Hippocampal Glutathione Depletion and pH Increment in Alzheimer’s Disease: An in vivo MRS Study

Abstract

Background:

Oxidative stress plays a major role in Alzheimer’s disease (AD) pathogenesis, and thus, antioxidant glutathione (GSH) has been actively investigated in mitigating the oxidative load. Significant hippocampal GSH depletion has been correlated with cognitive impairment in AD. Furthermore, postmortem studies indicated alterations in cellular-energy metabolism and hippocampal pH change toward alkalinity in AD.

Objective:

Concurrent analysis of hippocampal GSH and pH interplay in vivo on the same individual is quite unclear and hence requires investigation to understand the pathological events in AD.

Methods:

Total 39 healthy old (HO), 22 mild cognitive impairment (MCI), and 37 AD patients were recruited for hippocampal GSH using ¹H-MRS MEGA-PRESS and pH using 2D ³¹P-MRSI with dual tuned (¹H/³¹P) transmit/receive volume head coil on 3T-Philips scanner. All MRS data processing using KALPANA package and statistical analysis were performed MedCalc, respectively and NINS-STAT package.

Results:

Significant GSH depletion in the left and right hippocampus (LH and RH) among MCI and AD study groups as compared to HO was observed, whereas pH increased significantly in the LH region between HO and AD. Hippocampal GSH level negatively correlated with pH in both patient groups. The ROC analysis on the combined effect of GSH and pH in both hippocampal regions give accuracy for MCI (LH: 78.27%; RH: 86.96%) and AD (LH: 88%; RH: 78.26%) groups differentiating from HO.

Conclusion:

Outcomes from this study provide further insights to metabolic alterations in terms of concurrent assessment of hippocampal GSH and pH levels in AD pathogenesis, aiding in early diagnosis of MCI and AD.

Keywords

Alzheimer’s disease glutathione hippocampus pH mild cognitive impairment

INTRODUCTION

Alzheimer’s disease (AD) is the most common progressive and irreversible neurodegenerative disorder contributing 60–70% of global dementia cases [1]. Moreover, amnestic mild cognitive impairment (MCI) is the transitional stage between healthy old (HO) and AD with 10–15% annual conversion rate to AD [2 –4]. AD is characterized by impairment in an individual’s cognitive abilities to perform activities of daily living. The manifestation of AD starts with memory problems and disease progression leading to severe impairments in other cognitive domains including decision making and planning [5].

Various independent hypotheses were postulated to explain the etiology and pathogenesis of AD, which included amyloid-β (Aβ) deposition [6], neurofibrillary tangles of hyperphosphorylated tau protein [7], and oxidative stress (OS) [8, 9]. Contrary to these hypotheses, postmortem studies have found significant Aβ deposition in the brain of elderly individuals without any history of cognitive deficits [10, 11]. On the other hand, tau tangles have been reported in the brain of very mild dementia cases without any amyloid load [12]. Along with Aβ and tau hypotheses, the OS hypothesis gained considerable attention [8 , 14].

OS arises due to imbalance between the production of reactive oxygen species (ROS) (free radicals) and antioxidant defense system [8, 15]. It potentially serves as an early event that triggers the development of AD pathology [14]. Numerous studies have demonstrated that oxidative damage occurs in the brain areas responsible for higher level cognition in both MCI and AD [16, 17]. The antioxidant defense system includes several essential vitamins, enzymes, and small peptides. Glutathione (GSH) is a tripeptide, which has a profound role in modulating intracellular responses to OS, and its depletion aggravates oxidative damage in AD [18]. The redox imbalance due to OS has been known to alter the activity of enzymes that may cause an imbalance in H⁺ ion concentrations [19 –21]. Although studies have attempted to monitor neuronal pH shifts in the brain, the causes of pH alteration remain unknown [22 –24].

Our earlier work showed significant depletion in the hippocampal GSH level among HO, MCI, and AD groups [25], whereas another study on hippocampal pH level changes included only healthy participants [26]. However, the underlying interrelation between hippocampal GSH and pH levels in AD pathology has not been investigated till now. The altered GSH and pH levels in the brain can be measured noninvasively using advanced neuroimaging techniques to understand the brain structural, neurochemical, and microenvironment changes associated with AD pathology [27]. In vivo high-field MRS demonstrated its promising role in identifying metabolic changes through quantitative measures of metabolite levels and their ratios in healthy and diseased brains [27]. These imaging studies are particularly important in identifying the hallmark areas of an affected brain region and ruling out other causes of progressive cognitive decline.

We hypothesize that due to increased OS, GSH is depleted in hippocampal region and is subsequently associated with cellular microenvironment changes (e.g., pH alteration). To establish our hypothesis, we have undertaken a multinuclear noninvasive cross-sectional study design to assess both concurrent hippocampal GSH and pH levels using proton (¹H) MRS and phosphorus (³¹P) MRS, respectively, in HO, MCI, and AD study groups. Outcomes from this study provided further insights into metabolic alterations in terms of concurrent assessment of hippocampal GSH and pH levels in AD pathogenesis, aiding in the early diagnosis of MCI and AD.

MATERIALS AND METHODS

Study participants

Total 98 (39 HO, 22 MCI, and 37 AD) participants were recruited for this cross-sectional study (Table 1). The HO participants were recruited from National Capital Region, New Delhi, India, in association with HelpAge India. All MCI and AD patients were recruited from the Outpatient Department, Department of Neurology (MT), All India Institute of Medical Sciences (AIIMS), New Delhi (Dr. Manjari Tripathi, MD, DM). The MCI patients were diagnosed as per the revised Petersen criteria [28, 29]. Similarly, the AD patients were diagnosed as per the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s disease and Related Disorders Association revised criteria for the clinical diagnosis of probable AD [30, 31]. The eligibility criteria for all participants were ≥55 years old, without any history of psychiatric/other neurological illness or any other medical condition known to impact brain function and associated to study outcome. The participants with any known contradiction toward MRI due to metallic implants or claustrophobia were excluded. All the eligible participants went through ¹H MRI/MRS for GSH and ³¹P MRS for pH measurements in both left (LH) and right hippocampal (RH) regions.

