Abstract
Purpose
Liposarcomas are the most common soft tissue sarcomas of adults. The identification of lipoblastic cells in soft tissue sarcomas is mandatory for the diagnosis of most subtypes of liposarcomas but may be difficult in conventional histology. The present study focuses on the expression and possible diagnostic impact of two PAT family proteins, perilipin 1/perilipin and perilipin 2/adipophilin in human liposarcomas.
Methods
Eighty-seven cases of liposarcomas and 30 cases of non-lipomatous sarcomas were investigated immunohistochemically for perilipin 1 and 2 using entire tissue sections. Statistical analyses were performed using appropriate tests.
Results
Most liposarcomas and non-lipomatous sarcomas displayed positivity for perilipin 2. In contrast, while more than two-thirds of liposarcomas presented perilipin 1 positivity, all non-lipomatous sarcomas studied were negative for this marker, with statistical significance (p < 0.001). Perilipin 1 expression increased with adipocytic differentiation of liposarcoma subtypes showing statistical significance (p < 0.001). Non-lipomatous sarcomas demonstrated variable expression levels of perilipin 2. The expression level of perilipin 2 appeared to be correlated with tumor cell degeneration, e.g., through hypoxia.
Conclusions
Perilipin 2 is not well suitable for distinction between liposarcomas and non-lipomatous sarcomas. However, perilipin 1 appeared to be a highly specific marker for liposarcoma and adipocytic differentiation in sarcomas with intermediate sensitivity.
Keywords: Liposarcoma, Soft tissue sarcoma, Perilipin 1, Perilipin 2, Adipophilin, Differential expression, PAT family protein, Degeneration, Hypoxia
Introduction
Liposarcomas are the most common soft tissue sarcoma of adults (Goldblum et al. 2014). They are mostly located in the deep soft tissues, e.g., thigh or retroperitoneum (Goldblum et al. 2014). Histological subtypes include myxoid, dedifferentiated and pleomorphic liposarcomas (Fletcher et al. 2013). Atypical lipomatous tumor is a locally aggressive neoplasm, synonymous to well-differentiated liposarcoma and categorized as intermediate adipocytic tumor (Fletcher et al. 2013). The diagnosis of an atypical lipomatous tumor/well-differentiated liposarcoma relies on the detection of nuclear atypia of adipocytic cells or stromal cells (Katenkamp 2011). For other subtypes, identification of lipoblasts is mandatory to document the adipocytic differentiation of the tumor (Katenkamp 2011).
Essential for diagnosis is the conventional hematoxylin and eosin (H&E) morphology, but correct identification of liposarcoma may be difficult. Examples are liposarcomas with sparse adipocytic differentiation or small biopsies with limited tumor material. Also, the correct identification of true lipoblasts may pose problems. For example, lipoblast-like cells may occur in various non-lipomatous sarcomas (e.g., myxoid undifferentiated pleomorphic sarcomas, rhabdomyosarcomas with glycogen-loaden cells), in signet ring melanomas, in certain adenocarcinomas and high-grade lymphomas or simply as fixation artifact (Goldblum et al. 2014). Immunohistochemistry may help in the identification of liposarcomas, but the panel of available markers useful in this context is still insufficient. For example, S100 protein may sometimes highlight lipoblasts in liposarcomas, but this marker is not specific for liposarcomas and its utility in this regard is limited (Dabbs 2014). Molecular methods, such as FISH analysis, for gene amplification such as MDM2 or CDK4 amplification in atypical lipomatous tumor/well-differentiated liposarcoma or dedifferentiated liposarcoma (Katenkamp 2011; Goldblum et al. 2014; Fletcher et al. 2013) or gene translocation such as t(12;16) translocation resulting in fusion of DDIT3 and FUS genes or t(12;22) translocation generating fusion of DDIT3 and EWSR1 genes in myxoid liposarcomas (Dei Tos 2014; Fletcher et al. 2013; Goldblum et al. 2014; Katenkamp 2011) may aid to assure certain liposarcoma subtypes. However, suitable molecular methods such as FISH, PCR or array CGH (Fletcher et al. 2013) are time-consuming, not available in every laboratory and are not relevant for cases such as pleomorphic liposarcomas. Therefore, establishing new and specific immunohistochemical markers for adipocytic differentiation would be of great interest and potential use when differentiating liposarcoma from other soft tissue sarcomas.
