Large-scale CSF proteome profiling identifies biomarkers for accurate diagnosis of frontotemporal dementia

Yanaika S Hok-A-Hin; Lisa Vermunt; Carel FW Peeters; Emma L van der Ende; Sterre CM de Boer; Lieke H Meeter; Julie de Houwer; Harro Seelaar; John C van Swieten; William T Hu; Alberto Lleó; Daniel Alcolea; Sebastiaan Engelborghs; Anne Sieben; Alice Chen-Plotkin; David J Irwin; Wiesje M van der Flier; Yolande AL Pijnenburg; Charlotte E Teunissen; Marta del Campo

doi:10.1186/s13024-025-00882-5

. 2025 Aug 27;20:93. doi: 10.1186/s13024-025-00882-5

Large-scale CSF proteome profiling identifies biomarkers for accurate diagnosis of frontotemporal dementia

Yanaika S Hok-A-Hin ^1,^✉, Lisa Vermunt ^1,², Carel FW Peeters ³, Emma L van der Ende ¹, Sterre CM de Boer ^2,⁴, Lieke H Meeter ⁵, Julie de Houwer ⁵, Harro Seelaar ⁵, John C van Swieten ⁵, William T Hu ⁶, Alberto Lleó ^7,⁸, Daniel Alcolea ^7,⁸, Sebastiaan Engelborghs ^9,^10,¹¹, Anne Sieben ¹², Alice Chen-Plotkin ¹³, David J Irwin ¹³, Wiesje M van der Flier ², Yolande AL Pijnenburg ², Charlotte E Teunissen ^1,^#, Marta del Campo ^1,^14,^15,^#

PMCID: PMC12392567 PMID: 40866991

Abstract

Background

Diagnosis of Frontotemporal dementia (FTD) and its specific underlying neuropathologies (frontotemporal lobar degeneration; FTLD-Tau and FTLD-TDP) are challenging, and thus, fluid biomarkers are needed to improve diagnostic accuracy.

Methods

We used proximity extension assays to analyze 665 proteins in cerebrospinal fluid (CSF) samples from a multicenter cohort, which included patients with FTD (n = 189), Alzheimer’s Disease dementia (AD; n = 232), and cognitively unimpaired individuals (n = 196). In a subset, FTLD neuropathology was determined based on phenotype or genotype (FTLD-Tau = 87 and FTLD-TDP = 67). Differences in protein expression profiles were analyzed using nested linear models. Penalized generalized linear modeling was used to identify classification protein panels, which were translated to custom multiplex assays and validated in two clinical cohorts (cohort 1: n = 161; cohort 2: n = 162), one autopsy-confirmed cohort (n = 100), and one genetic cohort (n = 55).

Results

Forty-three proteins were differentially regulated in FTD compared to controls and AD, reflecting axon development, regulation of synapse assembly, and cell-cell adhesion mediator activity pathways. Classification analysis identified a 14- and 13-CSF protein panel that discriminated FTD from controls (FTD diagnostic panel, AUC: 0.96) or AD (FTD differential diagnostic panel, AUC: 0.91). Custom multiplex panels confirmed the strong discriminative performancen between FTD and controls (AUCs > 0.96) and between FTD and AD (AUCs > 0.88) across three validation cohorts, including one with autopsy confirmation (AUCs > 0.90). Validation in genetic FTD (including C9orf72, GRN, and MAPT mutation carriers) revealed high accuracy of the FTD diagnostic panel in identifying both the presymptomatic (AUCs > 0.95) and symptomatic (AUC: 1) stages. Six proteins were differentially regulated between FTLD-TDP and FTLD-Tau. However, a reproducible classification model could not be generated (AUC: 0.80).

Conclusions

Overall, this study introduces novel FTD-specific biomarker panels with potential use in diagnostic settings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13024-025-00882-5.

Keywords: FTD, FTLD, Tau, TDP43, Biomarkers, Proteomics, CSF

Background

Frontotemporal dementia (FTD) is the second most common form of young-onset dementia (i.e., defined as dementia with symptom onset before the age of 65). FTD presents heterogeneously, comprising various clinical, neuropathological, and genetic forms. From a clinical perspective, FTD can present with either behavioral and social changes (behavioral variant FTD, bvFTD), language impairment (semantic variant primary progressive aphasia, svPPA; and non-fluent variant PPA, nfvPPA), or motor dysfunction (corticobasal syndrome, CBS; and progressive supranuclear palsy, PSP) [1–4]. FTD is caused by different underlying neuropathologies leading to frontal temporal lobar degeneration (FTLD). Approximately 50% of FTLD cases develop aggregates of the microtubule Tau protein (FTLD-Tau), while 45% are characterized by cytoplasmic inclusion of the TDP-43 protein (FTLD-TDP). A small percentage of FTLD cases (5%) develop aggregates of the FUS protein (FTLD-FET) [5–7]. These pathologies can be predicted in familial forms of FTLD, where autosomal dominant mutations in the microtubule-associated protein tau (MAPT) gene lead to FTLD-Tau pathology, while mutations in the progranulin (GRN) or chromosome 9 open reading frame 72 (C9orf72) genes are associated with FTLD-TDP pathology [5, 8]. However, genetic FTD accounts for only 10–20% of FTLD cases, and in many families, the mutation status remains unknown, and novel genetic variants continue to be identified [9–12]. In sporadic cases, which are more common, the different neuropathological subtypes poorly correlate with the clinical presentation, with the exception of FTD accompanied with motor neuron disease and svPPA that shows a clinicopathological correlation with TDP-43 [13]. Thus, the different clinical phenotypes can have overlapping pathologies despite diverse genetic backgrounds, stressing the complexity of the disease and the importance of developing pathology-specific biomarkers, which are needed for targeted treatment.

Currently, there are no fluid biomarkers that specifically capture underlying FTD pathogenesis [14]. Given the overlapping clinical features with Alzheimer’s disease (AD) [15, 16], the analysis of the core AD-CSF biomarkers (i.e., amyloid-beta 1–40 and 1–42, Aβ1–40 and Aβ1–42 or the Aβ1–42/Aβ1–40 ratio; phosphorylated Tau181, pTau181; and total Tau, tTau) is often used to exclude AD diagnosis [17]. The concentrations of Neurofilament light (NfL) in CSF and blood are strongly increased in FTD but also in other neurodegenerative disorders since it reflects neuroaxonal damage [18, 19]. Nevertheless, this biomarker is considered valuable for distinguishing FTD from non-neurodegenerative disorders, such as primary psychiatric diseases, and it also shows good prognostic potential [20, 21]. Several studies have shown that the CSF pTau/tTau ratio could demarcate the main FTLD subtypes but with limited diagnostic accuracy [22, 23]. However, promising findings have been recently obtained with plasma GFAP/NfL ratio (FTLD-Tau vs. FTLD-TDP; AUC: 0.89), which requires additional validation in external cohorts [24]. Overall, there is an unmet need for fluid biomarkers specifically identifying FTD and its underlying pathologies, which is essential for accurate diagnosis, clinical trial inclusion, and to monitor the effects of treatments [25].

CSF proteomics offer the opportunity to study brain changes and associated processes in vivo. As extensively explored in AD, such analysis could offer insights into the disease’s etiology and reveal novel biomarker candidates [26–29]. Previous FTD biomarker discovery studies using unbiased mass spectrometry (MS)-based technologies identified multiple biomarker candidates (e.g., YKL-40 in FTD and NPTXR in GRN mutation carriers) [30, 31]. However, subsequent validation efforts did not show sufficient diagnostic accuracy of these markers [32–35]. The clinical and pathological heterogeneity of FTD and low sample sizes could explain the lack of validated biomarkers [30, 31]. In addition, the translation of novel FTD biomarkers could be hampered due to the use of different technologies between the discovery (e.g., MS) and validation (immunoassay-based methods) phases. Therefore, we followed a similar strategy to that previously applied for AD and dementia with Lewy bodies (DLB) [29, 36]. We here used the high-throughput immune-based proximity extension assay (PEA) to analyze the CSF proteome of an extensive and well-characterized cohort of patients with FTD presenting at memory clinics. We first aimed to identify CSF protein changes specifically associated with FTD and its main pathological subtypes (i.e., FTLD-Tau and FTLD-TDP). Next, we aimed to translate these findings into clinically feasible CSF protein panels to discriminate FTD and its pathological subtypes. Finally, we validated these panels in four independent cohorts, including an autopsy-confirmed and a genetic cohort.

Method

Participants

As part of our previous work [29], the discovery cohort (total = 617; Table 1) included CSF samples from patients diagnosed with FTD (n = 189), AD dementia (n = 232), and cognitively normal individuals (CON, n = 196). Most of the samples were selected from the Amsterdam Dementia Cohort (ADC; 120 FTD, 214 AD, 190 CON) [37]. To enrich for samples from patients with confirmed FTD, additional cases from the Center for Neurodegenerative Disease Research at the University of Pennsylvania (Penn; 46 FTD, 18 AD, 6 CON), Erasmus Medical Center (19 FTD), and the Goizueta Alzheimer’s Disease Research Center at Emory University (4 FTD) were selected. For a subset of the FTD patients (82%), the underlying neuropathology was known or could be predicted based on specific clinical diagnosis (FTLD-Tau = 87, FTLD-TDP = 67). FTLD-Tau was confirmed based on autopsy (n = 17), MAPT mutation (n = 16), and further enriched with patients clinically diagnosed with PSP (n = 31) and CBS (n = 23), which primarily associates with Tau neuropathology [38]. The FTLD-TDP group included autopsy-confirmed cases (n = 25) and patients with GRN (n = 8) or C9orf72 (n = 23) mutations, and further enriched with patients clinically diagnosed with svPPA (n = 11), which have a high likelihood of having TDP pathology [13]. Four additional validation cohorts were included for validation of the custom panels: two clinical cohorts, from the ADC (51 FTD, 55 AD, and 55 controls) and Sant Pau Initiative on Neurodegeneration (SPIN; 54 FTD, 53 AD, and 55 controls) [39], one FTLD/AD autopsy confirmed cohort from the Reference Center for Biological Markers of Dementia (BIODEM) and the neurobiobank of the Institute Born-Bunge (IBB) from the University of Antwerp (UAntwerp) (BIODEM-UAntwerp; 41 autopsy-FTLD; aFTLD, including 7 aFTLD-Tau and 31 aFTLD-TDP, 3 aFTLD-UPS, and 30 autopsy-AD; aAD). Noteworthy, 29 cognitively unimpaired controls from BIODEM-UAntwerp were included in this cohort, though not autopsy-confirmed. Lastly, individuals from the Frontotemporal Dementia Risk Cohort (FTD RisC) at Erasmus Medical Center were included. This cohort comprises patients with FTD caused by autosomal dominant mutations (14 symptomatic mutation carriers, including 10 C9orf72, 3 GRN, and 1 MAPT). Additionally, their at-risk relatives were also included, consisting of 8 non-carriers and 33 presymptomatic mutation carriers (22 C9orf72, 7 GRN, and 4 MAPT).

Table 1.

Cohort characteristics

	Discovery Cohort			Validation Cohort 1			Validation Cohort 2			FTLD/AD neuropathological
	Control	FTD	AD	Control	FTD	AD	Control	FTD	AD	Control	aFTLD	aAD
n	196	189	232	55	51	55	55	54	53	29	41	30
FTD Clinical Phenotype		35 bvFTD			38 bvFTD			31 bvFTD			na
		23 CBS			0 CBS			18 CBS-PSP*			na
		31 PSP			0 PSP			1 lvPPA			na
		11 svPPA			5 svPPA			2 svPPA			na
		0 nfvPPA			2 nfvPPA			2 nfvPPA			na
		42 Autopsy-confirmed			0 Autopsy-confirmed			0 Autopsy-confirmed			41 Autopsy-confirmed
		47 Mutation Carriers			6 Mutation Carrier			0 Mutation Carriers			na
Sex, Female n (%)	72 (36.7)	93 (49)	96 (41)	22 (40)	17 (33)	23 (42)	32 (58)	21 (38)	34 (64)	13 (45)	20 (49)	15 (50)
Age	58 (53 - 63)^a,b	65 (59 - 70)^c	66 (59 - 72)^c	58 (56 - 62)^a,b	63 (59 - 69)^c	66 (62 - 70)^c	61 (58 - 67)^a,b	76 (71 - 79)^c	74 (68 - 76)^c	64 (61 - 66)^b	64 (55 - 71)^b	72 (64 - 76)^a,c
MMSE	28 (27 - 29)^a,b	25 (22 - 27)^b,c	21 (17 - 24)^a,c	29 (27 - 30)^a,b	25 (23 - 27)^b,c	20 (18 - 23)^a,c	29 (29 - 30)^a,b	21 (18 - 25)^c	22 (19 - 25)^c	29.5 (29 - 30)	20 (14 - 26)	14 (7 - 21)
CSF Aβ42, pg/mL	1120 (1023 - 1241)^a,b	985 (782 - 1193)^b,c	602 (535 - 657)^a,c	1283 (1085 - 1373)^a,b/ 1629 (1372 - 1700)^b	1086 (888 - 1253)^b,c/ 1336 (1038 - 1634)^b	566 (525 - 661)^a,c/ 489 (468 - 576)^a,c	1194 (899 - 1397)^a,b	871 (591 - 1268)^b,c	537 (441 - 646)^a,c	1108 (952 - 1216)^a,b	681 (494 - 866)^b,c	404 (306 - 497)^a,c
CSF pTau, pg/mL	38 (31 - 45)^a,b	44 (36 - 57)^b,c	92 (77 - 112)^a,c	54 (39 - 66)/ 15 (12 - 19)^b	46 (39 - 61)^b/ 17 (12 - 24)^b	65 (50 - 92^)a/ 35 (31 - 37)^a	38 (28 - 47)^b	41 (32 - 49)^b	117 (81 - 165)^a,c	47 (39 - 56)^b	39 (29 - 49)^b	72 (49 - 84)^a,c
CSF tTau, pg/mL	213 (167 - 257)^a,b	324 (239 - 414)^b,c	751 (581 - 1005)^a,c	292 (213 - 443)^b/ 182 (150 - 230)	362 (254 - 487)^b/ 230 (174 - 290)	479 (343 - 684)^a/ 336 (327 - 346)	257 (198 - 314)^b	324 (224 - 434)^b	740 (546 - 976)^a,c	203 (168 - 257)^a,b	330 (222 - 441)^b,c	535 (325 - 674)^a,c
APOE4 carrier, n (%)	48 (25)	46 (24)	132 (57)	15 (27)	13 (26)	39 (71)	14 (26)	8 (15)	22 (41.5)	na	na	na
FTLD-Subtypes		FTLD-Tau (87 total):			na			na			aFTLD-Tau (7 total)
		17 Autopsy-confirmed									aFTLD-TDP (31 total)
		16 MAPT									aFTLD-UPS (3 total)
		31 PSP
		23 CBS
		FTLD-TDP (67 total):
		25 Autopsy-confirmed
		8 GRN
		23 C9orf72
		11 svPPA