Table 1

Participant characteristics and regions (LH and RH) specific GSH concentration and pH level of HO, MCI, and AD participants

Characteristics	HO	MCI	AD	Test Statistics	p ^*
Demographic characteristics
No. of participants (N)	39	22	37	–	–
Age (y)^†	65.05±6.35	65.27±7.31	67.81±7.42	H = 3.44^‡	0.178
Gender (M/F)	24/15	19/3	18/19	χ² = 8.36^§	0.015
GSH concentration
LH GSH concentration^†	1.15±0.21 (31)	0.93±0.25 (19)	0.86±0.19 (24)	F = 13.39^¶	< 0.001
RH GSH concentration^†	1.02±0.21 (24)	0.84±0.21 (15)	0.85±0.20 (19)	F = 5.48^¶	0.007
pH values
LH pH^†	7.01±0.04 (18)	7.02±0.03 (15)	7.05±0.06 (25)	H = 6.65^‡	0.036
RH pH^†	7.02±0.04 (18)	7.05±0.04 (15)	7.04±0.07 (25)	F = 1.10^¶	0.340

MRI and multinuclear ¹H and ³¹P MRS data acquisition

In this study, both ¹H and ³¹P MRI/MRS data for humans and phantom have been acquired at NBRC using 3T MR scanner (Achieva, Philips) equipped with a dual-tuned (¹H/³¹P) transmit/receive volume head coil (Rapid Corporation).

MRI data acquisition

Two dimensional T2-weighted MRI images using TSE sequence were acquired in axial, sagittal, and coronal directions with total acquisition time of 9 min. The acquisition parameters were set as TR =3000 ms, TE = 80 ms, and turbo factor = 15. These MRI images were used for anatomical localization of hippocampi to place ¹H MRS voxels and 2D MRSI grid consisting of both hippocampi as ROI for ³¹P MRS data acquisition.

¹H MRS data acquisition for GSH quantitation

In vivo GSH data from both LH and RH regions were acquired using MEshcher–GArwood Point RESolved Spectroscopy (MEGA-PRESS) [32] pulse sequence with selective ON and OFF editing pulses (duration = 20 ms, sinc-Gaussian shape and frequency bandwidth = 1300 Hz) set at 4.40 ppm and 5.00 ppm, respectively. Other GSH data acquisition parameters were set to voxel size = 2.5×2.5×2.5 cm³ (volume 15.6 cc), TE = 120 ms, TR = 2500ms, sample points = 2048, spectral bandwidth =2000 Hz, and no. of signal acquisitions (NSA)=16 with a total of 20 interleaved ON and OFF edited scans (10 each) for each ROI. Chemical Shift Selective Suppression (CHESS) pulse sequence [33] was used for water suppression along with Pencil beam volume shimming approach of second order [34] resulting in water linewidth of ≤18 Hz for bilateral hippocampi. Scan time for MEGA-PRESS experiments for each voxel was approximately 14 minutes. Similar parameters for GSH MEGA-PRESS data acquisition have been used in our earlier work [25 , 36].

³¹P MRS data acquisition for pH measurement

A phantom study was performed to validate the accuracy of pH measured using ³¹P MRS compared to pH measured using Mettler Toledo pH meter as a standard reference. A 1 : 1 solution of 3 mM phosphocreatine disodium salt (PCr) (Sigma Aldrich) and 3 mM NaH₂PO₄ (Pi) (Merck Limited) was prepared by dissolving PCr and NaH₂PO₄ in 600 ml Tris buffer where the initial pH of the sample was set to 6.66. For calibration of pH values measured using ³¹P MRS, the pH of this solution was further adjusted using 1 M NaOH to get the subsequent pH values as 6.83, 6.95, 7.00, 7.13, and 7.26. Each of these solutions was transferred to a spherical eddy current canceling phantom bottle (ECC, Philips) for ³¹P MRS data acquisition. The phantom solution was maintained at room temperature during electrode pH measurement and ³¹P MRS experiment.

Multi-voxel ³¹P MRS data using the 2D MRSI sequence on both phantom and on the same human participants as in ¹H-MRS, were acquired with optimum excitation pulse flip angle of 35°, sample data points = 1024, spectral bandwidth = 4000 Hz, NSA = 16, TR = 1000 ms, TE = 1.4 ms, and slice thickness = 30 mm. ³¹P signals were obtained in ∼15 min from 576 voxels (matrix size of 24×24) within a FOV of 240×240 mm² for each participant. The Waltz-16 sequence [37] was used for ¹H/³¹P decoupling with an offset frequency set at 420 Hz. Nuclear overhauser effect [38] enhancement with B1 max of 1μT was used for proton decoupled ³¹P MRS (mixing time 350 ms). Pencil beam volume shimming approach of second order [34] was used for better magnetic field homogeneity in the hippocampal area. Post-acquisition of ³¹P spectra, an additional T2-weighted TSE image was acquired in the same axial orientation for co-registration of ³¹P 2D MRSI grid to enable accurate anatomical positioning during post-experimental analysis.

Multinuclear ¹H and ³¹P MRS data processing

All data acquired from ¹H and ³¹P MRS were processed using MATLAB (The MathWorks, Inc.)-based MRS signal processing and analysis toolbox KALPANA [39]. The MRS processing pipeline included a quantitative assessment of signal quality and visual inspection of each sub-spectrum, where the data with poor signal quality due to lipid contamination or head movement artifacts were rejected. Figure 1 shows the representative voxel placement with its corresponding ¹H edited-difference GSH MEGA-PRESS spectrum and 2D MRSI grid placement with ³¹P MRS multi-voxel spectrum from the LH region of an HO participant.

Fig. 1

In vivo measurement of brain left hippocampal GSH using MEGA-PRESS and pH using ³¹P-MRS (2D-MRSI) of a 70-year-old healthy male participant. GSH measurement was performed on the fitted GSH peak obtained at ∼2.80 ppm and the area under the peak was used for GSH concentration calculation. The hippocampal pH is shown in the color-coded map along with the acquired respective ³¹P-MRS spectra. Estimation of pH was based on Henderson-Hasselbalch (Eq. 2) using the chemical shift difference (δPi) between phosphocreatine (PCr) and inorganic phosphate (Pi) peaks.

¹H MRS MEGA-PRESS data processing for GSH quantitation

Each spectrum of the GSH MEGA-PRESS data was initially auto-phase corrected using a common zero-order phase based on the first data point of FID to get in-phase water peak aligned at 4.67 ppm. The interleaved sum of individual ON and OFF phase-corrected spectra were further averaged to give “averaged-ON” and “averaged-OFF” spectra, respectively. Subsequently, an “edited-difference” spectrum obtained as the difference between the “averaged ON” and “averaged OFF” was used for desired GSH peak area estimation and quantitation. In MEGA-PRESS edited-difference spectrum, GSH Cys-β-CH₂ peak was observed at ∼2.80 ppm [25 , 36] and analyzed for both LH and RH brain regions of each study participant.

