The Anti-Apoptotic Protein Lifeguard Is Expressed in Osteosarcoma, Chondrosarcoma, and Soft Tissue Sarcoma.

Introduction
Sarcomas are a heterogeneous group of rare malignant tumors of mesenchymal origin, with a distinct tendency of hematogenous metastasis [1], accounting for approximately 1% of all adult malignancies but representing a significant burden in pediatric oncology, where they comprise about 14% of all childhood cancers [2]. They are broadly classified into bone sarcomas and soft tissue sarcomas. Osteosarcoma, the most common primary malignant bone tumor, has a peak incidence during adolescence [3]. The prognosis is strongly dependent on the tumor stage at diagnosis, which is determined using the TNM (Tumor, Metastasis) classification system established by the American Joint Committee on Cancer (AJCC) [4]. For osteosarcoma, staging is crucial: stage IIA refers to a high-grade, localized tumor that has not spread beyond the bone, while stage IIB indicates a high-grade tumor that has extended into the surrounding soft tissues. This distinction is critical, as the 5-year survival rate for patients with localized disease is approximately 60–80% but drops dramatically to below 20% for patients with metastatic disease [2, 5]. Despite multimodal treatment approaches, including surgery and chemotherapy, the 10-year survival rates have reached a plateau, highlighting the urgent need for more effective therapeutic strategies [6].
One possible explanation for this plateau could be that the enormous pathological diversity of more than 100 different sarcoma subtypes that is not yet adequately addressed in standardized sarcoma chemotherapy [6]. Additionally to the rareness of sarcoma, their pathologic heterogeneity with more than 100 different subtypes makes modern multimodal treatment in specialized sarcoma centers very challenging [7]. Furthermore, secondary cardiac, pulmonary, and neurocognitive disorders, as well as secondary malignancies, have been described in long-term sarcoma survivors after successful chemotherapeutic treatment [8]. Current research on osteosarcoma has identified more possible therapeutic targets with high potential for future targeted tumor therapy and immunotherapy [9, 10], based on a more precise understanding of the molecular mechanisms behind cell-growth signaling, the sarcoma microenvironment, and the apoptosis-evading strategies of different sarcoma subtypes.
One of the key hallmarks of cancer is the evasion of programmed cell death, or apoptosis [11]. Apoptosis can be initiated via two major pathways: the intrinsic (mitochondrial) pathway and the extrinsic (death receptor-mediated) pathway. The extrinsic pathway is often triggered by the Fas receptor (Fas) and its ligand (FasL) [12]. The binding of FasL to Fas leads to the formation of the death-inducing signaling complex and subsequent activation of caspase-8, initiating a caspase cascade that results in cell death [13]. It has been shown that a reduction in Fas expression is a mechanism for increased metastatic potential in osteosarcoma. The human osteosarcoma cell line LM7, which has low Fas expression, showed high metastatic potential in a nude mouse model [14].
Various apoptotic agents are reported to contribute to programmed cell death in osteosarcoma cell lines in vitro such as microRNA (miRNA) [15], natural compounds like chimaphilin [16] and icaritin [17], and numerous apoptotic proteins [11]. Pro-apoptotic proteins, for example, inhibitor of growth protein 4 (ING4), induced the apoptosis of osteosarcoma cells by the blockage of the nuclear factor kappa light-chain enhancer of activated B cells (NF-kB) signaling pathway and activation of the intrinsic pathway through decreasing the ratio of Bcl-2/Bax [18]. The Runx2 gene, an important transcription factor that is involved in osteoblast maturation and bone development, turned out to be overexpressed in many osteosarcoma cells [11]. Furthermore, the loss of Runx2 in osteosarcoma sensitized the cells to doxo-induced apoptosis in vitro and in vivo [19]. Another regulator of apoptosis in osteosarcoma can be found in the enhancer of zeste homolog 2 (EZH2) protein, which corresponds to the catalytic subunit of polycomb repressive complex 2 and showed higher expression in patients with advanced tumor stages [20]. Moreover, silencing of EZH2 in