Raman Spectrometric Detection Methods for Early and Non-Invasive Diagnosis of Alzheimer’s Disease

Abstract

The continuous increasing rate of patients suffering of Alzheimer’s disease (AD) worldwide requires the adoption of novel techniques for non-invasive early diagnosis and monitoring of the disease. Here we review the various Raman spectroscopic techniques, including Fourier Transform-Raman spectroscopy, surface-enhanced Raman scattering spectroscopy, coherent anti-Stokes Raman scattering spectroscopy, and confocal Raman microspectroscopy, that could be used for the diagnosis of AD. These techniques have shown the potential to detect AD biomarkers, such as the amyloid-β peptide and the tau protein, or the neurotransmitters involved in the disease (e.g., Glutamate and γ-Aminobutyric acid), or the typical structural alterations in specific brain areas. The possibility to detect the specific biomarkers in liquid biopsies and to obtain high resolution 3D microscope images of the affected area make the Raman spectroscopy a valuable ally in the early diagnosis and monitoring of AD.

Keywords

Alzheimer’s disease amyloid-β peptide brain tissue diagnosis Raman spectroscopy tau protein

INTRODUCTION

With 47 million people in the world actually suffering from Alzheimer’s disease (AD) and with the expectation that someone will join these ranks every 35 seconds, AD represents the neurodegenerative pathology with the greatest impact on the social cost in the healthcare systems worldwide [1]. In spite of the progress made in understanding the molecular mechanisms underlying the pathogenesis of the disease, the pharmacologic treatment fails essentially because the symptom-based diagnosis is made at a late stage and, quite often, is inaccurate and misleading [2]. The assessment of biochemical markers and the use of imaging diagnostic methods are of obvious utility in complementing the clinical criteria for a more accurate diagnosis of AD, yet their usefulness depends on the sensitivity of the method and the specificity of the biomarker.

Spectroscopic analyses of the vibrational mode of the molecule upon light excitation, such as InfraRed spectroscopy (IRS) and Raman spectroscopy (RS), provide unique signatures that identify unequivocally the biomarker within a complex mixture (so-called “fingerprint”). These methods are not invasive and non-destructive since they do not require manipulations (such as fixation or staining) of the sample, and can specifically detect the presence of several biomarkers in the patient’s specimens at once and in a relative short time. IRS measures the absorbance of photons as function of the dipole moment change in the molecules, while RS measures the inelastic spectra that originate from a change in the polarizability of the molecules. RS permits the fingerprinting of biomarkers in aqueous environment, which makes the technique of extreme interest for non-invasive diagnosis in liquid biopsies [3]. The technological improvements in laser excitation power, charge-coupled device detectors, and notch filters, the combination with advanced optical instruments and digital imaging, along with the implementation of theoretical calculations, have greatly expanded the range of applications of RS in the biological and medical sciences over the last decades [4 –9]. For instance, RS has proven of valuable assistance in the diagnosis of cancers [10 –13] and of inflammatory diseases [14].

In this article, we review the various RS techniques that have shown the potential to detect the molecular alterations in the biological specimens of AD patients, and that could assist the pre-symptomatic diagnosis and the monitoring of progressionof AD.

IMAGING-AND ANALYTICAL-BASED DIAGNOSIS OF ALZHEIMER’S DISEASE: A BRIEF OVERVIEW

AD is a progressive neurodegenerative disease and the leading cause of dementia in the aged population of 65–74-year-olds. Neuronal cell loss and defective inter-neuronal transmission characteristically involve the brain regions mediating memory and cognition in the hippocampus and cortex [15]. AD patients manifest progressive loss of short-term memory, temporal and spatial disorientation, speaking difficulties, behavioral disturbances, and dementia. An important issue that influences the effectiveness of the therapeutic approach relies on the early diagnosis and the clear discrimination of AD from other forms of dementia [16]. Clinical criteria rely on late onset neurological and cognitive symptoms and are not sufficient for differential diagnosis. Therefore, imaging and analytical methods are becoming fundamental for diagnosing and staging AD, as well as for monitoring the effectiveness of the therapeutic interventions. The accumulation of the amyloid-β peptide in the inter-neuronal space (Aβ plaques) and the intra-neuronal accumulation of microtubule-associated deposition of the protein tau (called neurofibrillary tau tangles) are believed to contribute to brain atrophy during development of AD, and are considered the two most reliable hallmarks of the disease [17].

In the last two decades, a variety of imaging methods have been implemented in order to detect the structural and functional alterations in the brain that associate with AD [18]. These include computed tomography (CT) [19, 20], magnetic resonance imaging (MRI) [21 –23], positron emission tomography (PET) [23 –25], and single photon emission computed tomography (SPECT) [26, 27]. In addition, clinicians can harness specific biomarkers to diagnose AD at the very early symptomatic stage, when cognitive impairment is still mild, or even at the asymptomatic stage. To this end, the presence of the fibrillar amyloid peptide Aβ₄₂ and of the tangle component tau can be assessed directly in the cerebrospinal fluid and even in blood plasma, given the physiological exchange of small peptides between these two fluids. Additional plasma biomarkers of potential assistance in the diagnosis of AD have been proposed based on proteomic analyses [28, 29]. Collectively, a number of biomarkers can be used to support the diagnosis and to discriminate AD from other forms of dementia [30 –42].

THE RAMAN SPECTROSCOPY IN A NUTSHELL

“Raman Effect” is named after the Indian physicist V.C. Raman, who first described in 1928 the inelastic scattering of the incident photons as they interact with the polarizable molecule of matter [43].