Quantitative assessment of GSH using edited-difference included the preprocessing steps of apodization using mixed Gaussian and exponential filter functions, followed by manual zero-order phase correction, ensuring the up-phase of GSH peak for the correct area estimation. Selective suppression of the residual water and lipid peaks was performed using Hankel–Lanczos Singular Value Decomposition filtering method [40]. Singular Spectrum Analysis was applied for spectral baseline estimation, and further, spectral fitting was performed using time-domain nonlinear least square cost function optimization [41].

In vivo GSH absolute quantitation (GSH_ABS) was computed from external reference based on GSH phantom experiments using solution of various concentrations. T1 and T2 relaxation time compensations of GSH were applied using formula given in Equation 1.

$[GS H_{ABS}] = \frac{Are a_{GSH} - k}{m} \times \frac{1 - exp (- \frac{TR}{T 1_{Phantom}})}{1 - exp (- \frac{TR}{T 1_{GSH}})} \times \frac{exp (- \frac{TE}{T 2_{Phantom}})}{exp (- \frac{TE}{T 2_{GSH}})}$ (1)

where TE = 120 ms and TR = 2500 ms are set from MEGA-PRESS experiment; Area_GSH is the calculated area under GSH peak; T1_GSH, T2_GSH, and T1_Phantom, T2_Phantom, are the relaxation time constants for GSH from human brain tissue and phantom, respectively, whereas m and k are the slope and intercept values obtained from the linearly fitted calibration curve of phantom data. Similar parameters and steps for GSH MEGA-PRESS data processing and quantitation have been detailed in our earlier work.[25 , 36] Illustrations of relative change in left and right hippocampal GSH concentration among the representative participants of HO, MCI, and AD groups using MEGA-PRESS sequence are presented in Fig. 2a and 2b, respectively.

Fig. 2

Illustration of relative change in GSH and pH among the HO, MCI, and AD participants using MEGA-PRESS sequence, and ³¹P 2D-MRSI sequence. (a) MEGA-PRESS edited spectra and MRS voxel placement to show the GSH depletion in left hippocampus, and (b) right hippocampus among HO, MCI, and AD. The absolute GSH concentration in the left and right hippocampus of one of the participants of each group (HO, MCI, and AD) have been shown along with their respective GSH peaks in the edited MEGA-PRESS spectra. Decrease in the peak area of the GSH is reflective of GSH depletion. (c) Change in hippocampal pH among the HO, MCI, and AD participants same as in a and b. Relative downfield shift of the Pi peak indicates increase of hippocampal pH in the AD. The color-coded pH value maps (Blue to Red: low to high) in the HO, MCI, and AD participants have been shown. A decrease in the GSH peak area in AD pathology, i.e., from HO to MCI and to AD relates to the decrease in left hippocampus GSH (from a and b), which can further be related with the respective increase in hippocampal pH (color-coded pH map on the left hippocampus in HO increasing towards MCI and further in AD as shown in c).

³¹P MRS data processing for pH measurement

The 3×3 voxels from the phantom center and the hippocampus regions of the human brain from the 2D MRSI grid were selected. The averaged MRS data were Fourier transformed into a frequency domain for the selected voxels and apodized with a combination of Gaussian and exponential filter functions. Every spectrum was referenced to PCr peak set at 0 ppm, and manual phase correction (with zeroth and first order) was applied to bring all detectable peaks in the same phase.

The chemical shifts of H₂PO₄^–2 and H₂PO₄^–1 [inorganic phosphate (Pi)] are characterized by two resonance peaks at 5.68 and 3.29 ppm, respectively. Due to fast chemical exchange between these two ³¹P moieties, an average resonance peak is observed for Pi, whose chemical shift is sensitive to pH change in the environment, whereas the PCr peak position remains unaffected. The difference between the chemical shifts of PCr and average Pi resonances was used to calculate the in vivo brain tissue pH using Equation 2 derived from the Henderson–Hasselbalch equation [42]. $pH = 6.77 + log [\frac{(δ Pi - 3.29)}{(5.68 - δ Pi)}]$ (2)

where δPi is the chemical shift difference between PCr and Pi resonances in ppm.

Six solutions with known pH (6.66, 6.83, 6.95, 7.00, 7.13, and 7.26) measured using pH meter, were further scanned using ³¹P MRS, the corresponding multi-voxel spectrum, and the Pi peak shift with respect to the PCr peak (set at 0 ppm) is represented in Supplementary Figure 1a. All data were processed, and the pH values were calculated from δPi. An agreement between pH measured from the same solution using ³¹P MRS and pH meter was observed (Supplementary Figure 1S). A measurement error of 0.03±0.019 [mean±standard deviation (SD)] was observed between pH value obtained using electrode and ³¹P MRS.

For in vivo average pH measurement from the LH and RH, 3×3 voxels covering the respective brain regions were selected from 2D MRSI overlay on MRI images obtained from the same participant to select the voxels representing hippocampi [26]. Data were further color-coded with heat-color map and superimposed on the MRI image to provide a visual representation of regional pH changes in bilateral hippocampus and phantom. Relative change in bilateral hippocampal pH level among the representative participants of HO, MCI, and AD groups using ³¹P 2D-MRSI sequence, respectively, are depicted in Fig. 2c.

Statistical analysis

The participant characteristics (age) and outcome measures (GSH concentration and pH level) from the LH and RH on a continuous scale are summarized using means with SD, whereas categorical variables (gender) are reported as frequency (M/F) ratio. Chi-square (χ²) test was used to assess gender differences among study participant groups (HO, MCI, and AD). All group-wise GSH concentrations and pH levels were assessed for normality using Shapiro–Wilk test and homoscedasticity using Levene’s test. The effect of different clinical status (MCI and AD) compared to HO on brain hippocampal GSH and pH levels was assessed by one-way ANOVA or Kruskal–Wallis, followed by Tukey–Kramer or Conover test for post-hoc multiple-comparison, respectively. The correlations between the hippocampal GSH concentration and pH level were obtained by calculating their Pearson’s correlation coefficient. Significance levels for all the statistical analyses were set at p < 0.05.

