Skip to main content
Download PDF

Investigating the association between the GAP-43 concentration with diffusion tensor imaging indices in Alzheimer’s dementia continuum

BMC Neurology volume 24, Article number: 397 (2024) Cite this article

Abstract

Background

Synaptic degeneration, axonal injury, and white matter disintegration are among the pathological events in Alzheimer’s disease (AD), for which growth-associated protein 43 (GAP-43) and diffusion tensor imaging (DTI) could be an indicator. In this study, the cerebrospinal fluid (CSF) GAP-43 clinical trajectories and their association with progression and AD hallmarks with white matter microstructural changes were evaluated.

Methods

A total number of 133 participants were enrolled in GAP-43 and DTI values were compared between groups, both cross-sectionally and longitudinally with two and four-year follow-ups. Subsequently, the correlation between GAP-43 levels in the CSF and DTI values was investigated using Spearman’s correlation.

Results

The CSF level of GAP-43 is negatively correlated with the mean diffusivity measures in Fornix (Cres)/Stria terminals in early and late MCI (rs=-0.478 p = 0.021 and rs=-0.425 p = 0.038). Additionally, the CSF level of GAP-43 is negatively correlated with fractional anisotropy in the cingulum in late MCI (rs=-0.437 p = 0.033). Moreover, the axial diffusivity in superior corona radiate (rs=-0.562 p = 0.005 and rs=-0.484 p = 0.036) and radial diffusivity in superior fronto-occipital fasciculus was negatively correlated with GAP-43 level in the early and mid-MCI participants (rs=-0.520 p = 0.011 and rs=-0.498 p = 0.030).

Conclusions

Presynaptic marker GAP-43 in combination with DTI can be used as a novel biomarker to identify microstructural synaptic degeneration in the early MCI. In addition, it can be used as a biomarker for tracking the progression of AD and monitoring treatment efficacy.

Peer Review reports

Introduction

Worldwide, the incidence of dementia is increasing due to the increasing number of elderly. It is estimated that the number of people with dementia will grow 2.6 times between 2019 and 2050 [1]. Alzheimer’s disease (AD) is the most common type of dementia, frequently affecting elderly people more than 65 years old throughout the world [2, 3]. Currently, there are no emerging treatments or approved drugs to reverse, slow, or stop the progressive disease process [4,5,6,7]. Mild cognitive impairment (MCI) is an intermediate state between dementia and normal cognitive aging which is associated with the long-lasting pathological deposition of misfolded proteins and synaptic dysfunction in the brain. One of the greatest obstacles researchers face is MCI classification [8, 9]. Although some MCI patients appear to remain stable or even improve over time, the majority of them change to AD dementia within five years [10]. MCI individuals who develop AD dementia are classified as having progressive MCI, and those who retain their MCI state are classified as having stable MCI [11].

Cognitive impairment and behavioral changes consequence from neurodegeneration are commonly observed about 10 years after the onset of proteinopathy. The presence of amyloid-β plaques (Aβ) and hyperphosphorylation of self-aggregating tau, which results in neurofibrillary tangles (NFTs), have effects on disease pathologic progression and serve as biomarkers [8, 12]. Studies have shown that AD-related pathological changes begin long before clinical symptoms appear. It is widely accepted that synaptic loss and dysfunction play a pivotal role in AD and are associated with cognitive decline [13, 14]. Thus, additional plasma biomarkers reflecting synaptic loss and dysfunction could further enhance differential diagnosis and are useful as outcome markers in clinical trials, providing new information about the stage of the disease, and preventative assessments.

Growth-associated protein 43 (GAP-43) has been known as a phosphoprotein of cerebrospinal fluid (CSF) that is encoded via the GAP-43 gene located in the nervous system [15, 16]. GAP-43 is essential for synaptic plasticity, reactive synaptogenesis formation, regulation of axonal outgrowth, and memory and learning functions [17]. The expression of presynaptic protein GAP-43 highly increases during synaptogenesis and neuronal development and after that in the hippocampus and associated cortex of the human brain [18]. Moderate GAP-43 expression can enhance memory, while overexpression of GAP-43 may cause the dysfunction of neuronal synapses and culminate in loss of memory [1, 19]. CSF GAP-43 levels have seldom been researched in AD patients [10, 20]. Current evidence suggests a substantial increase in the CSF level of GAP-43 in the early stages of AD [21].

Diffusion Tensor Imaging (DTI) is an MRI-based modality that allows the decoding of the water molecule’s mobility in vivo and provides information about the white matter (WM) changes, synaptic damage, and neurodegeneration in AD patients [22,23,24].

Emerging studies point to modifications of WM as a biological marker for pathological importance, which can be a potential target for early-stage dementia [25, 26]. Early accumulation of CSF and plasma p-tau has been reported in the corpus callosum of a brain with Alzheimer’s pathology, which is also associated with a decline in WM integrity [27]. There have been few studies on whether presynaptic CSF GAP-43 is associated with WM microstructure in MCI subjects, owing to the pivotal biomarker role of GAP-43 for AD [28]. In the current study, we carried out an evaluation to elucidate the relationship between GAP-43 levels and DTI-detected changes of WM in patients with MCI. Since GAP-43 is recently proposed to have potential biomarker utility in the AD continuum [14, 29], we aimed to investigate where this protein can indicate the microstructural changes in the brain, especially in the early stages of mild cognitive impairment.

Materials and methods

Participants and data

Data used in this research was extracted from a longitudinal multicenter study named Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). Since the time the project was initiated the ADNI3 content was not fully completed, the participant data were extracted from the ADNI1, ADNI-GO, and ADNI2 projects. Herein we evaluate a total number of 1269 GAP-43 samples from participants who took lumbar puncture. In addition, a total number of 977 DTI samples were extracted from the database. To perform a comparison between the GAP-43 and DTI values following screening was taken. After a glance at the data, 464 duplicated values for the ID of patients were detected; therefore 804 unique IDs were isolated. To compare DTI data with gap-43 data, datasets were combined based on ID and the year of the sample taken, therefore, 133 values remained. Moreover, to compare the amount of GAP-43 and DTI in different years, IDs containing repetitive GAP-43 values were analyzed separately (Fig. 1).