The PAT family of proteins comprises a group of amphiphilic proteins coating the surfaces of intracellular lipid storage droplets. The name derives from the first three identified members: perilipin 1 (formerly perilipin), perilipin 2 (formerly adipophilin) and perilipin 3 (formerly TIP47) (Bickel et al. 2009; Kimmel et al. 2010). PAT proteins regulate formation, structural maintainance and degradation of lipid droplets (Sztalryd and Kimmel 2014). Perilipin 1 exists in at least three protein isoforms (molecular weight around 56 kDa), is a marker of adipocyte differentiation and involved in regulating cellular hydrolysis of lipids. Perilipin 2 is a 50 kDa protein and a marker of early adipocyte differentiation. As differentiation progresses during adipogenesis, perilipin 2 protein levels decrease, while perilipin 1 protein levels increase. Early adipocytes have smaller lipid droplets coated by perilipin 2, whereas mature adipocytes show larger lipid droplets with perilipin 1 coating (Bickel et al. 2009; Sztalryd and Kimmel 2014).
The present study focuses on the expression and possible diagnostic impact of two PAT family proteins, perilipin 1/perilipin and perilipin 2/adipophilin in human liposarcomas in comparison with non-lipomatous sarcomas. Our results demonstrate that perilipin 1 might be suitable as a highly specific marker for liposarcoma and adipocytic differentiation in sarcomas with intermediate sensitivity.
Materials and methods
The archival files of the Institutes of Pathology of the Philipps University of Marburg and the Martin Luther University Halle-Wittenberg, Halle/Saale, were searched for lipomatous and non-lipomatous sarcomas between 1980 and 2004. One hundred and seventeen cases were selected for the present study, including 87 liposarcomas and 30 non-lipomatous sarcomas. Tumor types according to the WHO Classification of Soft Tissue Tumors (Fletcher et al. 2013) included in this study are listed in Table 1. Concerning the use of human material, the study was approved by the Ethics Committee of the Medical Faculty of the Philipps University of Marburg; the paper does not include any patient-identifying data.
Table 1.
Overview on the histological types of liposarcomas and non-lipomatous sarcomas studied and their classification in diagnosis groups according to WHO Classification 2013 (Fletcher et al. 2013)
| Entity | No. of cases (% of N = 117) |
|---|---|
| Liposarcomas | 87 (74.4 %) |
| Atypical lipomatous tumor/wdls | 20 (17.1 %) |
| Myxoid liposarcoma (incl. cases with round cell morphology) | 47 (40.2 %) |
| Pleomorphic liposarcoma | 16 (13.7 %) |
| Dedifferentiated liposarcoma | 4 (3.4 %) |
| Non-lipomatous sarcomas | 30 (25.6 %) |
| Rhabdomyosarcoma | 6 (5.1 %) |
| Leiomyosarcoma | 6 (5.1 %) |
| Fibrosarcoma | 6 (5.1 %) |
| Infantile fibrosarcoma | 1 (0.9 %) |
| Intimal sarcoma | 1 (0.9 %) |
| Undifferentiated pleomorphic sarcoma (formerly MFH) | 10 (8.5 %) |
wdls well-differentiated liposarcoma, incl. including, MFH malignant fibrous histiocytoma
Tissues were fixed in 4 % formalin solution, embedded in paraffin, cut at a thickness of 4 µm and stained with hematoxylin and eosin (H&E) for routine purposes. Immunohistochemistry was performed using standard methods (avidin-biotin-complex (ABC) peroxidase method, and EnVision™+Dual Link System-HRP, each Dako, Hamburg, Germany, with 3,3′-diaminobenzidine (DAB) as chromogen). For perilipin 1, a guinea pig polyclonal antibody (N-terminus) as well as a murine monoclonal IgG1 antibody (clone PERI 112.17) was used (own development, described and characterized by (Heid et al. 2013, 2014), commercially available through Progen Biotechnik, Heidelberg, Germany). Perilipin 2 was detected by a murine monoclonal IgG1 antibody (clone AP125) as described previously (Heid et al. 1998). The immunostainings were run on semiautomated Thermo Scientific Shandon Coverplates, Fisher Scientific, Waltham, USA, or on an automated immunostaining apparatus (Dako Autostainer plus, Dako, Hamburg, Germany).
The immunoreactivity of both perilipin 1 and 2 was graded semiquantitatively as negative (0 %), 0.1–5 % positive tumor cells, 5.1–50 % positive tumor cells, 50.1–80 % positive tumor cells and 80.1–100 % positive tumor cells. Areas of necrosis were excluded from analysis. Percentages were assessed by two independent observers, discrepancies were discussed with a third independent observer and consensus was reached.
Statistical analysis was performed using R and compareGroups (R Development Core Team 2008; Subirana et al. 2014). Differences in immunoreactivity between sarcoma groups and different types of liposarcomas and non-lipomatous sarcomas were evaluated with Chi-square, Fisher’s exact or Kruskal–Wallis tests where appropriate.