Open in a new tab

Continuous variables are presented as median ± interquartile range and dichotomous data as the number of cases with a percentage of the total (%)

Differences between groups were determined using the Kruskal-Wallis test with Bonferroni correction or Chi-squared test

Analysis of covariance was performed for CSF biomarker analysis, adjusting for age and sex when appropriate

In the discovery cohort, biomarker values from the luminex were transformed to innotest values using passingbablok regression formulas (method)

For validation cohort 1, the first set of biomarker values corresponds to data obtained using Innotest, while the second set corresponds to data obtained using Elecsys

CSF Biomarker data from validation cohort 2 were obtained with Lumipulse

a, p< 0.05 compared to FT(L)D

b, p< 0.05 compared to AD

c, p< 0.05 compared to CON

* 18 Patients were on the spectrum of CBS and PSP, with 11 showing a clinical profile more consistent with CBS, and 4 exhibiting features more indicative of PSP

Abbreviations: FTD = frontotemporal dementia, AD = Alzheimer’s Disease, MMSE = mini-mental state examination, CSF = cerebrospinal fluid,

FTLD = frontotemporal lobar degeneration, na = not applicable, TDP = Transactive response DNA binding protein of 43, FUS = fused in sarcoma, aFTLD = autopsy confirmed FTLD, aAD = autopsy confirmed AD

All participants underwent standard neurological screening and cognitive testing. FTD and AD diagnoses were made according to international consensus criteria [1–4, 40]. The autopsy-confirmed cohort included cases with a definite diagnosis according to international neuropathological examination guidelines for FTLD [41] and AD [42]. The genetic FTD cohort included participants identified as symptomatic based on international guidelines [1, 2]. Symptomatic onset was defined as the point at which caregivers first reported the presence of symptoms. Mini-mental state examination (MMSE) was used as a measurement for global cognition in all groups. In addition, for a subset of FTD cases from the ADC (discovery cohort, n = 62; clinical cohort 1, n = 44), the FTLD CDR^® plus NACC score was used as a clinical measure specific for FTD disease severity [43]. The control group included individuals with subjective cognitive decline, who scored normally on cognitive examinations with a negative AD CSF biomarkers profile.

CSF was collected and biobanked according to established protocols [44]. Concentrations of CSF Aβ42, pTau, and tTau were used to support AD diagnosis and to evaluate the presence of AD co-pathology in the FTD cases. These were analyzed locally as part of the diagnostic work-up using commercially available kits (ADC, Erasmus MC, and BIODEM-UAntwerp: ELISA Innotest Aβ(1-42), hTAUAg, pTau (181P; Fujirebio, Ghent), or ADC: Elecsys Aβ42, tTau and pTau (181P) CSF assays (Roche Diagnostics); Penn and Emory: Luminex xMAP INNO-BIA AlzBio3 (Luminex, Bio-techne); SPIN: Lumipulse G600, Fujirebio). Predefined cut-offs were used to define a positive AD biomarker profile. (ADC: tTau/Aβ42 ratio > 0.46; Penn: tTau/Aβ42 ratio > 0.30; Emory: Aβ42/tTau ratio < 6; SPIN: Aβ42/Aβ40 ratio < 0.062, tTau > 456 pg/mL and pTau > 63 pg/mL). In the ADC, Innotest Aβ42 concentrations were adjusted for drift over time [45]. CSF biomarkers measured on Luminex were transformed to Innotest values using conversion formulas based on Passing-Bablok regression analysis using cases for which both Luminex and Innotest values were available, as described previously [29]. CSF NfL was measured in a subset of cases from the ADC (discovery cohort: 42 CON, 74 FTD, and 54 AD; clinical cohort 1: 8 CON, 19 FTD, and 20 AD) either by NF-light^® ELISA (Uman Diagnostics, Sweden) or with the single molecular array (Simoa^®) NF-light™ advantage kit (Quanterix, USA) [46]. NfL concentrations measured by ELISA were converted to Simoa values using Passing-Bablok regression analysis, as described previously [46].

CSF protein profiling

979 CSF proteins were quantified using 11 Olink Target 96 validated multiplex panels based on proximity extension assay (PEA) technology (Olink Proteomics, Uppsala, Sweden) that were available at the time of analysis (Cardiometabolic, Cardiovascular II and III, cell regulation, development, immune response, inflammation, metabolism, neurology, oncology II and organ damage). CSF samples were randomized across multiple plates containing intra- and inter-plate quality controls (QCs) from the manufacturer and measured in two different rounds. Each round of measurement contained 16 bridging samples covering different clinical groups, which were used as a reference to account for potential batch effects. For each protein, the lower limit of detection (LOD) was determined by the company and defined as three standard deviations above the background from the negative controls included on every plate. Proteins were excluded from further analysis if their levels were below the LOD in more than 15% of the samples. A total of 665 proteins (642 unique proteins) were ultimately included for statistical analysis of the discovery cohort, similar to our previous study [29]. Protein abundance was reported in normalized protein expression (NPX) values.

Development of custom PEA assays

Multiplex-PEA assays were custom-developed by the manufacturer following standardized protocols [47]. We developed assays to measure 18 out of 24 proteins selected by the classification analysis described below. CSF samples from validation cohorts were randomized across plates. Each plate additionally included: four CSF QC samples, a negative control, and three calibrators used for normalization. Each custom assay had a LOD determined by the company, defined as three standard deviations above the background from the negative controls. Precision (intra- and inter-assay CV) was calculated using the four CSF QC samples (Supplementary Table 1). No cross-reactivity between assays for the specific proteins was detected. Samples from the validation cohorts were randomized across plates and normalized for any plate effects using the built-in inter-plate controls according to the manufacturer’s recommendations. Protein levels are reported in NPX values.

Statistical analyses

All processing and statistical analysis were performed in R version 4.2.1. Baseline demographics were tested by Kruskal-Wallis test followed by Bonferroni post-hoc, or Peasons’s chi-square test for continuous and categorical variables, respectively. Differences in protein abundance of the CSF proteome data was tested with nested linear models including age and sex in the model for the comparison between different groups (FTD vs. controls, FTD vs. AD, FTLD-Tau vs. controls, FTLD-TDP vs. controls, and FTLD-Tau vs. FTLD-TDP) as previously performed [29, 36]. For each pairwise comparison, multiplicity was taken into account by controlling the False Discovery Rate (FDR) at q ≤ 0.05 based on the number of features analyzed [48].

We next evaluated which CSF protein combination (CSF panels) could best discriminate the groups of interest while keeping the number of markers to a minimum so that they can be ultimately translated into small, practical custom panels, as described previously [29]. For this purpose, binary classification signatures (FTD vs. controls: FTD Diagnostic Panel, FTD vs. AD: FTD Differential Diagnostic Panel, and FTLD-Tau vs. FTLD-TDP: FTLD Subtype Panel) were constructed by way of penalized generalized linear modeling (GLM) with an elastic net penalty (a linear combination of lasso and ridge penalties) in the discovery cohort using the glmnet package, including age and sex as covariates in the model [29, 49, 50]. This penalty enables estimation in settings where the feature-to-sample ratio is too high for standard generalized linear regression. Moreover, it performs automatic feature decorrelation as well as feature-selection. For each classification exercise, we compare multiple models that reflect (a) a grid of values for the elastic-net mixing parameter, reflecting strong decorrelation to a pure logistic lasso regression, and (b) a grid of values reflecting the maximum number of proteins that may be selected under each model (21 markers maximum). The former grid (a) considers that we have little information on the collinearity burden in the data. The latter grid (b) considers that we want to keep the number of selected proteins relatively low for the future development of customized panels. The optimal penalty parameters in the penalized models were determined based on (balanced) 10-fold cross-validation of the model likelihood [29, 49]. The cross-validation was performed with balanced folds, by which each fold has an outcome group ratio close to the corresponding ratio in the full data set, also referred to as stratified cross-validation. The predictive performance of all models was assessed by way of (the comparison of) Receiver Operating Characteristic (ROC) curves and Area Under the ROC Curves (AUCs). The model with the highest AUC and lowest number of markers for each classification signature was selected. The fold-based selection proportions for each marker were assessed to identify and select the most promising markers within each model (i.e., features that are stably selected across each individual fold, thereby minimizing potential overfitting). To reflect the manual selection pressure for these final marker sets, each final logistic signature was subjected to a ridge-regularization with a penalty parameter of 0.1. The performance (AUC) was evaluated by internal validation: repeated 5-fold cross-validation with 1000 repeats. The 95% confidence interval around the resulting AUCs was based on resampling quantiles (percentile method). External validations assessed the performance of the final models with the markers of interest in the validation cohorts. We additionally compared our identified classification models to CSF NfL to discriminate FTD from controls and AD in the subset of cases for which this information was available.

Functional enrichment analysis was performed for all comparisons described above using Metascape, selecting GO biological Processes as an ontology source [51]. The total number of CSF proteins optimally analyzed was included as the enrichment background (n = 642). Default parameters were used for the analysis in which terms with a p-value < 0.01, a minimum count of 3, and an enrichment factor > 1.5 were collected and grouped into clusters based on their membership similarities.

Partial non-parametric correlation analysis was performed to understand the association among proteins within the CSF panels with cognitive function (MMSE score) and disease severity (FTLD-CDR^® plus NACC scores) [43]. This analysis was corrected for age, sex, and the clinician who performed the FTLD-CDR examination (included as a dummy variable), and was conducted both in the total discovery cohort and specifically within the FTD group.

Results

Subjects

An overview of the study is presented in Fig. 1A. The discovery cohort included a total of 617 samples, all measured as part of previous CSF proteomic studies [29, 36]. Custom multiplex panels were developed and validated in two independent clinical cohorts (validation cohort 1: n = 161; validation cohort 2: n = 162), one FTLD/AD autopsy-confirmed cohort (n = 100), and one genetic FTD cohort (n = 55). The demographic characteristics for the FT(L)D and AD cohorts are described in Table 1 and the genetic FTD cohort is described in Supplementary Table 1. Controls were younger compared to other diagnostic groups in the discovery cohort, clinical cohort 1, and clinical cohort 2. Patients with AD were older, and patients with FTLD were overall younger in the FTLD/AD autopsy cohort compared to the other cohorts. In the genetic FTD cohort, presymptomatic individuals were significantly younger compared to the symptomatic group (Supplementary Table 1).

CSF proteins are differentially regulated in FTD compared to controls and AD

CSF proteome profiling revealed 92 proteins differentially regulated between FTD and controls, of which 42 proteins were decreased and 50 proteins were increased (q < 0.05; Fig. 1B, Supplementary Table 2). The top 5 differentially regulated proteins related to FTD (median q: 2.5^− 07) are involved in the regulation of the Wnt signaling pathway (WIF1), tissue remodeling (MMP7), neuronal biosynthesis (APP), cell proliferation (NPDC1), and innate immunity (IL1RL2). Sensitivity analysis showed that excluding FTD patients with AD co-pathology (n = 19, 10%) had minimal effect on the CSF proteome profiling results (Supplementary Fig. 1). Next, we sought to determine whether the protein changes identified were especially related to FTD. We show that 36 proteins (40%) were uniquely dysregulated in FTD compared to controls and AD (e.g., WIF1, ROBO2, and SLITRK2; Fig. 1C). Nineteen proteins (20%) were dysregulated in FTD and AD compared to controls, which likely represent general dementia processes (e.g., CHIT1, MMP10, and MMP12; Fig. 1C). Twenty-eight proteins (30%) were dysregulated across all comparisons: controls versus FTD and AD, and FTD versus AD. Most of these proteins (21 out of 28; e.g., MIF, ITGB2, and sTREM1) were to some extent increased in FTD but were more prominently dysregulated in AD, indicating that they are likely more related to AD pathophysiological processes. However, 7 out of the 28 proteins (e.g., CHL1, GPC1, CNTN5, HBEGF, CLSTN2, GCP5) showed an opposite direction of change in FTD and AD and are likely differently involved in the pathophysiology of these types of dementias. Functional enrichment analysis of the proteins related to FTD pathophysiology (i.e., those that were increased in FTD as well as those that were specifically dysregulated in FTD compared to AD, n = 43) revealed associations with axon development, regulation of synapse assembly, and cell-cell adhesion mediator activity (Fig. 1D).

