Introduction

Progranulin is a secreted glycosylated protein with critical functions in numerous biological and pathological processes, including cell growth, tumorigenesis, wound healing, inflammation, immunity, infection, and diabetes (Cenik et al. 2012; Eriksen and Mackenzie 2008; Jian et al. 2013a; Toh et al. 2011; Abella et al. 2017). In the central nervous system (CNS), progranulin acts as a neurotrophic and neuroprotective factor. Recently, changes in progranulin expression that are related to pathological conditions have been reported in various neurological diseases (Fig. 1). Mutations in the progranulin gene (GRN) were recently linked to certain forms of frontotemporal lobar degeneration (FTLD) (Baker et al. 2006; Cruts et al. 2006; Gass et al. 2006). The GRN-related form of FTLD is neuropathologically characterized by the appearance of neuronal inclusions containing ubiquitinated and fragmented TAR DNA binding protein-43 (TDP-43). Measurement of progranulin in blood and cerebrospinal fluid (CSF) can identify FTLD patients and asymptomatic carriers of GRN mutations, with progranulin haploinsufficiency leading to decreased progranulin levels (Ghidoni et al. 2008, 2012a, b; Finch et al. 2009; Sleegers et al. 2009; Carecchio et al. 2011; Galimberti et al. 2018).

Fig. 1
figure 1

Progranulin expression levels in central nervous system diseases

In addition to progranulin deficiency, there are conditions in which protein expression is increased. Progranulin levels in biological fluids are generally low, but are upregulated in the inflammatory state, strongly supporting its use as a biomarker of disease onset and progression in several pathologies (Abella et al. 2017). Progranulin stimulates cell division and promotes tumor formation (Serrero 2003; Ong and Bateman 2003; Serrero and Ioffe 2003), and it is highly expressed in aggressive cancer cell lines and many malignancies. Changes in circulating progranulin levels have been observed in breast cancer (Koo et al. 2012), ovarian cancer (Han et al. 2011), hematological malignancies (Göbel et al. 2013; Yamamoto et al. 2017), and non-small-cell lung cancer (Edelman et al. 2014), as assessed by enzyme immunoassay (EIA). Therefore, progranulin may have potential as a prognostic biomarker of malignancy recurrence.

Progranulin is also associated with the pathophysiology of several autoimmune diseases. Progranulin binds to tumor necrosis factor (TNF) receptors (TNFRs), and disrupts TNFα–TNFR interactions (Liu and Bosch 2012; Jian et al. 2013b; Tang et al. 2011). Progranulin-deficient mice are susceptible to collagen-induced arthritis, while administration of progranulin alleviates inflammatory arthritis (Tang et al. 2011). Moreover, there are several reports demonstrating significantly higher concentrations of serum progranulin in autoimmune diseases (including rheumatoid arthritis [RA] and systemic lupus erythematous [SLE]), compared with healthy controls (Tanaka et al. 2012; Yamamoto et al. 2014).

In this chapter, we summarize recent advances on the use of progranulin as a potential biomarker of CNS diseases, including malignancies, and neurodegenerative and autoimmune neurological disorders.

Progranulin as a Biomarker of Neurodegenerative Diseases

Progranulin was first reported as a growth factor associated with tumor growth (He and Bateman 1999). In the CNS, progranulin functions as a neurotrophic and neuroprotective factor (Chitramuthu et al. 2017), and recent studies show that GRN mutations cause several neurodegenerative diseases. The first GRN mutations were discovered in FTLD families with ubiquitin- and TDP43-positive pathologies (Baker et al. 2006; Cruts et al. 2006; Gass et al. 2006). While heterozygosity for the mutations results in FTLD, homozygosity leads to neuronal ceroid lipofuscinosis, a lysosomal storage disease (Smith et al. 2012; Almeida et al. 2016). The clinical symptoms associated with FTLD are diverse, including behavioral and personality changes, language disorders of expression and comprehension, cognitive impairment, and occasionally, motor neuron disease (McKhann et al. 2001). Intriguingly, missense GRN mutations are also observed in patients with clinically diagnosed Alzheimer’s disease (Perry et al. 2013) and amyotrophic lateral sclerosis (ALS) (Schymick et al. 2007). Thus, patients with GRN mutations can present with a variety of neurodegenerative diseases and a broad spectrum of clinical phenotypes.

