Reversal of LTP-Like Cortical Plasticity in Alzheimer’s Disease Patients with Tau-Related Faster Clinical Progression

Abstract

Cerebrospinal fluid (CSF) concentrations of amyloid-β (Aβ), total tau (t-tau), and phosphorylated tau proteins are associated with different clinical progression in Alzheimer’s disease (AD). We enrolled forty newly diagnosed AD patients, who underwent lumbar puncture, and carried out a K-means cluster analysis based on CSF biomarkers levels, resulting in two AD patient groups: Cluster 1 showed relatively high levels of Aβ and low levels of tau; Cluster 2 showed relatively low levels of Aβ and high levels of tau. Cortical plasticity was tested using the intermittent and continuous theta burst stimulation (iTBS and cTBS) protocols evoking respectively long-term potentiation (LTP) and depression (LTD). Cholinergic transmission was tested by the short-latency afferent inhibition protocol. Neurophysiological evaluation showed that the two AD groups differed in terms of cortical plasticity: after iTBS, Cluster 2 patients showed a remarkable reversal of LTP toward LTD that was not observed in Cluster 1. LTD and central cholinergic transmission did not differ between groups. Patients were assessed longitudinally with Mini-Mental State Examination at 6, 12, and 18 month follow-ups. Cluster 2 AD had a faster cognitive decline already evident at the 12 month follow-up. High tau CSF levels were associated with LTD-like cortical plasticity and faster clinical progression. These results suggest that more aggressive tau pathology is associated with prominent LTD-like mechanisms of cortical plasticity and faster cognitive decline.

Keywords

Alzheimer’s disease amyloid-beta cortical plasticity long-term depression long-term potentiation tau transcranial magnetic stimulation

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by a remarkable clinical heterogeneity that, although still scarcely investigated and debated, accounts for the current amyloid cascade hypothesis, which is insufficient to account for such variability [1]. Cerebrospinal fluid (CSF) concentrations of amyloid-β (Aβ), total tau (t-tau), and phosphorylated tau (p-tau) proteins have been recently put forward as a promising tool to identify different groups of AD patients [2, 3]. Cluster analyses of Aβ_1 - 42, t-tau, and p-tau CSF levels have recently revealed the existence of different AD patients’ populations, each associated with dissimilar levels of memory, mental speed, and executive functions [4]. Indeed, these groups of AD patients face a different disease progression [5], with patients with very high levels of CSF t-tau and p-tau exhibiting a more malignant disease course [5, 6]. In animal models of AD, Aβ peptides and tau proteins are known to interfere with physiological mechanisms of neuronal synaptic plasticity. In particular it has been demonstrated that these molecules influence hippocampal long-term potentiation (LTP), an electrophysiological correlate of learning and memory [1 , 7–11]. These altered mechanisms have been related to spine shrinkage, neuronal network disarrangement, and cell death [12]. Despite this evidence, the impact of CSF Aβ_1 - 42, t-tau, andp-tau levels on synaptic transmission in AD patients has been scarcely investigated. In humans, cortical plasticity measures such as LTP and long term depression (LTD) can be obtained by applying non-invasive repetitive transcranial magnetic stimulation (TMS) over the primary motor cortex, using theta burst stimulation (TBS) protocols that mimic those described in animal models [13]. When tested with these methods, AD patients typically show a marked impairment of long-term potentiation-like cortical plasticity [14, 15]. Indeed, central cholinergic dysfunction can be tested in vivo in AD patients through the short-latency afferent inhibition (SAI) protocol by applying TMS over primary motor cortex [16 –19]. The motor cortex is considered a reliable model to investigate early changes in cortical plasticity and central cholinergic transmission occurring in AD patients who are affected only at later stages of disease, when AD becomes clinically manifest [20]. Therefore, in this study we aimed at investigating the neurophysiological features of cortical plasticity and cholinergic transmission in groups of AD patients characterized by heterogeneouspatterns of CSF concentrations of Aβ_1 - 42, t-tau, and p-tau proteins. We hypothesized that different groups of AD patients classified according to the CSF levels of biomarkers would have not only peculiar clinical traits [4, 5] but also specific features of cortical plasticity eventually associated with worse clinical progression.

METHODS

Subjects

Forty consecutive patients, admitted for complaining symptoms, were recruited at the memory clinic of the University Hospital Tor Vergata. After the first visit to our Centre, all patients underwent a complete clinical investigation, for diagnostic purposes, during a period not longer than 60 days, including medical history, neurological examination, Mini-Mental State Examination (MMSE), a complete blood screening, neuropsychological assessment, a neuropsychiatric evaluation,magnetic resonance or CT imaging, PET/CT, and lumbar puncture for CSF analysis (Table 1). Patients fulfilled the clinical criteria of dementia as defined by the DSM-IV and probable or possible AD according to the criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s disease and Related Disorders Association [21]. Disease duration was calculated using standardized semi-structured questions [22]. Neurophysiological examinations were performed at theSanta Lucia Foundation within 30 days from CSF sampling. In the 90 days prior to TMS evaluation, none of the patients were treated with drugs that could have modulated cerebral cortex excitability such as Acetylcholinesterase inhibitors [23], antidepressants, or any other neuroactive drugs (i.e., benzodiazepines, anti-epileptic drugs, or neuroleptics). Exclusion criteria were the following: patients with isolated cognitive deficits, patients with clinically manifest acute stroke in the last 6 months showing a Hachinsky scale score >4, and a radiological evidence of ischemic lesions, Aβ_1 - 42 CSF values >600 pg/mL. After the neurophysiological assessment patients started treatment with rivastigmine patch (n = 23) or donepezil (n = 17) and were followed longitudinally with clinical assess-ments and MMSE at 6, 12, and 18 months. Twenty-four age-, gender-, and education-matched healthy subjects (HS) were recruited as controls. Each participant or their legal guardian provided written informed consent after receiving an extensive description of the study. The study was performed according to the Declaration of Helsinki. The ethics committee of the Santa Lucia Foundation IRCSS approved this protocol (Prot. CE/AG4/PROG.392-08).