Results
Immunohistochemically, most liposarcomas studied (82/87 cases; 94.3 %) displayed positivity for perilipin 2 (Table 2). Non-lipomatous sarcomas were also mostly perilipin 2 positive (28/30 cases; 93.3 %). There was no statistically significant difference for perilipin 2 between the two groups (p = 0.961). As to the perilipin 2 staining pattern on the cellular level, in the liposarcomas some but not all lipoblasts showed positive macro- and microvesicular structures while immature tumor cells often exhibited granular cytoplasmic positivity (Fig. 1a, b). In non-lipomatous sarcomas, variable proportions of tumor cells (usually <50 %) revealed partly microvesicular but mainly granular immunostaining for perilipin 2 (Fig. 1d, e). Rhabdomyosarcomas showed in general more abundant perilipin 2 expression (Fig. 1d, e) as compared to leiomyosarcomas, fibrosarcomas and undifferentiated pleomorphic sarcomas, some fibrosarcomas being perilipin 2 negative (Table 3). Altogether, perilipin 2 expression in the subgroups of non-lipomatous sarcomas displayed a statistically significant difference (p = 0.043). The expression of perilipin 2 in tumor cells appeared not only to depend on the tumor type but, within individual tumors, also on zonal characteristics in that tumor cells close to necrotic areas were more frequently positive, suggesting a correlation of perilipin 2 expression and hypoxic tumor cell degeneration. Next to tumor cells, also intratumoral macrophages were perilipin 2 positive, and these cells also were more abundant around necrotic tumor areas and included prominently stained foamy macrophages.
Table 2.
Semiquantitative assessment of adipophilin/perilipin 2 and perilipin 1 in tumor cells of all cases of the two sarcoma groups; p values of the χ 2 or Fisher’s exact tests for association between the two groups
| Liposarcoma N = 87 |
Non-lipomatous sarcoma N = 30 |
p overall | |
|---|---|---|---|
| Adipophilin/ perilipin 2 | 0.961 | ||
| Negative | 5 (5.7 %) | 2 (6.7 %) | |
| 0.1–5 % positive cells | 29 (33.3 %) | 11 (36.7 %) | |
| 5.1–50 % positive cells | 29 (33.3 %) | 11 (36.7 %) | |
| 50.1–80 % positive cells | 14 (16.1 %) | 4 (13.3 %) | |
| 80.1–100 % positive cells | 10 (11.5 %) | 2 (6.7 %) | |
| Perilipin 1 | <0.001 | ||
| Negative | 25 (28.7 %) | 30 (100 %) | |
| 0.1–5 % positive cells | 15 (17.2 %) | 0 (0.00 %) | |
| 5.1–50 % positive cells | 24 (27.6 %) | 0 (0.00 %) | |
| 50.1–80 % positive cells | 11 (12.6 %) | 0 (0.00 %) | |
| 80.1–100 % positive cells | 12 (13.8 %) | 0 (0.00 %) |
Fig. 1.
Staining patterns of adipophilin/perilipin 2 and perilipin 1 in two selected cases of lipomatous and non-lipomatous sarcomas (a, d H&E stain; b, e adipophilin/perilipin 2 immunohistochemistry; c, f perilipin 1 immunohistochemistry). a–c A pleomorphic liposarcoma with groups of pleomorphic lipoblasts next to areas lacking lipoblastic differentiation (a H&E stain, ×200). Adipophilin/perilipin 2 stains some lipoblasts with positive microvesicular structures while immature, histologically non-lipogenic cells tumor cells often exhibit granular cytoplasmic positivity (b ×200). Perilipin 1 extensively marks the lipoblasts with macro- and microvesicular staining pattern corresponding to the fat vacuoles, while histologically non-lipogenic cells remain negative (c ×200). Arrows point out adipophilin/perilipin 2 negative, but perilipin 1 positive lipoblasts (b, c), while arrowheads emphasize adipophilin/perilipin 2 and 1 positive lipoblasts in the same tumor (b, c). d–f An embryonal rhabdomyosarcoma with vacuolated tumor cells (d H&E stain, ×200) reveals an extensive, mainly granular immunostaining for adipophilin/perilipin 2 (e, ×200), but no staining for perilipin 1 (f, ×200), indicating that the cytoplasmic vacuoles do not represent adipocytic lipid droplets
Table 3.