The energy of the incident photon promotes the vibrational transition bringing the molecule to a virtual level of energy. In Raman scattering, the energy of the emitted (scattered) photon is different from that of the exciting photon. The Raman scatter is called Stokes if originates from the interaction with a molecule that was in its basal (ground) energetic level and that after excitation falls to a vibrational state with energy above the ground level. In this case, the energy of the Raman scatter photon is lower (its wavelength is shifted to a higher value, i.e., lower frequency) than that of the exciting photon (Fig. 1a, b). On the other hand, if the molecule was already in an excited energetic level and after excitation falls to the ground level, the Raman photon has an increased energy (i.e., the wavelength of the emitted photon shifts to lower values). This type of Raman radiation is called anti-Stokes (Fig. 1a, b). Since at room temperature most of the molecules are in their ground energetic level, Raman Stokes scattering is easily detected, whereas the intensity of the anti-Stokes scatter photons is very low. However, only one of a thousand or ten thousand of the scattered photons is inelastic (i.e., Raman scatter of either Stokes or anti-Stokes type), while the majority of them leaves the matter with the same energy (and thus, unchanged wavelength) of the incident photon. The latter is an elastic scatter that is known as Rayleigh (Fig. 1a, b). Typically, the intensity of the Raman signal is a millionth of that of the incident light, while the Rayleigh band has an intensity up to one thousand that of the Raman bands. Raman intensity measures the number of scattered photons detected per second (the scatter photon is registered within 10^–15 second). This value depends on the intensity of the incident photon (i.e., the power of the laser), the cross-section, the length and density of the traversed material, and on other apparatus-specific parameters (e.g., the optical efficiency of the detector). In a typical diagram (Fig. 1c), the ordinate reports the Raman (relative) intensity commonly expressed in arbitrary units and the abscissa reports the energy of the vibrational transition expressed as wavenumber (cm^–1), which is inversely proportional to the wavelength. The Raman shift indicates the difference between the wavelengths of scattered and incident photons, and is given as difference in wavenumber (cm^–1). The cross-section measures the sectional areas of the torus in which the Raman spectrum can be detected, and this is a function of the solid angle (measured in units of Steradians). A typical cross-section of Raman scatter is 10^–29 cm² sr^–1 [44]. Raman spectrum is detectable in a wide wavelength region (400–2000 cm^–1), in which fall the biomolecules. The vibrational mode is characteristic for each molecule (or chemical bond) and is therefore unequivocally associated with a specific Raman spectrum, thus allowing the molecular fingerprinting of single amino acid, proteins and lipids in biological samples (Fig. 1c). Two major drawbacks of RS are the interference of fluorophores in the biological sample and the low intensity of the Raman peaks. Improvements in the Raman set-up instruments [45] and in the preparation of the sample have now made possible the Raman fingerprinting of molecular biomarkers for clinical diagnostic purposes. Here, we briefly introduce three of these principal technological advancements, namely Fourier Transform (FT) RS, Surface Enhanced RS (SERS), and Coherent Anti-Stokes RS (CARS). In 1986, Hirschfeld and Chase described the feature of FT-RS, resulting from coupling a diode-pumped neodymium-doped yttrium aluminum garnet (Nd:YAG) laser with a Michelson interferometer [46]. Thanks to the near-IR (NIR) wavelength of the laser as an exciting source, the fluorescence background of the sample is drastically reduced, thus avoiding interference with the Raman spectra. Furthermore, the reduced thermal effects and the high pump stability of the diode-pumped solid-state laser reduce flicker noise, thus increasing the overall Signal/Noise ratio [47]. These characteristics make the FT-RS well suited for the analysis of biological samples [48]. In the SERS technique, the signal is amplified by electromagnetic enhancement (EME) and chemical enhancement (CE) [49, 50]. Here, the target analyte is adsorbed on metal substrates (possibly, nanostructured) with a rough surface (Fig. 2). Generally, the SERS substrate is made of gold, silver, or copper. EME is contributed from hot spots that depend on the distance between the analytes and the metal surface. The metal surface roughness, core-shell metal nanoparticles and nanoparticle aggregation can yield large EME (factor: 10¹²). CE is a result of the interaction between the chemisorbed analytes and the metal surface, which increases the Raman scattering cross-section, and can yield an enhancement factor up to a magnitude of 10² [51]. Therefore, SERS is an excellent ultrasensitive platform for the simultaneous detection of multiple biomolecules for diagnostic purposes [52]. In 1965, Maker and Terhune reported the far stronger vibrational signals by using CARS. An overview of the theory and applications of CARS in biological sciences can be found in excellent article reviews [53 –55]. Essentially, the system uses “a pump photon at frequency ω_p and a Stokes photon at frequency ω_s to interact with a sample via a wave-mixing process. When the beat frequency (ω_p –ω_s) matches the frequency of a Raman active molecular vibration, the resonant oscillators are coherently driven by the excitation fields, thereby generating a strong anti-Stokes photon at ω_as = 2ω_p - ω_s.

Fig.1

The energy level diagram of scattering process and fingerprint of molecules.

Fig.2

The typical SERS enhancement spectra. SERS signals are enhanced through the interaction of nanoparticles and molecules.

Notably, RS in its implemented technical variants can be coupled with the microscope when the analysis is focused on a defined area of a solid optically transparent material, as it is the case of histological biopsies [6 , 54].

In the following paragraphs, we describe the Raman signatures and Raman mapping images obtained through various RS detection methods as applied in AD patients’ specimens and in AD experimental models.