CSF protein panels discriminate with high accuracy FTD from controls and AD

Next, we aimed to identify CSF biomarkers specific for FTD. One of the strongest proteins specifically dysregulated in FTD was WIF1, however, its performance as a single marker to discriminate FTD from controls was moderate (AUC: 0.794, 95% CI: 0.75–0.84). Thus, we next performed a classification analysis, followed by internal cross-validation, to investigate whether a combination of proteins could discriminate FTD from controls with higher accuracy (CSF panels, Fig. 1A). We identified a panel of 14 CSF proteins that discriminated FTD from controls with high accuracy (FTD diagnostic panel; AUC: 0.96, 95% CI: 0.91–0.99; Fig. 2A), which was comparable to the performance of CSF NfL (AUC: 0.96, 95% CI: 0.92-1; Fig. 2B). The model included proteins that were dysregulated specifically in FTD (i.e., WIF1, MMP7, GAL, VEGFA, NPDC1, APP, and CCL11; Supplementary Fig. 2), as wel as proteins related to common neurodegenerative processes and other neurodegenerative dementias (i.e., MMP1, MMP10, CHIT1, CCL3, PRDX1, and DDC; Supplementary Fig. 2), or reported to be associated with AD (i.e., CLEC5A; Supplementary Fig. 2). The performance of the FTD diagnostic panel to discriminate FTD from AD was considerably lower (FTD diagnostic panel AUC: 0.77, 95% CI: 0.67–0.86; for comparison: CSF NfL AUC: 0.80, 95% CI: 0.72–0.88, Fig. 2B). We thus next investigated whether we could find a combination of proteins that optimally discriminated FTD from AD. A panel consisting of 13 CSF proteins was identified that discriminated FTD from AD with high accuracy (FTD differential diagnostic panel; AUC: 0.91, 95% CI: 0.85–0.96; Fig. 2C). This panel contained proteins associated with FTD (i.e., GZMB, MMP7, CCL11, NPDC1, PLTP, and APEX1; Supplementary Fig. 3) but also proteins associated with AD (i.e., ABL1, ENO2, ITGB2, SMOC2, and THOP1; Supplementary Fig. 3), as also reported in our previous AD focused PEA study [29]. The model also included two proteins with no differences across groups, likely due to the model’s adjustment for inter-individual variability (i.e., VEGFR3 and KAZALD1; Supplementary Fig. 3) [52]. Of note, three proteins (i.e., MMP7, NPDC1, and CCL11) were included in both the FTD diagnostic and FTD differential diagnostic panel. The FTD-related proteins from these panels are associated with diverse biological pathways including inflammatory processes (GZMB), tissue remodeling (MMP1 and MMP7), cell proliferation (NPDC1), oxidative stress (APEX1), vascular functioning (VEGFA), and neuronal biosynthesis and functioning (APP and GAL) [53–57].

Fig. 2 — CSF biomarker panels for the diagnosis of FTD. (A) Receiver operating characteristic (ROC) curves depict the performance of 14-CSF biomarker panel discriminating FTD (n = 189) from controls (n = 196). Black line is the mean area under the curve (AUC) overall re-samplings (1000 repeats of 5-fold cross-validation, grey lines). (B) Forest plot shows the different AUC and 95% confidence interval for the 14 and 13 CSF biomarker panels and NfL to discriminate between FTD and controls (blue) or AD (red). (C) ROC curves depict the performance of 13-CSF biomarker panel discriminating FTD (n = 189) from AD (n = 232). (D) Correlation matrix heatmap representing the Spearman’s correlation coefficient in-between the proteins selected in each panel, MMSE score, and FTLD-CDR scores in the total cohort, and in a subset of the FTD group (n = 62)

Next, we sought to determine if the proteins included in our FTD panels are associated with disease severity and cognition as measured by FTLD-CDR and MMSE scores. In the FTD patients, we observed that PRDX1 (Rho = 0.47, p < 0.001) had a moderate to strong correlation while NPDC1 (Rho = -0.37, p < 0.001), APEX (Rho = 0.35, p < 0.01), PLTP (Rho = 0.33, p < 0.01), and CHIT1 (Rho = 0.28, p < 0.05) showed a moderate correlation with FTLD-CDR scores (Fig. 2D). In the total cohort, we show that the proteins associated with AD (ABL1, ENO2, ITGB2, SMOC2, and THOP1) together with MMP10 had the strongest correlations with MMSE scores (Rho’s between − 0.19 and − 0.34, all p < 0.001; Fig. 2D).

CSF proteins are differentially regulated in FTLD-Tau or FTLD-TDP subtypes

Our second aim was to investigate which protein dysregulations characterized the FTLD-Tau and FTLD-TDP subtypes. Compared to controls, we observed 60 proteins dysregulated in FTLD-Tau (top 5: MMP10, DDC, CCL3, MMP7, and WIF1; median q 4.2^− 08, q < 0.05; Fig. 3A, Supplementary Table 3) and 120 proteins that were dysregulated in FTLD-TDP (top 5: APP, NPDC1, WIF1, B4GAT1, and ROBO2; median q 1.2^− 06; Fig. 3B, Supplementary Table 4). When comparing FTLD-Tau to FTLD-TDP, we observed that 6 proteins (COCH, Siglec9, VSIG4, GRN, CD84, and C1QTNF1) were dysregulated between these groups (q < 0.05; Fig. 3C, Supplementary Table 5). This number increased to 185 dysregulated proteins when nominal significance was considered (i.e., p < 0.05, Supplementary Table 5). To show which proteins are specifically associated with one of the FTLD subtypes, we visualized the outcomes of three comparisons in an upset plot (Fig. 3D). Here, we identified three proteins uniquely related to FTLD-Tau (i.e., all increased in FTLD-Tau compared to controls and FTLD-TDP; Fig. 3E). These proteins play a role in various processes, including phagocytosis and the NLRP3 inflammasome (VSIG4), cell adhesion processes (Siglec9), and involvement in both the innate and adaptive immune responses (CD84) [58–60]. Two proteins were uniquely related to FTLD-TDP (i.e., both decreased in FTLD-TDP compared to controls and FTLD-Tau; Fig. 3E), which are related to lysosomal functioning (GRN) and the immune response (C1QTNF1) [61, 62]. Interestingly, one protein was uniquely dysregulated between FTLD-Tau and FTLD-TDP (COCH; Fig. 3E), with higher levels in FTLD-Tau. This protein plays a role in the modulation of cell shapes [63].

Fig. 3 — Differential abundance of CSF proteins in FTLD subtypes and FTLD biomarker panel. Volcano plots show CSF proteins differentially regulated between patients with FTLD-Tau (n = 87; A) or FTLD-TDP (n = 67; B) and controls (n = 196) and between these neuropathological subtypes (C). Each dot represents a protein. The beta coefficients (log2 fold-change) are plotted versus q values (-log10-transformed). Proteins significantly dysregulated after adjusting for false discovery rate (FDR, q < 0.05) are depicted in blue. The total number of proteins that are down-regulated (left) or up-regulated (right) is indicated. D) UpSet plot depicts proteins dysregulated between the FTLD-Tau, FTLD-TDP, and control groups. E) Violins represent the abundance (log2 NPX) of the CSF proteins that were uniquely dysregulated in FTLD-Tau and FTLD-TDP or between the subtypes. Boxplot within the violin indicates the median and interquartile range of the protein abundance. F) ROC curves depict the performance of 10-CSF biomarker panel discriminating FTLD-Tau (n = 87) from FTLD-TDP (n = 67). G) Forest plot shows the different AUC and 95% confidence interval for the 10 CSF biomarker panels, and pTau/tTau ratio to discriminate between neuropathological subtypes (purple; 87 FTLD-Tau and 67 FTLD-TDP), subtypes that are genetic and/or autopsy confirmed (brown; 33 FTLD-Tau and 56 FTLD-TDP), subtypes which are genetically confirmed (coral; 16 FTLD-Tau and 31 FTLD-TDP), and subtypes which are autopsy confirmed (beige; 17 FTLD-Tau and 25 FTLD-TDP). $ shows the comparison in a subset of cases (NfL: 76 FTD, 47 controls, and 54 AD; pTau/tTau ratio: 74 FTLD-Tau and 37 FTLD-TDP. # p < 0.05, *q < 0.05, **q < 0.01, ***q < 0.001. Abbreviations: CON, cognitively unimpaired controls; TDP, transactive response DNA binding protein 43

CSF protein panel to discriminate the FTLD subtypes

We next sought to identify CSF biomarkers specific to the FTLD subtypes. A panel of 10 CSF proteins that discriminated between FTLD-Tau and FTLD-TDP with moderate performance was identified (AUC: 0.797, 95% CI: 0.65–0.93; Fig. 3F). Despite the performance being superior to the pTau/tTau ratio (AUC: 0.68, 95% CI: 0.57–0.79; Fig. 3G), the large CI reflects high variability in the cross-validation procedures and thus poor robustness. This model included proteins that were specifically dysregulated in FTLD-Tau (i.e., VSIG4, CD84, and Siglec9) or FTLD-TDP (i.e., GRN) and those that were nominally or significantly changed between subtypes (i.e., COCH, GH, CST5, and PRSS8; Supplementary Fig. 3), as well as proteins that were similar across groups (i.e., IFNLR1 and ANGPT2; Supplementary Fig. 4). The performance of this model did not improve after stratifying the FTLD groups for genetic or sporadic cases (AUCs: 0.37–0.64; Fig. 3G).

Validation of FTD panels using custom PEA assays in independent cohorts

We next aimed to translate our proteomic findings into clinically useful tools. Thus, we developed custom multiplex-PEA assays containing 18 of the 24 proteins from the FTD panels described above (i.e., FTD diagnostic panel: 9 proteins; FTD differential diagnostic panel: 12 proteins; 3 proteins present in both panels). Six proteins could not be included due to technical limitations. The custom PEA assays (custom panels) showed over 90% detectability of the proteins in patient samples, with low coefficients of variation (CVs; <5% intra-CV and < 3% inter-CV; Supplementary Table 6). No custom multiplex-PEA assay was developed for the FTLD subtype panel.

We compared the protein's fold changes obtained in the custom panels for three validation cohorts (including one with autopsy confirmation) to those in the discovery cohort. The protein´s fold changes obtained in the comparisons between FTD and controls significantly correlated between those detected in the discovery and two validation cohorts (clinical cohort 1: Rho = 0.45, p > 0.05, clinical cohort 2: Rho = 0.65, p < 0.05 and FTLD/AD autopsy cohort: Rho = 0.93, p < 0.001; Supplementary Fig. 5A). In addition, the fold changes of WIF1 and MMP10 were strong and replicated across all cohorts (Fig. 4A). Similarly, strong correlations of the protein fold changes for the difference between FTD and AD patients were observed between the discovery cohort and the three validation cohorts (clinical cohort 1: Rho = 0.69, p < 0.05; clinical cohort 2: Rho = 0.78, p < 0.01; FTLD/AD autopsy cohort: Rho = 0.87, p < 0.001, Supplementary Fig. 5B). Noteworthy, we observed that the fold changes of some proteins lost significance after correction for multiple testing in clinical cohort 1 and clinical cohort 2, whereas several proteins showed their strongest effects in the FTLD/AD autopsy cohort (e.g., THOP1, MMP7, ENO2, CCL11, and KAZALD1; Figs. 4A and 5A).

Fig. 4 — Development and validation of custom CSF biomarker panels for FTD diagnosis in independent cohorts. (A) Lollipop plots depict the beta-coefficients obtained in the discovery phase in parallel to the beta-coefficients of the custom assays in clinical validation cohorts 1 and 2 and the FTLD/AD autopsy cohort. Grey dots show proteins that did not remain significant after correction for multiple testing. (B) Receiver operating characteristic (ROC) curves showing the performance of the CSF biomarker panel discriminating FTD from controls using the custom assays across the two clinical and one autopsy validation cohort. Inserts outline corresponding AUC and 95% CI. (C) Forest plots depict the different AUC and 95% CI obtained with the CSF FTD biomarker panels or CSF NfL in the comparison between FTD and controls (blue) or AD (red). (D) ROC curves showing the performance of the FTD diagnostic panel in the FTD genetic cohort to discriminate non-carriers from presymptomatic (PreSymp) and symptomatic (Symp) mutation carriers, and presymptomatic from symptomatic

Fig. 5 — Development and validation of custom CSF biomarker panels for FTD differential diagnosis in independent cohorts. (A) Lollipop plots depict the beta-coefficients obtained in the discovery phase in parallel to the beta-coefficients of the custom assays in clinical validation cohorts 1 and 2 and the FTLD/AD autopsy cohort. Grey dots show proteins that did not remain significant after correction for multiple testing. (B) Receiver operating characteristic (ROC) curves showing the performance of the CSF biomarker panel discriminating FTD from AD using the custom assays across the two clinical and one autopsy validation cohort. Inserts outline corresponding AUC and 95% CI. (C) Forest plots depict the different AUC and 95% CI obtained with the CSF FTD biomarker panels or CSF NfL in the comparison between FTD and AD (red) or controls (blue). (D) ROC curves showing the performance of the FTD differential diagnostic panel in the FTD genetic cohort to discriminate non-carriers from presymptomatic (PreSymp) and symptomatic (Symp) mutation carriers, and presymptomatic from symptomatic

The FTD diagnostic panel (containing 9 out of the 14 proteins) confirmed discrimination of FTD from controls with high accuracy in all validation cohorts (clinical cohort 1: AUC = 0.96, 95% CI: 0.92-1; clinical cohort 2: AUC = 0.96, 95% CI: 0.92-1; FTLD/AD autopsy cohort: AUC = 0.98, 95% CI: 0.95-1; Fig. 4B) and was similar to CSF NfL (clinical cohort 1: AUC = 0.95, 95% CI: 0.87-1; Fig. 4C). Exclusion of the FTD patients with AD co-pathology (n = 8) in clinical cohort 1 had minimal impact on the performance of the FTD diagnostic panel (AUC = 0.960, 95% CI: 0.91-1; Supplementary Fig. 6A). In addition, this FTD diagnostic panel showed strong performance for discriminating symptomatic mutation carriers from non-carriers (AUC = 1, 95% CI: 1–1) and presymptomatic carriers (AUC = 0.95, 95% CI: 0.90-1) as wel as in differentiating presymptomatic carriers from non-carriers (AUC = 1, 95% CI: 1–1; Fig. 4D; Supplementary Fig. 7). The FTD differential diagnostic panel (containing 12 out of the 13 proteins) could again discriminate FTD from AD with high accuracy in the three validation cohorts (clinical cohort 1: AUC = 0.88, 95% CI: 0.81–0.95; clinical cohort 2: AUC = 0.94, 95% CI: 0.90–0.98; FTLD/AD autopsy cohort: AUC = 0.90, 95% CI: 0.83–0.97; Fig. 5B), which was similar to NfL (clinical cohort 1, AUC = 0.93, 95% CI: 0.84-1; Fig. 5C). Exclusion of the FTD patients with AD co-pathology (n = 8) in clinical cohort 1 increased the diagnostic performance of the FTD differential diagnostic panel (AUC = 0.939, 95% CI: 0.87–0.99; Supplementary Fig. 6B). In addition, the FTD differential diagnostic panel showed good performance for discriminating symptomatic mutation carriers from non-carriers (AUC = 1, 95% CI: 1–1) and presymptomatic carriers (AUC = 0.93, 95% CI: 0.86-1) as wel as in differentiating presymptomatic carriers from non-carriers (AUC = 0.96, 95% CI: 0.90-1; Fig. 5D; Supplementary Fig. 8). To summarize, we replicated the discriminative performance as we observed in the discovery cohort, both for the FTD diagnostic and differential diagnostic panel.