Cognitive evaluation

At the time of enrollment, all recruited patients underwent a neuropsychological battery (Table 2), which included the following cognitive domains: general cognitive efficiency: MMSE [24, 25]; verbal episodic long-term memory: Rey auditory verbal long-term memory (15-Word List Immediate and 15-min Delayed recall) [26]; visuo-spatial abilities and visuo-spatial episodic long-term memory: Complex Rey’s Figure (copy and 10-min Delayed recall) [27]; executive functions: phonological word fluency [28]; analogic reasoning: Raven’s Colored Progressive Matrices [28]. For all employed tests, we used the Italian normative data for both score adjustment (gender, age, and education), and to define cut-off scores of normality, we determined a 95% tolerance interval as the lower limit. For each test, normative data are reported in the corresponding references.

CSF biomarkers analysis and APOE genotype

The first 12 mL of CSF were collected in a polypropylene tube and directly transported to the local laboratory for centrifugation at 2000×g at + 4°C for 10 min. The supernatant was pipetted off, gently stirred and mixed to avoid potential gradient effects, and aliquoted in 1 mL portions in polypropylene tubes that were stored at –80°C pending biochemical analyses, without being thawed and re-frozen. CSF t-tau and p-tau phosphorylated at Thr181 concentrations were determined using a sandwich ELISA (Innotest hTAU-Ag, Innogenetics, Gent, Belgium). Aβ_1 - 42 levels were determined using a sandwich ELISA (Innotest^® β-amyloid (1–42), Innogenetics, Gent, Belgium), specifically constructed to measure Aβ-amyloid containing both the first and 42nd amino-acid, as previously described [29].

Transcranial magnetic stimulation

All patients and healthy controls underwent continuous TBS (cTBS), intermittent TBS (iTBS), and SAI protocol in three different sessions, with at least a three day interval between each session. The order of the sessions was pseudo-randomized across patients and healthy controls. Motor evoked potentials (MEP) were recorded from the right first dorsal interosseous muscle using 9 mm diameter, Ag–AgCl surface cup electrodes. Responses were amplified with a Digitimer D360 amplifier (Digitimer Ltd, UK) and filtered (20 Hz-2 kHz), then recorded by computer using SIGNAL software with a sampling rate of 5 kHz per channel (Cambridge Electronic Devices, UK). A monophasicMagstim 200 device (Magstim Co, UK) was used to define the motor hot spot and to assess MEP size using standard 70 mm figure-of-eight shaped coil. The motor hot-spot was defined as the location where TMS consistently produced the largest MEP size at 120% of resting motor threshold (RMT) in the targetmuscle [30]. A second coil was connected to a biphasic Super Rapid Magstim stimulator (Magstim Co, UK) to deliver TBS. In the cTBS protocol bursts at 80% active motor threshold were repeated at 5 Hz (i.e., every 200 ms), while each burst consisted of three stimuli repeating at 50 Hz, for 40 s (600 pulses). In the iTBS protocol, a 2 s train of TBS was repeated 20 times, every 10 s for a total of 190 s (600 pulses) [13]. Twenty MEPs were collected and averaged at baseline. The intensity of the test pulse was adjusted so that it evoked a motor potential of about 1 mV peak-to-peak amplitude in each individual. Then, over the same hot-spot, twenty MEPs were recorded at 1–5, 6–10, 11–15, 16–20, and 21–25 min after TBS and averaged. SAI was studied using the technique that has been recently described [31, 32]. Conditioning stimuli were single pulses (200μs) of electrical stimulation applied through bipolar electrodes to the right median nerve at the wrist (cathode proximal). The intensity of the conditioning stimulus was set at just over motorthreshold for evoking a visible twitch of the thenar muscles. The intensity of the test cortical magnetic stimulus was adjusted to evoke a MEP in the relaxed right first dorsal interosseous with amplitude of approximately 1 mV peak to peak. The conditioning stimuli to the peripheral nerve preceded the magnetic test stimulus by different interstimulus intervals (ISIs), ranging from –4 to +8 ms from the N20 in steps of 4 ms [33]. Ten stimuli were delivered at each ISI. The subject was given audiovisual feedback at high gain to assist in maintaining complete relaxation. The inter-trial interval was set at 5 s (±10%), for a total duration of approximately five minutes. Measurements were made on each individual trial. The mean peak-to peak amplitude of the conditioned motor evoked potential at each ISI was expressed as a percentage of the mean peak-to-peak amplitude size of the unconditioned test pulse in that block.