Moreover, to assess the diagnostic utility of GSH concentration and pH levels over the bilateral hippocampi to differentiate three study groups (HO, MCI, and AD), receiver operator characteristic (ROC) analyses were performed using DeLong criteria [43]. Multivariate ROC curve analysis was also performed from the predicted probability values obtained by binary logistic regression for the combined effects of GSH concentration and pH level changes. For each ROC curve, the area under the curve (AUC), sensitivity, specificity, likelihood ratio (LR+/LR–), and accuracy were reported. To assess the discriminative power of hippocampal GSH and pH levels as an independent index test, their AUCs were estimated with significance level and 95% confidence interval (CI). All statistical analyses were performed using MedCalc software (version 15.8).

RESULTS

Participant characteristics

The study participant’s mean age was found to be 65.05±6.35 years for HO, 65.27±7.31 years for MCI, and 67.81±7.42 years for AD with no significant difference between groups (p = 0.178) with an overall 62.24% male participants (Table 1).

GSH concentration and pH level alteration in MCI and AD compared to HO

The effect of GSH concentration and pH level within HO, MCI, and AD groups are shown in Table 1, and its graphical representation is depicted using box and whiskers plot in Fig. 3. The mean hippocampal GSH concentration, when compared among the three groups, was found to be significantly different in LH (p < 0.001) and RH (p =0.007). The mean pH value also varied significantly in LH (p = 0.036) but not in RH (p = 0.340) among the three groups. Post-hoc analysis in LH region indicates significant decrease in the mean GSH concentration between HO (1.15±0.21 mM) and MCI (0.93±0.25 mM) (p = 0.002) as well as HO (1.15±0.21 mM) and AD (0.86±0.19 mM) (p < 0.001). Conversely, the mean pH value in the LH region did not differ significantly between HO (7.01±0.04) and MCI (7.02±0.03) but significantly increased toward alkalinity between HO (7.01±0.04) and AD (7.05±0.06) (p = 0.018). In RH region, GSH concentration decreased significantly between HO (1.02±0.21 mM) and MCI (0.84±0.21 mM) (p = 0.022) as well as when HO (1.02±0.21 mM) was compared to AD (0.85±0.20 mM) (p = 0.018). However, no group reflected significant changes in pH values of the RH region.

Fig. 3

The Box and Whisker plots for absolute GSH concentration values (left axis) and pH values (right axis) among HO (green), MCI (blue), and AD (red) groups in the (a) left hippocampus, and (b) right hippocampus. (^*p < 0.05, ^**p < 0.01, ^***p < 0.001).

Correlation between hippocampal GSH concentration and pH level among HO, MCI, and AD

The inter-relationship between GSH concentration with pH levels in the LH and RH were graphically represented using a scatter plot (Fig. 4) for all the three study groups (HO, MCI, and AD). We found a mild positive correlation of GSH concentration and pH level for the HO group in both LH (r = 0.28) and RH (r = 0.08) regions. However, in case of MCI group, a negative correlation (r = –0.59) between GSH concentration and pH level was found in the LH region and mild pH increment was observed with GSH depletion (r = –0.18) in the RH region. However, in case of AD, a moderate negative correlation (r = –0.39) between GSH concentration and pH level was found in both hippocampal regions.

Fig. 4

Illustration of the group-wise scatter plot with fitted line between hippocampal GSH concentration and pH level in HO, MCI and AD measured using non-invasive ¹H/³¹P MR Spectroscopy in (a) left hippocampus, (b) right hippocampus regions. The data points can be depicted by a green triangle (HO), blue diamond (MCI) and red circle (AD), and corresponding fitted lines are represented by a short dash (HO), long dash-short dash (MCI), and long dash (AD).

Diagnostic accuracy for GSH concentration and pH level to differentiate MCI and AD from HO

The diagnostic accuracy parameters for GSH, pH, and their combined effect from both LH and RH regions to differentiate diseased state (MCI and AD) from HO are tabulated in Table 2, and the corresponding ROC curve is shown in Fig. 5. GSH concentration in the bilateral hippocampus region showed higher AUCs to distinguish AD from HO in LH (0.85) and RH regions (0.73) as compared to other groups (HO versus MCI and MCI versus AD). LH GSH was found to be more sensitive (84.21%) to differentiate MCI from HO; similarly to differentiate AD from HO, higher specificity (90.32%) and accuracy (80.01%) was observed, whereas RH region showed high sensitivity (94.74%) with low specificity (45.83%) and accuracy (67.45%) in differentiating AD from HO. This region also showed low sensitivity (73.33%), specificity (70.83%), and accuracy (71.79%) to differentiate MCI from HO.

Table 2

Diagnostic accuracy characteristics (AUC, sensitivity, specificity, LR+/LR–and accuracy) for GSH concentrations, pH level for the bilateral hippocampus regions to differentiate among three groups (HO, MCI, and AD)

Regions	LH			RH
Parameters	Impact of GSH concentrations
Classification group	HO versus MCI	HO versus AD	MCI versus AD	HO versus MCI	HO versus AD	MCI versus AD
N (Neg./Pos. group)	50 (31/19)	55 (31/24)	43 (19/24)	39 (24/15)	43 (24/19)	34 (15/19)
AUC [95% CI]	0.74^** [0.59 – 0.85]	0.85^*** [0.72 – 0.93]	0.58 [0.42 – 0.73]	0.73^* [0.55 – 0.85]	0.73^** [0.58 – 0.86]	0.52 [0.34 – 0.69]
Sensitivity [95% CI]	84.21 [60.4 – 96.6]	66.67 [44.7 – 84.4]	79.17 [57.8 – 92.9]	73.33 [44.9 – 92.2]	94.74 [74.0 – 99.9]	52.63 [28.9 – 75.6]
Specificity [95% CI]	54.84 [36.0 – 72.7]	90.32 [74.2 – 98.0]	47.37 [24.4 – 71.1]	70.83 [48.9 – 87]	45.83 [25.6 – 67.2]	60.00 [32.3 – 83.7]
LR+/LR–	1.86/0.29	6.89/0.37	1.50/0.44	2.51/0.38	1.75/0.11	1.32/0.79
Accuracy	66.00	80.01	65.11	71.79	67.45	55.88
	Impact of pH
N (Neg./Pos. group)	33 (18/15)	43 (18/25)	40 (15/25)	33 (18/15)	43 (18/25)	40 (15/25)
AUC [95% CI]	0.59 [0.41 – 0.76]	0.72^** [0.55 – 0.84]	0.67^* [0.50 – 0.81]	0.66 [0.47 – 0.81]	0.57 [0.41 – 0.72]	0.56 [0.40 – 0.72]
Sensitivity [95% CI]	80.00 [51.9 – 95.7]	64.00 [42.5 – 82.0]	52.00 [31.3 – 72.2]	40.00 [16.3 – 67.7]	44.00 [24.4 – 65.1]	80.00 [59.3 – 93.2]
Specificity [95% CI]	50.00 [26.0 – 74.0]	77.78 [52.4 – 93.6]	86.67 [59.5 – 98.3]	94.44 [72.7 – 99.9]	77.78 [52.4 – 93.6]	40.0 [16.3 – 67.7]
LR+/LR–	1.60/0.40	2.88/0.46	3.90/0.55	7.20/0.64	1.00/–	1.33/0.50
Accuracy	63.65	69.77	65.00	69.67	58.15	65.00
	Impact of combined effect of GSH and pH
N (Neg./Pos. group)	23(11/12)	25 (11/14)	26 (12/14)	23 (11/12)	23 (11/12)	24 (12/12)
AUC [95% CI]	0.80^** [0.58 – 0.93]	0.91^*** [0.73 – 0.99]	0.58 [0.37 – 0.77]	0.89^*** [0.70 – 0.98]	0.80^** [0.59 – 0.94]	0.604 [0.39 – 0.80]
Sensitivity [95% CI]	91.67 [61.5 – 99.8]	85.71 [57.2 – 98.2]	28.57 [8.4 – 58.1]	91.67 [61.5 – 99.8]	75.00 [42.8 – 94.5]	75.00 [42.8 – 94.5]
Specificity [95% CI]	63.64 [30.8 – 89.1]	90.91 [58.7 – 99.8]	100.00 [73.5 – 100.0]	81.82 [48.2 – 97.7]	81.82 [48.2 – 97.7]	58.33 [27.7– 84.8]
LR+/LR–	2.52/0.13	9.43/0.16	–/0.71	5.04/0.10	4.13/0.31	1.80/0.43
Accuracy	78.27	88.00	61.57	86.96	78.26	66.67