Discussion

This study identified novel and specific protein changes for FTD and its main pathological subtypes. These findings were translated into two CSF biomarker panels that discriminated FTD from controls and AD patients with high accuracy (AUCs > 0.9). These results were validated by clinically feasible panels that measured the selected proteins in several independent cohorts, including an autopsy-confirmed and a genetic FTD cohort. We identified proteins specifically associated with FTLD-Tau or FTLD-TDP groups, but we could not identify a marker or combination of markers to robustly discriminate between these FTLD subgroups. Moreover, the panels identified the presymptomatic stage of FTD mutation carriers with high accuracy, while the further increase in the symptomatic stage suggests their potential as an objective tool for monitoring disease progression. The FTD proteins and biomarker panels identified in this study reflect a broad range of different biological processes associated with FTD and its biological subtypes including inflammatory processes, regulation of synapses, lysosomal functioning, tissue remodeling, and oxidative stress.

Fluid biomarkers specifically associated with FT(L)D pathophysiology are needed to improve diagnostic accuracy, for clinical trial inclusion, and to monitor treatment effects [14]. We have performed a large multicenter FTD proteomics study, including patients with different FTLD neuropathologies as well as a group with AD dementia. Up to 92 proteins were dysregulated between FTD and controls. By comparing the protein profiles to AD patients, we observed that 40% of these proteins were uniquely dysregulated in FTD (e.g., WIF1, ROBO2, and SLITRK2). Some proteins have been associated with FTD in previous CSF antibody-based proteomics (VEGFA [64]), or FTLD brain proteomics studies (ADAM23 [65–67] and WIF1 [68]). We also identified a subset of CSF proteins (30%) that were differentially regulated in both patients with FTD and AD compared to controls but with a different protein abundance between FTD and AD. Despite most of these being more prominently dysregulated in AD (e.g., SDC4, ITGB2, MIF, and sTREM1) [29, 69, 70], a small subset of proteins showed an opposite effect between these dementias (e.g., CHL1, GPC1, and CNTN5). Previous studies additionally detected decreased levels of CNTN5 and CHL1 in FTD and PSP using alternative platforms, which further supports the validity of our findings [71–73]. In addition, we detected a subset of proteins (20%) likely reflecting an overall neurodegenerative process, as they were similarly dysregulated in AD and FTD compared with controls. For instance, CHIT1 and MMP10 protein levels were increased in FTD and AD, as reported previously [29, 64, 74–79]. This study identifies numerous CSF proteins dysregulated in FTD. However, due to the inclusion of an AD group, we uncovered that half (43 out of 92) of the proteins are specifically associated with FTD, highlighting the importance of comparing to both controls and similar neurological diseases in biomarker studies. Among the proteins specifically associated with FTD, the ones showing the strongest effects were: WIF1, MMP7, APP, NPDC1, and IL1RL2. WIF1 is an inhibitory protein involved in the Wnt signaling cascade, a dysregulation of this pathway has been implicated in tau phosphorylation and other neuronal processes (e.g., neurogenesis, synaptic health and plasticity) [80, 81]. In addition, WIF1 was dysregulated in FTLD frontal cortex tissue [68], supporting its role within FTD pathophysiology, however, its specific function in relation to FTD remains to be elucidated. Enrichment analysis showed that proteins specifically dysregulated in FTD were enriched in biological pathways associated with axon development, regulation of synapse assembly, and cell-cell adhesion mediator activity. This is in line with previous unbiased brain and CSF proteomics studies, and multiplex CSF analysis showing a dysregulation of similar pathways in FTLD [31, 65, 66, 82, 83].

Considering that biomarkers to discriminate the FTLD-subtypes are currently lacking. We additionally investigated the CSF proteome in the main FTLD pathological subtypes. Despite many CSF proteins that were dysregulated in FTLD-Tau (60 proteins) or FTLD-TDP (120 proteins) compared to controls, overlapping analysis across all comparisons revealed three proteins related to FTLD-Tau (VSIG4, Siglec9, and CD84), two proteins related to FTLD-TDP (GRN and C1QTNF1), and one protein uniquely changed between the neuropathological subtypes (COCH). Given the established association between GRN mutations and TDP pathology, the decrease of GRN (progranulin) protein levels observed in FTLD-TDP aligns with previous GRN reductions in CSF and plasma of FTD-GRN mutation carriers [84, 85]. It should be noted that the FTLD-TDP group included a low number of GRN mutation carriers (11%). Considering that we observed decreased GRN levels in the total FTLD-TDP group, these results highlight that lysosomal dysfunction is an important feature in TDP pathology, which is additionally supported by post-mortem analysis [86, 87]. Furthermore, we also identified novel markers such as VSIG4. This protein was also found to be elevated in MAPT mutation carriers compared to non-carriers in a recent MS-proteomics study [88]. As a phagocytic receptor, it plays a key role in inflammatory responses and can mediate activation of the NLRP3 inflammasome [58]. Since the NLRP3 inflammasome plays a role within tau pathophysiology, this protein seems promising to reflect ongoing activation of the NLRP3 inflammasome in FTD patients [89]. To the best of our knowledge, many of the subtype-specific proteins identified in this study are novel and have not yet been linked to the main FTD pathologies or genetic subtypes in previous proteomics studies [30, 31, 65, 66]. Thus, these proteins require further investigation in CSF and brain tissue, to examine their relevance to ongoing tau or TDP43 pathophysiologies.

Next, we identified a classification signature of 10 CSF proteins to discriminate the two main FTLD subtypes (AUC: 0.8), with higher performance than the pTau/tTau ratio (AUC: 0.68). However, the moderate performance together with the large confidence intervals suggests that the model is rather unstable and may vary depending on the samples included. This variation could be explained by the inclusion of different FTD biochemical profiles, which likely differ between sporadic and familial cases with the same proteopathy, or even between genetic backgrounds of the same pathological subtypes (e.g., GRN and C9orf72) [14, 71, 88, 90]. Despite the unprecedented number of FTLD samples and proteins analyzed in this study, we identified relatively few robust markers associated with specific pathological subtypes. This limited discriminative power underscores both the biological complexity of FTLD and the inherent difficulty of distinguishing subtypes in vivo, a challenge that remains a critical hurdle for the field. While previous studies have reported promising results in differentiating broader diagnostic categories such as AD and FTD, the finer stratification of the FTLD subtypes requires deeper molecular insights. Notably, we observed a number of proteins with nominally significant differences across the FTLD subtypes, suggesting that larger, more powered studies may reveal additional subtype-specific biomarkers. These findings also reflect the heterogeneity within each FTLD subtype, emphasizing the importance of analyzing well-characterized cohorts, ideally with pathological and/or genetic confirmed individuals [14].

To translate the CSF proteome findings into practical biomarker tools for routine diagnostics or clinical trials, we applied classification analyses and identified two panels of 14 and 13 CSF proteins that can discriminate FTD from controls and AD dementia with high accuracies (AUCs of 0.96 and 0.91, respectively). Thereafter, custom multiplex assays were developed and validated in several independent cohorts. The fold changes for each protein correlated well between discovery and validation cohorts (Rho’s between 0.45 and 0.93). The high discriminative values for FTD were confirmed in three cohorts for both panels (AUCs > 0.877), supporting the robustness of our findings. In addition, these panels show added value by identifying FTD mutation carriers approaching symptom onset, and could thus be relevant to detect early pathological changes. Noteworthy, our novel FTD panels have similar discriminative performance compared to NfL. However, while NfL reflects neuroaxonal damage, our CSF panels offer a more comprehensive depiction of the underlying biological processes in FTD. The relevance of these FTD panels is additionally supported by the association of some markers (e.g., MMP10, PLTP, PRDX1, and NPDC1) with clinical parameters such as cognitive functioning or disease severity. For example, we detected positive correlations of APEX, PRDX1, and CHIT1 with FTLD-CDR scores. These proteins are related to oxidative stress and reactive microglial activity, processes that underlie FTD pathogenesis, and may therefore be relevant for disease staging [91, 92]. In addition, in agreement with previous studies, we observed that increased MMP10 levels were associated with worse cognitive performance [29, 74]. This could potentially be explained by the expression of MMP10 by microglial cells in the brain, as elevated MMP10 levels contribute to ongoing inflammatory responses leading to axonal damage, potentially leading to functional deficits such as cognitive impairment [93]. Overall, these associations provide further support for the potential value of these panels in clinical settings and trial contexts.

This study is not without limitations. The targeted proteomic approach employed in this study misses relevant proteins that might have been measured using unbiased MS methods (e.g., GFAP and neuropentraxins). However, our study still covered > 600 proteins covering a wide range of biological mechanisms and the workflow employed allowed us to swiftly translate our proteomic discovery findings into custom immunoassays for subsequent validation. Noteworthy, potential FTD misdiagnosis may influence biomarker results. However, diagnoses were made in specialized memory clinics and 47% of FTD individuals were either pathologically or genetically confirmed, and results were further validated in multiple independent cohorts. Furthermore, we observed that a subset of confirmed FTD patients in the discovery cohort exhibited a positive AD biomarker profile indicative of AD-co-pathology. Exclusion of these cases minimally affected the differential protein expression analyses, but increased the discriminative potential in the validation cohorts, which can probably be attributed to the association of several proteins in the panel with AD pathology (e.g., ABL1, THOP1, SMOC2, and ITGB2). Nevertheless, the potential influence of AD co-pathology on the FTD proteome should be explored in future studies involving larger sample sizes. Noteworthy, validation in the FTD genetic cohort demonstrated a high discriminative performance of our CSF biomarker panels in both presymptomatic and symptomatic FTD carriers compared to non-carriers. Although the smaller sample size in this cohort, relative to our other cohorts, may impact the reliability of the analysis, the strong performance of the panel remains promising for detecting FTD across different disease stages. However, these biomarker panels are not yet ready for clinical implementation. Additional validation in independent FTD cohorts is required to confirm their robustness and generalizability. Furthermore, it would be valuable to assess the performance of these novel panels in other disorders that share clinical and/or neuropathological features with FTD, such as primary psychiatric disorders or amyotrophic lateral sclerosis.

Conclusions

This study identified CSF proteome changes specifically associated to FTD. We have translated these findings into CSF biomarker panels that can accurately discriminate FTD from controls and AD in multiple cohorts, and in the presymptomatic stage of the disease. Furthermore, we detect protein changes specifically associated to FTLD-Tau or FTD-TDP, although larger and more homogeneous cohorts will have more power to discriminate between these main pathological subtypes. The panels developed within this study could prove valuable for diagnosis or to monitor treatment effects in clinical trials. The antibody-based technology employed in this study allowed us to efficiently translate our discovery findings into custom multiplex panels for further clinical validation showing reproducible findings. This workflow applied here could also be helpful for the development of fluid biomarkers in other human matrices (e.g., blood) and other biomedical fields beyond neurodegenerative dementias.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1^{(13.1KB, xlsx)}

Supplementary Material 2^{(44.9KB, xlsx)}

Supplementary Material 3^{(46.2KB, xlsx)}

Supplementary Material 4^{(44.9KB, xlsx)}

Supplementary Material 5^{(43.7KB, xlsx)}

Supplementary Material 6^{(12.5KB, xlsx)}

Supplementary Material 7^{(1.9MB, docx)}

Acknowledgements

Not applicable.