Fig. 1
figure 1

Patient selection flow chart. Patient data were extracted from ADNI-1, ADNI-GO, and ADNI-2 datasets. Groups indicated by red color were excluded from the analysis

Full size image

CSF GAP-43 measurements

To measure the GAP-43 level in the CSF, in-house ELISA techniques were implemented. The performing procedure is briefly described as follows: admixture of the anti-GAP-43 antibody of the mouse (0.77 µg/ml NM4) with Nunc-Immuno Polysorp microwell modules (Thermo Fisher Scientific, Massachusetts, US) in the pH of 9.6 provided by carbonate buffer in 4 °C temperatures. Moreover, in the measurement procedure, three containers were labeled as CSF, blank, and control samples. Subsequently, the samples were colorized with 3, 3´, 5, 5´-tetramethylbenzidine (TMB, KemEnTech Diagnostics). The color density at 450 nm was measured through a SunriseTM microplate absorbance reader (Tecan Group, Männedorf, Switzerland) by considering a reference value of 650 nm. The C-terminal of GAP-43 was a cue for two mouse monoclonal and polyclonal GAP-43 antibodies; the first was NM4 as a coating antibody, and the second was a polyclonal GAP-43 antibody as a detector antibody. Finally, all laboratory data (1268 samples) were analyzed to determine the exact concentration of GAP-43 indicated by pg/mg in the range of 312 − 20.000 pg/mL. Quality control samples named QC1 and QC2 were taken from patients’ CSF in collaboration with the Clinical Neurochemistry Laboratory, Sahlgrenska University Hospital, and Mölndal, Sweden. Detailed information was available in the previously published article [14].

Imaging data acquisition

The detailed data regarding the DTI protocol were previously mentioned in the Nir et al. 2013 study [30]. Whole-brain MRI scanning on 3 Tesla GE Medical Systems scanners was taken from participants and DTI values of participants were derived from raw MRI data according to the ADNI protocol. Besides, to assess the quality of MRI images in finding possible artifacts, visualization was performed. The image acquisition properties for Anatomical T1-weighted SPGR (spoiled gradient echo) sequences were as follows. The flip angle was selected as 11° in a 256 × 256 matrix. The voxel size was selected with the values of 1.0 × 1.0 × 1.2 mm3. Moreover, three values for Echo time (TE), Repetition Time (TR), and Inversion time (TI) were selected as 2.85 ms, 6.98 ms, and 400 ms, respectively. Besides, the value for diffusion-weighted images (DWI) was as follows: A matrix size of 256 × 256, total scan time of 9 min, 9000 ms for TR, and voxel size of 2.7 × 2.7 × 2.7 mm3. To achieve an improved signal-to-noise ratio, for every DTI scan, 46 images were acquired: 5 T2-weighted images without diffusion sensitization (b0 images), and 41 diffusion-weighted images for which the b-value was set at 1000 s/mm². To ensure quality assurance, all T1-weighted MR, FLAIR, and DWI images underwent a visual review to identify and exclude those with substantial motion, artifacts, or high white matter hyperintensities (± 2.5 standard deviations from the mean in the AD group).

Preprocessing steps

Head motion and eddy current distortions were then corrected for each subject’s raw DWI volumes by aligning them to the mean b0 image, the volume without diffusion sensitization, using the FSL tool eddy_correct (www.fmrib.ox.ac.uk/fsl). In addition, in the T1-weighted anatomical scans, extra-cerebral tissue was roughly removed by employing a range of software packages, with significant processing reliance on ROBEX-an automated program that has been previously trained on manually “skull-stripped” MRI data [31] and FreeSurfer [32]. The skull-stripped volumes were visually assessed, and the most appropriate one was chosen, occasionally undergoing additional manual editing. Anatomical scans were then subjected to intensity inhomogeneity normalization via the MNI nu_correct tool (www.bic.mni.mcgill.ca/software/). In addition, non-brain tissue was removed from the DWI through the application of the Brain Extraction Tool (BET) from FSL [33]. To achieve alignment of data from different subjects within the same 3D coordinate system, each T1-weighted anatomical image was subjected to linear alignment with a standard brain template, (downsampled Colin27), using FSL flirt [34] with six degrees of freedom (dof) to permit translations and rotations in three dimensions. In order to mitigate susceptibility artifacts resulting from echo-planar imaging (EPI), which can produce distortions at tissue-fluid boundaries, skull-stripped b0 images were linearly aligned (FSL flirt with 9 dof) and then elastically registered to their respective T1-weighted structural images via an inverse-consistent registration algorithm that utilized a mutual information cost function [35]. Following the generation of the 3D deformation fields, these were applied to the other 41 DWI volumes prior to the assessment of diffusion parameters. To ensure proper alignment of the DWI images with the structural T1-weighted scan, a corrected gradient table was computed.