All significant values were set at p < 0.05 (^*p < 0.05, ^**p < 0.01, ^***p < 0.001).

Fig. 5

Diagnostic accuracy test using ROC curves for absolute GSH concentration (green), hippocampal pH (blue), and combined effect of GSH-pH (red) among HO, MCI, and AD groups for the (a) left hippocampus and (b) right hippocampus between negative versus positive groups considered as HO versus MCI, HO versus AD respectively.

Similarly, in case of pH, HO versus AD showed better AUC (0.72) for LH region as compared to other groups and regions. To differentiate MCI from HO, pH showed 80% sensitivity, whereas to differentiate AD from HO, specificity of 77.78% was found in the LH region. To differentiate both patient groups from HO, RH pH showed high specificity (MCI versus HO: 94.44%; AD versus HO: 77.78%) but low sensitivity (MCI versus HO: 40%; AD versus HO: 44%) and accuracy (MCI versus HO: 69.67%; AD versus HO: 58.15%).

In another set of analyses, both outcomes (GSH and pH) were assessed together using binary logistic regression to obtain their combined effect in LH and RH regions separately. The combined effect resulted in higher and acceptable AUC in the range of 0.80–0.91 as compared to the individual tests in both the LH and RH regions to differentiate both disease groups from HO. This combined effect of GSH and pH was more sensitive (91.67%) in both the regions to differentiate MCI from HO; however, to specifically differentiate AD from HO, LH region showed better specificity (LH: 91.91%; RH: 81.82%). We also found the potential utility of combined GSH–pH effect to rule out MCI from HO (0.13 LR in LH and 0.10 in RH regions).

DISCUSSION

Hippocampal structural and functional changes in AD pathology

Hippocampi are bilateral structures within the medial temporal lobe, an essential part of the neural networks for encoding of short- and long-term memory, learning, and spatial navigation [44 –46]. Hippocampal structural, functional, and neurochemical changes in both MCI and AD pathological conditions have been measured using non-invasive neuroimaging techniques. These studies have demonstrated that LH is affected differently by pathological conditions than RH, reflecting impairment in semantic and episodic memory formation in AD [47]. Volumetric atrophy in the hippocampus has been associated with significant neuronal loss and amnestic cognitive impairment [48, 49]. Functional neuroimaging studies employing PET–MRI fusion and fMRI have reported hypometabolism and hyperactivity, respectively, in the hippocampus of amnestic MCI [50] and AD patients [51, 52]. High-field MRS has proven to be a promising non-invasive technique in identifying and quantifying metabolite levels and microenvironmental changes in brains of healthy and clinical population. Various studies have reported alterations in the hippocampal levels of metabolites (N-acetyl aspartate, creatine, myo-inositol, and GSH) in MCI and AD, when compared to healthy controls (HC) [25 , 53–55]. Studies have also demonstrated alterations in high-energy phosphates and tissue pH changes in the hippocampal region of AD compared to HC [26, 56].

Alteration of hippocampal GSH concentration and pH level

Hippocampal GSH concentration is non-invasively measured using ¹H MRS and acts as an oxidative stress (OS) marker. The mean GSH concentration showed a significant decrease in bilateral hippocampi of both MCI and AD groups when compared to HO (Fig. 3, Table 1). These findings are concurrent with those of a previously conducted MRS-based study [25] and a postmortem study [57]. ³¹P MRS is a multinuclear spectroscopy technique used to measure in vivo brain pH by calculating the chemical shift difference between PCr and Pi resonances. A cross-sectional ³¹P MRS healthy aging study has reported brain tissue pH decline with age [58]. Contrary to this, in AD pathology, studies reported significant increase in hippocampal tissue pH as compared to HO [26 , 59]. Our present findings also showed that the mean pH value in LH increases significantly toward alkalinity between HO and AD, whereas between HO and MCI groups, the increase is not significant. Moreover, in RH region, pH increases non-significantly in transition from HO to MCI and HO to AD (Fig. 3).

Oxidative stress in AD, and GSH and pH level alterations

OS occurs due to an imbalance between pro-oxidant and antioxidant levels. Excessive OS leads to free radical production, and these are further neutralized by the active sulfhydryl (-SH) group of Cys moiety of GSH. Thus, increased OS becomes the major driving factor for GSH depletion. Our findings (Fig. 3) indicating significant GSH depletion in the LH of MCI patients compared to HO are in concurrence with postmortem reports confirming GSH depletion [57]. Mechanisms underlying the rise in pH in the hippocampus are not fully understood, but most likely involved factors are presynaptic Ca²⁺/H⁺-ATPase [24], extracellular carbonic anhydrase [22], and GABA_A-receptor mediated bicarbonate efflux [23].