RAMAN SPECTROSCOPIC SIGNATURE OF AD LIQUID BIOPSIES

FT-RS has been tested for its capability to discriminate the normal and diseased plasma protein signature (secondary and tertiary structure) of aged healthy controls and of AD patients, respectively [56]. The authors also performed the FT-IR spectroscopy to obtain complementary data on the oxidation status of plasma lipids and proteins. The study included plasma samples from patients with mild, moderate, or severe AD symptoms, and from healthy elderly controls. An increase of the intensity of the peak at 1672 cm^–¹ (I₁₆₇₂) along with a concomitant decrease of the intensity of the peak at 1658 cm^–¹ (I₁₆₅₈), which respectively reflected the enrichment ofβ-sheet versus α-helix polypeptide backbone, were observed in the samples from moderate AD patients compared to the controls. Further analyses indicated that the increased I₁₆₇₂ was largely contributed by Aβ_1 - 40 and Aβ_1 - 42 peptides. In confirmation, the relative intensity of the 409 cm^–¹ peak (I₄₀₉), which associates with the amyloid peptides, also was increased in AD samples. Notably, the Raman peak region around 980–910 cm^–¹, which refers to the protein α-helices, showed a lower intensity in AD plasma compared to that of controls. Further, an increased ratio of the intensities I₇₅₈/I₇₄₃ and upshifting of the 743 cm^–¹ peak, attributed to changes in tryptophan and indicative of alterations in the tertiary protein structure, were observed in the AD plasma. These data were corroborated by the FT-IR spectroscopy, which revealed in AD plasma the presence of oxidized lipids and proteins in the same spectral region of the β-sheet proteins. Taken together, the spectroscopy data allowed distinguishing the plasma from mild AD patients and healthy donors with 91% sensitivity and 100% specificity, and with a slight less sensitivity (89%) and specificity (92%) it was possible to discriminate AD patients all together from age-matched healthy controls. Although the number of samples analyzed was limited (about 8–12 per group), this study supports the usefulness of vibrational spectroscopy for AD diagnosis. A few years later, this same group applied these techniques to study the changes in the plasma proteins composition of a larger number of AD patients versus age-matched healthy controls [57]. The ratio of the intensity of peaks at 1671 (attributable to β-sheet containing proteins) and 1658 (attributable to α helical-containing proteins) (I₁₆₇₁/I₁₆₅₈) consistently increased in the plasma from healthy to mild, moderate, and severe AD plasma. A similar conclusion was drawn from the decreased intensity of the spectral profile in the 980–910 cm^–¹ region, which correlates with α helical-containing proteins. Conversely, the band at 409 cm^–¹ mainly contributed by the Aβ-amyloid peptides increased along with AD progression in the patients’ plasma. Taken together, these data supported the view that progression in AD reflected in an increase of β-sheet containing globulins (such as α₂-macroglobulin and α₁-antitrypsin, two proteins found in amyloid plaques) and Aβ peptides, and in a concomitant decrease of albumin (the principal contributor of the α helicalsignal).

Great efforts have been made to exploit SERS for the development of in vitro assay of AD biomarkers. For instance, Zengin et al. have developed a SERS-based sandwich assay for the detection of tau protein in AD samples [58]. The system employs hybrid magnetic nanoparticles functionalized with monoclonal anti-tau antibodies as capture probes and the polyclonal anti-tau antibodies plus 5,5-dithiobis(2-dinitrobenzoic acid) (that binds to tau) immobilized on gold nanoparticles as SERS tags. The tau peptide is first isolated from the solution by using the magnetic nanoparticles and thereafter is sandwiched with the SERS substrate. The sandwich is then subjected to excitation with a laser of 785 nm and the Raman spectra recorded. Major peaks were registered at 1053 cm^–1, 1332 cm^–1, and 1553 cm^–1. In particular, the I₁₃₃₂ was proportional to tau concentration, showing a linear correlation in the range from 25 fM to 500 nM.

Development of a SERS platform for the rapid and selective detection of Aβ oligomers and fibril species in patient’s fluids could greatly assist the neurologist to monitor the progression of the disease [59]. Coté’s group reported the SERS spectroscopic analysis of the Aβ_1 - 40 peptide labeled with Congo red and adsorbed on a sialic acid-modified Au nanoshell platform [60]. The detection limit of Aβ concentration was 1 pM in Congo red SERS spectra. More recently, a new SERS platform for the rapid detection of Aβ_1 - 42 and tau proteins has been reported [61]. The authors have fabricated ion magnetic core-gold plasmonic shell nanoparticles that were attached on a hybrid graphene oxide-based multifunctional nanoplatform. Then, they conjugated anti-Aβ and anti-tau antibodies on hybrid graphene oxides. This platform permitted to selectively isolate the Aβ and tau proteins from the plasma sample. The SERS spectra were collected by using a 785 nm excitation laser with 2 mW power. The multiplex nanoplatform could identify and detect both Aβ and tau proteins when their concentration was as low as 100 fg/mL.

The SERS technique can also be combined with nanofluidic biosensors. These devices consist of nano-sized capillary, in which the bio-analyte flows, and of a detection system that collects the SERS spectra emitted by the bio-analyte upon excitation with a NIR laser. One such SERS-nanofluidic sensor was first designed by Chou and collaborators in 2008 [62]. Here, the Aβ peptides are first adsorbed on anti-Aβ complexed Au nano-spheres of 60 nm that are forced to flow in the nano-channels. The nano-spheres remain trapped in the detection area (2.2×0.2 μm²), where the nano-channel diameter is of 40 nm. The SERS spectra is then collected upon excitation by a 785 nm laser with spot size 2.2 μm. With such device, the authors could detect three different SERS peaks that could be associated with the concentration and configuration of the Aβ peptide. When the nano-spheres were loaded with Aβ at the concentration of 1.15 pM, the SERS peak was registered at 1266 cm^–¹. This peak was very sharp, indicating that practically all the Aβ peptides were in the same α-helix conformation. At the concentration of 1.15 nM (i.e., one thousand higher), the SERS peak located at 1266 cm^–¹ diminished while a new band appeared at 1244 cm^–¹. This phenomenon implicated that the proteins were in two different conformational states. When the concentration of Aβ was further increased of ten times (i.e., at 11.5 nM), the SERS peak located at 1266 cm^–¹ completely disappeared and only the band at 1244 cm^–¹ was recorded. The data reported in this work indicate that the secondary structure of the Aβ peptide can switch from the α-helix to the β-sheet conformation when its concentration increases from picoMolar to nano and microMolar, and that a SERS device can reveal these changes.