DTI maps

Utilizing FSL dtifit, a single diffusion tensor or ellipsoid was constructed at each voxel in the brain from the DWI scans that had undergone eddy and EPI correction. The eigenvalues of the diffusion tensor (λ1, λ2, λ3) were then used to generate scalar maps of anisotropy and diffusivity, indicating the lengths of the ellipsoid’s longest, middle, and shortest axes. Fractional anisotropy (FA) values were obtained from the following formula:

$$\:FA=\sqrt{\frac{3}{2}\:}\frac{\sqrt{{({{\uplambda\:}}_{1}-<{\uplambda\:}>)}^{2}+{({{\uplambda\:}}_{2}-<{\uplambda\:}>)}^{2}+{({{\uplambda\:}}_{3}-<{\uplambda\:}>)}^{2}\:}}{\sqrt{{{\uplambda\:}}_{1}^{2}+{{\uplambda\:}}_{2}^{2}+{{\uplambda\:}}_{3}^{2}}}\in\:\left[01\right]$$
$$\:<{\uplambda\:}>=\frac{{{\uplambda\:}}_{1}+\:{{\uplambda\:}}_{2}+{{\uplambda\:}}_{3}}{3}$$

In the mentioned formula < λ > was defined as mean diffusivity (MD). In addition, Axial diffusivity (AxD) was equal to λ1 (largest eigenvalue). Radial diffusivity (RD) values were defined through the following formula:

$$\:RD=\frac{{{\uplambda\:}}_{2}+{{\uplambda\:}}_{3}}{2}$$

White matter tract atlas and ROI summary measures

The FA images of the Johns Hopkins University (JHU) DTI atlas (ICBM-DTI-81) were used as the template to register the FA image of each subject [36]. In this regard, the following regions of interest (ROIs) were evaluated: Genu of the corpus callosum, Posterior thalamic radiation, Body of corpus callosum, Sagittal stratum, Splenium of the corpus callosum, External capsule, Cingulum (cingulate gyrus), Corticospinal tract, Cingulum (hippocampus), Cerebral peduncle, Fornix (crus)/Stria terminalis, Anterior limb of internal capsule, Superior longitudinal fasciculus, Posterior limb of internal capsule, Superior fronto-occipital fasciculus, Retrolenticular part of internal capsule, Inferior fronto-occipital fasciculus, Anterior corona radiata, Uncinate fasciculus, Superior corona radiata, Tapetum, and Posterior corona radiata. Beside JHU labels, additional ROIs including the entire corpus callosum, bilateral genu, body, and splenium of the corpus callosum were assessed.

TBSS tract atlas

The implementation of tract-based spatial statistics (TBSS) [37], was carried out using the FSL software package (http://www.fmrib.ox.ac.uk/fsl/), in accordance with the protocols delineated by the ENIGMA-DTI group. The corrected fractional anisotropy (FA) maps for each subject were first linearly registered and then elastically registered to the ENIGMA-DTI template in ICBM space. The resulting 3D deformation fields were applied to the three diffusivity maps. Subsequently, the spatially normalized data for FA, mean diffusivity (MD), radial diffusivity (RD), and axial diffusivity (AxD) from all subjects were projected onto the skeletonized ENIGMA-DTI template. The tissue-specific, smoothing-compensated method (T-SPOON) was employed to enhance tissue specificity and mitigate confounding factors arising from morphometric variations that remain inadequately addressed by elastic registration [38].

Statistical analysis

A one-way ANOVA and Chi-square tests were applied to the demographic data of the participants. Data normality was assessed using the Kolmogorov-Smirnov and Shapiro-Wilk tests. To assess the difference between the baseline values of GAP-43 in five groups, Kruskal-Wallis H was performed. To compare the value of the GAP-43 between three groups of normal, early and late MCI groups with 2 years and 4 years of follow-up, data were log-transformed and analyzed through repeated measure ANOVA. In addition, the rate of GAP-43 alternation after one year was visualized in a box plot. In evaluating the changes in the DTI values, the Friedman test was implemented with a two-year follow-up. Finally, to assess whether there is a correlation between the GAP-43 and DTI values, the Spearman test was conducted. In all of the above tests, the p-value below 0.05 was indicated as significant.

Results

Participants age range from 55 to 91 (mean: 73.08 years old). Data were allocated into five groups normal, early and late MCI, and AD considering the ADNI classification, which was based on the Mini-Mental State Examination (MMSE) score and the physician report on the state of the disease which was based on the clinical presentation, imaging findings and cognitive tests, including Clinical Dementia Rating Scale (Table 1).

Table 1 Demographic information of participants
Full size table

First, the amount of GAP-43 in CSF was evaluated in normal, early, and late MCI, and AD participants. The changes in the value of GAP-43 in each group were investigated in two and four-year follow-ups. Subsequently, the DTI values in all parameters of all brain areas were evaluated between normal, early, and late MCI subgroups in three consequence visits over two years. Eventually, the correlation between the affected areas in the brain and the GAP-43 levels was assessed.

GAP-43

Using the Kruskal-Wallis H test, the GAP-43 level in CSF significantly increases as the disease progresses (P-value = 0.039). By performing a longitudinal analysis, it could be perceived that GAP-43 changes are mainly revealed after four years. Although both groups of normal and early MCI experience significant changes with a p-value less than 0.05, the rate of the changes is higher compared to normal individuals according to mean differences. The GAP-43 measurements in the late MCI with the four-year follow-up were excluded due to the extremely limited sample size (Table 2).

Table 2 Repeated measure ANOVA test results were conducted on normal, early and late MCI groups with two and four years follow-ups
Full size table

The increase in the level of CSF GAP-43 can be defined by taking a glimpse of each group’s GAP-43 level changes per year. As is illustrated in Fig. 2, normal participants experienced a slight alternation in the GAP-43 value, whereas the early MCI group experienced a high rate of fluctuation even up to approximately 1200 pg/mL. The fluctuation value becomes more remarkable in the late MCI group with around 400 pg/mL compared to 200 pg/mL in the normal group (Fig. 2).

Fig. 2
figure 2

GAP-43 value changes per year in normal, early and late MCI. Although the values were not significant between groups, it indicates a rise in the mean GAP-43 value in one year if the disease stage progressed

Full size image

DTI measurement

In this section, a comparison was performed between different brain structural pathways of normal, early, and late MCI participants detected by DTI with two-year follow-ups. Since our data contains right-handed participants, we implement our analysis on the left hemisphere of the individuals.