GRN null mutations cause protein haploinsufficiency, leading to a significant reduction in progranulin levels in the plasma, serum and CSF of mutation carriers (Ghidoni et al. 2008, 2012a, b; Finch et al. 2009; Sleegers et al. 2009; Carecchio et al. 2011; Galimberti et al. 2018). The measurement of circulating progranulin levels enables screening of GRN mutation carriers quickly and inexpensively. Several reports show that plasma progranulin levels predict GRN mutation status in FTLD patients and asymptomatic family members (Ghidoni et al. 2008, 2012a; Finch et al. 2009; Galimberti et al. 2018). Finch et al. investigated progranulin levels in plasma samples from FTLD patients, including symptomatic and asymptomatic relatives of patients with GRN mutations (Finch et al. 2009). All FTLD patients with GRN mutations showed significantly reduced levels of progranulin in plasma, to approximately one-third of the levels observed in non-GRN carriers and control individuals. These researchers also found low progranulin levels in asymptomatic GRN mutation carriers. Galimberti et al. investigated whether plasma progranulin levels are predictors of GRN null mutations in FTLD family members in a cohort including FTLD patients, asymptomatic carriers, and non-carriers (Galimberti et al. 2018). They found that plasma progranulin levels in FTLD patients and asymptomatic carriers were significantly decreased compared with non-carriers. At a threshold of 61.55 ng/mL, the test showed a sensitivity of 98.8% and a specificity of 97.5% for predicting the presence of GRN null mutations, independent of symptoms. Thus, circulating progranulin levels may be a reliable biomarker, with high sensitivity and specificity, for the diagnosis and early detection of GRN-related neurodegenerative diseases. Measuring circulating progranulin levels may become an indispensable tool for preventing or delaying the onset of GRN-related neurodegenerative diseases in the near future.

Progranulin as a Biomarker of CNS Malignancies

Recent studies suggest that progranulin may be a potential clinical biomarker of various malignancies. Progranulin is associated with cell proliferation, migration, invasion, malignant transformation, angiogenesis, resistance to anticancer drugs, and immune evasion (Arechavaleta-Velasco et al. 2017). Progranulin is highly expressed in aggressive cancer cell lines and specimens from many malignancies (Table 1) (Serrero 2003; Serrero and Ioffe 2003; Han et al. 2011; Göbel et al. 2013; Yamamoto et al. 2017; Edelman et al. 2014; Frampton et al. 2012; Kim et al. 2010, 2012; Lovat et al. 2009; Selmy et al. 2010; Li et al. 2012; Tkaczuk et al. 2011; Lu et al. 2014; Wei et al. 2015a; Yang et al. 2015; Chen et al. 2008; Wang et al. 2012; Bandey et al. 2015; Ho et al. 2008; Cuevas-Antonio et al. 2010; Matsumura et al. 2006; Pan et al. 2004; Donald et al. 2001). Regardless of tumor type, progranulin is overexpressed in cancer cells and has growth-promoting and chemoresistant actions. In patients with malignancies, increased circulating progranulin levels have been observed by EIA. Moreover, increased circulating progranulin levels correlate with pathological grading and prognosis in several types of cancer (Table 2) (Koo et al. 2012; Han et al. 2011; Göbel et al. 2013; Yamamoto et al. 2017; Edelman et al. 2014; Kim et al. 2010, 2012; Selmy et al. 2010; Wang et al. 2012; Bandey et al. 2015; Ho et al. 2008; Cuevas-Antonio et al. 2010; Donald et al. 2001; Serrero et al. 2012; Li et al. 2011; Carlson et al. 2013).