Abbreviations

AD: Alzheimer’s dementia
ADC: Amsterdam dementia cohort
AUC: Area under the curve
CSF: Cerebrospinal fluid
CON: Cognitively normal individuals
CBS: Corticobasal syndrome
FDR: False discovery rate
FTD: Frontotemporal dementia
FTLD: Frontal temporal lobar degeneration
LOD: Lower limit of detection
MMSE: Mini–mental state examination
MS: Mass spectrometry
NPX: Normalized protein expression
PEA: Proximity extension assay
PPA: Primary progressive aphasia
PSP: Progressive supranuclear
QC: Quality controls

Author contributions

Y.H., M.C., and C.T. conceived and designed the study. Y.H., L.V., and C.P. performed the statistical analysis. Y.H., L.V., E.E., S.B., L.M., J.S., J.H., H.R., A.L., D.A., S.E., A.S., A.C.-P., D.I., W.F., Y.P., M.C., and C.T. recruited participants and collected clinical data and samples. Y.H. and M.C. arranged and prepared samples for proteomics analysis. Y.H., M.C., and C.T. drafted de manuscript. All authors contributed to the revision and editing of the manuscript.

Funding

This project was funded by the memorable project PRIDE (Project No. WE.03-2018-05), supported by Alzheimer Nederland. This research is part of the neurodegeneration research program of Amsterdam Neuroscience. Alzheimer Center Amsterdam is supported by Stichting Alzheimer Nederland and Stichting VUmc fonds. MdC is supported by the attraction talent fellowship from Comunidad de Madrid (2018-T2/BMD-11885). Data from Penn were contributed by the program project grants P01-AG-066597, U19 AG062418 (formerly P50 NS053488), and P30AG072979 (formerly AG010124).

Data availability

The source data generated in this study are available within this study (Supplementary Table 2, Supplementary Table 3, Supplementary Table 4, Supplementary Table 5). All models were built using publicly available packages and functions in R.

Declarations

Ethics approval and consent to participate

Approval was given by the institutional ethical review boards of each center. Written informed consent was obtained from all participants or their authorized representatives.

Consent for publication

Not applicable.

Competing interests

Y.H., C.P., E.E., S.B., L.M, J.S., J.H., H.R., W.H., A.S., A.C.-P, and D.I. declare no competing interest. M.C. has been an invited speaker at Eisai, is an associate editor at Alzheimer´s Research & Therapy and has been an invited writer for Springer Healthcare. L.V. received a grant for CORAL consortium by Olink proteomics. D.I. is a Scientific Advisory Board Member for Denali Therapeutics. D.A. participated in advisory boards from Fujirebio-Europe and Roche Diagnostics and received speaker honoraria from Fujirebio-Europe, Roche Diagnostics, Nutricia, Krka Farmacéutica S.L., Zambon S.A.U. and Esteve Pharmaceuticals. S.A. D.A., and A.L. declare a filed patent application (Title: Markers of synaptopathy in neurodegenerative disease; Applicant: Fundació Institut de Recerca de l’Hospital de la Santa Creu i Sant Pau, Inventors: Olivia BELBIN; Alberto LLEÓ; Alejandro BAYÉS; Juan FORTEA; Daniel ALCOLEA; Application number: PCT/EP2019/056535; International Publication Number: WO 2019/175379 A1; Current status: Active. Licensed to ADx Neurociences NV (Ghent, Belgium); this patent is not related to any specific aspect of the current manuscript). S.E. received personal fees from Eisai (paid to institution), personal fees from icometrix (paid to institution), personal fees from Novartis (paid to institution), personal fees from Roche (paid to institution) and personal fees from Roche and personal fees from Biogen, all outside the submitted work. W.F. has performed contract research for Biogen MA Inc, and Boehringer Ingelheim. W.F. has been an invited speaker at Boehringer Ingelheim, Biogen MA Inc, Danone, Eisai, WebMD Neurology (Medscape), Springer Healthcare. W.F. is consultant to Oxford Health Policy Forum CIC, Roche, and Biogen MA Inc. WF participated in advisory boards of Biogen MA Inc and Roche. All funding is paid to her institution. W.F. is a member of the steering committee of PAVE, and Think Brain Health. W.F. was associate editor of Alzheimer’s Research & Therapy in 2020/2021 and is currently an associate editor at Brain. C.E.T. has a collaboration contract with ADx Neurosciences, Quanterix, and Eli Lilly, performed contract research or received grants from AC-Immune, Axon Neurosciences, Bioconnect, Bioorchestra, Brainstorm Therapeutics, Celgene, EIP Pharma, Eisai, Grifols, Novo Nordisk, PeopleBio, Roche, Toyama, and Vivoryon. She serves on editorial boards of Medidact Neurologie/Springer, Alzheimer’s Research & Therapy, Neurology: Neuroimmunology & Neuroinflammation, and is editor of a Neuromethods book Springer. She had speaker contracts for Roche, Grifols, and Novo Nordisk.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Charlotte Teunissen and Marta del Campo shared senior authors.