Inter-relation of GSH concentration and pH level in AD pathology

Our concurrent assessment of hippocampal GSH and pH showed decreasing trend of GSH and increasing trend of pH when HO was compared with MCI and AD in both LH and RH regions. To investigate the inter-relation of the two outcomes for their relative changes, the correlation analysis between the GSH concentration and pH level (Fig. 4) reveals their negative correlation in MCI group in both LH and RH regions. Similarly, in case of AD, increase in pH was reflected with respect to GSH depletion in both regions. However, in contrast to the disease pathology, the control group showed very nominal correlation indicating no change in hippocampal pH with GSH in both LH and RH regions.

Hippocampal GSH and pH as early diagnostic biomarkers for MCI and AD

The independent analysis of GSH concentration in our previous study reported LH region to be a better model with higher sensitivity (87.50%) in differentiating MCI patients from HC [25]. Our present results also support these findings for LH region with sensitivity of 84.21% and 66.0% diagnostic accuracy. Furthermore, to differentiate HO from AD, LH GSH was observed with 66.67% sensitivity, 90.32% specificity, and 80.01% accuracy. For the hippocampal pH, our result findings showed LH region to be more sensitive (80.0%) in differentiating MCI from HO. However, none of the regional GSH or pH independently were found to be effective for differentiating MCI from AD [25]. However, while combining the effects of both GSH and pH, it provides higher diagnostic accuracy performance. A significant area under ROC curve in both RH and LH regions was observed in the range of 0.80–0.90 to differentiate both MCI and AD from HO with sensitivity >75%. The results of ROC analyses reflected both GSH and pH to be diagnostic biomarkers for both MCI onset and progression to AD.

CONCLUSION AND FUTURE PERSPECTIVES

This is the first study to non-invasively investigate antioxidant GSH and tissue pH changes in hippocampal region of HO, MCI, and AD using a multinuclear MRS technique. The present outcomes reveal GSH depletion and pH increment in the hippocampus of AD brain as compared to controls. OS-induced common molecular events contributing to both GSH depletion and pH increment in hippocampus region in AD pathology are still largely unknown, creating a need for further in vitro studies to develop a working model. The present study was performed with a small sample size; therefore, for establishing an evidence-based test for clinical practice, these MRS findings need to be conducted longitudinally with a larger cohort. These findings will provide evidence for GSH and pH to be a potential non-invasively measured outcome, which may be included in upcoming clinical trials aimed at aiding in modifying the disease progression.

LIMITATIONS

In our cross-sectional study, we have generated data from modest number of patients with MCI and AD (N = 22 MCI and N = 37 AD). We have planned for a longitudinal study with larger patients’ pool involving MCI and AD patients in the coming months.

AUTHOR’S CONTRIBUTION

Prof. Pravat Kumar Mandal (Principal Investigator) conceptualized the idea, experimental design, recruitment of study participants, performed experiments (MEGA-PRESS for GSH detection and 31P MRS for pH) on control and patient groups, participated in the pH calibration experiment, preliminary analysis of data, literature search, writing and editing the manuscript. Dr. Deepika Shukla was involved in processing and analysis of both 1H and 31P MRS data using Kalpana package, writing and editing the manuscript, figure preparation, participated in the pH calibration experiment, literature search, data management, statistical analysis, result interpretation and discussion. Mr. Ritwick Mishra participated in statistical data analysis, data management, figure preparation, literature search, result interpretation and discussion, writing and editing the manuscript. Ms. Khushboo Punjabi participated in the pH calibration experiment, literature search, figure preparation, discussion, writing and editing manuscript. Ms. Divya Dwivedi participated in literature search, discussion, writing and editing the manuscript. Dr. Manjari Tripathi participated in planning and patient recruitment from AIIMS, New Delhi and participated in discussions. Ms. Vaishali was involved in data management.

Footnotes

ACKNOWLEDGMENTS

Dr. Pravat K. Mandal (Principal Investigator) is thankful to the Department of Biotechnology, Government of India (Award no. BT/PR7361/MED/30/953/2013), Indo-Australia Biotechnology Fund (Grant no. BT/Indo-Aus/10/31/2016), and Ministry of Information Technology, Govt. of India (Grant No 4(5)2019 ITEA) for funding this project. Partial financial support in the form of Tata Innovation Fellowship (No. BT/HRD/35/01/05/2014) awarded to Dr. Pravat K. Mandal from the Department of Biotechnology, Ministry of Science and Technology, Government of India is highly appreciated. We thank all participants, senior citizens, and patients for their voluntary participation, cooperation, and interest in this study. Thanks to Prof. Peter Barker and Prof. Richard Edden (Radiology, Johns Hopkins Medicine, Baltimore, Maryland, USA) for providing the MEGA-PRESS patch and long-standing collaboration. We thank Prof. Anirban Basu (Scientist, NBRC), Dr. Kanhaiya Lal Kumawat (Lab Technician, NBRC), and Prof. Ranjit Giri (Scientist, NBRC) for their help and laboratory support for phantom sample preparation related to pH measurement. Thanks to Dr. Gayatri Viswakarma (Head, Department of Biostatistics, Indian Spinal Injuries Centre, New Delhi, India) for her comments. Thanks to Ms. Avantika Samkaria (Project Assistant, NINS Lab) for proofreading the article.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Dementia, https://www.who.int/news-room/fact-sheets/detail/dementia.

Farias

, Mungas

, Reed

, Harvey

, DeCarli

(2009) Progression of mild cognitive impairment to dementia in clinic-vs community-based cohorts. Arch Neurol 66, 1151–1157.

Michaud

, Su

, Siahpush

, Murman

(2017) The risk of incident mild cognitive impairment and progression to dementia considering mild cognitive impairment subtypes. Dement Geriatr Cogn Dis Extra 7, 15–29.

Ronald

(2009) Early diagnosis of Alzheimer’s disease: Is MCI too late? Curr Alzheimer Res 6, 324–330.

Marshall

, Amariglio

, Sperling

, Rentz

(2012) Activities of daily living: Where do they fit in the diagnosis of Alzheimer’s disease? Neurodegener Dis Manag 2, 483–491.

Hardy

, Higgins

(1992) Alzheimer’s disease: The amyloid cascade hypothesis. Science 256, 184–185.

Goedert

(1993) Tau protein and the neurofibrillary pathology of Alzheimer’s disease. Trends Neurosci 16, 460–465.

Markesbery

(1997) Oxidative stress hypothesis in Alzheimer’s disease. Free Radic Biol Med 23, 134–147.

, Wang

, Geng

(2018) Alzheimer’s disease hypothesis and related therapies. Transl Neurodegener 7, 2.

