effects of mycoplasma infection mediated by downregulated TFAM on the metastasis of HCC cells, wound healing and Transwell assays were carried out. Our results indicated that mycoplasma infection (almost all the cells were positive for dsDNA signals on cell membrane) obviously enhanced the migration ability of HCC cells when compared to uninfected HCC cells (Figures 6A,B,D,E and S4A,B,E,F,G). The Transwell migration and invasion assay further confirmed this phenomenon (Figure 6C,F), indicating that mycoplasma infection promoted the metastasis of HCC cells.

FIGURE 6.

Mycoplasma infection promoted the metastasis of hepatocellular carcinoma (HCC) cells. (A, D) Representative confocal microscopy images of immunofluorescence staining with anti‐mitochondrial transcription factor A (TFAM), anti‐dsDNA Abs and DAPI in mycoplasma infected (M (+)) and (M (−)) uninfected HCC cells as indicated. Blue, DAPI; green, TFAM; red, dsDNA. Scale bar, 40 μm. (B, E) Representative images and statistical analysis of wound healing migration assay for mycoplasma‐infected HCC cells as indicated. (C, F) Representative images and statistical analysis of Transwell migration and invasion assays for mycoplasma‐infected SNU‐739 and HLE cells as indicated. (G, J) Representative confocal microscopy images of immunofluorescence staining with anti‐TFAM, anti‐dsDNA Abs, and DAPI in mycoplasma‐infected HCC cells with treatment as indicated. Blue, DAPI; green, TFAM; red, dsDNA. Scale bar, 40 μm. (H, K) Representative images and statistical analysis of wound healing migration assay for mycoplasma‐infected HCC cells with treatment as indicated. (I, L) Representative images and statistical analysis of Transwell migration and invasion assays for mycoplasma‐infected HCC cells with treatment as indicated. Scale bar, 100 μm. Statistical difference was determined by t‐test. *p < 0.05; **p < 0.01. Ctrl, control; EV, empty vector

Next, we investigated the effect of mycoplasma infection enhanced by TFAM downregulation on the metastatic ability of human HCC cells. As shown in Figure 6G–I, TFAM knockdown obviously enhanced the metastatic ability of HCC cells, except for increasing mycoplasma infection. In contrast, TFAM overexpression dramatically inhibited mycoplasma infection and the metastatic ability of HCC cells (Figures 6J–L and S4C,D,H).

3.7. Mycoplasma infection activated NF‐κB signaling to facilitate the metastasis of HCC cells

Previous research has shown that mycoplasma infection in gastric cancer cells induces the activation of NF‐κB signaling, which plays an important role in cell migration. ⁹ , ²² Moreover, it has also been reported that ANXA2 facilitated the phosphorylation of p65 (subunit of NF‐κB) at Ser536 and nuclear translocation of p65 in gastric cancer cells. ⁹ Therefore, we further investigated this mechanism in HCC cells. Western blot analysis was undertaken to examine the effects of TFAM knockdown on NF‐κB signaling in HCC cells with or without mycoplasma infection. The data showed that TFAM knockdown significantly upregulated the expression of ANXA2, but not Ser536 phosphorylation of p65 in mycoplasma‐free HCC cells, implying that NF‐κB signaling is downstream of ANXA2 in HCC cells only in the context of mycoplasma infection (Figure 7A). Knockdown of ANXA2 reversed this effect, and Ser536 phosphorylation of p65 was attenuated in HCC cells infected with mycoplasma (Figure 7B). We next investigated the functional role of the NF‐κB pathway in the migration and invasion of HCC cells enhanced by TFAM‐related mycoplasma infection. As shown in Figures 7C,D, the treatment with the NF‐κB inhibitor Bay11‐7082 (Figure S5A) significantly decreased the metastasis of HCC cells following TFAM knockdown. Consistently, the knockdown of p65 with siRNA (Figure S5B) significantly decreased the metastasis of HCC cells with TFAM knockdown (Figures 7E,F, 8).

FIGURE 7.