References

1.Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez M, Cappa SF et al. Classification of primary progressive aphasia and its variants. Am Acad Neurol. 2011;76. [DOI] [PMC free article] [PubMed]
2.Rascovsky K, Hodges JR, Knopman D, Mendez MF, Kramer JH, Neuhaus J, et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain. 2011;134(Pt 9):2456–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Armstrong MJ, Litvan I, Lang AE, Bak TH, Bhatia KP, Borroni B et al. Criteria for the diagnosis of corticobasal degeneration. Am Acad Neurol. 2013;80. [DOI] [PMC free article] [PubMed]
4.Hoglinger GU, Respondek G, Stamelou M, Kurz C, Josephs KA, Lang AE, et al. Clinical diagnosis of progressive supranuclear palsy: the movement disorder society criteria. Mov Disord. 2017;32(6):853–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Irwin DJ, Cairns NJ, Grossman M, McMillan CT, Lee EB, Van Deerlin VM, et al. Frontotemporal lobar degeneration: defining phenotypic diversity through personalized medicine. Acta Neuropathol. 2015;129(4):469–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Mann DMA, Snowden JS. Frontotemporal lobar degeneration: pathogenesis, pathology and pathways to phenotype. Brain Pathol. 2017;27(6):723–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Neumann M, Rademakers R, Roeber S, Baker M, Kretzschmar HA, Mackenzie IR. A new subtype of frontotemporal lobar degeneration with FUS pathology. Brain. 2009;132(Pt 11):2922–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Lashley T, Rohrer JD, Mead S, Revesz T. Review: an update on clinical, genetic and pathological aspects of frontotemporal lobar degenerations. Neuropathol Appl Neurobiol. 2015;41(7):858–81. [DOI] [PubMed] [Google Scholar]
9.Rohrer JD, Guerreiro R, Vandrovcova J, Uphill J, Reiman D, Beck J et al. The heritability and genetics of frontotemporal lobar degeneration. Neurology. 2009;73. [DOI] [PMC free article] [PubMed]
10.Greaves CV, Rohrer JD. An update on genetic frontotemporal dementia. J Neurol. 2019;266(8):2075–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Pozzi FE, Aprea V, Giovannelli G, Lattuada F, Crivellaro C, Bertola F, et al. Clinical and neuroimaging characterization of the first frontotemporal dementia family carrying the MAPT p.K298E mutation. Neurogenetics. 2024;25(3):215–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Cheng HR, Lin RR, Li HL, Xue YY, Gao PR, Chen DF, et al. Identification and functional characterization of novel variants of MAPT and GRN in Chinese patients with frontotemporal dementia. Neurobiol Aging. 2023;123:233–43. [DOI] [PubMed] [Google Scholar]
13.Neumann M, Lee EB, Mackenzie IR. Frontotemporal lobar degeneration TDP-43-immunoreactive pathological subtypes: clinical and mechanistic significance. Adv Exp Med Biol. 2021;1281:201–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Del Campo M, Zetterberg H, Gandy S, Onyike CU, Oliveira F, Udeh-Momoh C, et al. New developments of biofluid-based biomarkers for routine diagnosis and disease trajectories in frontotemporal dementia. Alzheimers Dement. 2022;18(11):2292–307. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Hutchinson AD, Mathias JL. Neuropsychological deficits in frontotemporal dementia and alzheimer’s disease: a meta-analytic review. J Neurol Neurosurg Psychiatry. 2007;78(9):917–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Naasan G, Rabinovici GD, Ghosh P, Elofson JD, Miller BL, Coppola G, et al. Amyloid in dementia associated with Familial FTLD: not an innocent bystander. Neurocase. 2016;22(1):76–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Irwin DJ, Trojanowski JQ, Grossman M. Cerebrospinal fluid biomarkers for differentiation of frontotemporal lobar degeneration from alzheimer’s disease. Front Aging Neurosci. 2013;5:6. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Ashton NJ, Janelidze S, Al Khleifat A, Leuzy A, van der Ende EL, Karikari TK, et al. A multicentre validation study of the diagnostic value of plasma neurofilament light. Nat Commun. 2021;12(1):3400. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Bridel C, van Wieringen WN, Zetterberg H, Tijms BM, Teunissen CE, and the, NFLG et al. Diagnostic value of cerebrospinal fluid neurofilament light protein in neurology: a systematic review and meta-analysis. JAMA Neurol. 2019;76(9):10354–8. [DOI] [PMC free article] [PubMed]
20.Ducharme S, Dols A, Laforce R, Devenney E, Kumfor F, van den Stock J, et al. Recommendations to distinguish behavioural variant frontotemporal dementia from psychiatric disorders. Brain. 2020;143(6):1632–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Gendron TF, Heckman MG, White LJ, Veire AM, Pedraza O, Burch AR, et al. Comprehensive cross-sectional and longitudinal analyses of plasma neurofilament light across FTD spectrum disorders. Cell Rep Med. 2022;3(4):100607. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Meeter LHH, Vijverberg EG, Del Campo M, Rozemuller AJM, Donker Kaat L, de Jong FJ, et al. Clinical value of neurofilament and phospho-tau/tau ratio in the frontotemporal dementia spectrum. Neurology. 2018;90(14):e1231–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Pijnenburg YA, Verwey NA, van der Flier WM, Scheltens P, Teunissen CE. Discriminative and prognostic potential of cerebrospinal fluid phosphotau/tau ratio and neurofilaments for frontotemporal dementia subtypes. Alzheimers Dement (Amst). 2015;1(4):505–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Cousins KAQ, Shaw LM, Chen-Plotkin A, Wolk DA, Van Deerlin VM, Lee EB, et al. Distinguishing frontotemporal lobar degeneration Tau from TDP-43 using plasma biomarkers. JAMA Neurol. 2022;79(11):1155–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Desmarais P, Rohrer JD, Nguyen QD, Herrmann N, Stuss DT, Lang AE, et al. Therapeutic trial design for frontotemporal dementia and related disorders. J Neurol Neurosurg Psychiatry. 2019;90(4):412–23. [DOI] [PubMed] [Google Scholar]
26.Higginbotham L, Ping L, Dammer EB, Duong DM, Zhou M, Gearing M, et al. Integrated proteomics reveals brain-based cerebrospinal fluid biomarkers in asymptomatic and symptomatic alzheimer’s disease. Sci Adv. 2020;6:43. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Bader JM, Geyer PE, Muller JB, Strauss MT, Koch M, Leypoldt F, et al. Proteome profiling in cerebrospinal fluid reveals novel biomarkers of alzheimer’s disease. Mol Syst Biol. 2020;16(6):e9356. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Dammer EB, Ping L, Duong DM, Modeste ES, Seyfried NT, Lah JJ, et al. Multi-platform proteomic analysis of alzheimer’s disease cerebrospinal fluid and plasma reveals network biomarkers associated with proteostasis and the matrisome. Alzheimers Res Ther. 2022;14(1):174. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Del Campo M, Peeters CFW, Johnson ECB, Vermunt L, Hok-A-Hin YS, van Nee M, et al. CSF proteome profiling across the alzheimer’s disease spectrum reflects the multifactorial nature of the disease and identifies specific biomarker panels. Nat Aging. 2022;2(11):1040–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Teunissen CE, Elias N, Koel-Simmelink MJ, Durieux-Lu S, Malekzadeh A, Pham TV, et al. Novel diagnostic cerebrospinal fluid biomarkers for pathologic subtypes of frontotemporal dementia identified by proteomics. Alzheimers Dement (Amst). 2016;2:86–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.van der Ende EL, Meeter LH, Stingl C, van Rooij JGJ, Stoop MP, Nijholt DAT, et al. Novel CSF biomarkers in genetic frontotemporal dementia identified by proteomics. Ann Clin Transl Neurol. 2019;6(4):698–707. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Del Campo M, Galimberti D, Elias N, Boonkamp L, Pijnenburg YA, van Swieten JC, et al. Novel CSF biomarkers to discriminate FTLD and its pathological subtypes. Ann Clin Transl Neurol. 2018;5(10):1163–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.van der Ende EL, Xiao M, Xu D, Poos JM, Panman JL, Jiskoot LC, et al. Neuronal pentraxin 2: a synapse-derived CSF biomarker in genetic frontotemporal dementia. J Neurol Neurosurg Psychiatry. 2020;91(6):612–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Bolsewig K, Hok AHYS, Sepe FN, Boonkamp L, Jacobs D, Bellomo G, et al. A combination of neurofilament light, glial fibrillary acidic protein, and neuronal Pentraxin-2 discriminates between frontotemporal dementia and other dementias. J Alzheimers Dis. 2022;90(1):363–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Llorens F, Thune K, Tahir W, Kanata E, Diaz-Lucena D, Xanthopoulos K, et al. YKL-40 in the brain and cerebrospinal fluid of neurodegenerative dementias. Mol Neurodegener. 2017;12(1):83. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Del Campo M, Vermunt L, Peeters CFW, Sieben A, Hok AHYS, Lleo A, et al. CSF proteome profiling reveals biomarkers to discriminate dementia with lewy bodies from alzheimer s disease. Nat Commun. 2023;14(1):5635. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.van der Flier WM, Scheltens P. Amsterdam dementia cohort: performing research to optimize care. J Alzheimers Dis. 2018;62(3):1091–111. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Ingelsson M, Ramasamy K, Russ C, Freeman SH, Orne J, Raju S, et al. Increase in the relative expression of Tau with four microtubule binding repeat regions in frontotemporal Lobar degeneration and progressive supranuclear palsy brains. Acta Neuropathol. 2007;114(5):471–9. [DOI] [PubMed] [Google Scholar]
39.Alcolea D, Clarimon J, Carmona-Iragui M, Illan-Gala I, Morenas-Rodriguez E, Barroeta I, et al. The Sant Pau initiative on neurodegeneration (SPIN) cohort: A data set for biomarker discovery and validation in neurodegenerative disorders. Alzheimers Dement (N Y). 2019;5:597–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-Gateau P, Cummings J, et al. Research criteria for the diagnosis of alzheimer’s disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 2007;6(8):734–46. [DOI] [PubMed] [Google Scholar]
41.Cairns NJ, Bigio EH, Mackenzie IR, Neumann M, Lee VM, Hatanpaa KJ, et al. Neuropathologic diagnostic and nosologic criteria for frontotemporal lobar degeneration: consensus of the consortium for frontotemporal lobar degeneration. Acta Neuropathol. 2007;114(1):5–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Montine TJ, Phelps CH, Beach TG, Bigio EH, Cairns NJ, Dickson DW, et al. National institute on aging-alzheimer’s association guidelines for the neuropathologic assessment of alzheimer’s disease: a practical approach. Acta Neuropathol. 2012;123(1):1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Miyagawa T, Brushaber D, Syrjanen J, Kremers W, Fields J, Forsberg LK, et al. Utility of the global CDR plus NACC FTLD rating and development of scoring rules: data from the ARTFL/LEFFTDS consortium. Alzheimers Dement. 2020;16(1):106–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Hok-A-Hin YS, Willemse EAJ, Teunissen CE, Del Campo M. Guidelines for CSF processing and biobanking: impact on the identification and development of optimal CSF protein biomarkers. Methods Mol Biol. 2019;2044:27–50. [DOI] [PubMed] [Google Scholar]
45.Tijms BM, Willemse EAJ, Zwan MD, Mulder SD, Visser PJ, van Berckel BNM, et al. Unbiased approach to counteract upward drift in cerebrospinal fluid Amyloid-beta 1–42 analysis results. Clin Chem. 2018;64(3):576–85. [DOI] [PubMed] [Google Scholar]
46.Vermunt L, Otte M, Verberk IMW, Killestein J, Lemstra AW, van der Flier WM, et al. Age- and disease-specific reference values for neurofilament light presented in an online interactive support interface. Ann Clin Transl Neurol. 2022;9(11):1832–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Proteomics O. Development and validation of customized PEA biomarker panels with clinical utility.
48.Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J Royal Stat Soc Ser B (Methodological). 1995;57.
49.de Leeuw FA, Peeters CFW, Kester MI, Harms AC, Struys EA, Hankemeier T, et al. Blood-based metabolic signatures in alzheimer’s disease. Alzheimers Dement (Amst). 2017;8:196–207. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33(1):22. [PMC free article] [PubMed] [Google Scholar]
51.Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Karlsson L, Vogel J, Arvidsson I, Åström K, Janelidze S, Blennow K et al. Cerebrospinal fluid reference proteins increase accuracy and interpretability of biomarkers for brain diseases. BioRxiv. 2023. [DOI] [PMC free article] [PubMed]
53.Poe M, Blake JT, Boulton DA, Gammon M, Sigal NH, Wu JK, et al. Human cytotoxic lymphocyte granzyme B. Its purification from granules and the characterization of substrate and inhibitor specificity. J Biol Chem. 1991;266(1):98–103. [PubMed] [Google Scholar]
54.Mroczko B, Groblewska M, Barcikowska M. The role of matrix metalloproteinases and tissue inhibitors of metalloproteinases in the pathophysiology of neurodegeneration: a literature study. J Alzheimers Dis. 2013;37(2):273–83. [DOI] [PubMed] [Google Scholar]
55.Spencer ML, Theodosiou M, Noonan DJ. NPDC-1, a novel regulator of neuronal proliferation, is degraded by the ubiquitin/proteasome system through a PEST degradation motif. J Biol Chem. 2004;279(35):37069–78. [DOI] [PubMed] [Google Scholar]
56.Pei DS, Jia PP, Luo JJ, Liu W, Strauss PR. AP endonuclease 1 (Apex1) influences brain development linking oxidative stress and DNA repair. Cell Death Dis. 2019;10(5):348. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L. VEGF receptor signalling - in control of vascular function. Nat Rev Mol Cell Biol. 2006;7(5):359–71. [DOI] [PubMed] [Google Scholar]
58.Huang X, Feng Z, Jiang Y, Li J, Xiang Q, Guo S et al. VSIG4 mediates transcriptional Inhibition of Nlrp3 and Il-1b in macrophages. Sci Adv. 2019;5. [DOI] [PMC free article] [PubMed]
59.Zhang JQ, Nicoll G, Jones C, Crocker PR. Siglec-9, a novel Sialic acid binding member of the Immunoglobulin superfamily expressed broadly on human blood leukocytes. J Biol Chem. 2000;275(29):22121–6. [DOI] [PubMed] [Google Scholar]
60.Yan Q, Malashkevich VN, Fedorov A, Fedorov E, Cao E, Lary JW, et al. Structure of CD84 provides insight into SLAM family function. PNAS. 2007;104:25. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Tanaka Y, Suzuki G, Matsuwaki T, Hosokawa M, Serrano G, Beach TG, et al. Progranulin regulates lysosomal function and biogenesis through acidification of lysosomes. Hum Mol Genet. 2017;26(5):969–88. [DOI] [PubMed] [Google Scholar]
62.Kishore U, Gaboriaud C, Waters P, Shrive AK, Greenhough TJ, Reid KB, et al. C1q and tumor necrosis factor superfamily: modularity and versatility. Trends Immunol. 2004;25(10):551–61. [DOI] [PubMed] [Google Scholar]
63.Goel M, Sienkiewicz AE, Picciani R, Lee RK, Bhattacharya SK. Cochlin induced TREK-1 co-expression and Annexin A2 secretion: role in trabecular meshwork cell elongation and motility. PLoS ONE. 2011;6(8):e23070. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Bostrom G, Freyhult E, Virhammar J, Alcolea D, Tumani H, Otto M, et al. Different inflammatory signatures in alzheimer’s disease and frontotemporal dementia cerebrospinal fluid. J Alzheimers Dis. 2021;81(2):629–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Mol MO, Miedema SSM, Melhem S, Li KW, Koopmans F, Seelaar H, et al. Proteomics of the dentate gyrus reveals semantic dementia specific molecular pathology. Acta Neuropathol Commun. 2022;10(1):190. [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Miedema SSM, Mol MO, Koopmans FTW, Hondius DC, van Nierop P, Menden K, et al. Distinct cell type-specific protein signatures in GRN and MAPT genetic subtypes of frontotemporal dementia. Acta Neuropathol Commun. 2022;10(1):100. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Umoh ME, Dammer EB, Dai J, Duong DM, Lah JJ, Levey AI, et al. A proteomic network approach across the ALS-FTD disease spectrum resolves clinical phenotypes and genetic vulnerability in human brain. EMBO Mol Med. 2018;10(1):48–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Mofrad RB, Del Campo M, Peeters CFW, Meeter LHH, Seelaar H, Koel-Simmelink M, et al. Plasma proteome profiling identifies changes associated to AD but not to FTD. Acta Neuropathol Commun. 2022;10(1):148. [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Hok-A-Hin YS, Del Campo M, Boiten WA, Stoops E, Vanhooren M, Lemstra AW, et al. Neuroinflammatory CSF biomarkers MIF, sTREM1, and sTREM2 show dynamic expression profiles in alzheimer’s disease. J Neuroinflammation. 2023;20(1):107. [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Zhang S, Zhao J, Zhang Y, Zhang Y, Cai F, Wang L, et al. Upregulation of MIF as a defense mechanism and a biomarker of alzheimer’s disease. Alzheimers Res Ther. 2019;11(1):54. [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Remnestal J, Oijerstedt L, Ullgren A, Olofsson J, Bergstrom S, Kultima K, et al. Altered levels of CSF proteins in patients with FTD, presymptomatic mutation carriers and non-carriers. Transl Neurodegener. 2020;9(1):27. [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Wise A, Li J, Yamakawa M, Loureiro J, Peterson B, Worringer K, et al. CSF proteomics in patients with progressive supranuclear palsy. Neurology. 2024;103(3):e209585. [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Muqaku B, Anderl-Straub S, Werner L, Nagl M, Otto M, Teunissen CE, et al. Contactin proteins in cerebrospinal fluid show different alterations in dementias. J Neurol. 2024;271(12):7516–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
74.Martino Adami PV, Orellana A, Garcia P, Kleineidam L, Alarcon-Martin E, Montrreal L, et al. Matrix metalloproteinase 10 is linked to the risk of progression to dementia of the alzheimer’s type. Brain. 2022;145(7):2507–17. [DOI] [PubMed] [Google Scholar]
75.Abu-Rumeileh S, Steinacker P, Polischi B, Mammana A, Bartoletti-Stella A, Oeckl P, et al. CSF biomarkers of neuroinflammation in distinct forms and subtypes of neurodegenerative dementia. Alzheimers Res Ther. 2019;12(1):2. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Steinacker P, Verde F, Fang L, Feneberg E, Oeckl P, Roeber S, et al. Chitotriosidase (CHIT1) is increased in microglia and macrophages in spinal cord of amyotrophic lateral sclerosis and cerebrospinal fluid levels correlate with disease severity and progression. J Neurol Neurosurg Psychiatry. 2018;89(3):239–47. [DOI] [PubMed] [Google Scholar]
77.van der Ende EL. Veld S, Hanskamp I, van der Lee S, Dijkstra JIR, Hok AHYS, CSF proteomics in autosomal dominant Alzheimer’s disease highlights parallels with sporadic disease. Brain. 2023. [DOI] [PMC free article] [PubMed]
78.Whelan CD, Mattsson N, Nagle MW, Vijayaraghavan S, Hyde C, Janelidze S, et al. Multiplex proteomics identifies novel CSF and plasma biomarkers of early alzheimer’s disease. Acta Neuropathol Commun. 2019;7(1):169. [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Guo Y, Chen SD, You J, Huang SY, Chen YL, Zhang Y et al. Multiplex cerebrospinal fluid proteomics identifies biomarkers for diagnosis and prediction of Alzheimer’s disease. Nat Hum Behav. 2024. [DOI] [PubMed]
80.Bovolenta P, Esteve P, Ruiz JM, Cisneros E, Lopez-Rios J. Beyond Wnt inhibition: new functions of secreted Frizzled-related proteins in development and disease. J Cell Sci. 2008;121(Pt 6):737–46. [DOI] [PubMed] [Google Scholar]
81.Hadi F, Akrami H, Shahpasand K, Fattahi MR. Wnt signalling pathway and Tau phosphorylation: A comprehensive study on known connections. Cell Biochem Funct. 2020;38(6):686–94. [DOI] [PubMed] [Google Scholar]
82.Sogorb-Esteve A, Nilsson J, Swift IJ, Heller C, Bocchetta M, Russell LL, et al. Differential impairment of cerebrospinal fluid synaptic biomarkers in the genetic forms of frontotemporal dementia. Alzheimers Res Ther. 2022;14(1):118. [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Saloner R. Large-scale network analysis of the cerebrospinal fluid proteome identifies molecular signatures of frontotemporal Lobar degeneration. Preprint Biorxv. 2024. [DOI] [PMC free article] [PubMed]
84.Goossens J, Bjerke M, Van Mossevelde S, Van den Bossche T, Goeman J, De Vil B, et al. Diagnostic value of cerebrospinal fluid tau, neurofilament, and progranulin in definite frontotemporal Lobar degeneration. Alzheimers Res Ther. 2018;10(1):31. [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Meda F, Simren J, Borroni B, Cantoni V, Archetti S, Biasiotto G et al. Analytical and clinical validation of a blood progranulin ELISA in frontotemporal dementias. Clin Chem Lab Med. 2023. [DOI] [PMC free article] [PubMed]
86.Davis SE, Cook AK, Hall JA, Voskobiynyk Y, Carullo NV, Boyle NR, et al. Patients with sporadic FTLD exhibit similar increases in lysosomal proteins and storage material as patients with FTD due to GRN mutations. Acta Neuropathol Commun. 2023;11(1):70. [DOI] [PMC free article] [PubMed] [Google Scholar]
87.Mackenzie IR, Baker M, Pickering-Brown S, Hsiung GY, Lindholm C, Dwosh E, et al. The neuropathology of frontotemporal Lobar degeneration caused by mutations in the progranulin gene. Brain. 2006;129(Pt 11):3081–90. [DOI] [PubMed] [Google Scholar]
88.Sogorb-Esteve A, Weiner S, Simrén J, Swift IJ, Bocchetta M, Todd EG, et al. Proteomic analysis reveals distinct cerebrospinal fluid signatures across genetic frontotemporal dementia subtypes. Sci Transl Med. 2025;17(784):eadm9654. [DOI] [PubMed] [Google Scholar]
89.Ising C, Venegas C, Zhang S, Scheiblich H, Schmidt SV, Vieira-Saecker A, et al. NLRP3 inflammasome activation drives Tau pathology. Nature. 2019;575(7784):669–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
90.Lleo A, Irwin DJ, Illan-Gala I, McMillan CT, Wolk DA, Lee EB, et al. A 2-Step cerebrospinal algorithm for the selection of frontotemporal Lobar degeneration subtypes. JAMA Neurol. 2018;75(6):738–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
91.Palluzzi F, Ferrari R, Graziano F, Novelli V, Rossi G, Galimberti D, et al. A novel network analysis approach reveals DNA damage, oxidative stress and calcium/camp homeostasis-associated biomarkers in frontotemporal dementia. PLoS ONE. 2017;12(10):e0185797. [DOI] [PMC free article] [PubMed] [Google Scholar]
92.Malpetti M, Cope TE, Street D, Jones PS, Hezemans FH, Mak E, et al. Microglial activation in the frontal cortex predicts cognitive decline in frontotemporal dementia. Brain. 2023;146(8):3221–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
93.Nuttall RK, Silva C, Hader W, Bar-Or A, Patel KD, Edwards DR, et al. Metalloproteinases are enriched in microglia compared with leukocytes and they regulate cytokine levels in activated microglia. Glia. 2007;55(5):516–26. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1^{(13.1KB, xlsx)}