By performing Friedman analyses considering two-year follow-ups in three visits, the following results were obtained. In the early MCI group, a significant change was observed in the cerebral peduncle (p = 0.022) and posterior limb of the internal capsule (PLIC) (p = 0.009), in the AxD and FA values, respectively. Unlike FA, for the AxD and MD parameters of DTI, the Retrolenticular part of the internal capsule (RPIC) revealed significant change among late MCI participants (p = 0.001 and p = 0.011). Posterior thalamic radiation revealed significant changes in the normal, early and late MCI group. Moreover, the cingulum cortex depicted significant changes in late MCI compared to normal participants in the FA parameter (p = 0.004), nevertheless, in the RD parameter there is a significant change in both normal and late MCI groups (p = 0.013 and p = 0.019). The significant values of the cerebellar peduncle, Genu of the corpus callosum, and Medial lemniscus are hypothesized to be the consequence of normal aging rather than MCI state. Fornix (cres)/Stria terminalis is another important pathway that is damaged with a remarkable value among MCI participants (FA: early MCI p = 0.022, late MCI p = 0.002; MD: early MCI: p = 0.016). AxD, MD, and RD parameters in the late MCI group specified significant changes in the sagittal stratum (p = 0.005, p = 0.005, and p = 0.007), and Superior fronto-occipital fasciculus (p = 0.003, p = 0.001, and p = 0.001). The cingulum (hippocampus) tract is highlighted for the early MCI signal alternation in AxD and MD (p = 0.033 and p = 0.032). The AxD anterior and Superior corona radiata neural tract has remarkable changes in the late MCI stage (p = 0.002 and p = 0.004), in contrast to the Posterior corona radiata. The FA value of the Inferior fronto-occipital fasciculus is not suggested to be involved in the MCI brain changes, since the value is too close to the p-value threshold (p = 0.047). Eventually, the changes in DTI values started from the posterior part of the brain (Posterior corona radiata) and continued to the anterior part of the brain (Anterior corona radiata) (Fig. 3). More information regarding the p-values is available in the supplementary Table 1.

Fig. 3
figure 3

Statistical maps depicted AxD, FA, MD, and RD values of brain tracts which changes in two years in three groups of normal, early MCI, and late MCI

Full size image

GAP-43 and DTI

Spearman correlation analyses were performed to investigate possible relationships between GAP-43 and DTI values (Fig. 4). In this analysis, the participants were divided into five groups, normal (N = 48), early MCI (N = 23), MCI (N = 19), late MCI (N = 24), and AD (N = 19). The DTI tracts for this section were selected based on the tracts with significant changes during MCI.

If we take a glimpse of normal group results, it can be concluded that Cingulum (hippocampus) has a significant negative correlation with GAP-43 in two AxD and MD parameters in the normal group (r=-0.331, p = 0.021), whereas this tract has a positive correlation in the late MCI (r = 0.476, p = 0.019) and AD groups (r = 0.468, p = 0.043). Moreover, the GAP-43 value was significantly correlated with anterior corona radiate in early (MD: r=-0.471, p = 0.023; RD: r=-0.421, p = 0.046). On the contrary, among early MCI and MCI groups in AxD value of superior corona radiate (r=-0.562, p = 0.005 and r=-0.484, p = 0.036) and AxD (r=-0.515, p = 0.012; r = 0.034, p = 0.034), MD (r=-0.521, p = 0.011; r=-0.518, p = 0.023), and RD (r=-0.520, p = 0.011; p = 0.030, r=-0.498) values of superior fronto-occipital fasciculus, a significant negative correlations with GAP-43 were detected. Eventually, in the early and late MCI group by considering the fornix (cres)/Stria terminalis pathway, there is a significant correlation between GAP-43 and DTI in MD parameter (r=-0.478, p = 0.021; r=-0.425, p = 0.038). Based on the results it could be concluded that the GAP-43 value highly correlated with the DTI parameters in the early MCI groups (Suppl Table 2).

Fig. 4
figure 4

GAP-43 correlation with DTI values in five groups of normal, early MCI, MCI, late MCI, and AD

Full size image

Discussion

The current study performed a longitude analysis on the level of CSF GAP-43 over two and four years of follow-ups, in which the level of GAP-43 revealed a significant increase after four years. Moreover, changes in the DTI values in different brain areas were assessed after two years of follow-up. Finally, an analysis to reveal the correlation between the GAP-43 and the areas of the brain with high changes in DTI values in the disease group was performed.

In the Zhu et al. study, the project initiated with a hypothesis mentioning an increase in CSF GAP-43 levels in AD and MCI subjects. The results acknowledge the fact that the GAP-43 level had a direct correlation with the AD stage determined by the MMSE score. Furthermore, GAP-43 was specified as a potential biomarker for AD follow-up since MCI to AD progression became predictable by tracking the level of CSF GAP-43 [39]. Accordingly, the result of our study depicted an enhancement in the GAP-43 value in the different stages of AD, which became significant after four years of follow-ups. Another study with a huge sample size of 787 participants enrolled from the ADNI database implemented a regression model to link CSF GAP-43 to AD progression and other biomarkers. The results illustrated a correlation between CSF GAP-43 and AD stage as well as tau level (r = 0.768, P < 0.001) and neuroimaging hallmarks such as the volume of the medial temporal lobe and hippocampus [40]. In this regard, the current study demonstrated a correlation between CSF GAP-43 and DTI parameters in AD patients and revealed a significant correlation between GAP-43 value and DTI parameters in the early MCI. Just a short time before the results of Qiang et al. study became online; Öhrfelt et al. published a similar article with the same sample size of 787 participants from the ADNI database searching for the correlation between GAP-43 and Amyloid-beta (Aβ) level. The study result reported a higher risk of AD progression in participants with positive Aβ and a high amount of CSF GAP-43, represented by a hazard ratio of 8.56. Meanwhile, the study reported a correlation between GAP-43 and Aβ measured by positron emission tomography (PET) scan [41]. Likewise, it could be mentioned in our study that the mean level of GAP-43 was higher in late MCI in comparison to the normal and early MCI groups. Besides, our result related to GAP-43 was justified by the Zhang et al. study. Similar to our study, Zhang et al. mentioned the potential role of GAP-43 in the prediction of dementia onset in follow-ups over a period of 2 or 4 years [28].