Semiquantitative assessment of adipophilin/perilipin 2 and perilipin 1 in different types of non-lipomatous sarcomas
| Rhabdomyosarcoma N = 6 |
Leiomyosarcoma N = 6 |
Fibrosarcoma N = 6 |
Infantile fibrosarcoma N = 1 |
Intimal sarcoma N = 1 |
Undifferentiated pleomorphic sarcoma (formerly MFH) N = 10 |
|
|---|---|---|---|---|---|---|
| Adipophilin/perilipin 2 | ||||||
| Negative | 0 (0.0 %) | 0 (0.0 %) | 2 (33.3 %) | 0 (0.0 %) | 0 (0.0 %) | 0 (0.0 %) |
| 0.1–5 % positive cells | 0 (0.0 %) | 4 (66.7 %) | 2 (33.3 %) | 0 (0.0 %) | 0 (0.0 %) | 5 (50.0 %) |
| 5.1–50 % positive cells | 4 (66.7 %) | 1 (16.7 %) | 2 (33.3 %) | 0 (0.0 %) | 0 (0.0 %) | 4 (40.0 %) |
| 50.1–80 % positive cells | 1 (16.7 %) | 1 (16.7 %) | 0 (0.0 %) | 1 (100 %) | 0 (0.0 %) | 1 (10.0 %) |
| 80.1–100 % positive cells | 1 (16.7 %) | 0 (0.0 %) | 0 (0.0 %) | 0 (0.0 %) | 1 (100 %) | 0 (0.0 %) |
| Perilipin 1 | ||||||
| Negative | 6 (100 %) | 6 (100 %) | 6 (100 %) | 1 (100 %) | 1 (100 %) | 10 (100 %) |
MFH malignant fibrous histiocytoma
More than two-thirds of liposarcomas (62/87 cases; 71.3 %) presented perilipin 1 positivity (Table 2), evident predominantly in lipoblasts with macro- and microvesicular staining pattern corresponding to the fat vacuoles (Figs. 1c, 2). In contrast, all 30 non-lipomatous sarcomas studied were negative for perilipin 1 (Table 2; Fig. 1f), resulting in a statistically significant difference between the two groups (p < 0.001) and suggesting that among soft tissue sarcomas perilipin 1 is highly specific for liposarcomas.
Fig. 2.
Typical heterogeneous perilipin 1 immunostaining in myxoid liposarcoma. Note mostly microvesicular staining in lipoblasts of different stages of maturation as well as granular staining in scattered small tumor cells indicating incipient adipocytic differentiation (×200)
Among subtypes of liposarcomas, especially myxoid liposarcomas showed strong perilipin 2 expression (Fig. 2; Table 4). In fact, perilipin 2 expression in more than 50 % of tumor cells was almost only observed in myxoid liposarcomas. Also, pleomorphic liposarcomas displayed a comparably more abundant expression of perilipin 2, whereas atypical lipomatous tumor/well-differentiated liposarcomas and dedifferentiated liposarcomas showed lower expression levels. Statistical analysis revealed a significant difference in perilipin 2 expression between the liposarcoma subtypes (p = 0.003). When comparing the liposarcoma types with more pronounced perilipin 2 staining (myxoid and pleomorphic liposarcoma) with the other liposarcomas, a statistically significant difference was shown (p < 0.001). Perilipin 1 positivity was most apparent among atypical lipomatous tumors/well-differentiated liposarcomas, as most of their tumor cells (mature adipocyte-like cells and lipoblasts) contained large perilipin 1 positive fat vacuoles. One case of this subtype, which belonged to the sclerosing variety, was perilipin 1 negative. Myxoid liposarcomas showed highly variable proportions of perilipin 1 positive tumor cells, while pleomorphic liposarcomas contained mostly low levels (up to 5 % of tumor cells). Thus, perilipin 1 expression increased with adipocytic differentiation of liposarcoma subtypes, revealing a statistically significant difference (p < 0.001).
Table 4.
Semiquantitative assessment of adipophilin/perilipin 2 and perilipin 1 in subtypes of liposarcomas; p values of the Kruskal–Wallis test for association between the different subtypes
| Atypical lipomatous tumor/wdls N = 20 |
Myxoid liposarcoma N = 47 |
Pleomorphic liposarcoma N = 16 |
Dedifferentiated liposarcoma N = 4 |
p overall | |
|---|---|---|---|---|---|
| Adipophilin/perilipin 2 | 0.003 | ||||
| Negative | 0 (0.0 %) | 5 (10.6 %) | 0 (0.0 %) | 0 (0.0 %) | |
| 0.1–5 % positive cells | 12 (60.0 %) | 9 (19.1 %) | 7 (43.8 %) | 1 (25.0 %) | |
| 5.1–50 % positive cells | 8 (40.0 %) | 10 (21.3 %) | 8 (50.0 %) | 3 (75.0 %) | |
| 50.1–80 % positive cells | 0 (0.0 %) | 13 (27.7 %) | 1 (6.3 %) | 0 (0.0 %) | |
| 80.1–100 % positive cells | 0 (0.0 %) | 10 (21.3 %) | 0 (0.0 %) | 0 (0.0 %) | |
| Perilipin 1 | <0.001 | ||||
| Negative | 1 (5.0 %) | 18 (38.3 %) | 4 (25.0 %) | 2 (50.0 %) | |
| 0.1–5 % positive cells | 0 (0.0 %) | 5 (10.6 %) | 9 (56.2 %) | 1 (25.0 %) | |
| 5.1–50 % positive cells | 3 (15.0 %) | 18 (38.3 %) | 2 (12.5 %) | 1 (25.0 %) | |
| 50.1–80 % positive cells | 4 (20.0 %) | 6 (12.8 %) | 1 (6.3 %) | 0 (0.0 %) | |
| 80.1–100 % positive cells | 12 (60.0 %) | 0 (0.0 %) | 0 (0.0 %) | 0 (0.0 %) |
wdls well-differentiated liposarcoma
Discussion