The above results could be essentially reproduced in other SERS nanodevices even more sensitive in terms of concentration of the Aβ peptide [63, 64]. Here, the researchers have immobilized the Aβ_1 - 40 peptide at different concentrations (starting from 10 fM and increasing by a factor of 100×) on 80 nm Au nanoparticles. The nanoparticles were trapped in the detection area at the junction between the micro- and nano-channels, and here they were scanned with a 785 nm excitation laser. The device could detect the Aβ peptide (bound to the nano-spheres) at a concentration as low as 10 fM, revealed as a SERS peak at 1264 cm^–¹, which indicated a full conformation of the peptide as α-helix. The main peak shifted from 1264 cm^–¹ to 1241 cm^–¹ (the latter is typical of the β-sheet secondary structure) along with the increase in the concentration of the Aβ peptide present on the nano-spheres from 10 fM up to 1 μM. Again, these experiments proved that when at very low concentration (i.e., in the range of femto Molar), the Aβ peptide is dispersed as a monomer in the α-helix conformation, whereas it switches to the aggregate-prone β-sheet conformation as soon as its concentration increases above the 100 pM. Most importantly, the above data highlight the capability of SERS-based devices to detect these changes in the secondary structure of the Aβ peptide, being able to detect when the β fibrils form and to determine at which (range of) concentration this occurs.

Clearly, these devices have the potential to be implemented for the laboratory diagnosis of AD using the plasma or the cerebrospinal fluid as a source of the Aβ or tau protein, where the concentration of these proteins could be in the range of femtomoles.

Excessive GABAergic stimulation, due to abnormal conversion of glutamate (GLU) into γ-aminobutyric acid (GABA), has been implicated in AD progression [65]. A SERS-based analytical method has been employed to detect these neurotransmitters in serum [66]. The SERS spectroscopic profiles of GABA and GLU had been previously characterized, with a detection limit of 10^–⁴ M and of 10^–⁷ M, respectively [67, 68]. The researchers used Ag nanoparticles as SERS substrate. The SERS spectra were collected via laser power excitation (power 450 mW, wavelength 785 nm). The SERS technique could discriminate and determine the respective concentration of GLU and GABA when mixed in cow blood serum at a concentration of 8 μM [66].

The RS coupled with microscopy can make easy to analyze tiny amounts of blood-derived samples deposited as a drop on cover-slips. As an example, the sera from healthy controls, from AD patients with mild to moderate symptoms, and from patients suffering of dementia not related to AD were investigated by confocal Raman microspectroscopy [69]. The serum sample was placed and let dry on a microscope glass slide (covered with aluminum foil), and the SERS spectrum was collected upon excitation with a laser of 50 mW and 785 nm excitation wavelength. Using genetic algorithm variable analysis combined with artificial neural network to discriminate Raman signal differences in the four groups of samples, the authors could define a fingerprint for the diagnosis of AD at the mild/moderate stage with an accuracy of 95% [69].

In another study, the platelets from transgenic mice have been analyzed by micro-RS as a proof-of principle for a non-invasive differential diagnosis of AD [70, 71]. The researchers compared the Raman spectra of the platelets isolated from mice modeling AD, Parkinson’s disease, and vascular dementia versus control age-matched mice. The platelets were excited with a 17 mW laser at 633 nm and the Raman spectra were recorded in a micro-RS system. Two significant peaks, located at 740 cm^–1 and 1654 cm^–1, were recorded in the platelets isolated from 4-month-old and 12-month-old AD mice and from their respective controls. Noteworthy, the I₇₄₀ showed a proportional increase with the age of AD mice. On the other hand, the I₁₆₅₄ could differentiate the AD-derived platelets from the control counterparts [71].

The above examples demonstrate the versatility of RS techniques to detect AD biomarkers present in biofluids at very low range of concentration, which support their potential utilization for the non-invasive diagnosis of AD.

MICROSCOPE-RAMAN SPECTROSCOPY OF HISTOLOGIC SPECIMENS

Micro-RS integrates the signals from spectroscopy with imaging data from the microscope. Thus, it provides the molecular vibrational fingerprint of the bio-analyte along with its tissue and subcellular mapping by controlling the X-Y scanning of the microscope.

FT-RS coupled with microscopy was employed to investigate the structural changes caused by paraffin fixation and solvents in cerebral tissue of AD patients [72]. It was found that the standard procedure for tissue fixation results in loss of nucleic acid and lipid components and in conformational changes of the proteins [72].

Differences in the micro-Raman spectra were reported when comparing the hippocampus tissue section isolated from normal control rats and from AD-model rats [73]. The researchers used a laser spot (power: 21 mW; excitation wavelength: 785 nm) of 1.5 μm focused on brain hippocampus tissues. A Raman peak with wavenumber of 1670 cm^–¹ contributed from β sheets aggregations was recorded in AD affected areas. Additional Raman bands at 1065, 1300, and 1440 cm^–¹ in the same areas were assigned to cholesterol, suggestive of increased lipid accumulation along with the aggregation of Aβ. The Raman band at 1088 cm^–¹ was assigned to PO₂^– stretching vibration, which likely was contributed by hyperphosphorylated tau protein. In this study, the Raman spectra helped to discriminate healthy and AD affected hippocampus tissues with 93.8% sensitivity and 89.6% specificity.

FT-RS combined with microscopy allowed to detect the presence of Aβ aggregates after staining with Congo red [74]. In this study, matched brain sections from healthy controls and from AD patients were de-paraffinized, labelled with Congo red, and analyzed by FT-RS. Healthy and diseased area could be discriminated based on the shift of Raman peaks located at 1589 cm^–¹ and 1573 cm^–¹, respectively.