It is worth mentioning the result of the Nabizadeh et al. study in which the correlation of the neurofilament light (Nfl) was calculated in different brain white matter tracks in AD patients from the ADNI database. The result revealed a negative correlation of Nfl with FA value in the fronto-occipital fasciculus, uncinate fasciculus, fornix, corpus callosum, corona radiata, and internal capsule, whereas there was a positive association between the sagittal stratum, hippocampal cingulum, cingulum, and thalamic radiation in AxD, MD, and RD values of DTI [27]. In our study, the AxD value of superior corona radiate and RD value of superior fronto-occipital fasciculus had a negative correlation with GAP-43 level in the early and mid-MCI participants. In addition, the MD value of Fornix (cres) / Stria terminalis has a negative correlation in early and late MCI, and the FA value of cingulum (hippocampus) has a positive correlation in late MCI. In addition to the correlation of DTI values with Nfl, there is a positive correlation between CSF values of Nfl and GAP-43 (r = 0.36, p = 0.015) [42]. Finally, in Bergamino et al. research, the DTI imaging technique was specified as a potent application in the detection of microstructural changes in AD. The main imaging hallmark to differentiate AD, MCI, and normal participants was reported as the FA value of fornix and corpus callosum [43]. Accordingly, based on our results, Superior fronto-occipital fasciculus, Retrolenticular part of the internal capsule, fornix, and FA value of cingulum were potential DTI biomarkers for categorizing normal, MCI, and AD participants with two years follow-ups.

In the Velazquez et al. research analyzing the ADNI data, the hippocampal volume was determined as a biomarker for predicting the progression of MCI to AD [44]. For the early stage of AD, Mondragón et al., who worked on the ADNI database, documented the anterior cingulate cortex as the indispensable part involved in this stage [45]. Similarly, in our study, the AxD, MD value of the connection of the cingulum to the hippocampus was a biomarker to differentiate the early and late stages of MCI due to its significant changes over two years. Moreover, in a previously published study, the MD value of the hippocampus was significantly higher in the MCI and AD groups [46]. Since the MD value is sensitive to necrosis as well as edema and an increase in AxD value could be an indication of higher CSF density [47], it could be concluded that the positive correlation of cingulum (hippocampus) MD and AxD value with GAP-43 detected in the late MCI and AD groups were suggested to be due to necrosis and CSF replacement in these areas. Nir et al. study, which performed analysis on 155 ADNI participants, specified the FA parameter of the DTI as the least sensitive in defining the AD stages [30]. Conversely, in our study, the FA value of the cingulum could be used as a marker of AD stages since it had a remarkable difference in following up on the early and late stages of AD.

A combination of GAP-43 and DTI measurements could be a diagnostic and monitoring biomarker to differentiate the early stage of MCI from normal participants. Previous studies mentioned different imaging biomarkers for AD severity. For instance, evaluating microstructural changes detected by DTI highlighted a reduction in the fiber number of perforant pathway along with a reduction in FA values of the entorhinal layer II. These results specified the involvement of the entorhinal-hippocampal pathway not only in AD participants but also the preclinical AD [48]. Another study applied Quantitative susceptibility mapping (QSM) in the AD group to detect iron accumulation in the tissue as an indicator of AD pathology landmark [49]. Blood-brain barrier (BBB) imaging tools are another novel imaging technique to assess the pathology of AD. These techniques including ASL-based measuring BBB water exchange rate and DCE-MRI measuring BBB permeability provide precious data regarding the BBB dysfunction in AD [50]. Moreover, the combination of QSM and DP-pCASL to analyze brain iron dynamics as well as BBB function to monitor the AD severity along with AD pathology provides promising biomarkers for AD [51].

The main limitation in performing this study was the low number of participants after applying filters for patients who took DTI and lumbar punctures for GAP-43. Express differently, the inclusion criteria of our protocol only included patients who took lumbar puncture for GAP-43 evaluation along with DTI imaging in the same year to minimize the possible bias in finding a correlation between the level of GAP-43 and DTI values.

Conclusion

Our results suggest a relationship between the GAP-43 biomarker and white matter integrity in certain tracts in patients with a diagnosis of MCI and AD. In recent studies, CSF GAP-43 was illustrated as a reliable synaptic biomarker to detect MCI progression to AD. The level of GAP-43 highly increased over 4 years in both normal, as a result of aging, and early MCI groups with higher intensity due to disease progression. While for the DTI values only two years of follow-ups revealed remarkable changes between normal and MCI patients. Our findings reflected a correlation between WM changes in multiple brain tracks corresponding to AD with the CSF level of GAP-43.

Data availability

Data collection and sharing for this project is accessible through the ADNI database (adni.loni.usc.edu).

References

  1. Kumar D, Sharma A, Sharma L. A comprehensive review of Alzheimer’s association with related proteins: pathological role and therapeutic significance. Curr Neuropharmacol. 2020;18(8):674–95.

    Article  PubMed Central  CAS  Google Scholar 

  2. Aye S, Jönsson L, Gustavsson A, Tate A, P SG, Eriksdotter M. Effect of mortality in cost-effectiveness modeling of disease-modifying treatment for Alzheimer’s disease. Alzheimer’s Dementia: Diagnosis Assess Disease Monit. 2023;15(1):e12422.