10.

Swerdlow

(2007) Is aging part of Alzheimer’s disease, or is Alzheimer’s disease part of aging? Neurobiol Aging 28, 1465–1480.

11.

Swerdlow

(2007) Pathogenesis of Alzheimer’s disease. Clin Interv Aging 2, 347–359.

12.

de Paula

VJR

, Guimaraes

, Diniz

, Forlenza

(2009) Neurobiological pathways to Alzheimer’s disease: Amyloid-beta, tau protein or both? Dement Neuropsychol 3, 188–194.

13.

Nunomura

, Castellani

, Zhu

, Moreira

, Perry

, Smith

(2006) Involvement of oxidative stress in Alzheimer disease. J Neuropathol Exp Neurol 65, 631–641.

14.

Nunomura

, Perry

, Aliev

, Hirai

, Takeda

, Balraj

, Jones

, Ghanbari

, Wataya

, Shimohama

, Chiba

, Atwood

, Petersen

, Smith

(2001) Oxidative damage is the earliest event in Alzheimer disease. J Neuropathol Exp Neurol 60, 759–767.

15.

Markesbery

(1999) The role of oxidative stress in Alzheimer disease. Arch Neurol 56, 1449–1452.

16.

Keller

, Schmitt

, Scheff

, Ding

, Chen

, Butterfield

, Markesbery

(2005) Evidence of increased oxidative damage in subjects with mild cognitive impairment. Neurology 64, 1152–1156.

17.

Ding

, Markesbery

, Chen

, Li

, Keller

(2005) Ribosome dysfunction is an early event in Alzheimer’s disease. J Neurosci 25, 9171–9175.

18.

Aksenov

, Markesbery

(2001) Changes in thiol content and expression of glutathione redox system genes in the hippocampus and cerebellum in Alzheimer’s disease. Neurosci Lett 302, 141–145.

19.

Bessman

, Carpenter

(1985) The creatine-creatine phosphate energy shuttle. Annu Rev Biochem 54, 831–862.

20.

McLeish

, Kenyon

(2005) Relating structure to mechanism in creatine kinase. Crit Rev Biochem Mol Biol 40, 1–20.

21.

Wallimann

, Wyss

, Brdiczka

, Nicolay

, Eppenberger

(1992) Intracellular compartmentation, structure and function of creatine kinase isoenzymes in tissues with high and fluctuating energy demands: The ‘phosphocreatine circuit’ for cellular energy homeostasis. Biochem J 281 (Pt 1), 21–40.

22.

Chen

, Chesler

(1992) pH transients evoked by excitatory synaptic transmission are increased by inhibition of extracellular carbonic anhydrase. Proc Natl Acad Sci U S A 89, 7786–7790.

23.

Kaila

, Voipio

(1987) Postsynaptic fall in intracellular pH induced by GABA-activated bicarbonate conductance. Nature 330, 163–165.

24.

Makani

, Chesler

(2010) Rapid rise of extracellular pH evoked by neural activity is generated by the plasma membrane calcium ATPase. J Neurophysiol 103, 667–676.

25.

Mandal

, Saharan

, Tripathi

, Murari

(2015) Brain glutathione levels–a novel biomarker for mild cognitive impairment and Alzheimer’s disease. Biol Psychiatry 78, 702–710.

26.

Mandal

, Akolkar

, Tripathi

(2012) Mapping of hippocampal pH and neurochemicals from in vivo multi-voxel 31P study in healthy normal young male/female, mild cognitive impairment, and Alzheimer’s disease. J Alzheimers Dis 31 Suppl 3, S75–86.

27.

Wang

, Tan

, Wang

, Liu

, Yin

, Wang

, Chang

, Jiang

, Yu

(2015) Magnetic resonance spectroscopy in Alzheimer’s disease: Systematic review and meta-analysis. J Alzheimers Dis 46, 1049–1070.

28.

Gauthier

, Reisberg

, Zaudig

, Petersen

, Ritchie

, Broich

, Belleville

, Brodaty

, Bennett

, Chertkow

, Cummings

, de Leon

, Feldman

, Ganguli

, Hampel

, Scheltens

, Tierney

, Whitehouse

, Winblad

(2006) Mild cognitive impairment. Lancet 367, 1262–1270.

29.

Petersen

, Smith

, Waring

, Ivnik

, Tangalos

, Kokmen

(1999) Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 56, 303–308.

30.

McKhann

, Drachman

, Folstein

, Katzman

, Price

, Stadlan

(1984) Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34, 939–944.

31.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr. , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

32.

Terpstra

, Henry

, Gruetter

(2003) Measurement of reduced glutathione (GSH) in human brain using LCModel analysis of difference-edited spectra. Magn Reson Med 50, 19–23.

33.

Haase

, Frahm

, Hanicke

, Matthaei

(1985) 1H NMR chemical shift selective (CHESS) imaging. Phys Med Biol 30, 341–344.

34.

Gruetter

(1993) Automatic, localized in vivo adjustment of all first- and second-order shim coils. Magn Reson Med 29, 804–811.

35.

Shukla

, Mandal

, Ersland

, Gruner

, Tripathi

, Raghunathan

, Sharma

, Chaithya

, Punjabi

, Splaine

(2018) A multi-center study on human brain glutathione conformation using magnetic resonance spectroscopy. J Alzheimers Dis 66, 517–532.

36.

Shukla

, Mandal

, Tripathi

, Vishwakarma

, Mishra

, Sandal

(2020) Quantitation of in vivo brain glutathione conformers in cingulate cortex among age-matched control, MCI, and AD patients using MEGA-PRESS. Hum Brain Mapp 41, 194–217.

37.

Shaka

, Keeler

, Freeman

(1983) Evaluation of a new broadband decoupling sequence: WALTZ-16. J Magn Reson 53, 313.

38.

van Sluis

, Payne

, Leach

(1995) Increased NOE enhancement in 1H decoupled 31P MRS. Magn Reson Med 34, 893–897.

39.

Mandal

, Shukla

(2020) KALPANA: Advanced spectroscopic signal processing platform for improved accuracy to aid in early diagnosis of brain disorders in clinical setting. J Alzheimers Dis 75, 397–402.

40.

Pijnappel

, van den Boogaart

, van Ormondt

(1992) SVD-based quantification of magnetic resonance signals. J. Mag Resonance 97, 171–178.

41.

Golyandina

, Shlemov

(2015) Variations of Singular Spectrum Analysis for separability improvement: Nnon-orthogonal decompositions of time SERIES. Stat Interface 8, 277–294.