Data analysis

Data were analyzed using SPSS for Windows version 11.0. A hierarchical cluster analysis (Ward method) was applied to assess the different solutions based on minimum variance within clusters and relatively equal cluster sizes as in Wallin [5]. The sequence of mergers in the dendrogram suggested a 2-cluster solution. A K-mean cluster analysis defining 2 clusters based on the CSF biomarkers Aβ_1 - 42, t-tau, and p-tau was performed. The K-mean cluster analysis forms groups to obtain maximal differences in cluster variables between clusters and minimal differences within groups and is preferred when non-normally distributed data are analyzed. To further characterize the cognitive profile of each cluster, one-way ANOVAs with GROUP (Cluster 1 vs. Cluster 2 AD patients) as between subjects main factor were performed separately on MMSE scores, immediate and delayed Rey auditory verbal long term memory scores, copy and delayed copy of complex Figure’s Rey scores; phonological word fluency scores and Raven’s Colored Progressive Matrices scores. When a statistically significant effect was observed, Unequal N HSD tests were used for post-hoc analyses. The threshold of significance was set at p < 0.05. The cognitive outcome at 2 months and the annual change in MMSE were analyzed with general linear models between the CSF clusters, adjusting for the baseline MMSE level, age, gender, and presence of APOE 4 allele.

For TMS experiments, two-way repeated measure ANOVAs were performed on MEP amplitude, expressed as percentage of change in comparison to baseline for each TBS protocol (cTBS and iTBS) with TIME (1–5, 6–10, 11–15,16–20, and 21–25 min after TBS) as within subjects’ factors and GROUP (Cluster 1, Cluster 2, Healthy controls) as between subjects’ factors. For short-latency afferent inhibition the electrophysiological parameters of AD patients were compared by means of repeated measures ANOVA with ISI (–4, 0, +4, and +8 ms plus the latency of the N20) as within subjects’ factors and GROUP (Cluster 1, Cluster 2, Healthy controls) as between subjects’ factors. The Greenhouse-Geisser correction was used for non-spherical data. When a significant main effect was reached, paired t-tests with Bonferroni correction were employed to characterize the different effects of the specific ISIs. Mauchley’s test examined for sphericity. For all statistical analyses, a p value of <0.05 was considered to be significant.

To further explore the inter-individual variability of the response to the iTBS, cTBS and SAI, we conducted an additional cluster analysis based initially on individual neurophysiological (MEP) data and not on the CFS biomarkers in AD patients (as in the main analysis described above). SPSS TwoStep cluster analysis was performed to determine if there are patters of response for each TMS protocol as described in Lopez-Alonso et al. [34]. This clustering method determines the optimal number of clusters that best explains variance in the data automatically. MEP amplitudes of each block, normalized to the baseline, were used for this analysis. Separate ANOVAs of repeated measurement were conducted for eachcluster (when established) and each TMS protocol over the absolute MEP values. Finally, “t” student for independent measurements were conducted to compare the CSF biomarkers between clusters.

RESULTS

CSF biomarkers

Two groups of AD patients were classified using the K-means cluster analysis. Cluster 1 consisted of 20 AD patients with relatively high levels of Aβ_1 - 42 (358.8±127.5 mean±SD, pg/mL) and low levels of t-tau (357.5±190.2 mean±SD, pg/mL) and p-tau (63.8±31.1 mean±SD, pg/mL). Cluster 2 consisted of 20 AD patients with lower levels of Aβ_1 - 42 (245.6±122.1 mean±SD, pg/mL) and high levels of t-tau (989.05±183.1 mean±SD, pg/mL) and p-tau (114.15±44.8 mean±SD, pg/mL) (Table 1). The 2 clusters differed in CSF Aβ_1 - 42 values (F(1, 38) = 8.42, p = 0.006), t-tau values (F(1, 36) = 82.40, p = 0.000001), and p-tau values (F(1, 38) = 17.61, p = 0.00016). The 2 clusters did not differ in gender, APOE genotype, education, age at disease onset, disease duration at baseline, or autonomies of daily living levels at baseline (Table 1). The two clusters showed differences in their cognitive profile (Table 2). There were higher MMSE scores in Cluster 1 when compared to Cluster 2 (F(1, 38) = 7.96, p = 0.007) and higher scores at the Raven’s Colored Progressive Matrices in Cluster 1 when compared to Cluster 2 (F(1, 38) = 8.46, p = 0.006).

Transcranial magnetic stimulation

The TMS procedures were well tolerated in all subjects. The mean (SD) RMT to TMS was not different between the two AD groups (Cluster 1:39.8±5.2 % MSO versus Cluster 2:36.2±6.4 % MSO) of maximum stimulator output. However, Cluster 2, but not Cluster 1 AD patients, had lower RMT compared to HS (36.2±6.4 % MSO versus 42.2±5.4 % MSO; p = 0.005). Baseline mean motor evoked potentials amplitude (SD) did not differ between Cluster 1 and 2 AD patients and HS (Cluster 1:1.18±0.44 mV; Cluster 2:1.21±0.29 mV; HS: 1.11±0.39 mV). For the intermittent TBS protocol there was an effect for the GROUP (F(2, 61) = 23.36; p = 0.00001) but not the TIME (4, 152) = 1.50; p = 0.22) main factor; the interaction GROUP×TIME was significant (F(4, 152) = 2.14; p = 0.03). Cluster 2 AD patients showed an overall more altered LTP-like cortical plasticity, with a reversal of LTP-like cortical plasticity towards LTD in comparison with Cluster 1 (Fig. 1), as confirmed by post-hoc analysis with Bonferroni correction at 5, 10, 15, and 20 min time points (all p < 0.005). For the continuous TBS protocol, the repeated measure ANOVA performed on the percentage changes of the mean motor evoked potential amplitude did not show any effect for the GROUP main factor: F(2, 61) = 0.21; p = 0.81 and for the GROUP×TIME interaction (F(8, 244) = 0.73; p = 0.66). Only the TIME main factor was significant (F(8, 244) = 3.12; p = 0.03) (Fig. 2). The ANOVA analysis performed on short-latency afferent inhibition measurements at baseline showed a significant ISIs (F(6, 183) = 11.4; p = 0.00001) and GROUP (F(2, 61) = 3.55; p = 0.031) main factor. Post hoc analysis showed that both AD clusters had weaker SAI in comparison with healthy controls independently from ISI (all p < 0.01). The interaction GROUP×ISIs was not significant (F(6, 183) = 1.39; p = 0.21) (Fig. 3).