Table 1 Malignancies showing enhanced progranulin expression
Table 2 Malignancies with an association between progranulin level and pathological grading or prognosis

In the CNS, progranulin is often highly expressed in gliomas (Wang et al. 2012; Bandey et al. 2015). Progranulin plays a role in astrocytoma progression and is a prognostic biomarker for glioblastoma, with overexpression predicting decreased survival (Wang et al. 2012). Progranulin is overexpressed in tumors from patients with glioblastoma multiforme, and is associated with tumorigenicity and temozolomide resistance (Bandey et al. 2015). It is also implicated in the growth of intracranial meningioma (Kim et al. 2010). Recently, we reported that increased CSF progranulin levels are found in patients with CNS lymphomas (primary and secondary CNS lymphoma) and carcinomas with CNS metastasis (carcinomatous meningitis and brain metastasis) (Kimura et al. 2018). We examined CSF progranulin levels in various CNS diseases by EIA. Specifically, we compared progranulin levels among disease groups in CSF samples from 230 patients, including 18 with lymphomas (12 with CNS metastasis and 6 without CNS metastasis), 21 with carcinomas (10 with CNS metastasis and 11 without CNS metastasis), and 191 control patients with non-cancer neurological diseases. Median CSF progranulin levels were significantly higher in the lymphoma with CNS metastasis group compared with the lymphoma without CNS metastasis and control non-cancer groups. Additionally, levels were also significantly higher in the carcinoma with CNS metastasis group compared with the carcinoma without CNS metastasis and control non-cancer groups, except for patients with infectious neurological disorders (Fig. 2). Importantly, increased CSF progranulin levels were observed in lymphomas and carcinomas with metastasis regardless of tumor type. Using receiver operator characteristic (ROC) curves, we determined the suitability of CSF progranulin as a biomarker for lymphomas and carcinomas with CNS metastasis. The area under the ROC curve (AUC) was 0.969 for differentiating lymphoma with CNS metastasis (compared with lymphoma without CNS metastasis and non-cancer neurological diseases), and 0.918 for differentiating carcinoma with CNS metastasis (compared with carcinoma without CNS metastasis and non-cancer neurological diseases) (Fig. 3). These findings are clinically important because diagnosing CNS metastases can be difficult in patients with lymphomas and carcinomas as well as in those with histories of these diseases and whose neurological symptoms (such as headache, gait disturbance, sensory disturbance, and cognitive impairment) are also observed in other inflammatory and non-inflammatory neurological diseases. Diagnosis is also difficult in patients with lymphomas and carcinomas without CNS metastasis with paraneoplastic neurological syndromes or side effects of chemotherapy.

Fig. 2
figure 2

Progranulin levels in cerebrospinal fluid (CSF PGRN) of patients with lymphoma and carcinoma with and without central nervous system (CNS) metastasis. CSF PGRN levels were significantly higher in patients with lymphoma and carcinoma with CNS involvement (CNS+) compared with those without CNS involvement (CNS−), as well as controls consisting of patients with non-cancer neurological diseases (such as autoimmune neurological disorders [ANDs], functional neurological disorders [FNDs], infectious neurological disorders [INDs], and non-inflammatory neurological disorders [NINDs]). Black dots: outliers

Fig. 3
figure 3

Receiver operator characteristic (ROC) curve analysis of progranulin levels in cerebrospinal fluid (CSF PGRN) of (a) lymphoma with central nervous system (CNS) involvement (CNS+ lymphoma); and (b) carcinoma with CNS involvement (CNS+ carcinoma). ROC curve analyses of CSF PGRN levels could distinguish with high sensitivity and specificity, patients with CNS+ lymphoma from those without CNS involvement (CNS− lymphoma) or non-cancer neurological diseases. Similarly, CSF PGRN levels could distinguish with high sensitivity and specificity, patients with CNS+ carcinoma from those without CNS involvement (CNS− carcinoma) or non-cancer neurological diseases. AUC, area under the curve