osteosarcoma by siRNA reduced osteosarcoma cell growth, invasion, and lung metastasis [21]. Another pathway resulting in extrinsic programmed cell death is the Fas/Fas ligand system, which results in activation of caspase-8 and subsequent caspase cascades, initialized by ligation of Fas to its agonistic antibody and formation of death-inducing signaling complex [13, 22]. In addition to the intrinsic mitochondria pathway, Fas-mediated programmed cell death inhibition seems to be not only highly relevant for apoptosis but also for metastasis in human osteosarcoma, shown by reduced Fas expression levels in LM7 cell lines with high metastatic potential [14]. The studies of Koshkina et al. [23] gave further evidence to the idea that osteosarcoma cells that reduced their Fas expression got promoted and positively selected during the process of metastasis in the Fas-ligand (FasL)-positive microenvironment of the lung [24]. Osteosarcoma cells with higher Fas expression might be eliminated by FasL-induced apoptosis in the lung, resulting in the selection of Fas-negative cell populations [24]. It was shown that Fas-/FasL-mediated apoptosis signaling in osteosarcoma during lung metastasis did not depend only on the amount of Fas expression [23]. In addition, downregulation of Fas signaling might serve as a gatekeeper for metastasis [23]. Supporting this hypothesis, osteosarcoma K7 cells were transfected with Fas-associated protein with death domain-dominant negative, resulting in K7/Fas-associated protein with death domain-dominant negative cells, that showed the ability to form Fas-positive metastases in a mice model with higher metastatic potential than untransfected K7 cells [23]. The expression of FasL, although being lower than in carcinoma, was detected in many subtypes of sarcoma, being predominantly high in rhabdomyosarcoma, malignant schwannoma, and Ewing sarcoma [25]. Taking these findings together, Fas-mediated cell death seems to be essential for both evading apoptosis and enabling metastasis of osteosarcoma by altering Fas expression and signaling [23–25]. Therefore, it is necessary to examine the expression of more anti-apoptotic proteins in different sarcoma tissues that interfere with Fas-mediated apoptosis, such as Lifeguard protein (LFG) also known as Fas apoptosis inhibitory molecule 2 (FAIM2) [26]. The endogenous expression of LFG in human cancer tissue was already shown for many breast cancer tissues and cell lines, being positively correlated with tumor stage and causing a reduced sensitivity to Fas signaling without changing the levels of Fas expression [27]. In contrast to past research, that indicated a physical association between LFG and Fas in healthy human tissue [26, 28], results in Fas-antibody immunofluorescence showed no co-localization between LFG and Fas in MCF-7 breast cancer cell lines, supposing there could be an altered function of LFG in human cancer tissues [27, 29]. Currently, the data regarding LFG in sarcoma tissue is scarce. Therefore, the expression and exact function of LFG in clinical related phenotypes of sarcoma needs to be investigated.
The protein LFG, also known as Fas apoptosis inhibitory molecule 2 (FAIM2), is a member of the Bax-inhibitor-1 (BI-1) family that specifically inhibits Fas-mediated cell death [26, 30, 31]. LFG is a highly conserved, multi-pass transmembrane protein located in the plasma membrane and intracellular membranes [26, 32, 33]. It has been shown to be highly expressed in several breast cancer cell lines and tissues, where its expression level correlated with higher tumor grades and reduced Fas sensitivity [27]. Downregulation of LFG sensitized breast cancer and sarcoma cell lines (MCF-7 and SW872, respectively) to Fas-induced cell death [29]. Recent research also suggested that LFG expression, triggered by serum albumin, could promote the viability of osteosarcoma cells during intravasation [34].
However, the expression and anti-apoptotic function of LFG in a broader range of sarcoma subtypes and its correlation with clinical stages remain unclear so far. Therefore, the aim of the present study was to analyze LFG expression in a comprehensive cohort of human osteosarcoma, chondrosarcoma, and soft tissue sarcoma samples using immunohistochemistry and to compare these expression levels with those in healthy tissues.