In a separate analysis, we employed Pearson’s r correlation coefficient in univariate correlations in order to explore any influence the CSF values of Aβ_1 - 42,t-tau, and p-tau could have on the individual amount of mean change induced by the iTBS and cTBS protocol and by SAI protocol at ISI = 0 across all AD patients. We found that t-tau levels correlated with both iTBS (p = 0.026; r² = 0.122) and cTBS (p = 0.042; r² = 0.104) LTD-like induced effects; similarly, p-tau levels correlated with iTBS (p = 0.023; r² = 0.127) LTD-like induced effects (Fig. 4). CSF values of Aβ_1 - 42 did not correlate with any iTBS nor cTBS measure (Fig. 4). We did not find any correlation for SAI parameters with CSF values.

Additional two-step cluster analysis performed on individual MEP amplitude data revealed the existence of two different clusters of AD patients for the iTBS but not for cTBS and SAI protocols. For the iTBS data, Cluster 1 and Cluster 2 consisted of eight and thirty-two AD patients, respectively. The ANOVA analysis revealed that AD patients in Cluster 1 (n = 8) showed MEP facilitation, although not significant in comparison with baseline, in response to the iTBS. However, AD patients in Cluster 2 (n = 32) showed a significant MEP inhibition (F(1, 38) = 16.00; p < 0.001). Post-hoc analysis revealed that all the post-stimulation MEPs amplitude were significant lower in comparison with the baseline MEP amplitude (p < 0.001 for all the time points) (Fig. 5). Interestingly, the tau and p-tau levels were significantly lower in Cluster 1 in comparison with Cluster 2 (t = 2.86; p = 0.007 and t = 2.23; p = 0.032, respectively), confirming the findings obtained in the main analysis in which AD patients were first divided in two groups depending on the individual CSF biomarkers in which higher tau CSF levels were associated with LTD (i.e. MEP inhibition).

Clinical follow-up

During follow-up AD patients in both clusters were all treated with standard acetylcholinesterase inhibitors therapy. The two clusters did not differ in the number of patients undergoing different cholinesterase-inhibitors treatment (Cluster 1:11 patients under rivastigmine and 9 patients under donepezil treatment; Cluster2:12 patients under rivastigmine and 8 patients underdonepezil treatment). Follow-up evaluation, with MMSE scores performed at 6, 12, and 18 months, revealed that Cluster 2 progressed faster than Cluster 1 AD patients as shown by ANOVA (significant MMSE score×TIME interaction (F(3, 114) = 3.97, p = 0.009) (Fig. 6). Post Hoc analyses revealed that in Cluster 2 AD patients’ MMSE scores were lower than baseline evaluation as early as at 12 months (p = 0.0002) and showed a further step of decline between the 12 months and the 18 months follow-up (p = 0.0001). On the other hand, in Cluster 1 AD patients’ MMSE scores became different (lower) from baseline only after 18 months (p = 0.02). We then employed Pearson’s r correlation coefficient in univariate correlations in order to explore any influence the CSF values of Aβ_1 - 42, t-tau, and p-tau could have on MMSE scores at baseline and at follow-up across all patients. We found that only t-tau, but not p-tau or Aβ_1 - 42 levels, correlated with MMSE scoresat baseline (p = 0.014; r² = 0.147). T-tau, but not p-tau or Aβ_1 - 42 levels, correlated with the severity of cognitive decline (measured in delta MMSE scores at 18 months) (p = 0.034; r² = 0.112). We also performed Spearman correlation analyses between cognitive decline (delta score with baseline evaluation) and iTBS and cTBS induced cortical plasticity (individual mean value). For both protocols, AD patients presenting with more evident LTD-like cortical plasticity had more severe cognitive decline at 12 months (cTBS: p = 0.026; r² = 0.123; iTBS: p = 0.018; r² = 0.137) and at 18 months (cTBS: p = 0.044; r² = 0.102; iTBS: p = 0.020; r² = 0.133) (Fig. 7). No correlations were found for SAI values. These results indicate that AD patients with greater tau pathology and more pronounced LTD-like cortical plasticity tend to have faster cognitive decline.

DISCUSSION

The CSF-based cluster analyses allowed us to identify two groups of newly diagnosed AD patients showing clear-cut differences in their clinical and neurophysiological presentation. Critical for the study, the 2 groups of patients were completely undistinguishable for all demographic variables, genetic risk (apoe4 distribution across groups), vascular risk factors (diabetes, hypertension, hypercholesterol), clinical onset (all amnestic forms), and disease duration calculated using standardized semi-structured questions [22].