A more accurate investigation of the sample fingerprints can be achieved using confocal Raman microspectroscopy, which allows scanning the sample at different depths. High resolution and optimization of the Raman peaks are obtained through a fine modulation of the pinhole by using objectives with high numerical aperture [75]. For instance, confocal Raman microspectroscopy was employed to search for the presence of Aβ in the lens of AD patients [76].

CARS coupled with microscopy can further provide the advantage of a deep penetration of thick samples and high spatial resolution, which makes the system a valid tool for tissue imaging [54 , 77], particularly of those containing a high proportion of lipid content like the brain tissue [55]. CARS microscopy can provide the Raman signals and 3D images of interaction between lipid and Aβ plaques in human AD tissues without using any fluorescence labeling [78, 79]. The combination of CARS with gradient-index lens is an alternative endoscopy technique that can overcome the limitation of superficial tissue. This technique allows space-resolved spectral information studies of AD brain tissue in vivo. For instance, with the combination of CARS and microendoscope, it was possible to obtain a non-linear optical imaging and molecular vibrational spectroscopy of neuronal membrane lipids of AD brain tissue [80]. In this study, the membrane lipids-associated CARS signals showed peaks at different positions in normal and AD brain tissues. In addition, the Raman intensity I₂₈₅₀ in AD brain tissues was about four times higher than the intensity of the corresponding peak (I₂₈₄₅) in normal brain tissues, and this was attributed to the high presence of GABA in dentate gyrus of diseasedtissues [80].

CONCLUDING REMARKS AND PERSPECTIVES

The number of people that will be affected by AD is expected to rise along with increased life expectancy. The economic cost for the medical care at home or in public or private hospices adds to the social cost that already burden on the AD patient and the family. These costs will impact on the general budget for Medicare and, as a chain reaction, eventually will affect the entire health welfare policy in every country.

Unfortunately, AD is commonly diagnosed too late, when irreversible damages in the brain have occurred. Easy to perform and non-invasive methods for the early detection of AD, and its discrimination from other forms of dementia, would improve the quality of patients’ life and reduce the healthcare costs.

Here we have reviewed the Raman spectrometric detection methods that have shown the potential of being translated into clinics for the diagnosis of AD (see also Table 1). Along with the fingerprinting of the biological sample, which allows the discrimination of healthy and diseased specimens, RS coupled with confocal microscopy can also provide a 3D image of affected areas (Fig. 3). The continuous advancements in the technology of physical instruments, such as fibers, lasers, confocal system, microscopes, and spectrometers, will further contribute to increase the resolution of spectroscopic information and of images. RS detection have the potentiality to combine with MRI/CT/PET/SPECT imaging techniques to provide more accurate and precise information on the pathology. We can envisage the introduction of vibrational spectroscopy into clinical setting in the near future [81]. Therefore, AD patients could soon take advantage of non-invasive methods for early diagnosis that will facilitate the designing of more effective therapies, and at the same time will drive the adoption of adequate healthcarestrategies.

Table 1

Raman spectrometric detection methods: potential application for Alzheimer’s disease diagnosis

RS Technique	Target	Method/instrument	Advantage/main outcome	Reference
	molecule/specimens
FT-RS	Plasma blood of aged healthy controls and of AD patients	Combination of discriminant analysis and the ratio of Raman peak intensity	Non-invasive, rapid, and relatively inexpensive procedure	[56]
FT-RS	Brain tissue	Congo red labeling	Possibility of in vivo diagnosis	[74]
SERS	Tau protein	Sandwich assay of nanoparticles and Raman report	Detection limit of tau was less than 25 fM	[58]
SERS	Aβ_1 - 40 peptide	Congo Red labeling of Aβ_1 - 40 peptide labeled	Detection limit of Aβ_1 - 40 concentration was 1 pM	[60]
SERS	Aβ_1 - 42 and tau proteins	Plasmonic-magnetic hybrid graphene oxide platform	Aβ_1 - 42 and tau detection in concentrations as low as 100 fg/mL	[61]
SERS	Aβ peptides	SERS-nanofluidic sensor	Demonstrates that Aβ peptide can switch from the α-helix to the β-sheet conformation.	[62]
SERS	Aβ peptides	SERS-nanofluidic device	Detection of Aβ peptides at concentration as low as10 fM	[63, 64]
SERS	GLU and GABA	GLU adsorbed on Ag nanoparticles/GABA adsorbed on Ag nanoparticles	Detection of GLU and GABA at concentration as low as 8 μM	[66]
SERS	GABA	GABA adsorbed on Ag nanoparticles	Detection of GABA at concentration as low as 0.1 mM	[67]
SERS	GLU and GABA	GLU adsorbed on Ag nanoparticles/GABA adsorbed on Ag nanoparticles	GABA and GLU detected at 0.1 mM and 0.1 μM, respectively	[68]
Micro-RS	hippocampus tissue	Raman data analyzed by principal component analysis	Demonstrates simultaneous changes of cholesterol and Aβ	[73]
Micro-RS	platelets of AD mice model	Raman spectroscopic data combined with multilayer perceptron networks classification method	Discriminates AD cognitive symptoms from Parkinson’s disease and vascular dementia	[70]
Confocal Raman microspectroscopy	blood serum	Genetic algorithm variable analysis combined with artificial neural network to discriminate Raman signal differences	Non-invasive; differential diagnosis	[69]
Coherent anti-Stokes Raman scattering with gradient-index lens	mice brain tissue	a multimodal multiphoton nonlinear optical microspectroscopy imaging system	Normal and AD brain were clearly distinguished based on CARS spectra difference	[80]

AD, Alzheimer’s disease; Aβ, amyloid-β; CARS, coherent anti-Stokes Raman spectroscopy; FT-RS, Fourier Transform Raman spectroscopy; GABA, γ-aminobutyric acid; GLU, glutamate; RS, Raman spectroscopy; SERS, surface-enhanced Raman spectroscopy.