Mycoplasma infection activated nuclear factor‐κB (NF‐κB) signaling to promote the migration and invasion of hepatocellular carcinoma (HCC) cells. (A) Western blot analysis for the expression of annexin A2 (ANXA2), p65, phosphorylated p65 subunit of NF‐κB (p‐p65), and p37 in HCC cells with mitochondrial transcription factor A (TFAM) knockdown. Tubulin was used as internal loading. M. Hyo, Mycoplasma hyorhinis. (B) Western blot analysis for the expression of TFAM, ANXA2, p65, p‐p65, and p37 in HCC cells with mycoplasma infection, and ANXA2 was transiently knocked‐down by siRNA against ANXA2. (C, D) Representative images and statistical analysis of migration and invasion assays for SNU‐368 cells treated with the NF‐κB inhibitor BAY 11‐7082 (12.5 μM) for 12 h. (E, F) Representative images and statistical analysis of migration and invasion assays were carried out using SNU‐368 cells 48 h after p65 knockdown (n = 3 independent experiments). **p < 0.01. Ctrl, control

FIGURE 8.

Schematic representation depicting the mechanism of mitochondrial transcription factor A (TFAM) knockdown‐mediated mycoplasma infection enhancement in the promotion of hepatocellular carcinoma (HCC) metastasis. Briefly, decreased TFAM expression upregulated the transcription factor specificity protein 1 (Sp1) by elevating mitochondrial reactive oxygen species (ROS) production, which increased the expression of annexin A2 (ANXA2), leading to the increasement of mycoplasma infection. Moreover, TFAM downregulation‐mediated mycoplasma infection enhancement promoted HCC migration and invasion by activating the nuclear factor‐κB signaling pathway. ROS, reactive oxygen species

4. DISCUSSION

Several key findings emerged from this study. First, we found that the downregulation of TFAM facilitated mycoplasma infection in HCC cells. Second, mycoplasma infection promoted the metastasis of HCC cells and was significantly correlated with the prognosis of HCC patients. Third, decreased TFAM expression upregulated the transcription factor Sp1 to increase ANXA2 expression and then promoted mycoplasma infection. Moreover, mycoplasma infection mediated by decreased TFAM expression promoted the migration and invasion of HCC cells by activating the NF‐κB signaling pathway.

Clinically, previous studies have suggested that characteristics of tissue‐resident mycoplasma correlate with cancer risks, pathological types, and cancer prognosis. ²⁶ , ²⁷ Our study showed that lipoprotein p37 of mycoplasma was positive in 76% of HCC cases. Moreover, the overall survival and disease‐free survival of p37‐positive HCC patients were obviously lower than those of p37‐negative patients. These results were partially supported by a previous investigation, in which p37‐positive CTCs were detected in 42 of 47 HCC patients (89%). ²³ In the Kaplan–Meier survival analysis, p37 expression of CTCs was significantly correlated with poor disease‐free survival in HCC patients. ²³ Similarly, p37 was shown to be positive in 52% of prostate cancer tissues and in 39.5%–44.3% of gastric cancer tissues. ⁹

To date, the host factors that mediate mycoplasma infection remain unclear. In the present study, we found for the first time that downregulation of TFAM expression facilitates mycoplasma infection in HCC cells. This was supported by evidence that mycoplasma infection is negatively associated with TFAM expression in HCC tissues and positively associated with poor prognosis in HCC patients. Similarly, a recent study showed that mitochondria exert an antibacterial role by generating ROS, hosting signaling regulators, such as mitochondrial antiviral signaling protein, and modulating host innate immunity. ¹¹ , ²⁸ Moreover, TFAM deficiency‐induced mtDNA stress is involved in interferon‐stimulated gene expression and broad microorganism resistance. ²⁹ It has also been reported that intravenous administration of TFAM in healthy animals upregulates circulating levels of pro‐inflammatory cytokines interleukin‐6 and tumor necrosis factor‐α, which are required for efficient host clearance of pathogens. ³⁰