This means that the differences in cognitive performances observed at baseline cannot be regarded as expression of different disease stages. All of our AD patients were studied at the time of the first diagnosis at the memory clinic. Thus a more altered CSF biomarkers profile at the time of clinical assessment likely reflected an underlying different and more aggressive pathology. This is in line with previous studies showing that AD patients with relatively high levels of t-tau and p-tau and low levels of Aβ_1 - 42 are generally more cognitively impaired [4]. The first main finding of our study is that AD patients with altered CSF biomarkers profile show remarkable neurophysiological features. When tested with cTBS both groups showed the expected strong MEP inhibition, as already described in previous studies [35] (spared LTD). When tested with the iTBS protocol Cluster 1 AD patients showed the expected absence of LTP [14, 35]. Cluster 2 AD patients exhibited a surprising result, showing a clear reversal of LTP toward LTD. Therefore, these data revealed that AD patients with more pathological CSF tau biomarkers are characterized by a strong tendency of cortical plasticity to form LTD. These findings were supported by another independent cluster analysis conducted on the MEP data (and not on CSF biomarkers). Thirty-two AD patients (80% of the total sample) showed an inhibitory response to iTBS, a protocol that is expected to induce MEP facilitation at least in 43% of control subjects [34]. Therefore, this abnormal response to iTBS cannot be explained by the variability associated to this TMS protocol [34, 36]. Importantly, correlation analyses performed across all patients revealed that this neurophysiological feature is associated with t-tau, but not CSF Aβ1-42 values. In addition, AD patients with an inhibitory effect in response to iTBS were also characterized by higher levels of tau and p-tau values in comparison with AD patients who did not show a similar inhibition. These data confirm the findings that CSF tau levels are strongly related with the burdened synaptic activity in AD patients [37] in which glutamatergic [38], cholinergic, and dopaminergic [39] transmission is deeply altered by the occurring synaptic remodeling.

These findings find support on the idea that tau is not considered a mere constituent of the cell cytoskeleton, but has been suggested to play a role as regulator of physiological synaptic function similarly to Aβ peptides [40 –43]. The localization of tau in pre- and post-synaptic sites, and its relationship with post-synaptic cell signaling machinery, strongly supports this hypothesis [42]. Moreover Aβ peptides and tau proteins exist also in soluble forms (oligomers) that can be released in the extracellular space where they may induce direct toxic effects on neuronal transmission [43]. For instance tau oligomers derived from brains of postmortem AD patients are potent inhibitors of long-term potentiation in hippocampal brain slices [11]. In this regard, these oligomers can be considered as acute toxic tau species affecting in vivo both mitochondrial and synaptic function [10]. It is also possible that the synaptic dysfunction associated with tau could be modulated by other interrelated factors, such as the inflammation cascade products [44] interleukins (ILs) IL-1β or IL-6 [45].

The second important finding of our study is that AD patients presenting with more pathological CSF tau levels undergo to a faster cognitive decline. These data are in line with the novel idea that AD patients with more pathological CSF tau biomarkers face a more aggressive disease progression [4 , 48]. In particular correlation analyses revealed that faster cognitive decline was associated with more pronounced LTD-like cortical plasticity. This presents the possibility that an altered cortical plasticity, eventually caused by tau pathology, could be linked to the clinical worsening. In this regard, recent studies of the molecular mechanisms of LTD suggest a crucial role for the signaling pathways of apoptosis (programmed cell death) in the weakening and elimination of synapses and dendritic spines [49]. LTD has been repeatedly indicated as an electrophysiological correlate of excitatory synapse elimination. Given that apoptosis mechanisms seem to underlie LTD [50], it is appealing to consider LTD and the associated spine shrinkage and elimination as different facets of an overall process of synapse involution. Further studies are needed to test this fascinating hypothesis in AD.

We did not find any correlation between the SAI values with CSF values nor with cognitive progression, despite both AD clusters showed impairment of cholinergic transmission in comparison with healthy controls. On the other hand, there was a strong correlation among TBS-induced plasticity at the time of the first neurophysiological evaluation and clinical progression. To the best of our knowledge, the current study is the first to put in relation in AD patients neurophysiological parameters of cortical plasticity and cholinergic function (SAI) with clinical outcome. Thus, these findings suggest that cortical plasticity assessed by TBS could represent the most useful neurophysiological biomarker in AD patients in terms of clinical evolution prediction.

The two identified clusters show different levels of CSF biomarkers, with Cluster 2 showing more severe clinical progression. It is possible that cognitive reserve could have played a role in determining faster cognitive decline. Further studies are needed to clarify the interaction among CSF biomarkers, cortical plasticity, and cognitive reserve [51 –54].

A limitation of the current study is that, due to methodological constrains, TMS measures were limited to the motor system/motor cortex while the pathology and dysfunction in AD is obviously more widespread. It thus stands to wonder whether such TMS neurophysiologic measures of the motor system/cortex may not simply be missing alterations in other cortical regions. Combining TMS with electroencephalography could allow the testing of similar mechanisms of cortical plasticity in other non-motor cognitive areas such as the posterior parietal cortex [55]. In this regard, additional studies combining C-Pittsburgh compound B (PiB)-PET [7] and ¹⁸F-FDG-PET [20] or with novel tau tracers with our combined CSF-based cluster analysis and with neurophysiology could be important to clarify these unsolved questions and to further differentiate these groups [56].