Fig.3

Raman spectroscopy imaging applied for Alzheimer’s disease diagnosis. Molecular vibrational spectroscopy of biomarkers (molecular fingerprint) and 3D image of affected brain areas.

DISCLOSURE STATEMENT

Authors’ disclosures available online (http://j-alz.com/manuscript-disclosures/16-1238r1).

References

Reitz

, Mayeux

(2014) Alzheimer disease: Epidemiology, diagnostic criteria, risk factors and biomarkers. Biochem Pharmacol 88, 640–651.

Dos Santos Picanco

, Ozela

, de Fatima de Brito Brito

, Pinheiro

, Padilha

, Braga

, de Paula da Silva

, Dos

Santos CB

, Rosa

, da Silva Hage-Melim

(2016) Alzheimer’s disease: A review from the pathophysiology to diagnosis, new perspectives for pharmacological treatment. Curr Med Chem 23. doi: 10.2174-0929867323666161213101126

Mitchell

, Gajjar

, Theophilou

, Martin

, Martin-Hirsch

(2014) Vibrational spectroscopy of biofluids for disease screening or diagnosis: Translation from the laboratory to a clinical setting. J Biophotonics 7, 153–165.

Popp

, Krafft

, Mayerhöfer

(2011) Modern Raman spectroscopy for biomedical applications. Optik Photonik 6, 24–28.

Ramoji

, Neugebauer

, Bocklitz

, Foerster

, Kiehntopf

, Bauer

, Popp

(2012) Toward a spectroscopic hemogram: Raman spectroscopic differentiation of the two most abundant leukocytes from peripheral blood. Anal Chem 84, 5335–5342.

Nallala

, Piot

, Diebold

, Gobinet

, Bouche

, Manfait

, Sockalingum

(2014) Infrared and Raman imaging for characterizing complex biological materials: A comparative morpho-spectroscopic study of colon tissue. Appl Spectrosc 68, 57–68.

Huang

C-C

(2016) Applications of Raman spectroscopy in herbal medicine. Appl Spectrosc Rev 51, 1–11.

Byrne

, Knief

, Keating

, Bonnier

(2016) Spectral pre and post processing for infrared and Raman spectroscopy of biological tissues and cells. Chem Soc Rev 45, 1865–1878.

Pence

, Mahadevan-Jansen

(2016) Clinical instrumentation and applications of Raman spectroscopy. Chem Soc Rev 45, 1958–1979.

10.

Hsu

C-W

, Huang

C-C

, Sheu

J-H

, Lin

C-W

, Lin

L-F

, Jin

J-S

, Chau

L-K

, Chen

(2016) Novel method for differentiating histological types of gastric adenocarcinoma by using confocal Raman microspectroscopy. PloS One 11, e0159829–e0159841.

11.

Hsu

C-W

, Huang

C-C

, Sheu

J-H

, Lin

C-W

, Lin

L-F

, Jin

J-S

, Chen

(2016) Differentiating gastrointestinal stromal tumors from gastric adenocarcinomas and normal mucosae using confocal Raman microspectroscopy. J Biomed Opt 21, 075006.

12.

Kirsch

, Schackert

, Salzer

, Krafft

(2010) Raman spectroscopic imaging for in vivo detection of cerebral brain metastases. Anal Bioanal Chem 398, 1707–1713.

13.

Kong

, Kendall

, Stone

, Notingher

(2015) Raman spectroscopy for medical diagnostics — From in-vitro biofluid assays to in-vivo cancer detection. Adv Drug Deliv Rev 89, 121–134.

14.

Bielecki

, Bocklitz

, Schmitt

, Krafft

, Marquardt

, Gharbi

, Knosel

, Stallmach

, Popp

(2012) Classification of inflammatory bowel diseases by means of Raman spectroscopic imaging of epithelium cells. J Biomed Opt 17, 076030.

15.

Sabuncu

, Desikan

, Sepulcre

, Yeo

BTT

, Liu

, Schmansky

, Reuter

, Weiner

, Buckner

, Sperling

(2011) The dynamics of cortical and hippocampal atrophy in Alzheimer disease. Arch Neurol 68, 1040–1048.

16.

Jack

Jr , Albert

, Knopman

, McKhann

, Sperling

, Carrillo

, Thies

, Phelps

(2011) Introduction to the recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 257–262.

17.

Selkoe

(2001) Alzheimer’s disease: Genes, proteins, and therapy. Physiol Rev 81, 741–766.

18.

Johnson

, Fox

, Sperling

, Klunk

(2012) Brain imaging in Alzheimer disease. Cold Spring Harb Perspect Med 2, a006213.

19.

Smith

, Jobst

, Szatmari

, Jaskowski

, King

, Smith

, Molyneux

, Esiri

, McDonald

, Wald

(1992) Detection in life of confirmed Alzheimer’s disease using a simple measurement of medial temporal lobe atrophy by computed tomography. Lancet 340, 1179–1183.

20.

Fujisawa

, Tsunoda

, Hino

, Shibuya

, Takeda

, Aoki

(2015) Alzheimer’s disease or Alzheimer’s syndrome?: A longitudinal computed tomography neuroradiological follow-up study of 56 cases diagnosed clinically as Alzheimer’s disease. Psychogeriatrics 15, 255–271.

21.

Knight

, McCann

, Kauppinen

, Coulthard

(2016) Magnetic resonance imaging to detect early molecular and cellular changes in Alzheimer’s disease. Front Aging Neurosci 8, 139–148.

22.

Dawe

, Yu

, Leurgans

, Schneider

, Buchman

, Arfanakis

, Bennett

, Boyle

(2016) Postmortem MRI: A novel window into the neurobiology of late life cognitive decline. Neurobiol Aging 45, 169–177.

23.