Materials and Methods

LFG Protein Structure
The three-dimensional model of the protein structure of LFG was created using the SWISS-MODEL® software (page accessed on March 10, 2023). SWISS-MODEL® is an automated protein homology server. The LFG protein structure was based on comparative modeling and needed a given amino acid sequence, deposited in the protein data bank “uniprot” (uniprot PDB) for the modeling process. The structural template was identified at uniprot and added to the SWISS-MODEL® server. The alignment of target sequence and template structures, model building, and model quality evaluation were done automatically by the server. The modeling results of the predicted LFG protein structure were compared to other modeling processes and evaluated using global model quality estimate values.

Osteosarcoma and Chondrosarcoma Samples
For LFG expression analysis in osteosarcoma and chondrosarcoma, the commercially available tissue array (OS802c, hematoxylin-eosin stain; Biomax US, Derwood, MD, USA), containing 80 different patient samples (1 core/case), was used. According to the manufactures’ specifications, 50 samples of osteosarcoma, 28 samples of chondrosarcoma, and 2 samples of healthy human bone tissue were included in the analysis and were prepared for immunohistochemistry. For each tumor sample, pathological grading and clinical findings, such as TNM classification, were included in the analysis. The healthy human bone samples served as controls.

Soft Tissue Sarcoma Samples
For LFG expression analysis in soft tissue sarcoma, the commercially available soft tissue sarcoma array (SO208a, hematoxylin-eosin stain; Biomax US, Derwood, MD, USA) was used. According to the manufacturers’ specifications, 55 samples of different soft tissue sarcoma subtypes, 4 samples of healthy human smooth muscle tissue, and 5 samples of healthy human skeletal muscle tissue were analyzed and prepared for immunohistochemistry. For each tumor sample, pathological grading and clinical findings such as TNM classification were included in the analysis. The healthy human skeletal and smooth muscle samples served as controls.

Sample Preparation
The osteosarcoma and the soft tissue sarcoma tissue arrays were deparaffinized in xylen (Lab Alley, Spicewood, TX, USA) and rinsed with phosphate-buffered saline (PBS) (Thermo Fisher Scientific, Schwerte, Germany) for 5 min (min) each. The slides were immersed in a descending alcohol concentration series (100% ethanol, 80% ethanol, 70% ethanol) for 2 min each.

Immunohistochemistry
For antigen retrieval, the slides were pressure cooked in 6.5 mm sodium citrate (pH 6.0) (Thermo Fisher Scientific, Schwerte, Germany) and washed thrice with PBS (Thermo Scientific, Schwerte, Germany). Specific background staining was reduced by incubating the slides in 2.5% bovine serum albumin/1× Tris-buffered saline (Lab Alley, Spicewood, TX, USA) for 60 min. To optimize primary antibody binding, the slides were incubated at 4°C overnight with rabbit anti-hLFG primary antibody (Santa Cruz Biotechnology, Dallas, TX, USA; final dilution 1:100) dissolved in 1% bovine serum albumin/PBS solution (Thermo Fisher Scientific, Schwerte, Germany). After incubation, the slides were washed with PBS thrice for 5 min. For secondary antibody binding, the slides were incubated with goat anti-rabbit 800CW conjugated secondary antibody (IRDye® Li-Cor, Lincoln, USA; final dilution 1:100) at room temperature for 30 min. The Li-Cor infrared imaging system Odyssey® (Li-Cor Biosciences, Lincoln, NE, USA) was used to detect the signals. For further analysis, the Image Studio Software Odyssey® (Li-Cor Biosciences, Lincoln, NE, USA) was used to visualize LFG protein expression. The relative fluorescent units, measured by Li-Cor Infra-Red Imaging System, were used to quantify LFG protein expression.