Numerous potential biomarkers for CNS malignancies have been reported. However, none are currently in clinical use for monitoring CNS metastasis (Berghoff et al. 2013). Diagnosis of CNS metastasis is usually based on brain magnetic resonance imaging studies and cytological examinations of CSF, but these methods have limited sensitivity and specificity. We therefore proposed that measuring CSF progranulin levels may help screen for CNS metastasis of lymphomas and carcinomas, regardless of pathological diagnosis. In particular, high CSF progranulin level might be a novel indicator for CNS lymphoma. While several potential diagnostic and prognostic markers for CNS lymphoma have previously been reported (Aviles et al. 1991; Hansen et al. 1992; Lee et al. 2005; Roy et al. 2008; Baraniskin et al. 2011; Wei et al. 2015b; Yu et al. 2016; Viaccoz et al. 2015; Strehlow et al. 2016; Rubenstein et al. 2013; Fischer et al. 2009; Ahluwalia et al. 2012), there is presently no reliable biomarker with high sensitivity and specificity for diagnosing CNS lymphoma. For diagnosis CNS lymphoma, it is not uncommon to perform brain biopsies, which are invasive and, in some cases, histologically inconclusive. CSF progranulin can be easily and inexpensively quantified by EIA. Further studies are needed to clarify whether CSF progranulin levels can indeed be used as a diagnostic biomarker of CNS lymphoma.

It is unclear why CSF progranulin levels in patients with CNS metastasis of lymphomas and carcinomas are elevated. Previous immunohistochemical analysis of lymphoid malignancies in patients with diffuse large B cell lymphoma (the most common type of CNS lymphoma) showed progranulin expression in lymphoma cells and in tumor-associated activated macrophage cells (TAMs) surrounding these cells (Yamamoto et al. 2017). We speculate that increased CSF progranulin levels in patients with CNS metastasis of lymphomas and carcinomas is caused by the secretion of progranulin from tumor cells and TAMs in the CNS.

Progranulin as a Biomarker of Autoimmune Neurological Disorders

There is emerging evidence that progranulin may also be associated with various autoimmune diseases, including RA, Sjögren’s syndrome, SLE, and systemic sclerosis (Jian et al. 2018). Progranulin has been shown to have therapeutic effectiveness in inflammatory arthritis by functioning as an endogenous antagonist of TNFα signaling by competitively binding to TNFR (Liu and Bosch 2012; Jian et al. 2013b; Tang et al. 2011). It was also reported that progranulin exerts its anti-inflammatory action through multiple pathways, including induction of regulatory T cell differentiation and IL-10 expression, and by inhibiting chemokine release from macrophages (Jian et al. 2018). Serum progranulin levels are significantly higher in RA patients compared with age-matched healthy controls (Yamamoto et al. 2014). Moreover, circulating progranulin in RA patients is related to TNFα and soluble TNFR2 concentrations, and the progranulin/TNFα ratio correlates with disease stage in RA patients. High progranulin levels are also detected in serum samples from SLE patients (Tanaka et al. 2012), and serum progranulin levels are significantly associated with clinical symptoms and laboratory parameters in SLE, which are in turn related to disease activity. Importantly, serum progranulin levels are significantly decreased after successful treatment of SLE. Collectively, these observations suggest that the measurement of serum progranulin may be a useful approach for monitoring disease activity in patients with RA and SLE.