Supplementary Material 2^{(44.9KB, xlsx)}

Supplementary Material 3^{(46.2KB, xlsx)}

Supplementary Material 4^{(44.9KB, xlsx)}

Supplementary Material 5^{(43.7KB, xlsx)}

Supplementary Material 6^{(12.5KB, xlsx)}

Supplementary Material 7^{(1.9MB, docx)}

Data Availability Statement

[CR1] 1.Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez M, Cappa SF et al. Classification of primary progressive aphasia and its variants. Am Acad Neurol. 2011;76. [DOI] [PMC free article] [PubMed]

[CR2] 2.Rascovsky K, Hodges JR, Knopman D, Mendez MF, Kramer JH, Neuhaus J, et al. Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain. 2011;134(Pt 9):2456–77. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Armstrong MJ, Litvan I, Lang AE, Bak TH, Bhatia KP, Borroni B et al. Criteria for the diagnosis of corticobasal degeneration. Am Acad Neurol. 2013;80. [DOI] [PMC free article] [PubMed]

[CR4] 4.Hoglinger GU, Respondek G, Stamelou M, Kurz C, Josephs KA, Lang AE, et al. Clinical diagnosis of progressive supranuclear palsy: the movement disorder society criteria. Mov Disord. 2017;32(6):853–64. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Irwin DJ, Cairns NJ, Grossman M, McMillan CT, Lee EB, Van Deerlin VM, et al. Frontotemporal lobar degeneration: defining phenotypic diversity through personalized medicine. Acta Neuropathol. 2015;129(4):469–91. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Mann DMA, Snowden JS. Frontotemporal lobar degeneration: pathogenesis, pathology and pathways to phenotype. Brain Pathol. 2017;27(6):723–36. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Neumann M, Rademakers R, Roeber S, Baker M, Kretzschmar HA, Mackenzie IR. A new subtype of frontotemporal lobar degeneration with FUS pathology. Brain. 2009;132(Pt 11):2922–31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Lashley T, Rohrer JD, Mead S, Revesz T. Review: an update on clinical, genetic and pathological aspects of frontotemporal lobar degenerations. Neuropathol Appl Neurobiol. 2015;41(7):858–81. [DOI] [PubMed] [Google Scholar]

[CR9] 9.Rohrer JD, Guerreiro R, Vandrovcova J, Uphill J, Reiman D, Beck J et al. The heritability and genetics of frontotemporal lobar degeneration. Neurology. 2009;73. [DOI] [PMC free article] [PubMed]

[CR10] 10.Greaves CV, Rohrer JD. An update on genetic frontotemporal dementia. J Neurol. 2019;266(8):2075–86. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Pozzi FE, Aprea V, Giovannelli G, Lattuada F, Crivellaro C, Bertola F, et al. Clinical and neuroimaging characterization of the first frontotemporal dementia family carrying the MAPT p.K298E mutation. Neurogenetics. 2024;25(3):215–23. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Cheng HR, Lin RR, Li HL, Xue YY, Gao PR, Chen DF, et al. Identification and functional characterization of novel variants of MAPT and GRN in Chinese patients with frontotemporal dementia. Neurobiol Aging. 2023;123:233–43. [DOI] [PubMed] [Google Scholar]

[CR13] 13.Neumann M, Lee EB, Mackenzie IR. Frontotemporal lobar degeneration TDP-43-immunoreactive pathological subtypes: clinical and mechanistic significance. Adv Exp Med Biol. 2021;1281:201–17. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Del Campo M, Zetterberg H, Gandy S, Onyike CU, Oliveira F, Udeh-Momoh C, et al. New developments of biofluid-based biomarkers for routine diagnosis and disease trajectories in frontotemporal dementia. Alzheimers Dement. 2022;18(11):2292–307. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Hutchinson AD, Mathias JL. Neuropsychological deficits in frontotemporal dementia and alzheimer’s disease: a meta-analytic review. J Neurol Neurosurg Psychiatry. 2007;78(9):917–28. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Naasan G, Rabinovici GD, Ghosh P, Elofson JD, Miller BL, Coppola G, et al. Amyloid in dementia associated with Familial FTLD: not an innocent bystander. Neurocase. 2016;22(1):76–83. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Irwin DJ, Trojanowski JQ, Grossman M. Cerebrospinal fluid biomarkers for differentiation of frontotemporal lobar degeneration from alzheimer’s disease. Front Aging Neurosci. 2013;5:6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Ashton NJ, Janelidze S, Al Khleifat A, Leuzy A, van der Ende EL, Karikari TK, et al. A multicentre validation study of the diagnostic value of plasma neurofilament light. Nat Commun. 2021;12(1):3400. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Bridel C, van Wieringen WN, Zetterberg H, Tijms BM, Teunissen CE, and the, NFLG et al. Diagnostic value of cerebrospinal fluid neurofilament light protein in neurology: a systematic review and meta-analysis. JAMA Neurol. 2019;76(9):10354–8. [DOI] [PMC free article] [PubMed]

[CR20] 20.Ducharme S, Dols A, Laforce R, Devenney E, Kumfor F, van den Stock J, et al. Recommendations to distinguish behavioural variant frontotemporal dementia from psychiatric disorders. Brain. 2020;143(6):1632–50. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Gendron TF, Heckman MG, White LJ, Veire AM, Pedraza O, Burch AR, et al. Comprehensive cross-sectional and longitudinal analyses of plasma neurofilament light across FTD spectrum disorders. Cell Rep Med. 2022;3(4):100607. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Meeter LHH, Vijverberg EG, Del Campo M, Rozemuller AJM, Donker Kaat L, de Jong FJ, et al. Clinical value of neurofilament and phospho-tau/tau ratio in the frontotemporal dementia spectrum. Neurology. 2018;90(14):e1231–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Pijnenburg YA, Verwey NA, van der Flier WM, Scheltens P, Teunissen CE. Discriminative and prognostic potential of cerebrospinal fluid phosphotau/tau ratio and neurofilaments for frontotemporal dementia subtypes. Alzheimers Dement (Amst). 2015;1(4):505–12. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Cousins KAQ, Shaw LM, Chen-Plotkin A, Wolk DA, Van Deerlin VM, Lee EB, et al. Distinguishing frontotemporal lobar degeneration Tau from TDP-43 using plasma biomarkers. JAMA Neurol. 2022;79(11):1155–64. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Desmarais P, Rohrer JD, Nguyen QD, Herrmann N, Stuss DT, Lang AE, et al. Therapeutic trial design for frontotemporal dementia and related disorders. J Neurol Neurosurg Psychiatry. 2019;90(4):412–23. [DOI] [PubMed] [Google Scholar]

[CR26] 26.Higginbotham L, Ping L, Dammer EB, Duong DM, Zhou M, Gearing M, et al. Integrated proteomics reveals brain-based cerebrospinal fluid biomarkers in asymptomatic and symptomatic alzheimer’s disease. Sci Adv. 2020;6:43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Bader JM, Geyer PE, Muller JB, Strauss MT, Koch M, Leypoldt F, et al. Proteome profiling in cerebrospinal fluid reveals novel biomarkers of alzheimer’s disease. Mol Syst Biol. 2020;16(6):e9356. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR28] 28.Dammer EB, Ping L, Duong DM, Modeste ES, Seyfried NT, Lah JJ, et al. Multi-platform proteomic analysis of alzheimer’s disease cerebrospinal fluid and plasma reveals network biomarkers associated with proteostasis and the matrisome. Alzheimers Res Ther. 2022;14(1):174. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Del Campo M, Peeters CFW, Johnson ECB, Vermunt L, Hok-A-Hin YS, van Nee M, et al. CSF proteome profiling across the alzheimer’s disease spectrum reflects the multifactorial nature of the disease and identifies specific biomarker panels. Nat Aging. 2022;2(11):1040–53. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Teunissen CE, Elias N, Koel-Simmelink MJ, Durieux-Lu S, Malekzadeh A, Pham TV, et al. Novel diagnostic cerebrospinal fluid biomarkers for pathologic subtypes of frontotemporal dementia identified by proteomics. Alzheimers Dement (Amst). 2016;2:86–94. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.van der Ende EL, Meeter LH, Stingl C, van Rooij JGJ, Stoop MP, Nijholt DAT, et al. Novel CSF biomarkers in genetic frontotemporal dementia identified by proteomics. Ann Clin Transl Neurol. 2019;6(4):698–707. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Del Campo M, Galimberti D, Elias N, Boonkamp L, Pijnenburg YA, van Swieten JC, et al. Novel CSF biomarkers to discriminate FTLD and its pathological subtypes. Ann Clin Transl Neurol. 2018;5(10):1163–75. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.van der Ende EL, Xiao M, Xu D, Poos JM, Panman JL, Jiskoot LC, et al. Neuronal pentraxin 2: a synapse-derived CSF biomarker in genetic frontotemporal dementia. J Neurol Neurosurg Psychiatry. 2020;91(6):612–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Bolsewig K, Hok AHYS, Sepe FN, Boonkamp L, Jacobs D, Bellomo G, et al. A combination of neurofilament light, glial fibrillary acidic protein, and neuronal Pentraxin-2 discriminates between frontotemporal dementia and other dementias. J Alzheimers Dis. 2022;90(1):363–80. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Llorens F, Thune K, Tahir W, Kanata E, Diaz-Lucena D, Xanthopoulos K, et al. YKL-40 in the brain and cerebrospinal fluid of neurodegenerative dementias. Mol Neurodegener. 2017;12(1):83. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Del Campo M, Vermunt L, Peeters CFW, Sieben A, Hok AHYS, Lleo A, et al. CSF proteome profiling reveals biomarkers to discriminate dementia with lewy bodies from alzheimer s disease. Nat Commun. 2023;14(1):5635. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.van der Flier WM, Scheltens P. Amsterdam dementia cohort: performing research to optimize care. J Alzheimers Dis. 2018;62(3):1091–111. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Ingelsson M, Ramasamy K, Russ C, Freeman SH, Orne J, Raju S, et al. Increase in the relative expression of Tau with four microtubule binding repeat regions in frontotemporal Lobar degeneration and progressive supranuclear palsy brains. Acta Neuropathol. 2007;114(5):471–9. [DOI] [PubMed] [Google Scholar]

[CR39] 39.Alcolea D, Clarimon J, Carmona-Iragui M, Illan-Gala I, Morenas-Rodriguez E, Barroeta I, et al. The Sant Pau initiative on neurodegeneration (SPIN) cohort: A data set for biomarker discovery and validation in neurodegenerative disorders. Alzheimers Dement (N Y). 2019;5:597–609. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-Gateau P, Cummings J, et al. Research criteria for the diagnosis of alzheimer’s disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 2007;6(8):734–46. [DOI] [PubMed] [Google Scholar]

[CR41] 41.Cairns NJ, Bigio EH, Mackenzie IR, Neumann M, Lee VM, Hatanpaa KJ, et al. Neuropathologic diagnostic and nosologic criteria for frontotemporal lobar degeneration: consensus of the consortium for frontotemporal lobar degeneration. Acta Neuropathol. 2007;114(1):5–22. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Montine TJ, Phelps CH, Beach TG, Bigio EH, Cairns NJ, Dickson DW, et al. National institute on aging-alzheimer’s association guidelines for the neuropathologic assessment of alzheimer’s disease: a practical approach. Acta Neuropathol. 2012;123(1):1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Miyagawa T, Brushaber D, Syrjanen J, Kremers W, Fields J, Forsberg LK, et al. Utility of the global CDR plus NACC FTLD rating and development of scoring rules: data from the ARTFL/LEFFTDS consortium. Alzheimers Dement. 2020;16(1):106–17. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Hok-A-Hin YS, Willemse EAJ, Teunissen CE, Del Campo M. Guidelines for CSF processing and biobanking: impact on the identification and development of optimal CSF protein biomarkers. Methods Mol Biol. 2019;2044:27–50. [DOI] [PubMed] [Google Scholar]

[CR45] 45.Tijms BM, Willemse EAJ, Zwan MD, Mulder SD, Visser PJ, van Berckel BNM, et al. Unbiased approach to counteract upward drift in cerebrospinal fluid Amyloid-beta 1–42 analysis results. Clin Chem. 2018;64(3):576–85. [DOI] [PubMed] [Google Scholar]

[CR46] 46.Vermunt L, Otte M, Verberk IMW, Killestein J, Lemstra AW, van der Flier WM, et al. Age- and disease-specific reference values for neurofilament light presented in an online interactive support interface. Ann Clin Transl Neurol. 2022;9(11):1832–7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Proteomics O. Development and validation of customized PEA biomarker panels with clinical utility.

[CR48] 48.Benjamini Y, Hochberg Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J Royal Stat Soc Ser B (Methodological). 1995;57.