Del Sole

, Malaspina

, Magenta Biasina

(2016) Magnetic resonance imaging and positron emission tomography in the diagnosis of neurodegenerative dementias. Funct Neurol 31, 205–215.

24.

Duara

, Grady

, Haxby

, Sundaram

, Cutler

, Heston

, Moore

, Schlageter

, Larson

, Rapoport

(1986) Positron emission tomography in Alzheimer’s disease. Neurology 36, 879–879.

25.

Schöll

, Lockhart Samuel

, Schonhaut Daniel

, O’Neil James

, Janabi

, Ossenkoppele

, Baker Suzanne

, Vogel Jacob

, Faria

, Schwimmer Henry

, Rabinovici Gil

, Jagust William

(2016) PET imaging of tau deposition in the aging human brain. Neuron 89, 971–982.

26.

Ono

, Watanabe

, Kitada

, Matsumura

, Ihara

, Saji

(2016) Highly selective tau-SPECT imaging probes for detection of neurofibrillary tangles in Alzheimer’s disease. Sci Rep 6, 34197–34208.

27.

Colloby

, Field

, Wyper

, O’Brien

, Taylor

(2016) A spatial covariance 123I-5IA-85380 SPECT study of alpha4beta2 nicotinic receptors in Alzheimer’s disease. Neurobiol Aging 47, 83–90.

28.

Hye

, Lynham

, Thambisetty

, Causevic

, Campbell

, Byers

, Hooper

, Rijsdijk

, Tabrizi

, Banner

, Shaw

, Foy

, Poppe

, Archer

, Hamilton

, Powell

, Brown

, Sham

, Ward

, Lovestone

(2006) Proteome-based plasma biomarkers for Alzheimer’s disease. Brain 129, 3042–3050.

29.

Liao

, Yu

, Kuo

, Lin

, Kuo

(2007) Proteomics analysis of plasma for potential biomarkers in the diagnosis of Alzheimer’s disease. Proteomics Clin Appl 1, 506–512.

30.

Mehta

, Pirttila

, Mehta

, Sersen

, Aisen

, Wisniewski

(2000) Plasma and cerebrospinal fluid levels of amyloid beta proteins 1-40 and 1-42 in Alzheimer disease. Arch Neurol 57, 100–105.

31.

Song

, Poljak

, Smythe

, Sachdev

(2009) Plasma biomarkers for mild cognitive impairment and Alzheimer’s disease. Brain Res Rev 61, 69–80.

32.

Anoop

, Singh

, Jacob

, Maji

(2010) CSF Biomarkers for Alzheimer’s Disease Diagnosis. Int J Alzheimers Dis 2010, 606802.

33.

Schoonenboom

, Reesink

, Verwey

, Kester

, Teunissen

, van de Ven

, Pijnenburg

, Blankenstein

, Rozemuller

, Scheltens

, van der Flier

(2012) Cerebrospinal fluid markers for differential dementia diagnosis in a large memory clinic cohort. Neurology 78, 47–54.

34.

Rosen

, Hansson

, Blennow

, Zetterberg

(2013) Fluid biomarkers in Alzheimer’s disease - current concepts. Mol Neurodegener 8, 20–31.

35.

Forlenza

, Radanovic

, Talib

, Aprahamian

, Diniz

, Zetterberg

, Gattaz

(2015) Cerebrospinal fluid biomarkers in Alzheimer’s disease: Diagnostic accuracy and prediction of dementia. Alzheimers Dement (Amst) 1, 455–463.

36.

Blennow

, Dubois

, Fagan

, Lewczuk

, de Leon

, Hampel

(2015) Clinical utility of cerebrospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease. Alzheimers Dement 11, 58–69.

37.

Niemantsverdriet

, Goossens

, Struyfs

, Martin

, Goeman

, De Deyn

, Vanderstichele

, Engelborghs

(2016) Diagnostic impact of cerebrospinal fluid biomarker (pre-)analytical variability in Alzheimer’s disease. J Alzheimers Dis 51, 97–106.

38.

Olsson

, Lautner

, Andreasson

, Ohrfelt

, Portelius

, Bjerke

, Holtta

, Rosen

, Olsson

, Strobel

, Wu

, Dakin

, Petzold

, Blennow

, Zetterberg

(2016) CSF and blood biomarkers for the diagnosis of Alzheimer’s disease: A systematic review and meta-analysis. Lancet Neurol 15, 673–684.

39.

Szalárdy

, Zádori

, Klivényi

, L

(2016) The role of cerebrospinal fluid biomarkers in the evolution of diagnostic criteria in Alzheimer’s disease: Shortcomings in prodromal diagnosis. J Alzheimers Dis 53, 373–392.

40.

Keshavan

, Heslegrave

, Zetterberg

, Schott

(2017) Blood biomarkers for Alzheimer’s disease: Much promise, cautious progress. Mol Diagn Ther 21, 13–22.

41.

Rivero-Santana

, Ferreira

, Perestelo-Perez

, Westman

, Wahlund

, Sarria

, Serrano-Aguilar

(2017) Cerebrospinal fluid biomarkers for the differential diagnosis between Alzheimer’s disease and frontotemporal lobar degeneration: Systematic review, HSROC analysis, and confounding factors. J Alzheimers Dis 55, 625–644.

42.

Counts

, Ikonomovic

, Mercado

, Vega

, Mufson

(2017) Biomarkers for the early detection and progression of Alzheimer’s disease. Neurotherapeutics 14, 35–53.

43.

Raman

, Krishnan

(1928) A new type of secondary radiation. Nature 121, 501–502.

44.

Colles

, Griffiths

(1972) Relative and absolute Raman scattering cross sections in liquids. J Chem Phys 56, 3384–3391.

45.

Kim

, Owen

, Carey

(1993) High-performance Raman spectroscopic system based on a single spectrograph, CCD, notch filters, and a Kr+ laser ranging from the near-IR to near-UV regions. Appl Spectrosc 47, 1780–1783.