In conclusion, heterogeneity of AD appears to be one of the most important features to understand and treat cognitive decline progression. Our study, using for the first time neurophysiological tools associated with CSF biomarker levels, allowed us to determine that AD patients differ in terms of synaptic transmission and faster clinical progression depending on the tau CSF levels. The current study on one hand deepens our understanding on the AD-related synaptic mechanisms and on the other provides the basis for new therapeutic strategies directed to treat different clinical features in well-defined groups of AD patients [14, 57].

Footnotes

ACKNOWLEDGMENTS

This work was supported by grants of the Italian Ministry of Health to Giacomo Koch (239/GR-2009-1591859) and to Marco Bozzali (47/RF-2010-2311484).

Authors’ disclosures available online ().

References

Hardy

, Selkoe

(2002) The amyloid hypothesis of Alzheimer’s disease: Progress and problems on the road to therapeutics. Science 297, 353–356.

Mattsson

, Zetterberg

, Hansson

, Andreasen

, Parnetti

, Jonsson

, Herukka

, van der Flier

, Blankenstein

, Ewers

, Rich

, Kaiser

, Verbeek

, Tsolaki

, Mulugeta

, Rosán

, Aarsland

, Visser

, Schröder

, Marcusson

, de Leon

, Hampel

, Scheltens

, Pirttilä

, Wallin

, Jönhagen

, Minthon

, Winblad

, Blennow

(2009) CSF biomarkers and incipient Alzheimer disease in patients with mild cognitive impairment. JAMA 302, 385–393.

Schmidt

, Nixon

, Mathews

(2012) Tissue processing prior to analysis of Alzheimer’s disease associated proteins and metabolites, including Aβ. Methods Mol Biol 849, 493–506.

van der Vlies

, Verwey

, Bouwman

, Blankenstein

, Klein

, Scheltens

, van der Flier

(2009) CSF biomarkers in relationship to cognitive profiles in Alzheimer’s disease. Neurology 72, 1056–1061.

Wallin

, Blennow

, Zetterberg

, Londos

, Minthon

, Hansson

(2010) CSF biomarkers predict a more malignant outcome in Alzheimer disease. Neurology 74, 1531–1537.

Koch

, Belli

, Giudice

, Lorenzo

, Sancesario

, Sorge

, Bernardini

, Martorana

(2013) Frailty among Alzheimer’s disease patients. CNS Neurol Disord DrugTargets 12, 507–511.

Klyubin

, Betts

, Welzel

, Blennow

, Zetterberg

, Wallin

, Lemere

, Cullen

, Peng

, Wisniewski

, Selkoe

, Anwyl

, Walsh

, Rowan

(2008) Amyloid beta protein dimer-containing human CSF disrupts synaptic plasticity: Prevention by systemic passive immunization. J Neurosci 28, 4231–4237.

Shankar

, Li

, Mehta

, Garcia-Munoz

, Shepardson

, Smith

, Brett

, Farrell

, Rowan

, Lemere

, Regan

, Walsh

, Sabatini

, Selkoe

(2008) Amyloid-beta protein dimers isolated directly from Alzheimer’s brains impair synaptic plasticity and memory. Nat Med 14, 837–842.

, Hong

, Shepardson

, Walsh

, Shankar

, Selkoe

(2009) Soluble oligomers of amyloid Beta protein facilitate hippocampal long-term depression by disrupting neuronal glutamate uptake. Neuron 62, 788–801.

10.

Lasagna-Reeves

, Castillo-Carranza

, Sengupta

, Guerrero-Munoz

, Kiritoshi

, Neugebauer

, Jackson

, Kayed

(2012) Alzheimer brain-derived tau oligomers propagate pathology from endogenous tau. Sci Rep 2, 700.

11.

Lasagna-Reeves

, Castillo-Carranza

, Sengupta

, Clos

, Jackson

, Kayed

(2011) Tau oligomers impair memory and induce synaptic and mitochondrial dysfunction in wild-type mice. Mol Neurodegener 6, 39.

12.

Palop

, Mucke

(2010) Amyloid-beta-induced neuronal dysfunction in Alzheimer’s disease: From synapses toward neural networks. Nat Neurosci 13, 812–818.

13.

Huang

, Edwards

, Rounis

, Bhatia

, Rothwell

(2005) Theta burst stimulation of the human motor cortex. Neuron 45, 201–206.

14.

Koch

, Di Lorenzo

, Bonní

, Giacobbe

, Bozzali

, Caltagirone

, Martorana

(2014) Dopaminergic modulation of cortical plasticity in Alzheimer’s disease patients. Neuropsychopharmacology 39, 2654–2661.

15.

Brem

, Atkinson

, Seligson

, Pascual-Leone

(2013) Differential pharmacological effects on brain reactivity and plasticity in Alzheimer’s disease. Front Psychiatry 4, 124.

16.

Tokimura

, Di Lazzaro

, Tokimura

, Oliviero

, Profice

, Insola

, Mazzone

, Tonali

, Rothwell

(2000) Short latency inhibition of human hand motor cortex by somatosensory input from the hand. J Physiol 523, 503–513. Erratum in: J Physiol (Lond). 524, 942.

17.

Di Lazzaro

, Oliviero

, Profice

, Pennisi

, Di Giovanni

, Zito

, Tonali

, Rothwell

(2000) Muscarinic receptor blockade has differential effects on the excitability of intracortical circuits in the human motor cortex. Exp Brain Res 135, 455–461.