Statistical Analysis
The LFG protein expression was observed for each subtype of sarcoma in different tumor stages, matched to healthy tissue controls. The tumor stages were based on clinical findings, specified according to the TNM classification (World Health Organization 2017) and pathological grading. The means of LFG expression and the 95% confidence intervals (CI) were determined on the basis of measured LFG fluorescent units for each sarcoma subtype and stage. A p value of <0.05 was determined to be statistically significant. The results were calculated using Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) and visualized with Datatab software (Datatab, Seiersberg, Austria).

Results

LFG Protein Simulation
To visualize the three-dimensional structure of the LFG protein, the sequence-based SWISS-MODEL® software was used (Fig. 1). Figure 1a shows the expected LFG monomer protein structure that consists of seven transmembrane domains (blue) that pass multiple times through the plasma membrane and are linked to a separate protein tail (orange). Figure 1b shows the template alignment of LFG, as given by the uniprot protein data bank.

High Expression of LFG Protein in Osteosarcoma Tissues
As described above, osteosarcoma samples were analyzed using commercially available immunofluorescence arrays. Table 1 gives an overview of the different anatomical localizations of the samples, the corresponding TNM classification, tumor grading, tumor stage, number of samples, and mean LFG intensity. Osteosarcoma tissues from various anatomical regions were included, predominantly coming from femur and tibia (Table 1).
The mean LFG expression in osteosarcoma stage IIA (54.7; 95% CI = 47.1; 62.4) was significantly higher than in healthy bone marrow tissue (20.7; 95% CI = 18.6; 22.8) (p < 0.05). The mean LFG expression in osteosarcoma stage IIB (61.1; 95% CI = 56.9; 65.3) was significantly higher than in healthy bone marrow tissue (20.7; 95% CI = 18.6; 22.8) (p < 0.05) (Fig. 2I). Osteosarcoma stage IIB showed significantly higher LFG expression compared to osteosarcoma stage IIA (p < 0.05) (Fig. 2I). Samples with advanced osteosarcoma stage IIB had a larger tumor size, compared to IIA, and had the highest mean LFG expression of all osteosarcoma samples in the analysis. Both samples with osteosarcoma stages IIA and IIB had grade 3 tumors in which the sarcoma cells grew more aggressively and showed abnormal cell morphology. Figure 2II shows a representative selection of LFG antibody-stained fluorescent images of osteosarcomas. The increase in the LFG intensity in osteosarcoma stage IIB (Fig. 2IIc), compared to stage IIA (Fig. 2IIb) and the controls (Fig. 2IIa), can be seen in the constant increase of green immunofluorescence from Figure 2IIa to Figure 2IIc.

High Expression of LFG Protein in Chondrosarcoma

Table 2 gives an overview of the different tumor localizations, tumor grading, tumor stage, and mean LFG intensity of chondrosarcoma samples. A significantly higher LFG expression in chondrosarcoma stage IIA was observed (53.4; 95% CI = 42.4; 64.5) compared to chondrosarcoma stage IA (36.1; 95% CI = 32.4; 39.8) (p < 0.05) (Fig. 3I). A significant higher LFG expression in chondrosarcoma stage IIA was observed (53.4; 95% CI = 42.4; 64.5) compared to chondrosarcoma stage IA (36.1; 95% CI = 32.4; 39.8) (Fig. 3I). The LFG expression level detected in stage IIB chondrosarcoma (49.3; 95% CI = 39.5; 59.1) was lower than in stage IIA (53.4; 95% CI = 42.4; 64.5) (p > 0.05), although still significantly higher than in healthy bone marrow tissue (20.7; 95% CI = 18.6; 22.8) (p < 0.05). While LFG expression increased from stage IA to IIA, no consistent positive correlation with tumor stage was observed across all chondrosarcoma stages.