The identification of lipoblastic cells in soft tissue sarcomas is mandatory for most subtypes of liposarcoma (Katenkamp 2011). However, in conventional morphology it may be extremely difficult to discriminate lipoblasts from other forms of vacuolated or lipoblast-like tumor cells (with lipid droplets or non-fatty vacuoles) occurring in non-lipomatous mesenchymal tumors e.g., as intrinsic morphological feature or as a non-specific phenomenon of tumor cell degeneration (see Introduction). Histochemical fat stains might be helpful but cannot be applied with paraffin-embedded tissue and also would not distinguish adipocytic vacuoles from degenerative or resorptive lipid droplets. Also, current immunohistological markers are insufficient for this histodiagnostic problem. Thus, immunohistological markers specific for true lipid vacuoles formed by adipocytic cells and tumors might get highly relevant for the differential diagnosis of soft tissue sarcomas.
Already some time ago, we reported on preliminary data concerning the possible use of the PAT proteins adipophilin/perilipin 2 and perilipin/perilipin 1 in the pathohistological differential diagnosis of human soft tissue sarcomas, suggesting perilipin 1 as a highly specific marker for the recognition of lipomatous differentiation and the diagnosis of liposarcomas (Mrozinski et al. 1999). Meanwhile, monoclonal antibodies against perilipin 1 such as PERI 112.17 became available (Heid et al. 2013) rendering a diagnostic application more feasible. We now confirm and extend our previous data and demonstrate the very high specificity—as opposed to intermediate sensitivity—of perilipin 1 for liposarcomas.
Recently, in a paper characterizing myxoid liposarcomas with extensive lipoma-like changes, Iwasaki et al. (2015) described positivity of most liposarcomas (including conventional myxoid, well-differentiated and dedifferentiated liposarcomas) for both adipophilin/perilipin 2 and perilipin/perilipin 1. The authors suggested that the combined immunohistochemical detection of perilipin 1 and 2 may provide a useful ancillary tool for identification of lipoblastic cells in soft tissue sarcomas. However, they did not include non-lipomatous sarcomas in their analysis which as we here show also express perilipin 2 but not perilipin 1. We agree with their finding of tiny perilipin 2 positive fat droplets present in immature lipoblastic cells of myxoid liposarcoma but similar positivity in tiny cytoplasmic structures is a feature of many non-lipomatous sarcomas lacking true lipoblasts, indicating that perilipin 2 is not a specific lipoblast marker.
Perilipin 2, previously called adipophilin or adipose differentiation-related protein (ADRP), is a lipid droplet-coating protein expressed early in adipogenesis but down-regulated in mature adipocytes. In normal tissues, it has a broad cell type distribution (Heid et al. 1998; Straub et al. 2013). It is also expressed in many types of carcinomas, most prominently in certain lipid-storing tumors such as clear cell renal carcinoma (Straub et al. 2010). Among sarcomas, as our study demonstrates, perilipin 2 seems to be a sensitive marker for liposarcomas, as about 94 % of these tumors display perilipin 2 positivity and immunostaining for this protein may outline vacuolated tumor cells corresponding to immature microvesicular lipoblasts (Fig. 1b; see also (Iwasaki et al. 2015), for the expression of perilipin 1 versus perilipin 2 in cultured human adipocytes and during adipogenic differentiation see (Heid et al. 2014)). Moreover, perilipin 2 expression in more than 50 % of tumor cells was rarely found in non-lipomatous sarcomas and, among liposarcomas, was nearly exclusive for myxoid liposarcoma. Therefore, perilipin 2 might help as a surrogate marker in identifying certain lipoblastic cells in putative liposarcomas. However, obviously not all perilipin 2 positive tumor cells of liposarcomas represent lipoblasts, and, notably, most non-lipomatous soft tissue tumors are also at least partially positive for perilipin 2. Accordingly, there was no statistical difference between the liposarcoma and the non-lipomatous sarcoma group. Increased perilipin 2 expression was found in regressive tumor cells in ischemic zones close to necrotic areas, probably due to fatty degeneration. In addition, lipid-storing intratumoral macrophages, including foam cells, were also prominently perilipin 2 positive and sometimes were difficult to distinguish from perilipin 2 positive tumor cells. In previous studies, perilipin 2 mRNA was demonstrated in a foam cell subset in human arteriosclerosis via RT-PCR (Wang et al. 1999). Straub et al. reported on perilipin 2 expression in lipid-laden macrophages in cholesteatosis and cholesterol polyps of the gall bladder, in atherosclerotic lesions and in fat necrosis after trauma as well as in microvesicular hepatocyte steatosis (Straub et al. 2013). In a cell culture model of human preadipocytes, perilipin 2 positive lipid droplets appeared after exogenous uptake of oleic acid in the absence of adipocytic differentiation (Heid et al. 2014). Altogether, the diagnostic value of perilipin 2 for the typing of soft tissue sarcomas appears limited.