    Google Scholar 

  3. Shah-Abadi ME, Ariaei A, Moradi F, Rustamzadeh A, Tanha RR, Sadigh N, et al. In Silico Interactions of Natural and Synthetic compounds with Key proteins involved in Alzheimer’s Disease: prospects for Designing new therapeutics compound. Neurotox Res. 2023;41(5):408–30.

    Article  CAS  Google Scholar 

  4. Mukhopadhyay S, Banerjee D. A primer on the evolution of aducanumab: the first antibody approved for treatment of Alzheimer’s disease. J Alzheimers Dis. 2021;83(4):1537–52.

    Article  CAS  Google Scholar 

  5. Ariaei A, Ramezani F. The promising impact of Bemcentinib and Repotrectinib on sleep impairment in Alzheimer’s disease. J Biomol Struct Dyn. 2023:1–17.

  6. Rustamzadeh A, Tafakhori A, Ariaei A, Heydari M, Shah-Abadi ME, Seif F. Targeting Caspases 3/6 and cathepsins L/B May decrease Laminopathy-Induced apoptosis in Alzheimer’s Disease. J Alzheimers Dis. 2024;101(1):211–21.

    Article  CAS  Google Scholar 

  7. Usefi F, Rustamzadeh A, Ghobadi Z, Sadigh N, Mohebi N, Ariaei A et al. Rosuvastatin attenuates total-tau serum levels and increases expression of miR-124-3p in dyslipidemic Alzheimer’s patients: a historic cohort study. Metab Brain Dis. 2024:1–11.

  8. Choi HG, Kim J-H, Kim JH, Kim ES, Park HY, Min K-W, et al. Associations between Proton pump inhibitors and Alzheimer’s disease: a nested case–control study using a Korean nationwide health screening cohort. Alzheimers Res Ther. 2022;14(1):91.

    Article  PubMed Central  CAS  Google Scholar 

  9. Brandes Lourenço R, Machado de Campos B, Rizzi L, Sakzenian de Souza M, Forlenza OV, Giroud Joaquim H, et al. Functional connectome analysis in mild cognitive impairment: comparing Alzheimer’s Disease Continuum and suspected Non-alzheimer Pathology. Brain Connect. 2022;12(9):774–83.

    Article  Google Scholar 

  10. Lu Y, Initiative ADN. Cerebrospinal fluid growth-associated protein 43 levels in patients with progressive and stable mild cognitive impairment. Aging Clin Exp Res. 2022:1–8.

  11. Moscoso A, Silva-Rodríguez J, Aldrey JM, Cortés J, Fernández-Ferreiro A, Gómez-Lado N, et al. Prediction of Alzheimer’s disease dementia with MRI beyond the short-term: implications for the design of predictive models. NeuroImage: Clin. 2019;23:101837.

    Article  Google Scholar 

  12. Mohammadi H, Ariaei A, Ghobadi Z, Gorgich EAC, Rustamzadeh A. Which neuroimaging and fluid biomarkers method is better in theranostic of Alzheimer’s disease? An umbrella review. IBRO Neurosci Rep. 2024.

  13. Zhu Z, Lu J, Wang S, Peng W, Yang Y, Chen C, et al. Acrolein, an endogenous aldehyde induces synaptic dysfunction in vitro and in vivo: involvement of RhoA/ROCK2 pathway. Aging Cell. 2022;21(4):e13587.

    Article  PubMed Central  CAS  Google Scholar 

  14. Sandelius Å, Portelius E, Källén Å, Zetterberg H, Rot U, Olsson B, et al. Elevated CSF GAP-43 is Alzheimer’s disease specific and associated with tau and amyloid pathology. Alzheimer’s Dement. 2019;15(1):55–64.

    Article  Google Scholar 

  15. Patil AA. Characterization of GAP43-Gfp Fusion protein in PC12 and NIH3T3 stable transfectant cells. Texas A&M University-Kingsville; 2021.

  16. Maroto I, Costas-Insua C, Berthoux C, Moreno E, Ruiz-Calvo A, Montero-Fernández C et al. Control of a hippocampal recurrent Excitatory Circuit by Cannabinoid receptor-interacting protein Gap43. 2022.

  17. Wu Q-L, Cheng Y-Q, Liu A-J, Zhang W-D. Formononetin recovered injured nerve functions by enhancing synaptic plasticity in ischemic stroke rats. Biochem Biophys Res Commun. 2020;525(1):67–72.

    Article  CAS  Google Scholar 

  18. Holahan MR. A shift from a pivotal to supporting role for the growth-associated protein (GAP-43) in the coordination of axonal structural and functional plasticity. Front Cell Neurosci. 2017;11:266.

    Article  PubMed Central  Google Scholar 

  19. Zamzow DR, Elias V, Acosta VA, Escobedo E, Magnusson KR. Higher levels of protein palmitoylation in the frontal cortex across aging were associated with reference memory and executive function declines. Eneuro. 2019;6(1).

  20. Ramesh M, Govindaraju T. Multipronged diagnostic and therapeutic strategies for Alzheimer’s disease. Chem Sci. 2022;13(46):13657–89.

    Article  PubMed Central  CAS  Google Scholar 

  21. Andersson A, Remnestål J, Nellgård B, Vunk H, Kotol D, Edfors F, et al. Development of parallel reaction monitoring assays for cerebrospinal fluid proteins associated with Alzheimer’s disease. Clin Chim Acta. 2019;494:79–93.

    Article  CAS  Google Scholar 

  22. Pieri V, Trovatelli M, Cadioli M, Zani DD, Brizzola S, Ravasio G, et al. In vivo diffusion tensor magnetic resonance tractography of the sheep brain: an atlas of the ovine white matter fiber bundles. Front Veterinary Sci. 2019;6:345.