18.

Di Lazzaro

, Oliviero

, Tonali

, Marra

, Daniele

, Profice

, Saturno

, Pilato

, Masullo

, Rothwell

(2002) Noninvasive in vivo assessment of cholinergic cortical circuits in AD using TMS. Neurology 59, 392–397.

19.

Martorana

, Mori

, Esposito

, Kusayanagi

, Monteleone

, Codecá

, Sancesario

, Bernardi

, Koch

(2009) Dopamine modulates cholinergic cortical excitability in Alzheimer’s disease patients. Neuropsychopharmacology 34, 2323–2328.

20.

Suvá

, Favre

, Kraftsik

, Esteban

, Lobrinus

, Miklossy

(1999) Primary motor cortex involvement in Alzheimer’sdisease. J Neuropathol Exp Neurol 58, 1125–1134.

21.

Varma

, Snowden

, Lloyd

, Talbot

, Mann

, Neary

(1999) Evaluation of the NINCDS-ADRDA criteria in the differentiation of Alzheimer’s disease and frontotemporal dementia. J Neurol Neurosurg Psychiatry 66, 184–188.

22.

Sano

, Devanand

, Richards

, Miller

, Marder

, Bell

, Dooneief

, Bylsma

, Lafleche

, Albert

, et al. (1995) A standardized technique for establishing onset and duration of symptoms of Alzheimer’s disease. Arch Neurol 52, 961–966.

23.

Di Lazzaro

, Oliviero

, Pilato

, Saturno

, Dileone

, Marra

, Ghirlanda

, Ranieri

, Gainotti

, Tonali

(2005) Neurophysiological predictors of long term response to AChE inhibitors in AD patients. J Neurol Neurosurg Psychiatry 76, 1064–1069.

24.

Folstein

, Folstein

, McHugh

(2005) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

25.

Magni

, Binetti

, Padovani

, Cappa

, Bianchetti

, Trabucchi

(1996) The Mini-Mental State Examination in Alzheimer’s disease and multi-infarct dementia. Int Psychogeriatr 8, 127–134.

26.

Carlesimo

, Buccione

, Fadda

, Graceffa

, Mauri

, Lorusso

, Bevilacqua

, Caltagirone

(2002) Standardizzazione di due test di memoria per uso clinico: Breve Racconto e Figura di Rey. Nuova Riv Neurol 12, 1–13.

27.

Caffarra

, Vezzadini

, Dieci

, Zonato

, Venneri

(2002) Rey-Osterrieth complex figure: Normative values in an Italian population sample. Neurol Sci 22, 443–447.

28.

Carlesimo

, Caltagirone

, Gainotti

(1996) The Mental Deterioration Battery: Normative data, diagnostic reliability and qualitative analyses of cognitive impairment. The Group for the Standardization of the Mental Deterioration Battery. Eur Neurol 36, 378–384.

29.

Martorana

, Esposito

, Di Lorenzo

, Giacobbe

, Sancesario

, Bucchi

, Bonní

, Bernardini

, Sorge

, Sancesario

, Bernardi

, Caltagirone

, Koch

(2012) CSF levels of Aβ1-42 relationship with cholinergic cortical activity in Alzheimer’s disease patients. J Neural Transm 119, 771–778.

30.

Rossini

, Barker

, Berardelli

, Caramia

, Caruso

, Cracco

, Dimitrijević

, Hallett

, Katayama

, Lücking

(1994) Non-invasive electrical and magnetic stimulation of the brain, spinal cord and roots: Basic principles and procedures for routine clinical application. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol 91, 79–92.

31.

Sailer

, Molnar

, Paradiso

, Gunraj

, Lang

, Chen

(2003) Short and long latency afferent inhibition inParkinson’s disease. Brain 126, 1883–1894.

32.

Lang

, Hasan

, Sueske

, Paulus

, Nitsche

(2008) Cortical hypoexcitability in chronic smokers? A transcranial magnetic stimulation study. Neuropsychopharmacology 33, 2517–2523.

33.

Martorana

, Di Lorenzo

, Esposito

, Lo Giudice

, Bernardi

, Caltagirone

, Koch

(2013) Dopamine D2-agonist rotigotine effects on cortical excitability and central cholinergic transmission in Alzheimer’s disease patients. Neuropharmacology 64, 108–113.

34.

López-Alonso

, Cheeran

, Río-Rodríguez

, Fernández-Del-Olmo

(2014) Inter-individual variability in response to non-invasive brain stimulation paradigms. Brain Stimul 7, 372–380.

35.

Koch

, Di Lorenzo

, Bonní

, Ponzo

, Caltagirone

, Martorana

(2012) Impaired LTD- but not LTD-like cortical plasticity in Alzheimer’s disease patients. J Alzheimers Dis 31, 593–599.

36.

Hamada

, Murase

, Hasan

, Balaratnam

, Rothwell

(2013) The role of interneuron networks in driving human motor cortical plasticity. Cereb Cortex 23, 1593–1605.

37.

Koch

, Esposito

, Kusayanagi

, Monteleone

, Codecá

, Di Lorenzo

, Caltagirone

, Bernardi

, Martorana

(2011) CSF tau levels influence cortical plasticity in Alzheimer’s disease patients. J Alzheimers Dis 26, 181–186.

38.