Our results and data from Straub et al. (2010, 2013) demonstrate the widespread storage of lipid droplets in various types of tumors, with perilipin 2 representing one of the major lipid droplet-coating proteins. Along this line, accumulation of lipid droplets has also been described for malignant peripheral nerve sheath tumors (Patel et al. 2015) in which fatty acid synthase may be involved, and this enzyme of lipid metabolism, which is overexpressed in many cancers including soft tissue tumors (Menendez and Lupu 2007; Pandey et al. 2012), has been shown in vitro to support tumor growth in both MPNST, i.e., a non-lipomatous sarcoma (Patel et al. 2015) and in liposarcomas (Olsen et al. 2010). In fact, fatty acid synthase has been suggested as a potential therapeutic target (Olsen et al. 2010; Patel et al. 2015; Pandey et al. 2012). In hepatocellular carcinomas, expression of perilipin 2 was positively correlated with tumor cell proliferation (Straub et al. 2010). Future studies should exploit whether perilipin 2 might also be conceivable as an antineoplastic target protein.
Perilipin 1 is the major lipid droplet-coating protein of mature uni- and plurivacuolar adipocytes, which lack perilipin 2 due to down-regulation during adipogenesis (Brasaemle et al. 1997; Kimmel et al. 2010). Among normal tissues, perilipin 1 has a high specificity for adipocytes of white and brown fatty tissue; beyond, it only occurs in sebaceous gland cells and endocrine steroidogenic cells such as adrenal cortical and Leydig cells. In hepatocytes, perilipin 1 appears de novo upon steatogenesis, being correlated with chronic steatosis and the degree of steatosis (Pawella et al. 2014; Straub et al. 2008). Correspondingly, in benign and malignant tumors the expression of perilipin 1 is very restricted, notably to adipogenic, sebaceous and hepatocellular tumors (Straub et al. 2010). The high specificity of perilipin 1 for adipogenic differentiation in mesenchymal cells was demonstrated in vitro in human preadipocytes which in the undifferentiated state contain small lipid droplets including perilipin 2, but upon adipogenic differentiation induction specifically produce bigger perilipin 1 positive lipid droplets (Heid et al. 2014).
The expression of perilipin 1 in liposarcomas has been reported previously (Mrozinski et al. 1999; Iwasaki et al. 2015; Straub et al. 2010). In the present study, we demonstrate that among soft tissue sarcomas perilipin 1 is highly specific for liposarcomas, being expressed with variable frequency in all liposarcoma subtypes but being completely absent from all non-lipomatous sarcomas analyzed. This may be of particular diagnostic interest when vacuolated tumor cells (suggestive of lipoblasts) are present in non-lipomatous sarcomas. However, while perilipin 1 positivity in a sarcomatous tumor renders the diagnosis of a liposarcoma most probable, perilipin 1 negativity does not rule out a liposarcoma. In our study, the sensitivity of perilipin 1 as a marker for liposarcomas was 71.3 %. Whether heterologous lipomatous differentiation rarely observed among soft tissue sarcomas, e.g., in malignant peripheral nerve sheath tumors (Suresh et al. 2009), is accompanied by perilipin 1 expression remains to be determined.
Concerning the liposarcoma subtypes, with perilipin 1 positivity in more than 80 % of tumor cells, an atypical lipomatous tumor/well-differentiated liposarcoma can be assumed according to the present results. If more than 50 % of tumor cells display perilipin 1 positivity, an atypical lipomatous tumor/well-differentiated liposarcoma or a myxoid liposarcoma are more likely than a pleomorphic or dedifferentiated liposarcoma. In our study, also very focal expression of perilipin 1 was recorded since we stained representative whole tumor slides rather than tissue microarrays (TMAs), and even such focal expression appears to be highly specific and diagnostic since it was not observed in non-lipomatous sarcomas. Sparse perilipin 1 expression was a common finding in pleomorphic liposarcomas, the diagnosis of which may be challenging especially due to its relative rarity among liposarcomas and therefore limited availability of molecular genetic data (Goldblum et al. 2014). Thus, FISH diagnostic may not be as helpfully applied as with the other liposarcoma types (see Introduction), and the most important clue for the diagnosis of pleomorphic liposarcoma is the detection of even focal adipocytic differentiation (Dei Tos 2014): Here, even sparse perilipin 1 expression might be of particular diagnostic relevance.