Figure 3IIa–e shows fluorescent images of chondrosarcoma samples with different tumor stages. The LFG intensity is expressed by green immunofluorescence. Healthy bone marrow tissue showed the lowest LFG expression compared to all subtypes of chondrosarcoma (Fig. 3IIa). The LFG expression increases from chondrosarcoma stage IA (Fig. 3IIb) to stage IB (Fig. 3IIc), as shown by an increase in intensity of the green immunofluorescence. Figure 2IId shows the high LFG expression in chondrosarcoma stage IIA, significantly higher compared to controls (Fig. 3IIa) and stage IA chondrosarcoma (Fig. 3IIb). Figure 3IIe shows the LFG expression of chondrosarcoma stage IIB being significantly higher compared to controls (Fig. 3IIa) but lower compared to stage IIA (Fig. 3IId).

High Expression of LFG Protein in Soft Tissue Sarcoma
Taking into consideration that the anti-apoptotic protein LFG showed high expression levels in osteosarcoma and chondrosarcoma, the expression of LFG in different subtypes of soft tissue sarcoma was analyzed. Table 3 shows the different soft tissue sarcoma subtypes that were investigated regarding LFG expression. A total number of 55 soft tissue sarcoma tissue samples were analyzed, using 5 skeletal muscle samples and 4 endometrium smooth muscle samples as unmatched controls. Table 3 gives an overview of the different soft tissue subtypes, the corresponding TNM classification, tumor grading, tumor stage, and mean LFG intensity of the samples. The means of LFG expression were calculated for each soft tissue subtype as a whole by summarizing the data of all corresponding stages for each subtype (Table 3). Overall, the LFG expression variated strongly depending on the soft tissue subtype (Fig. 4a). Liposarcoma showed low LFG expression (47.4; 95% CI = 43.2; 51.6), not significantly different compared to skeletal muscle control (45.8; 95% CI = 41.9; 49.7) (p > 0.05), whereas malignant schwannoma (86.3; 95% CI = 57.8; 114.9), dermatofibrosarcoma (65.9; 95% CI = 56.9; 74.8), and leiomyosarcoma (90.6; 95% CI = 65.2; 116.0) had significantly higher LFG expression compared to healthy skeletal muscle control and healthy smooth muscle control (49.3; 95% CI = 43.5; 55.2) (Fig. 4a).
The LFG expression in angiosarcoma (60.6; 95% CI = 54.7; 66.4) and synovial cell sarcoma (59.0; n = 1) was higher compared to controls, but the difference was not statistically significant (p > 0.05) (Fig. 4a). Figure 4b shows that LFG expression in dermatofibrosarcoma stage IA (63.5; 95% CI = 54.7; 72.3) was significantly higher compared to the controls (p < 0.05), with an increase in LFG expression from stage IA to more advanced stage IB (82.3; n = 1). Due to the small sample size (n = 1 for each stage), no statistical comparison between stages could be performed for malignant schwannoma. Numerically, LFG expression increased from stage IA (56.2; n = 1) to stage IIA (74.7; n = 1) and stage IIB (124.9; n = 1), with stage III showing lower expression (89.6; n = 1) compared to stage IIB (Fig. 4c).
The highest LFG expression of all specimens was detected in leiomyosarcoma stage IA (139.3; n = 1) (Fig. 4d). The LFG expression in leiomyosarcoma stage IIB (77.3; 95% CI = 61.9; 92.7) was significantly higher compared to both healthy controls (Fig. 4d). Due to the small sample size in individual stages (n = 1 for stage IA and III, n = 3 for stage IIB), no statistical correlation between LFG expression and tumor stage could be established for leiomyosarcoma.
The LFG expression in leiomyosarcoma is exemplarily shown in Figure 5. Figure 5a shows an immunofluorescent image of endometrium smooth muscle as control with low LFG expression. The highest LFG expression of all specimens was detected in leiomyosarcoma stage IA presenting with multiple, widely distributed fluorescent signals (Fig. 5b). Figure 5c shows the high LFG expression of leiomyosarcoma stage IIB, being significantly higher (p < 0.05) compared to endometrium smooth muscle control (Fig. 5a). Figure 5d shows an image of the high LFG expression that was observed in leiomyosarcoma stage III.

