Idiopathic normal pressure hydrocephalus (iNPH) is treatable but can cause dementia, gait disturbance, and urinary incontinence in older patients with ventriculomegaly. The diagnosis of iNPH depends on successful cerebrospinal fluid (CSF) shunt surgery (Ishikawa et al., 2008; Mori et al., 2012; Nakajima et al., 2021). In addition to ventriculomegaly, brain imaging shows intraparenchymal white matter lesions, leukoaraiosis, and sometimes microbleeds, which may compromise brain function and the response to shunt therapy. Both leukoaraiosis and microbleeds are caused by impaired drainage of interstitial fluid and molecular clearance related to glymphatic (glia-lymphatic) system malfunction (Iliff et al., 2013; Weller et al., 2015; Simon and Iliff, 2016). Impaired glymphatic fluid transport resulting in the accumulation of intracellular waste products such as amyloid-β and tau aggregation is referred to as glymphopathy and may be involved in multiple pathological processes, including normal aging, Alzheimer’s disease, and iNPH (Benveniste and Nedergaard, 2022). We previously showed that the amount of lipocalin-type prostaglandin D synthase (L-PGDS) in CSF closely correlates with age-related brain white matter changes scale (Wahlund et al., 2001) detected via magnetic resonance imaging (MRI) of iNPH (Nishida et al., 2014).

The core symptoms of iNPH are non-specific. Thus, we considered diagnosing older adults with symptomatic ventriculomegaly using CSF tap tests (Ishikawa et al., 2010). Various biomarkers, including L-PGDS, were detected in CSF; however, the roles of these biomarkers in iNPH are unclear (Mase et al., 2003; Brettschneider et al., 2004). L-PGDS is a lipocalin family protein with multiple functions (Urade, 2021). In the central nervous system, L-PGDS is primarily synthesized in leptomeningeal cells and oligodendrocytes in rodents and humans and is secreted into the CSF as β-trace—a major human CSF protein (Chérasse et al., 2018; Urade, 2021). The L-PGDS expression is upregulated in oligodendrocytes in murine models of lysosomal disease (Mohri et al., 2006) and in the demyelinating focus of patients with multiple sclerosis (Kagitani-Shimono et al., 2006). L-PGDS promotes myelination by exocytosis from oligodendrocyte-lineage cells (Pan et al., 2023). In the context of neurodegeneration, L-PGDS acts as a chaperone, binding to amyloid β peptides to prevent fibril formation (Kanekiyo et al., 2007; Kannaian et al., 2019). We demonstrated that L-PGDS levels closely correlated with tau levels rather than amyloid β levels in CSF from patients with iNPH (Nishida et al., 2014).

The MRI diffusion tensor imaging (DTI) is frequently utilized to estimate white matter pathologies in vivo (Song et al., 2002; Alexander et al., 2007; Shimoji et al., 2013). Four parameters, including fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD), represent different aspects of white matter integrity and pathology, including axonal degeneration, and demyelination (Winklewski et al., 2018). The brain morphology of aged or hydrocephalic patients is not suitable for quantitative MRI analysis but methodological progress, including tract-based spatial statistics (TBSS), has been used to accomplish this task (Cader et al., 2007).

Thus, previous studies on normal pressure hydrocephalus reveal the close relationship between tau, L-PGDS, and MRI white matter damage scores. The recognized chaperone function of L-PGDS to amyloid-β—another key molecule in glymphopathy—indicates the relationship between tau and L-PGDS. We objectively assessed white matter damage by DTI analysis and identified relevant factors. Moreover, we confirmed whether tau directly interacts with L-PGDS. In this study, we analyzed the clinical, radiological, and CSF profiles of 60 consecutive patients who were referred to our institute with probable iNPH. We used TBSS parameters instead of age-related brain white matter changes to confirm the significant correlation of white matter integrity with L-PGDS levels in the CSF. Moreover, we demonstrate that L-PGDS binds to tau protein. Finally, we confirmed the colocalization of L-PGDS and tau protein in postmortem brain specimens. Based on our results, we hypothesize that L-PGDS works as a buffer for brain homeostasis by maintaining glymphatic drainage.