In conclusion, perilipin 2 can help in identifying lipoblastic cells in putative liposarcomas, but its expression is not specific for liposarcomas, i.e., it is not suitable as a marker of differentiation between liposarcomas and non-lipomatous sarcomas. However, perilipin 1, as detectable by well-characterized monoclonal antibodies, proved to be a highly specific marker of liposarcoma and adipocytic differentiation in sarcomas with intermediate sensitivity.
Acknowledgments
We thank Stefanie Winter-Simanowski, Anke Holzbach-Röper and Viktoria Wischmann for expert technical assistance. Dr. Raoul Hinze (Martin Luther University of Halle-Wittenberg, Institute of Pathology, Halle/Saale, present address HELIOS Clinics, Institute of Pathology, Schwerin) was of great help with the recruitment of cases. We also thank Hanna Daniel (Philipps University of Marburg, Institute of Medical Biometry and Epidemiology, Marburg) for helpful advice and comments regarding the statistical analysis and Prof. Dr. Werner W. Franke (Helmholtz Group for Cell Biology, German Cancer Research Center, Heidelberg) for continuous support. CCW is supported by the postdoctoral lecture qualification program of the Anneliese Pohl Foundation, Marburg.
Funding
The work was funded by Philipps University of Marburg, Marburg, Germany.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval
All procedures performed in studies involving human participants were in accordance with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Concerning the use of human material, the study was approved by the Ethics Committee of the Medical Faculty of the Philipps University of Marburg. This article does not contain any studies with animals performed by any of the authors.
References
- Bickel PE, Tansey JT, Welte MA (2009) PAT proteins, an ancient family of lipid droplet proteins that regulate cellular lipid stores. Biochim Biophys Acta 1791(6):419–440. doi:10.1016/j.bbalip.2009.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brasaemle DL, Barber T, Wolins NE, Serrero G, Blanchette-Mackie EJ, Londos C (1997) Adipose differentiation-related protein is an ubiquitously expressed lipid storage droplet-associated protein. J Lipid Res 38(11):2249–2263 [PubMed] [Google Scholar]
- Dabbs DJ (2014) Diagnostic immunohistochemistry: theranostic and genomic applications, 4th edn. Elsevier/Saunders, Philadelphia [Google Scholar]
- Dei Tos AP (2014) Liposarcomas: diagnostic pitfalls and new insights. Histopathology 64(1):38–52. doi:10.1111/his.12311 [DOI] [PubMed] [Google Scholar]
- Fletcher CDM, Bridge JA, Hogendoorn PCW, Mertens F (eds) (2013) WHO classification of tumours of soft tissue, 4th edn. World Health Organization Classification of Tumours, Lyon [Google Scholar]
- Goldblum JR, Folpe AL, Weiss SW, Enzinger FM (2014) Enzinger and Weiss’s soft tissue tumors, 6th edn. Saunders/Elsevier, Philadelphia [Google Scholar]
- Heid HW, Moll R, Schwetlick I, Rackwitz H-R, Keenan TW (1998) Adipophilin is a specific marker of lipid accumulation in diverse cell types and diseases. Cell Tissue Res 294(2):309–321. doi:10.1007/s004410051181 [DOI] [PubMed] [Google Scholar]
- Heid H, Rickelt S, Zimbelmann R, Winter S, Schumacher H, Dorflinger Y (2013) Lipid droplets, perilipins and cytokeratins—unravelled liaisons in epithelium-derived cells. PLoS ONE 8(5):e63061. doi:10.1371/journal.pone.0063061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heid H, Rickelt S, Zimbelmann R, Winter S, Schumacher H, Dorflinger Y, Kuhn C, Franke WW (2014) On the formation of lipid droplets in human adipocytes: the organization of the perilipin–vimentin cortex. PLoS ONE 9(2):e90386. doi:10.1371/journal.pone.0090386 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Iwasaki H, Ishiguro M, Nishio J, Aoki M, Yokoyama R, Yokoyama K, Taguchi K, Nabeshima K (2015) Extensive lipoma-like changes of myxoid liposarcoma: morphologic, immunohistochemical, and molecular cytogenetic analyses. Virchows Arch 466(4):453–464. doi:10.1007/s00428-015-1721-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Katenkamp D (2011) Histological classification of soft tissue tumors and staging according to the TNM system (Histologische Klassifikation von Weichgewebstumoren und Stadieneinteilung gemass TNM-System). Pathologe 32(1):8–13. doi:10.1007/s00292-010-1391-1 [DOI] [PubMed] [Google Scholar]