Discussion
The present study is among the first to describe LFG expression in different types of sarcoma. The data presented show convincing evidence that LFG expression is significantly higher in osteosarcoma and chondrosarcoma compared to healthy bone marrow tissue. More importantly, LFG expression was found to positively correlate with tumor stage in osteosarcoma (stage IIB > IIA). In chondrosarcoma, LFG expression increased from stage IA to IIA, but no consistent positive correlation across all stages was observed as stage IIB showed lower expression than stage IIA. In addition, significantly higher expression of anti-apoptotic LFG was observed in several soft tissue sarcoma subtypes, including dermatofibrosarcoma, malignant schwannoma, and leiomyosarcoma.
The correlation between LFG expression and tumor stage was reduced in subtypes of soft tissue sarcoma, compared to osteosarcoma and chondrosarcoma. Therefore, it is necessary to examine the expression of more anti-apoptotic proteins in different sarcoma tissues that interfere with Fas-mediated apoptosis, such as LFG also known as Fas apoptosis inhibitory molecule 2 (FAIM2) [26]. The endogenous expression of LFG in human cancer tissue was already shown for many breast cancer tissues and cell lines, being positively correlated with tumor stage and causing a reduced sensitivity to Fas signaling without changing the levels of Fas expression [27]. Currently, the data regarding LFG in sarcoma tissue are scarce. Therefore, the expression and exact function of LFG in clinical related phenotypes of sarcoma needs to be investigated.
The results of the present study show that LFG is significantly highly expressed in tissues of osteosarcoma and chondrosarcoma than in healthy bone marrow tissue. Additionally, the level of LFG expression was different in several subtypes of soft tissue sarcoma, with the highest expression in dermatofibrosarcoma, malignant schwannoma, and leiomyosarcoma. Furthermore, LFG expression positively correlated with tumor stages of osteosarcoma (IIB > IIA). In chondrosarcoma, LFG expression was significantly higher in stage IIA compared to stage IA, although the overall correlation with tumor stage was not consistent across all stages.
It is important to note that the chondrosarcoma cohort in this study included tumors of different histological grades (grades 1, 2, and 3), whereas the osteosarcoma cohort consisted exclusively of grade 3 tumors. Therefore, the observed variation in LFG expression across chondrosarcoma stages may be influenced not only by tumor stage but also by tumor grade. Future studies with larger sample sizes stratified by both stage and grade are needed to clarify the independent effects of these parameters on LFG expression in chondrosarcoma.
The high LFG expression in osteosarcoma can be seen as another possible explanation for the decrease of Fas-mediated apoptosis in advanced tumor diseases [25]. The correlation of LFG expression in osteosarcoma to the clinical stages underlines the importance of LFG for osteosarcoma cell growth and supports the concept of LFG-guided cell dedifferentiation within metastatic cell populations [23–25]. In addition, Pan et al. [34] stated that LFG overexpression enabled lower malignant osteosarcoma to survive in lower serum albumin environments that might act as a defense system against growing tumor lesions during hematogenous metastasis. However, only higher levels of serum albumin, occurring in nearby neovascularization, empower osteosarcoma cells with high expression of LFG via calcium-mediated interactions to enter the circulatory system and form metastasis in distant organs [34].
Unlike with Bax/Bcl-xL, for which high expression rates have been demonstrated in chondrosarcoma this is the first time that high LFG expression could be correlated to chondrosarcoma and special subtypes of soft tissue sarcoma [35–37]. The anti-apoptotic properties of LFG in advanced stages of chondrosarcoma might represent another explanation that chemotherapy resistance added to a higher expression of anti-apoptotic Bcl-2 proteins and the presence of multidrug resistance pumps [37]. Nevertheless, a more detailed research on LFG expression and function compared to related BI-1 expression in osteosarcoma and chondrosarcoma is needed.
This study has several limitations that should be acknowledged. First, the sample size for control tissues was small, with only two samples of healthy bone marrow tissue serving as controls for both osteosarcoma and chondrosarcoma. Similarly, for soft tissue sarcomas, the control groups consisted of only five skeletal muscle samples and four smooth muscle samples. These small control sample sizes may limit the generalizability of our findings, and future studies with larger control cohorts are warranted. Second, the sample sizes for individual stages within some soft tissue sarcoma subtypes were very small (n = 1–2 in many cases), which precluded meaningful statistical comparisons between stages. Therefore, the numerical trends observed in these subgroups should be interpreted with caution and require validation in larger cohorts. Third, the control tissues used in this study (bone marrow for bone sarcomas, skeletal and smooth muscle for soft tissue sarcomas) are not necessarily the direct tissues of origin for all sarcoma subtypes analyzed. For example, bone marrow is not the tissue of origin for osteosarcoma or chondrosarcoma, which arise from bone-forming and cartilage-forming cells, respectively. Similarly, skeletal and smooth muscles are not the tissues of origin for liposarcoma, angiosarcoma, dermatofibrosarcoma, or malignant schwannoma. Therefore, the comparisons made in this study are between sarcoma tissues and healthy tissues of similar anatomical regions, rather than true “corresponding” tissues of origin. Fourth, this was a retrospective study using commercially available tissue microarrays, and therefore detailed clinical follow-up data (e.g., survival, recurrence, treatment response) were not available. Prospective studies correlating LFG expression with clinical outcomes are needed to establish its prognostic and predictive value. Finally, this study focused solely on LFG protein expression by immunohistochemistry. Functional studies, such as LFG knockdown or overexpression experiments in sarcoma cell lines, are needed to establish a causal role for LFG in sarcoma cell survival, apoptosis resistance, and metastasis. Concluding the LFG expression analysis of the present study, significantly higher levels of LFG expression in leiomyosarcoma might be linked to the aggressive tumor growth and high resistance rates to chemotherapy of this sarcoma subtype [38]. Although no continuous correlation could be found between tumor stage and LFG expression in soft tissue sarcoma, significant higher LFG expression rates, compared to adjacent healthy tissue, were observed in all soft tissue sarcoma subtypes except liposarcoma and synovial cell sarcoma. More research, including higher numbers of soft tissue sarcoma specimen and cell lines, is needed to clarify the expression of LFG and its role regarding tumorigenesis in soft tissue sarcoma.