[CR49] 49.de Leeuw FA, Peeters CFW, Kester MI, Harms AC, Struys EA, Hankemeier T, et al. Blood-based metabolic signatures in alzheimer’s disease. Alzheimers Dement (Amst). 2017;8:196–207. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Friedman J, Hastie T, Tibshirani R. Regularization paths for generalized linear models via coordinate descent. J Stat Softw. 2010;33(1):22. [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Zhou Y, Zhou B, Pache L, Chang M, Khodabakhshi AH, Tanaseichuk O, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Karlsson L, Vogel J, Arvidsson I, Åström K, Janelidze S, Blennow K et al. Cerebrospinal fluid reference proteins increase accuracy and interpretability of biomarkers for brain diseases. BioRxiv. 2023. [DOI] [PMC free article] [PubMed]

[CR53] 53.Poe M, Blake JT, Boulton DA, Gammon M, Sigal NH, Wu JK, et al. Human cytotoxic lymphocyte granzyme B. Its purification from granules and the characterization of substrate and inhibitor specificity. J Biol Chem. 1991;266(1):98–103. [PubMed] [Google Scholar]

[CR54] 54.Mroczko B, Groblewska M, Barcikowska M. The role of matrix metalloproteinases and tissue inhibitors of metalloproteinases in the pathophysiology of neurodegeneration: a literature study. J Alzheimers Dis. 2013;37(2):273–83. [DOI] [PubMed] [Google Scholar]

[CR55] 55.Spencer ML, Theodosiou M, Noonan DJ. NPDC-1, a novel regulator of neuronal proliferation, is degraded by the ubiquitin/proteasome system through a PEST degradation motif. J Biol Chem. 2004;279(35):37069–78. [DOI] [PubMed] [Google Scholar]

[CR56] 56.Pei DS, Jia PP, Luo JJ, Liu W, Strauss PR. AP endonuclease 1 (Apex1) influences brain development linking oxidative stress and DNA repair. Cell Death Dis. 2019;10(5):348. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L. VEGF receptor signalling - in control of vascular function. Nat Rev Mol Cell Biol. 2006;7(5):359–71. [DOI] [PubMed] [Google Scholar]

[CR58] 58.Huang X, Feng Z, Jiang Y, Li J, Xiang Q, Guo S et al. VSIG4 mediates transcriptional Inhibition of Nlrp3 and Il-1b in macrophages. Sci Adv. 2019;5. [DOI] [PMC free article] [PubMed]

[CR59] 59.Zhang JQ, Nicoll G, Jones C, Crocker PR. Siglec-9, a novel Sialic acid binding member of the Immunoglobulin superfamily expressed broadly on human blood leukocytes. J Biol Chem. 2000;275(29):22121–6. [DOI] [PubMed] [Google Scholar]

[CR60] 60.Yan Q, Malashkevich VN, Fedorov A, Fedorov E, Cao E, Lary JW, et al. Structure of CD84 provides insight into SLAM family function. PNAS. 2007;104:25. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Tanaka Y, Suzuki G, Matsuwaki T, Hosokawa M, Serrano G, Beach TG, et al. Progranulin regulates lysosomal function and biogenesis through acidification of lysosomes. Hum Mol Genet. 2017;26(5):969–88. [DOI] [PubMed] [Google Scholar]

[CR62] 62.Kishore U, Gaboriaud C, Waters P, Shrive AK, Greenhough TJ, Reid KB, et al. C1q and tumor necrosis factor superfamily: modularity and versatility. Trends Immunol. 2004;25(10):551–61. [DOI] [PubMed] [Google Scholar]

[CR63] 63.Goel M, Sienkiewicz AE, Picciani R, Lee RK, Bhattacharya SK. Cochlin induced TREK-1 co-expression and Annexin A2 secretion: role in trabecular meshwork cell elongation and motility. PLoS ONE. 2011;6(8):e23070. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Bostrom G, Freyhult E, Virhammar J, Alcolea D, Tumani H, Otto M, et al. Different inflammatory signatures in alzheimer’s disease and frontotemporal dementia cerebrospinal fluid. J Alzheimers Dis. 2021;81(2):629–40. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Mol MO, Miedema SSM, Melhem S, Li KW, Koopmans F, Seelaar H, et al. Proteomics of the dentate gyrus reveals semantic dementia specific molecular pathology. Acta Neuropathol Commun. 2022;10(1):190. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR66] 66.Miedema SSM, Mol MO, Koopmans FTW, Hondius DC, van Nierop P, Menden K, et al. Distinct cell type-specific protein signatures in GRN and MAPT genetic subtypes of frontotemporal dementia. Acta Neuropathol Commun. 2022;10(1):100. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Umoh ME, Dammer EB, Dai J, Duong DM, Lah JJ, Levey AI, et al. A proteomic network approach across the ALS-FTD disease spectrum resolves clinical phenotypes and genetic vulnerability in human brain. EMBO Mol Med. 2018;10(1):48–62. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Mofrad RB, Del Campo M, Peeters CFW, Meeter LHH, Seelaar H, Koel-Simmelink M, et al. Plasma proteome profiling identifies changes associated to AD but not to FTD. Acta Neuropathol Commun. 2022;10(1):148. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR69] 69.Hok-A-Hin YS, Del Campo M, Boiten WA, Stoops E, Vanhooren M, Lemstra AW, et al. Neuroinflammatory CSF biomarkers MIF, sTREM1, and sTREM2 show dynamic expression profiles in alzheimer’s disease. J Neuroinflammation. 2023;20(1):107. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Zhang S, Zhao J, Zhang Y, Zhang Y, Cai F, Wang L, et al. Upregulation of MIF as a defense mechanism and a biomarker of alzheimer’s disease. Alzheimers Res Ther. 2019;11(1):54. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Remnestal J, Oijerstedt L, Ullgren A, Olofsson J, Bergstrom S, Kultima K, et al. Altered levels of CSF proteins in patients with FTD, presymptomatic mutation carriers and non-carriers. Transl Neurodegener. 2020;9(1):27. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR72] 72.Wise A, Li J, Yamakawa M, Loureiro J, Peterson B, Worringer K, et al. CSF proteomics in patients with progressive supranuclear palsy. Neurology. 2024;103(3):e209585. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR73] 73.Muqaku B, Anderl-Straub S, Werner L, Nagl M, Otto M, Teunissen CE, et al. Contactin proteins in cerebrospinal fluid show different alterations in dementias. J Neurol. 2024;271(12):7516–24. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR74] 74.Martino Adami PV, Orellana A, Garcia P, Kleineidam L, Alarcon-Martin E, Montrreal L, et al. Matrix metalloproteinase 10 is linked to the risk of progression to dementia of the alzheimer’s type. Brain. 2022;145(7):2507–17. [DOI] [PubMed] [Google Scholar]

[CR75] 75.Abu-Rumeileh S, Steinacker P, Polischi B, Mammana A, Bartoletti-Stella A, Oeckl P, et al. CSF biomarkers of neuroinflammation in distinct forms and subtypes of neurodegenerative dementia. Alzheimers Res Ther. 2019;12(1):2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR76] 76.Steinacker P, Verde F, Fang L, Feneberg E, Oeckl P, Roeber S, et al. Chitotriosidase (CHIT1) is increased in microglia and macrophages in spinal cord of amyotrophic lateral sclerosis and cerebrospinal fluid levels correlate with disease severity and progression. J Neurol Neurosurg Psychiatry. 2018;89(3):239–47. [DOI] [PubMed] [Google Scholar]

[CR77] 77.van der Ende EL. Veld S, Hanskamp I, van der Lee S, Dijkstra JIR, Hok AHYS, CSF proteomics in autosomal dominant Alzheimer’s disease highlights parallels with sporadic disease. Brain. 2023. [DOI] [PMC free article] [PubMed]

[CR78] 78.Whelan CD, Mattsson N, Nagle MW, Vijayaraghavan S, Hyde C, Janelidze S, et al. Multiplex proteomics identifies novel CSF and plasma biomarkers of early alzheimer’s disease. Acta Neuropathol Commun. 2019;7(1):169. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR79] 79.Guo Y, Chen SD, You J, Huang SY, Chen YL, Zhang Y et al. Multiplex cerebrospinal fluid proteomics identifies biomarkers for diagnosis and prediction of Alzheimer’s disease. Nat Hum Behav. 2024. [DOI] [PubMed]

[CR80] 80.Bovolenta P, Esteve P, Ruiz JM, Cisneros E, Lopez-Rios J. Beyond Wnt inhibition: new functions of secreted Frizzled-related proteins in development and disease. J Cell Sci. 2008;121(Pt 6):737–46. [DOI] [PubMed] [Google Scholar]

[CR81] 81.Hadi F, Akrami H, Shahpasand K, Fattahi MR. Wnt signalling pathway and Tau phosphorylation: A comprehensive study on known connections. Cell Biochem Funct. 2020;38(6):686–94. [DOI] [PubMed] [Google Scholar]

[CR82] 82.Sogorb-Esteve A, Nilsson J, Swift IJ, Heller C, Bocchetta M, Russell LL, et al. Differential impairment of cerebrospinal fluid synaptic biomarkers in the genetic forms of frontotemporal dementia. Alzheimers Res Ther. 2022;14(1):118. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR83] 83.Saloner R. Large-scale network analysis of the cerebrospinal fluid proteome identifies molecular signatures of frontotemporal Lobar degeneration. Preprint Biorxv. 2024. [DOI] [PMC free article] [PubMed]

[CR84] 84.Goossens J, Bjerke M, Van Mossevelde S, Van den Bossche T, Goeman J, De Vil B, et al. Diagnostic value of cerebrospinal fluid tau, neurofilament, and progranulin in definite frontotemporal Lobar degeneration. Alzheimers Res Ther. 2018;10(1):31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR85] 85.Meda F, Simren J, Borroni B, Cantoni V, Archetti S, Biasiotto G et al. Analytical and clinical validation of a blood progranulin ELISA in frontotemporal dementias. Clin Chem Lab Med. 2023. [DOI] [PMC free article] [PubMed]

[CR86] 86.Davis SE, Cook AK, Hall JA, Voskobiynyk Y, Carullo NV, Boyle NR, et al. Patients with sporadic FTLD exhibit similar increases in lysosomal proteins and storage material as patients with FTD due to GRN mutations. Acta Neuropathol Commun. 2023;11(1):70. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR87] 87.Mackenzie IR, Baker M, Pickering-Brown S, Hsiung GY, Lindholm C, Dwosh E, et al. The neuropathology of frontotemporal Lobar degeneration caused by mutations in the progranulin gene. Brain. 2006;129(Pt 11):3081–90. [DOI] [PubMed] [Google Scholar]

[CR88] 88.Sogorb-Esteve A, Weiner S, Simrén J, Swift IJ, Bocchetta M, Todd EG, et al. Proteomic analysis reveals distinct cerebrospinal fluid signatures across genetic frontotemporal dementia subtypes. Sci Transl Med. 2025;17(784):eadm9654. [DOI] [PubMed] [Google Scholar]

[CR89] 89.Ising C, Venegas C, Zhang S, Scheiblich H, Schmidt SV, Vieira-Saecker A, et al. NLRP3 inflammasome activation drives Tau pathology. Nature. 2019;575(7784):669–73. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR90] 90.Lleo A, Irwin DJ, Illan-Gala I, McMillan CT, Wolk DA, Lee EB, et al. A 2-Step cerebrospinal algorithm for the selection of frontotemporal Lobar degeneration subtypes. JAMA Neurol. 2018;75(6):738–45. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR91] 91.Palluzzi F, Ferrari R, Graziano F, Novelli V, Rossi G, Galimberti D, et al. A novel network analysis approach reveals DNA damage, oxidative stress and calcium/camp homeostasis-associated biomarkers in frontotemporal dementia. PLoS ONE. 2017;12(10):e0185797. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR92] 92.Malpetti M, Cope TE, Street D, Jones PS, Hezemans FH, Mak E, et al. Microglial activation in the frontal cortex predicts cognitive decline in frontotemporal dementia. Brain. 2023;146(8):3221–31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR93] 93.Nuttall RK, Silva C, Hader W, Bar-Or A, Patel KD, Edwards DR, et al. Metalloproteinases are enriched in microglia compared with leukocytes and they regulate cytokine levels in activated microglia. Glia. 2007;55(5):516–26. [DOI] [PubMed] [Google Scholar]

PERMALINK

Large-scale CSF proteome profiling identifies biomarkers for accurate diagnosis of frontotemporal dementia

Yanaika S Hok-A-Hin

Lisa Vermunt

Carel FW Peeters

Emma L van der Ende

Sterre CM de Boer

Lieke H Meeter

Julie de Houwer

Harro Seelaar

John C van Swieten

William T Hu

Alberto Lleó

Daniel Alcolea

Sebastiaan Engelborghs

Anne Sieben

Alice Chen-Plotkin

David J Irwin

Wiesje M van der Flier

Yolande AL Pijnenburg

Charlotte E Teunissen

Marta del Campo

Abstract

Background

Methods

Results

Conclusions

Supplementary Information

Background

Method

Participants

Table 1.

CSF protein profiling

Development of custom PEA assays

Statistical analyses

Results

Subjects

Fig. 1.

CSF proteins are differentially regulated in FTD compared to controls and AD

CSF protein panels discriminate with high accuracy FTD from controls and AD

Fig. 2.

CSF proteins are differentially regulated in FTLD-Tau or FTLD-TDP subtypes

Fig. 3.

CSF protein panel to discriminate the FTLD subtypes

Validation of FTD panels using custom PEA assays in independent cohorts

Fig. 4.

Fig. 5.

Discussion

Conclusions

Supplementary Information

Acknowledgements

Abbreviations

Author contributions

Funding

Data availability

Declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Footnotes

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases