Loss of function mutations in the progranulin gene are related to pro-inflammatory cytokine dysregulation in frontotemporal lobar degeneration patients

Progranulin (PGRN), also known as granulin-epithelin precursor, is a pleiotropic molecule promoting survival, cell cycle progression, proliferation, and migration of several cell types, and is recognized as an important regulator of the wound-healing response [1, 2]. In the periphery, PGRN is involved in inflammation regulation and, as a whole molecule, it possesses anti-inflammatory effects [3, 4]. Within the brain, PGRN is up-regulated in neuroinflammatory conditions and its increased expression by microglia may be relevant in responses to brain injury, neuroinflammation and neurodegeneration [5, 6]. PGRN is encoded by a single gene on chromosome 17q21 and, interestingly, loss of-function mutations in the PGRN gene have been identified as a monogenic cause of frontotemporal lobar degeneration (FTLD) [7‐9]. Up to now, almost 70 pathogenic PGRN mutations have been described, and all are expected to cause PGRN haploinsufficiency (Alzheimer Disease and Frontotemporal Dementia Database. http://​www.​molgen.​ua.​ac.​be/​FTDmutations/​). Patients with PGRN mutations may present with initial disease symptoms typically in their 50s or 60s, with a wide spectrum of disease onset and clinical presentation [10].

The heterogeneity of PGRN-related FTLD suggests that several other factors of both genetic and environmental nature, including inflammation, can influence disease expression. Indeed, similar to other neurodegenerative disorders, inflammatory mechanisms might participate in the pathophysiology of FTLD, since the disease is characterized by in vivo activation of microglia [11], and increased levels of intrathecal pro-inflammatory mediators [12, 13]. However, no data in FTLD patients establishing a link between PGRN mutations and inflammatory changes have been so far provided, and demonstration of an imbalance in inflammation pathways and establishment of the timing of inflammation abnormalities in PGRN-related disease might be of crucial help in identifying new therapeutic targets for this orphan disorder.

In the present study, we evaluated FTLD patients with and without PGRN haploinsufficiency for circulating levels of cytokines with pro-inflammatory action that are likely relevant to neurodegeneration, epitomized by IL-6, TNF-α and IL-18. In particular, IL-6 and TNF-α, which are measurable in human serum, were chosen here as inflammatory markers because of their implication in the animal model of PGRN deficiency-mediated neurodegeneration [14]. In addition, the cytokine IL-18 was also included in this study because of its pivotal role in neuroinflammation and its association with cognitive decline in demented patients with Alzheimer's disease [15]. Finally, asymptomatic PGRN mutation carriers were evaluated to assess the inflammatory cytokine profile in pre-symptomatic disease stages.

Patients were recruited who met clinical criteria for FTLD [16, 17]. All subjects underwent a somatic and neurological evaluation, routine laboratory tests, a brain structural imaging study, and blood sampling for genetic screening and serum PGRN dosage. The diagnostic assessment was accomplished by a review of full medical history, a semi-structured neurological examination, and a complete standardized neuropsychological status evaluation. Inclusion and exclusion criteria were applied as previously reported [18].

A group of healthy controls (HC) were further included. They were neither related to one another nor to FTLD patients; in particular, we selected only subjects in the same patients' age range, with a MMSE score ≥26 and a formal neurological examination that excluded the presence of a dementing condition.

Moreover, siblings belonging to families of FTLD patients bearing PGRN mutations were considered. Cognitive or behavioral deficits were excluded by a standardized neuropsychological assessment. These were subgrouped into asymptomatic subjects carrying PGRN mutation (ASYMP-PGRN+), to assess inflammatory profiles before PGRN-related FTLD onset, and those without PGRN mutations (ASYMP-PGRN-). A control group similar in age composition was recruited as well.

The study was performed with the understanding and written consent of each subject, and was reviewed and approved by the local ethical committee.

Venous blood samples were drawn from each patient for PGRN sequencing. Total genomic DNA was prepared from peripheral blood according to standard procedures, as previously published [9]. All the 12 exons plus exon 0 of PGRN and at least 30 base pairs (bp) of their flanking introns were sequenced. Serum PGRN was assessed by enzyme-linked immunosorbent assay (ELISA) kit (AdipoGen Inc), according to the manufacturer's instructions. Similarly, IL-6 and TNF-α were measured using specific high-sensitivity ELISA kits (Quantikine HS, R&DSystems), while IL-18 was determined using antibodies (coating: clone 125-2H; detecting: clone 159-12B) and standards from MBL (Nagoya, Japan). The limit of detection for these assays was 0.16 pg/ml for IL-6; 0.5 pg/ml for TNF-α and 15 pg/ml for IL-18. Comparisons among the subject groups on continuous variables were performed using univariate analyses of variance. Correlations between cytokine serum levels and demographic or clinical characteristics were performed by using correlation analyses (Pearson's r) and Fisher's r to z transformation. The level of statistical significance was set at P < 0.05.

Recent studies on PGRN-deficient mice suggest that endogenous PGRN plays a central role in the regulation of inflammatory cytokine production. In particular, PGRN deficiency causes progressive neurodegeneration through enhanced microglia activation and chronic inflammation characterized by increased production of pro-inflammatory molecules, such as IL-6 and TNF-α [14, 19, 20]. Indeed, the pleiotropic pro-inflammatory cytokines IL-6 and TNF-α have been proposed to exert both protective and detrimental effects in CNS, but their neurotoxic capability might prevail in chronic conditions [21, 22]. In accord with animal data, the main results of our study indicate that the pro-inflammatory cytokine IL-6 is significantly elevated in FTLD PGRN+ patients, as a consequence of both the loss of functional PGRN and the development of dementia. In the same direction, a previous study has shown that polymorphisms in the IL-6 gene can modify cognitive and behavioral performances in FTLD patients [23], supporting the potential specific role of this cytokine in influencing symptom occurrence in FTLD. However, in our sample of PGRN-mutated and non-mutated FTLD subjects, there was no correlation between serum levels of IL-6 and MMSE scores, suggesting that the relationship between peripheral cytokine elevation and cognitive impairment might be more complex than the association of cytokine gene variants with the disease.

Different from IL-6, in our study TNF-α tends, without reaching statistical significance, to be higher in FTLD patients without PGRN mutations, as compared to controls. Indeed, another study reported that in FTLD patients versus controls, cerebrospinal fluid levels of TNF-α were significantly increased, without statistically significant changes in circulating TNF-α [13]. Thus, it cannot be excluded that TNF-α might play a role in FTLD, likely independent upon PGRN haploinsufficiency, but this needs to be further investigated. Finally, although emerging data suggest that another pro-inflammatory cytokine, namely IL-18, might be associated with dementia of Alzheimer's type [15], in this study we did not observe any differences in IL-18 serum concentration between FTLD subjects and controls, suggesting that dissimilar inflammatory mechanisms might be implicated in the two different types of dementia.

This work has some limitations that need to be taken into account for discussion purposes. First of all, the number of subjects, namely patients carrying PGRN mutations, is low and the results should be further replicated in larger samples. Moreover, it might be of interest to study an extended panel of cytokines, to study peripheral versus intrathecal cytokine modulation, and to study ex-vivo cellular production of inflammatory mediators and post-mortem analysis of brain tissue to elucidate in depth the inflammatory response pathway changes that underlay PGRN insufficiency. Finally, although IL-6 apparently exerts a predominantly detrimental action in chronic brain disorders [21], the molecular mechanisms determining IL-6 effects on FTLD neurodegenerative pathways remain to be investigated.

Despite these limitations, this study demonstrates for the first time an increased inflammatory response in PGRN-defective humans, suggesting that inflammation might participate in leading to the onset of FTLD symptoms, and highlights possible new molecular targets for PGRN-related FTLD treatment.