We investigated the relationship between clinico-radiological factors and CSF profiles of iNPH patients using DTI analysis, with particular attention to parenchymal white matter integrity in vivo and the interaction of tau and L-PGDS ex vivo. We focused on L-PGDS due to its chaperone activity (including amyloid β), which may play a role in glymphopathy in iNPH. Our results highlight two main findings. First, we confirmed the significant correlation of several DTI parameters with clinico-demographical parameters and CSF L-PGDS levels. The main CSF trends were as follows: increased L-PGDS negatively correlated with overall white matter integrity (FA) and positively correlated with markers of demyelination (MD and RD). Most of the affected white matter tracts are long traveling association and commissural fibers. Second, we demonstrated that L-PGDS binds to tau protein, suggesting that L-PGDS functions as a tau chaperone in a similar fashion to how it functions with amyloid β (Kanekiyo et al., 2007; Kannaian et al., 2019). We also confirmed the colocalization of L-PGDS with tau in brain sections from patients with Alzheimer’s disease, a representative glymphopathy.

Many differential diagnoses exist for iNPH due to the non-specific symptoms, including gait, urination, and cognition frequently in aging. Tight high convexity and disproportionately enlarged subarachnoid-space hydrocephalus (DESH) are important features of ventricular enlargement distinguishing cerebral atrophy from impaired CSF circulation. The mechanism by which impaired CSF circulation leads to DESH morphology is still unknown, although a check valve phenomenon around the choroidal fissure has been proposed (Yamada et al., 2016). CSF shunting will reduce this tightness in high convexity. Of note, DESH indicates the contours of changes in the brain, but the acceptable extent of brain parenchymal white matter impairment is unclear. A mixture of axonal degeneration and demyelination are among the changes in the brain parenchyma of normal pressure hydrocephalus (Tullberg et al., 2002). FA clusters correlating with L-PGDS may reflect demyelination indicated by MD/RD rather than axonal degeneration indicated by AD. L-PGDS is secreted by oligodendrocytes and involved in demyelinating diseases and myelination (Kagitani-Shimono et al., 2006; Chérasse et al., 2018; Pan et al., 2023). Notably, L-PGDS concentration in CSF correlates with changes suggesting demyelination in TBSS. Indeed, shunt surgical intervention and subsequent recovery of brain parenchymal structure and function has several unexplored aspects. The reversibility of brain dysfunction by shunting should reflect the degree of brain parenchymal tissue damage, but cases with extensive white matter damage and microbleeds should be considered carefully, even if the brain contour is DESH. Increased microbleeds correlate with ventricular enlargement in Alzheimer’s disease (Kuroda et al., 2020). As for DESH brain, parenchymal atrophy seems to advance over the years, gradually losing its characteristic sulcal narrowness while symptoms progress (Toma et al., 2011; Jaraj et al., 2017; Isaacs and Hamilton, 2021).

Glymphopathy caused by malfunction of waste clearance in the central nervous system occurs in iNPH (Ringstad et al., 2017; Benveniste et al., 2019; Benveniste and Nedergaard, 2022). A well-known disease model of glymphopathy is Alzheimer’s disease, one of the major comorbidities for iNPH. Distinguishing between the two diseases is difficult, especially in the advanced stages. Although correlations with DTI parameters in both diseases have been reported, white matter fibers that correlate with CSF amyloid β and tau have been detected for Alzheimer’s disease but not for iNPH (Hattingen et al., 2010; Koyama et al., 2012; Kamiya et al., 2016; Alm and Bakker, 2019). These results are similar to the results from our study; we did not detect any white matter tract correlations with CSF biomarkers related to neurodegeneration, such as amyloid β and tau, in patients with iNPH. However, we did find fibers that correlated with L-PGDS, suggesting that this molecule plays a significant role in the disease.

In neurodegeneration, amyloid β fibril formation is prevented by the chaperone activity of L-PGDS (Kanekiyo et al., 2007; Kannaian et al., 2019). However, another important biomarker of neurodegenerative disease, tau, has not been studied. Tau is a hydrophilic protein but also forms neurotoxic aggregates in Alzheimer’s disease and in several tauopathies (Mandelkow and Mandelkow, 2012). We previously showed that tau levels were closely linked with L-PGDS levels in the CSF of patients with iNPH (Nishida et al., 2014). Tau carries many charges and can interact with many partners, yet its interaction with L-PGDS has not been clarified. Accordingly, we determined the interaction between the tau and L-PGDS ex vivo and found that the interactions were similar to those observed with amyloid β. Because L-PGDS can adsorb hydrophobic small molecules owing to the lipocalin capacity, tau is also adsorbed intracellularly. During a study involving the staining of monkey kidneys, fluorescent signals from labeled L-PGDS and lysosomes were merged (Nagata et al., 2009), suggesting that L-PGDS adsorbs tau, which is then degraded in lysosomes. Potentially, L-PGDS removes tau from the cell by adsorbing this for lysosomal degradation. Thus, L-PGDS may play a role in waste clearance pathways, such as the glymphatic system.