Annexin A2 is increased in multiple tumors and is associated with tumor grade and clinical outcome. ³¹ , ³² Recent studies have uncovered a relationship between ANXA2 and microbial infection. ³³ , ³⁴ For instance, ANXA2 contributes to human papilloma virus infection ³⁵ and participates in HCV replication. ³⁵ In addition, ANXA2 facilitates mycoplasma infection through its N‐terminal. ⁹ Our study further showed that ANXA2 played an important role in mycoplasma infection in HCC cells. Moreover, decreased TFAM expression upregulated the transcription factor Sp1 to increase ANXA2 expression, which is consistent with previous research showing that Sp1 upregulates ANXA2 transcription to promote migration and invasion of oral squamous cell carcinoma. ³⁶ Additionally, it has been proved that downregulation of TFAM promotes ROS production in several types of cells. ¹³ , ³⁷ Reactive oxygen species regulate Sp1 expression to promote progression of human tumors. ³⁸ , ³⁹

The NF‐κB signaling pathway has been shown to be activated in many types of cancer cells and significantly promotes their aggressive phenotype. ⁴⁰ In the present study, we found that TFAM knockdown promoted mycoplasma infection and NF‐κB signal‐mediated migration and invasion of HCC cells. Similarly, it has been reported that NF‐κB signaling is activated and maintains the aggressive cancer phenotype in gastric cancer with mycoplasma infection. ⁹ , ²² Mycoplasma infection has also been shown to inhibit p53, activate NF‐κB, and synergize with oncogenic Ras in rodent fibroblast transformation. ⁴¹ Moreover, knockdown of ANXA2, a mycoplasma binding protein, attenuated mycoplasma‐induced Ser536 phosphorylation of p65 in gastric cancer cells ⁹ and in HCC cells infected with mycoplasma. It is noteworthy that TFAM knockdown significantly upregulated ANXA2 expression, but not Ser536 phosphorylation of p65 in mycoplasma‐free HCC cells, implying that NF‐κB signaling is downstream of ANXA2 in HCC cells only in the context of mycoplasma infection. We therefore speculated that the interaction between mycoplasma and ANXA2 might be one of the key factors to activate NF‐κB signaling.

Additionally, the purified p37 or recombinant p37 protein has been proved to promote the invasiveness of prostate carcinoma cells or gastric cancer cells in a dose‐dependent manner in previous studies, ⁹ , ⁴² whereas the enhanced invasive potential of cancer cells could be completely reversed by preincubation of the cancer cells with an anti‐p37 Ab. These results indicated that p37 protein can, at least partially, substitute for mycoplasma infection in tumor cells.

In summary, our study illustrates that TFAM play important roles in mycoplasma infection of HCC cells, suggesting that TFAM could be a potential therapeutic target against aggressive HCC infected with mycoplasma.

AUTHOR CONTRIBUTIONS

Yinping Wang: data curation, formal analysis, methodology, supervision, and writing – original draft. Gang Wang: software, data curation, formal analysis, writing – original draft, methodology. Xin Hong: data curation and formal analysis. Jing Zhao: data curation and formal analysis. Dan Wu: data curation and formal analysis. Lin Chen: data curation and investigation. Xiaoli Liu: data curation and formal analysis. Deyu Kong: data curation and investigation. Qichao Huang: project administration, writing – review and editing. Jinliang Xing: project administration. Nan Wang: project administration, writing – review and editing. Yilin Zhao: conceptualization, resources, supervision, project administration, writing – review and editing.

FUNDING INFORMATION

This study was supported by the National Natural Science Foundation of China (82200040, 82002511, 82072722, 82273375, 82203509, and 81972590).

CONFLICTS OF INTEREST

The authors have no conflict of interest.

ETHICS STATEMENT

The study protocol was approved by the ethics committee of Fourth Military Medical University. Permit number: KY20213188‐1.

Informed consent: The tissue samples were obtained with written informed consent from each patient.

Animal studies: N/A.

Supporting information

Appendix S1.

Wang Y, Wang G, Hong X, et al. Downregulated mitochondrial transcription factor A enhances mycoplasma infection to promote the metastasis of hepatocellular carcinoma. Cancer Sci. 2023;114:1464‐1478. doi: 10.1111/cas.15715