, Okamoto

, Lipton

, Xu

(2014) Oligomeric Aβ-induced synaptic dysfunction in Alzheimer’s disease. Mol Neurodegener 14, 48.

39.

Martorana

, Koch

(2014) Is dopamine involved in Alzheimer’s disease?. Front Aging Neurosci 6, 252.

40.

Reynolds

, Garwood

, Wray

, Price

, Kellie

, Perera

, Zvelebil

, Yang

, Sheppard

, Varndell

, Hanger

, Anderton

(2008) Phosphorylation regulates tau interactions with Src homology 3 domains of phosphatidylinositol 3-kinase, phospholipase Cgamma1, Grb2, and Src family kinases. J Biol Chem 283, 18177–18186.

41.

Usardi

, Pooler

, Seereeram

, Reynolds

, Derkinderen

, Anderton

, Hanger

, Noble

, Williamson

(2011) Tyrosine phosphorylation of tau regulates its interactions with Fyn SH2 domains, but not SH3 domains, altering the cellular localization of tau. FEBS J 278, 2927–2937.

42.

Tai

, Serrano-Pozo

, Hashimoto

, Frosch

, Spires-Jones

, Hyman

(2012) The synaptic accumulation of hyperphosphorylated tau oligomers in Alzheimer’s disease is associated with dysfunction of the ubiquitin-proteasomesystem. Am J Pathol 181, 1426–1435.

43.

Pooler

, Noble

, Hanger

(2014) A role for tau at the synapse in Alzheimer’s disease pathogenesis. Neuropharmacology 76, 1–8.

44.

Krstic

, Knuesel

(2013) Deciphering the mechanism underlying late-onset Alzheimer’s disease. Nat Rev Neurol 9, 25–34.

45.

Darreh-Shori

, Brimijoin

, Kadir

, Almkvist

, Nordberg

(2006) Differential CFS butyrylcholinesterase levels in Alzheimer’s disease patients with the ApoE epsilon4 allele, in relation to cognitive function and cerebral glucose metabolism. Neurobiol Dis 24, 326–333.

46.

Darreh-Shori

, Vijayaraghavan

, Aeinehband

, Piehl

, Lindblom

, Nilsson

, Ekdahl

, Långström

, Almkvist

, Nordberg

(2013) Functional variability in butyrylcholinesterase activity regulates intrathecal cytokine and astroglial biomarker profiles in patients with Alzheimer’sdisease. Neurobiol Aging 34, 2465–2481.

47.

Brys

, Pirraglia

, Rich

, Rolstad

, Mosconi

, Switalski

, Glodzik-Sobanska

, De Santi

, Zinkowski

, Mehta

, Pratico

, Saint Louis

, Wallin

, Blennow

, de Leon

(2009) Prediction and longitudinal study of CSF biomarkers in mild cognitive impairment. Neurobiol Aging 30, 682–690.

48.

Seppälä

, Koivisto

, Hartikainen

, Helisalmi

, Soininen

, Herukka

(2011) Longitudinal changes of CSF biomarkers in Alzheimer’s disease. Alzheimers Dis 25, 583–594.

49.

Zhou

, Homma

, Poo

(2004) Shrinkage of dendritic spines associated with long-term depression of hippocampal synapses. Neuron 44, 749–757.

50.

Sheng

, Ertürk

(2013) Long-term depression: A cell biological view. Philos Trans R Soc Lond B Biol Sci 2, 369.

51.

Amieva

, Mokri

, Le Goff

, Meillon

, Jacqmin-Gadda

, Foubert-Samier

, Orgogozo

, Stern

, Dartigues

(2014) Compensatory mechanisms in higher-educated subjects with Alzheimer’s disease: A study of 20 years of cognitive decline. Brain 137, 1167–1175.

52.

Almeida

, Schultz

, Austin

, Boots

, Dowling

, Gleason

, Bendlin

, Sager

, Hermann

, Zetterberg

, Carlsson

, Johnson

, Asthana

, Okonkwo

(2015) Effect of cognitive reserve on age-related changes in cerebrospinal fluid biomarkers of Alzheimer disease. JAMA Neurol 72, 699–706.

53.

Bozzali

, Dowling

, Serra

, Spanò

, Torso

, Marra

, Castelli

, Dowell

, Koch

, Caltagirone

, Cercignani

(2015) The impact of cognitive reserve on brain functional connectivity in Alzheimer’s disease. J Alzheimers Dis 44, 243–250.

54.

Serra

, Musicco

, Cercignani

, Torso

, Spanò

, Mastropasqua

, Giulietti

, Marra

, Bruno

, Koch

, Caltagirone

, Bozzali

(Aging) (2015) Cognitive reserve and the risk for Alzheimer’s disease: A longitudinal study. Neurobiol Aging 36, 592–600.

55.

Veniero

, Ponzo

, Koch

(2013) Paired associative stimulation enforces the communication between interconnected areas. J Neurosci 33, 13773–13783.

56.

Lista

, Garaci

, Ewers

, Teipel

, Zetterberg

, Blennow

, Hampel

(2014) CSF Aβ1-42 combined with neuroimaging biomarkers in the early detection, diagnosis and prediction of Alzheimer’s disease. Alzheimers Dement 10, 381–392.

57.

Medina

, Avila

(2014) New perspectives on the role of tau in Alzheimer’s disease. Implications for therapy. Biochem Pharmacol 88, 540–547.