The CSF biomarker tau may be involved in the progression of iNPH (Kudo et al., 2000). Tau, a microtubule-associated protein, is primarily located in the axons of neuronal cells that promotes and stabilizes microtubule assembly (Weingarten et al., 1975; Mandelkow and Mandelkow, 2012). Increased tau levels in CSF correlate with the severity of neuronal damage and loss (Blennow et al., 1995). Tau increases in the CSF with age and the severity of clinical symptoms in iNPH (Kudo et al., 2000; Nishida et al., 2014). In contrast, tau levels tend to be low in patients with good cognitive recovery following shunt surgery (Kudo et al., 2000; Miyajima et al., 2013).

Decreased L-PGDS may be due to the arachnoidopathy (loss of arachnoid cells producing L-PGDS) secondary to NPH after subarachnoid hemorrhage (Mase et al., 2003; Brettschneider et al., 2004). Decreased L-PGDS correlated with a narrow callosal angle, which is a feature of uneven CSF distribution in DESH-type iNPH (Nishida et al., 2014). Paradoxically, patients with low L-PGDS concomitant with low tau levels exhibited high cognitive function despite severe arachnoidopathy. The role of arachnoidopathy in the development of DESH-type iNPH is unclear. One hypothesis is that L-PGDS bound to tau remains within the brain parenchyma due to stagnated glymphatic flow while the canopy is narrow in the prodromal to early phases of DESH in iNPH. After disease progression and brain atrophy progression, L-PGDS is released into the CSF and can be detected by lumber puncture.

Another discrepancy is the difference in location, with immunohistochemical studies showing colocalization of L-PGDS and tau usually in the gray matter, whereas MRI evaluation captures white matter damage partially due to the methodology of the MRI studies. However, this can be interpreted as immunohistochemical studies capturing the adsorption of pathological tau by L-PGDS in gray matter cells. Considerably, both substances increase in the CSF as the cellular degeneration progresses, causing progressive white matter damage on MRI and producing more severe symptoms.

Our study has several limitations. We did the histopathological study on brain specimens with Alzheimer’s disease instead of iNPH, due to difficulties in obtaining postmortem brain tissue from patients with iNPH. We demonstrated the interaction between L-PGDS and tau by SPR analysis, but the regulation of tau by L-PGDS is hard to demonstrate in vitro. Moreover, we included both definite iNPH and probable iNPH patients with various stages of disease for our in vivo analysis.

In conclusion, our findings support a role for L-PGDS in iNPH pathogenesis or progression, based on the concept of glymphopathy. The detailed role of L-PGDS is still unclear, but one possibility is that the chaperone activity of L-PGDS inhibits tau aggregation. Although iNPH is still an obscure heterogeneous disease concept, we hope our findings clarify the cause of this disease and aging brain homeostasis in general.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author/s.

The studies involving humans were approved by Institutional review boards of Kitano Hospital (P151100704) and the University of Tokyo (approval number: 17-256). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

NNi: Writing–original draft, Writing – review & editing. NNa: Writing–original draft, Writing – review & editing. KS: Methodology, Writing – review & editing, Writing–original draft. NJ: Data curation, Writing – review & editing, Writing – original draft. KU: Methodology, Writing – review & editing, Writing – original draft. AO: Methodology, Writing – review & editing, Writing – original draft. MT: Methodology, Writing – review & editing, Writing – original draft. YU: Conceptualization, Writing – review & editing, Writing – original draft. SM: Validation, Writing – review & editing, Writing – original draft. KI: Validation, Writing – review & editing, Writing – original draft. RO: Data curation, Writing – review & editing, Writing – original draft. MI: Supervision, Validation, Writing – review & editing, Writing – original draft. HT: Supervision, Validation, Writing – review & editing, Writing – original draft.