42.

Barbiroli

, Montagna

, Cortelli

, Iotti

, Lodi

, Barboni

, Monari

, Lugaresi

, Frassineti

, Zaniol

(1995) Defective brain and muscle energy metabolism shown by in vivo 31P magnetic resonance spectroscopy in nonaffected carriers of 11778 mtDNA mutation. Neurology 45, 1364–1369.

43.

DeLong

, DeLong

, Clarke-Pearson

(1988) Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics 44, 837–845.

44.

Hartley

, Bird

, Chan

, Cipolotti

, Husain

, Vargha-Khadem

, Burgess

(2007) The hippocampus is required for short-term topographical memory in humans. Hippocampus 17, 34–48.

45.

Eichenbaum

(2017) The role of the hippocampus in navigation is memory. J Neurophysiol 117, 1785–1796.

46.

Kumaran

(2008) Short-term memory and the human hippocampus. J Neurosci 28, 3837–3838.

47.

Ezzati

, Katz

, Zammit

, Lipton

, Zimmerman

, Sliwinski

, Lipton

(2016) Differential association of left and right hippocampal volumes with verbal episodic and spatial memory in older adults. Neuropsychologia 93, 380–385.

48.

Morra

, Tu

, Apostolova

, Green

, Avedissian

, Madsen

, Parikshak

, Hua

, Toga

, Jack

Jr. , Schuff

, Weiner

, Thompson

, Alzheimer’s Disease Neuroimaging Initiative (2009) Automated 3D mapping of hippocampal atrophy and its clinical correlates in 400 subjects with Alzheimer’s disease, mild cognitive impairment, and elderly controls. Hum Brain Mapp 30, 2766–2788.

49.

Frisoni

, Ganzola

, Canu

, Rub

, Pizzini

, Alessandrini

, Zoccatelli

, Beltramello

, Caltagirone

, Thompson

(2008) Mapping local hippocampal changes in Alzheimer’s disease and normal ageing with MRI at 3 Tesla. Brain 131, 3266–3276.

50.

Bakker

, Krauss

, Albert

, Speck

, Jones

, Stark

, Yassa

, Bassett

, Shelton

, Gallagher

(2012) Reduction of hippocampal hyperactivity improves cognition in amnestic mild cognitive impairment. Neuron 74, 467–474.

51.

Zilberter

, Zilberter

(2017) The vicious circle of hypometabolism in neurodegenerative diseases: Ways and mechanisms of metabolic correction. J Neurosci Res 95, 2217–2235.

52.

Choi

, Son

, Noh

, Lee

, Kim

, Park

(2018) Glucose hypometabolism in hippocampal subdivisions in Alzheimer’s disease: A pilot study using high-resolution (1)(8)F-FDG PET and 7.0-T MRI. J Clin Neurol 14, 158–164.

53.

Wang

, Zhao

, Yu

, Zhou

, Li

(2009) Regional metabolic changes in the hippocampus and posterior cingulate area detected with 3-Tesla magnetic resonance spectroscopy in patients with mild cognitive impairment and Alzheimer disease. Acta Radiol 50, 312–319.

54.

Wang

, Xiao

, Li

, Ding

, Ling

, Chen

, Fang

(2011) Using proton magnetic resonance spectroscopy to identify mild cognitive impairment. Int Psychogeriatr 24, 19–27.

55.

Caserta

, Ragin

, Hermida

, Ahrens

, Wise

(2008) Single voxel magnetic resonance spectroscopy at 3 Tesla in a memory disorders clinic: Early right hippocampal NAA/Cr loss in mildly impaired subjects. Psychiatry Res 164, 154–159.

56.

Rijpma

, van der Graaf

, Meulenbroek

, Olde Rikkert

MGM

, Heerschap

(2018) Altered brain high-energy phosphate metabolism in mild Alzheimer’s disease: A 3-dimensional (31)P MR spectroscopic imaging study. Neuroimage Clin 18, 254–261.

57.

Sultana

, Piroddi

, Galli

, Butterfield

(2008) Protein levels and activity of some antioxidant enzymes in hippocampus of subjects with amnestic mild cognitive impairment. Neurochem Res 33, 2540–2546.

58.

Forester

, Berlow

, Harper

, Jensen

, Lange

, Froimowitz

, Ravichandran

, Iosifescu

, Lukas

, Renshaw

, Cohen

(2010) Age-related changes in brain energetics and phospholipid metabolism. NMR Biomed 23, 242–250.

59.

Mecheri

, Marie-Cardine

, Sappey-Marinier

, Bonmartin

, Albrand

, Ferry

, Coppard-Meyer

, Courpron

(1997) In vivo hippocampal (31)P NMR metabolites in Alzheimer’s disease and ageing. Eur Psychiatry 12, 140–148.

Hippocampal Glutathione Depletion and pH Increment in Alzheimer’s Disease: An in vivo MRS Study

Abstract

Background:

Objective:

Methods:

Results:

Conclusion:

Keywords

INTRODUCTION

MATERIALS AND METHODS

Study participants

MRI and multinuclear 1H and 31P MRS data acquisition

MRI data acquisition

1H MRS data acquisition for GSH quantitation

31P MRS data acquisition for pH measurement

Multinuclear 1H and 31P MRS data processing

1H MRS MEGA-PRESS data processing for GSH quantitation

31P MRS data processing for pH measurement

Statistical analysis

RESULTS

Participant characteristics

GSH concentration and pH level alteration in MCI and AD compared to HO

Correlation between hippocampal GSH concentration and pH level among HO, MCI, and AD

Diagnostic accuracy for GSH concentration and pH level to differentiate MCI and AD from HO

DISCUSSION

Hippocampal structural and functional changes in AD pathology

Alteration of hippocampal GSH concentration and pH level

Oxidative stress in AD, and GSH and pH level alterations

Inter-relation of GSH concentration and pH level in AD pathology

Hippocampal GSH and pH as early diagnostic biomarkers for MCI and AD

CONCLUSION AND FUTURE PERSPECTIVES

LIMITATIONS

AUTHOR’S CONTRIBUTION

Footnotes

ACKNOWLEDGMENTS

References

MRI and multinuclear ¹H and ³¹P MRS data acquisition

¹H MRS data acquisition for GSH quantitation

³¹P MRS data acquisition for pH measurement

Multinuclear ¹H and ³¹P MRS data processing

¹H MRS MEGA-PRESS data processing for GSH quantitation

³¹P MRS data processing for pH measurement