    Article  Google Scholar 

  23. Contarino VE, Siggillino S, Arighi A, Scola E, Fumagalli GG, Conte G, et al. Association of superficial white matter alterations with cerebrospinal fluid biomarkers and cognitive decline in neurodegenerative dementia. J Alzheimers Dis. 2022;85(1):431–42.

    Article  CAS  Google Scholar 

  24. Ebrahimi Shah-abadi M, Ariaei A, Mohammadi H, Shabani A, Rahmani Tanha R, Tavakolian Ferdousie V, et al. Recent advances and future directions in imaging of Peripheral Nervous System: a comprehensive review for therapeutics Approach. J Adv Med Biomedical Res. 2023;31(148):415–31.

    Article  Google Scholar 

  25. Miller B, Kim S-J, Mehta HH, Cao K, Kumagai H, Thumaty N et al. Mitochondrial DNA variation in Alzheimer’s disease reveals a unique microprotein called SHMOOSE. Mol Psychiatry. 2022:1–14.

  26. Nabizadeh F, Pourhamzeh M, Khani S, Rezaei A, Ranjbaran F, Deravi N, et al. Plasma phosphorylated-tau181 levels reflect white matter microstructural changes across Alzheimer’s disease progression. Metab Brain Dis. 2022;37(3):761–71.

    Article  CAS  Google Scholar 

  27. Nabizadeh F, Balabandian M, Rostami MR, Kankam SB, Ranjbaran F, Pourhamzeh M, et al. Plasma neurofilament light levels correlate with white matter damage prior to Alzheimer’s disease: results from ADNI. Aging Clin Exp Res. 2022;34(10):2363–72.

    Article  Google Scholar 

  28. Zhang H, Lyu D, Jia J, Initiative AsDN. The trajectory of cerebrospinal fluid growth-associated protein 43 in the Alzheimer’s disease continuum: a longitudinal study. J Alzheimers Dis. 2022;Preprint:1–12.

    Google Scholar 

  29. McGrowder DA, Miller F, Vaz K, Nwokocha C, Wilson-Clarke C, Anderson-Cross M, et al. Cerebrospinal fluid biomarkers of Alzheimer’s disease: current evidence and future perspectives. Brain Sci. 2021;11(2):215.

    Article  PubMed Central  CAS  Google Scholar 

  30. Nir TM, Jahanshad N, Villalon-Reina JE, Toga AW, Jack CR, Weiner MW, et al. Effectiveness of regional DTI measures in distinguishing Alzheimer’s disease, MCI, and normal aging. NeuroImage: Clin. 2013;3:180–95.

    Article  Google Scholar 

  31. Iglesias JE, Liu CY, Thompson PM, Tu Z. Robust brain extraction across datasets and comparison with publicly available methods. IEEE Trans Med Imaging. 2011;30(9):1617–34.

    Article  Google Scholar 

  32. Fischl B, van der Kouwe A, Destrieux C, Halgren E, Ségonne F, Salat DH, et al. Automatically parcellating the human cerebral cortex. Cereb Cortex. 2004;14(1):11–22.

    Article  Google Scholar 

  33. Smith SM. Fast robust automated brain extraction. Hum Brain Mapp. 2002;17(3):143–55.

    Article  PubMed Central  Google Scholar 

  34. Jenkinson M, Bannister P, Brady M, Smith S. Improved optimization for the robust and accurate linear registration and motion correction of brain images. NeuroImage. 2002;17(2):825–41.

    Article  PubMed  Google Scholar 

  35. Leow AD, Yanovsky I, Chiang MC, Lee AD, Klunder AD, Lu A, et al. Statistical properties of jacobian maps and the realization of unbiased large-deformation nonlinear image registration. IEEE Trans Med Imaging. 2007;26(6):822–32.

    Article  Google Scholar 

  36. Mori S, Oishi K, Jiang H, Jiang L, Li X, Akhter K, et al. Stereotaxic white matter atlas based on diffusion tensor imaging in an ICBM template. NeuroImage. 2008;40(2):570–82.

    Article  Google Scholar 

  37. Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, Mackay CE, et al. Tract-based spatial statistics: voxelwise analysis of multi-subject diffusion data. NeuroImage. 2006;31(4):1487–505.

    Article  Google Scholar 

  38. Lee JE, Chung MK, Lazar M, DuBray MB, Kim J, Bigler ED, et al. A study of diffusion tensor imaging by tissue-specific, smoothing-compensated voxel-based analysis. NeuroImage. 2009;44(3):870–83.

    Article  Google Scholar 

  39. Zhu Y, Guo X, Zhu F, Zhang Q, Yang Y, Initiative ADN. Association of CSF GAP-43 and APOE ε4 with cognition in mild cognitive impairment and Alzheimer’s Disease. Cells. 2022;12(1):13.

    Article  PubMed Central  Google Scholar 

  40. Qiang Q, Skudder-Hill L, Toyota T, Wei W, Adachi H. CSF GAP-43 as a biomarker of synaptic dysfunction is associated with tau pathology in Alzheimer’s disease. Sci Rep. 2022;12(1):17392.

    Article  PubMed Central  CAS  Google Scholar 

  41. Öhrfelt A, Benedet AL, Ashton NJ, Kvartsberg H, Vandijck M, Weiner MW, et al. Association of CSF GAP-43 with the rate of Cognitive decline and progression to dementia in amyloid-positive individuals. Neurology. 2023;100(3):e275–85.

    Article  PubMed Central  Google Scholar 

  42. Saloner R, Fonseca C, Paolillo EW, Asken BM, Djukic NA, Lee S, et al. Combined effects of synaptic and Axonal Integrity on Longitudinal Gray Matter Atrophy in cognitively unimpaired adults. Neurology. 2022;99(20):e2285–93.