- Kimmel AR, Brasaemle DL, McAndrews-Hill M, Sztalryd C, Londos C (2010) Adoption of PERILIPIN as a unifying nomenclature for the mammalian PAT-family of intracellular lipid storage droplet proteins. J Lipid Res 51(3):468–471. doi:10.1194/jlr.R000034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Menendez JA, Lupu R (2007) Fatty acid synthase and the lipogenic phenotype in cancer pathogenesis. Nat Rev Cancer 7(10):763–777. doi:10.1038/nrc2222 [DOI] [PubMed] [Google Scholar]
- Mrozinski J, Schwetlick I, Heid HW, Winter-Simanowski S, Holzbach A, Hinze R, Franke WW, Moll R (1999) Adipophilin and perilipin, two lipid-droplet-associated proteins, in the pathohistological differential diagnosis of human soft tissue sarcomas. P14. J Cancer Res Clin Oncol 125(Suppl):S48–S49 [Google Scholar]
- Olsen AM, Eisenberg BL, Kuemmerle NB, Flanagan AJ, Morganelli PM, Lombardo PS, Swinnen JV, Kinlaw WB (2010) Fatty acid synthesis is a therapeutic target in human liposarcoma. Int J Oncol 36(5):1309–1314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pandey PR, Liu W, Xing F, Fukuda K, Watabe K (2012) Anti-cancer drugs targeting fatty acid synthase (FAS). Recent Pat Anticancer Drug Discov 7(2):185–197 [DOI] [PubMed] [Google Scholar]
- Patel AV, Johansson G, Colbert MC, Dasgupta B, Ratner N (2015) Fatty acid synthase is a metabolic oncogene targetable in malignant peripheral nerve sheath tumors. Neuro Oncol 17(12):1599–1608. doi:10.1093/neuonc/nov076 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pawella LM, Hashani M, Eiteneuer E, Renner M, Bartenschlager R, Schirmacher P, Straub BK (2014) Perilipin discerns chronic from acute hepatocellular steatosis. J Hepatol 60(3):633–642. doi:10.1016/j.jhep.2013.11.007 [DOI] [PubMed] [Google Scholar]
- R Development Core Team (2008) R: A language and environment for statistical computing. Austria, Vienna [Google Scholar]
- Straub BK, Stoeffel P, Heid H, Zimbelmann R, Schirmacher P (2008) Differential pattern of lipid droplet-associated proteins and de novo perilipin expression in hepatocyte steatogenesis. Hepatology 47(6):1936–1946. doi:10.1002/hep.22268 [DOI] [PubMed] [Google Scholar]
- Straub BK, Herpel E, Singer S, Zimbelmann R, Breuhahn K, Macher-Goeppinger S, Warth A, Lehmann-Koch J, Longerich T, Heid H, Schirmacher P (2010) Lipid droplet-associated PAT-proteins show frequent and differential expression in neoplastic steatogenesis. Mod Pathol 23(3):480–492. doi:10.1038/modpathol.2009.191 [DOI] [PubMed] [Google Scholar]
- Straub BK, Gyoengyoesi B, Koenig M, Hashani M, Pawella LM, Herpel E, Mueller W, Macher-Goeppinger S, Heid H, Schirmacher P (2013) Adipophilin/perilipin-2 as a lipid droplet-specific marker for metabolically active cells and diseases associated with metabolic dysregulation. Histopathology 62(4):617–631. doi:10.1111/his.12038 [DOI] [PubMed] [Google Scholar]
- Subirana I, Sanz H, Vila J (2014) Building bivariate tables: the compareGroups package for R. J Stat Softw 57(12):1–1625400517 [Google Scholar]
- Suresh TN, Harendra Kumar ML, Prasad CS, Kalyani R, Borappa K (2009) Malignant peripheral nerve sheath tumor with divergent differentiation. Indian J Pathol Microbiol 52(1):74. doi:10.4103/0377-4929.44971 [DOI] [PubMed] [Google Scholar]
- Sztalryd C, Kimmel AR (2014) Perilipins: lipid droplet coat proteins adapted for tissue-specific energy storage and utilization, and lipid cytoprotection. Biochimie 96:96–101. doi:10.1016/j.biochi.2013.08.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang X, Reape TJ, Li X, Rayner K, Webb CL, Burnand KG, Lysko PG (1999) Induced expression of adipophilin mRNA in human macrophages stimulated with oxidized low-density lipoprotein and in atherosclerotic lesions. FEBS Lett 462(1–2):145–150. doi:10.1016/S0014-5793(99)01521-5 [DOI] [PubMed] [Google Scholar]