Conclusion
In summary, high LFG expression could be found in osteosarcoma, chondrosarcoma, and several soft tissue sarcoma subtypes. LFG expression even correlated with tumor stage in osteosarcoma and chondrosarcoma and might function as a predictive marker for tumor progression. Further research on LFG expression and function, including apoptosis-related changes in tumor gene expression and LFG knockout experiments in vitro and in vivo, will help evaluate the potential of LFG as a novel target for sarcoma therapy.

Acknowledgment
The authors are grateful to Andrea Lazaridis for her excellent technical assistance.

Statement of Ethics
Human ethics and consent to participate declarations are not applicable.

Conflict of Interest Statement
The authors have no conflicts of interest to declare.

Funding Sources
This research received no external funding. This publication is funded by the Deutsche Forschungsgemeinschaft (DFG) as part of the “DFG-Publikationsfonds 23/25” program of the library of Hannover Medical School. Open Access funding was enabled and organized by Projekt DEAL.

Author Contributions
Conceptualization: Sebastian Senger, Sarah Strauß, Peter M. Vogt, Frederik Schlottmann, and Vesna Bucan. Data curation, formal analysis, and visualization: Sebastian Senger and Vesna Bucan. Funding acquisition: Peter M. Vogt and Vesna Bucan. Investigation: Sebastian Senger, Andrea Lazaridis, and Vesna Bucan. Methodology: Sebastian Senger, Sarah Strauß, and Vesna Bucan. Project administration and supervision: Sarah Strauß, Peter M. Vogt, and Vesna Bucan. Resources and validation: Sarah Strauß and Vesna Bucan. Writing – original draft: Sebastian Senger. Writing – review and editing: Sebastian Senger, Sarah Strauß, Frederik Schlottmann, Peter M. Vogt, and Vesna Bucan.