    Article  PubMed Central  CAS  Google Scholar 

  43. Bergamino M, Keeling EG, Walsh RR, Stokes AM. Systematic assessment of the impact of DTI methodology on fractional anisotropy measures in Alzheimer’s disease. Tomography. 2021;7(1):20–38.

    Article  PubMed Central  Google Scholar 

  44. Velazquez M, Lee Y. Multimodal ensemble model for Alzheimer’s disease conversion prediction from early mild cognitive impairment subjects. Comput Biol Med. 2022;151:106201.

    Article  Google Scholar 

  45. Mondragón JD, Maurits NM, De Deyn PP, Initiative ADN. Functional connectivity differences in Alzheimer’s disease and amnestic mild cognitive impairment associated with AT (N) classification and anosognosia. Neurobiol Aging. 2021;101:22–39.

    Article  Google Scholar 

  46. Hong YJ, Yoon B, Lim S-C, Shim YS, Kim J-Y, Ahn KJ, et al. Microstructural changes in the hippocampus and posterior cingulate in mild cognitive impairment and Alzheimer’s disease: a diffusion tensor imaging study. Neurol Sci. 2013;34:1215–21.

    Article  Google Scholar 

  47. Alexander AL, Hurley SA, Samsonov AA, Adluru N, Hosseinbor AP, Mossahebi P, et al. Characterization of cerebral white matter properties using quantitative magnetic resonance imaging stains. Brain Connect. 2011;1(6):423–46.

    Article  PubMed Central  Google Scholar 

  48. Uchida Y, Onda K, Hou Z, Troncoso JC, Mori S, Oishi K. Microstructural Neurodegeneration of the entorhinal-Hippocampus pathway along the Alzheimer’s Disease Continuum. J Alzheimers Dis. 2023;95(3):1107–17.

    Article  PubMed Central  CAS  Google Scholar 

  49. Uchida Y, Kan H, Sakurai K, Oishi K, Matsukawa N. Quantitative susceptibility mapping as an imaging biomarker for Alzheimer’s disease: the expectations and limitations. Front Neurosci. 2022;16:938092.

    Article  PubMed Central  Google Scholar 

  50. Uchida Y, Kan H, Sakurai K, Oishi K, Matsukawa N. Contributions of blood-brain barrier imaging to neurovascular unit pathophysiology of Alzheimer’s disease and related dementias. Front Aging Neurosci. 2023;15:1111448.

    Article  PubMed Central  CAS  Google Scholar 

  51. Uchida Y, Kan H, Furukawa G, Onda K, Sakurai K, Takada K, et al. Relationship between brain iron dynamics and blood-brain barrier function during childhood: a quantitative magnetic resonance imaging study. Fluids Barriers CNS. 2023;20(1):60.

    Article  PubMed Central  CAS  Google Scholar 

Download references

Acknowledgements

This project was supported by the Tehran University of Medical Sciences, Tehran, Iran, under the supervision of Neurotrack as a student research team.

Funding

No funding was received for conducting this study.

Author information

Authors and Affiliations

  1. School of Medicine, Iran University of Medical Science, Hemmat Highway, Next to Milad Tower, Tehran, 1449614535, Iran

    Armin Ariaei

  2. Advanced Diagnostic and Interventional Radiology Research Center (ADIR), Tehran University of Medical Sciences, Tehran, Iran

    Armin Ariaei

  3. Department of Biology, Islamic Azad University East Tehran Branch, Tehran, Iran

    Atousa Ghorbani

  4. School of Medicine, Tabriz University of Medical Science, Tabriz, Iran

    Elham Habibzadeh

  5. Department of Clinical Biochemistry, Islamic Azad University Shahrood Branch, Shahrood, Iran

    Nazanin Moghaddam

  6. Department of Advanced Sciences and Technology, Islamic Azad University Tehran Medical Sciences, Tehran, Iran

    Negar Chegeni Nezhad

  7. Department of Nuclear Medicine and Radiobiology, Faculty of Medicine and Health Sciences, Centre Hospitalier Universitaire de Sherbrooke (CHUS), Université de Sherbrooke, Sherbrooke, Canada

    Amirabbas Abdoli

  8. Department of Psychology, Islamic Azad University Semnan Branch, Semnan, Iran

    Samira Mazinanian

  9. School of Medicine, Tehran University of Medical Sciences, Tehran, Iran

    Mohammad Sadeghi & Mahsa Mayeli

Authors
  1. Armin Ariaei
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Atousa Ghorbani
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Elham Habibzadeh
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Nazanin Moghaddam
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Negar Chegeni Nezhad
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Amirabbas Abdoli
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Samira Mazinanian
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Mohammad Sadeghi
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Mahsa Mayeli
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

A.A. contributed to data clearing, data analysis, writing, and editing. A.G. contributed to the writing of the introduction and result section and editing. E.H. contributed to writing the discussion and method section and editing. N.M. contributed to writing the method section. S.M. contributed to data clearing. N.C.N. and A.Ab. contributed to writing the introduction section. M.S. contributed to reviewing and editing the manuscript. M.M. contributed to supervising and writing the first draft of the manuscript. All authors reviewed the final version of the manuscript.

Corresponding author

Correspondence to Armin Ariaei.

Ethics declarations

Ethical approval

Ethical approval was waived by the local Ethics Committee of Tehran University of Medical Sciences in view of the retrospective nature of the study and all the procedures being performed were part of the routine care.

Consent to participate

Informed consent was obtained from all individual participants included in the study based on the ADNI declaration (adni.loni.usc.edu).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

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

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ariaei, A., Ghorbani, A., Habibzadeh, E. et al. Investigating the association between the GAP-43 concentration with diffusion tensor imaging indices in Alzheimer’s dementia continuum. BMC Neurol 24, 397 (2024). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12883-024-03904-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12883-024-03904-9

Keywords

BMC Neurology

ISSN: 1471-2377

Contact us