There are several reports describing the association between progranulin and CNS autoimmune neurological disorders, including multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) (Fenoglio et al. 2010; De Riz et al. 2010; Vercellino et al. 2011, 2016; Kimura et al. 2017). Indeed, progranulin was recently reported to be strongly expressed in the brains of patients with MS, specifically, in macrophages/microglia in active lesions and in activated microglia in normal-appearing white matter (Vercellino et al. 2011). Comparison of progranulin levels in the CSF of MS patients, non-inflammatory controls and inflammatory controls revealed significantly higher progranulin concentrations during MS relapse and in patients with progressive MS compared with MS patients in remission and non-inflammatory controls. This suggests that CSF progranulin levels may be a promising marker for active MS, although one report showed unaltered CSF progranulin levels in MS patients compared with controls (De Riz et al. 2010). Previously, we compared CSF progranulin levels in 17 patients with relapsing-remitting type (RR)-MS and 20 patients with non-inflammatory neurological disorders. CSF progranulin levels were significantly higher in RR-MS patients during relapses compared with non-inflammatory controls (migraine and psychosomatic disorders) (Fig. 4). A recent study found that GRN polymorphisms influence the progression of disability and relapse recovery in MS, which may be related to circulating progranulin levels (Vercellino et al. 2016). It was suggested that the increased progranulin expression by microglia and macrophages in MS brain tissue might play a role in neuronal and axonal protection during brain inflammation.

Fig. 4
figure 4

Progranulin levels in cerebrospinal fluid (CSF PGRN) during the acute phase in 15 neuromyelitis optica spectrum disorder (NMOSD) patients, 17 relapsing-remitting type multiple sclerosis (RR-MS) patients, and 20 non-inflammatory controls (NIC) (4 migraine and 16 psychosomatic disorder). CSF PGRN levels were significantly higher in patients with NMOSD compared with patients with RR-MS and NIC. Similarly, CSF PGRN levels were significantly higher in patients with RR-MS compared with NIC

NMOSD is an inflammatory disorder of the CNS that was previously thought to be a clinical subtype of MS, but more recently has been shown to be a distinct clinical and pathophysiologic entity (Katz 2016). Discovery of a disease-specific serum autoantibody against aquaporin-4 (AQP4), which is a water channel protein abundant in astrocyte foot processes surrounding brain capillaries, increased our understanding of this diverse spectrum of disorders (Lennon et al. 2005). We previously reported that CSF progranulin levels are significantly higher in NMOSD patients compared with RR-MS patients and non-inflammatory controls (Fig. 4) (Kimura et al. 2017). The elevated CSF progranulin levels correlated with CSF IL-6 levels, CSF cell count, CSF protein levels, and were related to total spinal cord lesion length in NMOSD patients. There are several additional reports showing that CSF protein levels, CSF IL-6 levels and total spinal cord lesion length during the acute phase correlate with disease severity in NMOSD patients (Içöz et al. 2010; Jarius et al. 2011; Murchison et al. 2015). Therefore, CSF progranulin levels during the acute phase may reflect NMOSD disease severity. Moreover, CSF progranulin levels during the acute phase also correlate with improvements in expanded disability status scale (EDSS) score, which is a method for quantifying disability in MS and NMOSD patients. These findings suggest that the anti-inflammatory and neurotrophic effects of progranulin may impact recovery from relapse in NMOSD. Therefore, CSF progranulin level is a potential biomarker of disease severity and prognosis in NMOSD.

Conclusion

Progranulin levels are altered in various CNS diseases. Decreased progranulin indicates the presence of GRN mutations, and circulating progranulin is a useful biomarker for the rapid and inexpensive large-scale screening of GRN mutation carriers in FTLD, which may be initially clinically diagnosed as another neurodegenerative disease, such as Alzheimer’s or motor neuron disease. An upregulation of progranulin in the CNS is observed in various malignancies, including glioma, CNS lymphoma, carcinomatous meningitis, and brain metastasis. Hence, CSF progranulin levels could be used as a marker for monitoring CNS metastasis of lymphomas and carcinomas regardless of tumor type, which is often hard to diagnose clinically. Increased CSF progranulin levels are also observed in the acute phase of autoimmune CNS diseases such as MS and NMOSD. In the CNS, progranulin produced by microglia and macrophages might play a role in neuronal and axonal protection during the acute phase, and thereby affect recovery. CSF progranulin level might be a useful indicator of prognosis after relapse in MS and NMOSD. In addition to being a potential biomarker of CNS disease, progranulin may also hold promise as a neurotherapeutic agent.