46.

Hirschfeld

, Chase

(1986) FT-Raman spectroscopy: Development and justification. Appl Spectrosc 40, 133–137.

47.

Carter

RAB

, Martin

, Netto

, Soares

(2004) FT-Raman spectroscopy study of human breast tissue. In Proceedings of the SPIE - The International Society for Optical Engineering, pp. 190–197.

48.

Shim

, Wilson

(1997) Development of an in vivo Raman spectroscopic system for diagnostic applications. J Raman Spectrosc 28, 131–142.

49.

Kneipp

, Kneipp

(2008) SERS— a single-molecule and nanoscale tool for bioanalytics. Chem Soc Rev 37, 1052–1060.

50.

Sharma

, Frontiera

, Henry

A-I

, Ringe

, Van Duyne

(2012) SERS: Materials, applications, and the future. Mater Today 15, 16–25.

51.

Doering

, Nie

(2002) Single-molecule and single-nanoparticle SERS: Examining the roles of surface active sites and chemical enhancement. J Phys Chem B 106, 311–317.

52.

Laing

, Gracie

, Faulds

(2016) Multiplex in vitro detection using SERS. Chem Soc Rev 45, 1901–1918.

53.

Cheng

J-X

, Xie

(2004) Coherent anti-Stokes Raman scattering microscopy: Instrumentation, theory, and applications. J Phys Chem B 108, 827–840.

54.

Rodriguez

, Lockett

, Holtom

(2006) Coherent anti-stokes Raman scattering microscopy: A biological review. Cytometry A 69, 779–791.

55.

Evans

, Xie

(2008) Coherent anti-Stokes Raman scattering microscopy: Chemical imaging for biology and medicine. Annu Rev Anal Chem 1, 883–909.

56.

Carmona

, Molina

, Calero

, Bermejo-Pareja

, Martínez-Martín

, Toledano

(2013) Discrimination analysis of blood plasma associated with Alzheimer’s disease using vibrational spectroscopy. J Alzheimers Dis 34, 911–920.

57.

Carmona

, Molina

, López-Tobar

, Toledano

(2015) Vibrational spectroscopic analysis of peripheral blood plasma of patients with Alzheimer’s disease. Anal Bioanal Chem 407, 7747–7756.

58.

Zengin

, Tamer

, Caykara

(2013) A SERS-based sandwich assay for ultrasensitive and selective detection of Alzheimer’s tau protein. Biomacromolecules 14, 3001–3009.

59.

Guerrini

, Arenal

, Mannini

, Chiti

, Pini

, Matteini

, Alvarez-Puebla

(2015) SERS detection of amyloid oligomers on metallorganic-decorated plasmonic beads. ACS Appl Mater Interfaces 7, 9420–9428.

60.

Beier

, Cowan

, Chou

I-H

, Pallikal

, Henry

, Benford

, Jackson

, Good

, Coté

(2007) Application of surface-enhanced Raman spectroscopy for detection of beta amyloid using nanoshells. Plasmonics 2, 55–64.

61.

Demeritte

, Nellore

, Kanchanapally

, Sinha

, Pramanik

, Chavva

, Ray

(2015) Hybrid graphene oxide based plasmonic-magnetic multifunctional nanoplatform for selective separation and label-free identification of Alzheimer’s disease biomarkers. ACS Appl Mater Interfaces 7, 13693–13700.

62.

Chou

, Benford

, Beier

, Coté

, Wang

, Jing

, Kameoka

, Good

(2008) Nanofluidic biosensing for β-amyloid detection using surface enhanced Raman spectroscopy. Nano Lett 8, 1729–1735.

63.

Choi

, Huh

, Erickson

(2011) Size-selective concentration and label-free characterization of protein aggregates using a Raman active nanofluidic device. Lab Chip 11, 632–638.

64.

Choi

, Huh

, Erickson

(2012) Ultra-sensitive, label-free probing of the conformational characteristics of amyloid beta aggregates with a SERS active nanofluidic device. Microfluid Nanofluidics 12, 663–669.

65.

Lanctot

, Herrmann

, Mazzotta

, Khan

, Ingber

(2004) GABAergic function in Alzheimer’s disease: Evidence for dysfunction and potential as a therapeutic target for the treatment of behoural and psychological symptoms of dementia. Can J Psychiatry 49, 439–453 avi.

66.

Monfared

AMT

, Tiwari

, Trudeau

, Anis

(2015) Surface-enhanced Raman Scattering spectroscopy for the detection of glutamate and α-aminobutyric acid in serum by partial least squares analysis. IEEE Photon J 7, 1–16.

67.

Castro

, Sanchez-Cortes

, Ramos

JVG

, Otero

, Marcos

(1997) Surface-enhanced Raman spectroscopy of γ-aminobutyric acid on silver colloid surfaces. Biospectroscopy 3, 449–455.

68.

Tiwari

, Khetani

, Monfared

AMT

, Smith

, Anis

, Trudeau

(2012) Detection of amino acid neurotransmitters by surface enhanced Raman scattering and hollow core photonic crystal fiber. In SPIE BiOS International Society for Optics and Photonics, pp. 82330Q-82330Q-82336.

69.

Ryzhikova

, Kazakov

, Halamkova

, Celmins

, Malone

, Molho

, Zimmerman

, Lednev

(2014) Raman spectroscopy of blood serum for Alzheimer’s disease diagnostics: Specificity relative to other types of dementia. J Biophotonics 8, 584–596.

70.

Chen

, Tian

, Baek

, Shang

, Park

, Liu

, Yao

, Wang

, Cheng

, Peng

, Shen

, Hu

(2011) Laser Raman detection of platelet as a non-invasive approach for early and differential diagnosis of Alzheimer’s disease. Laser Phys Lett 8, 547–552.

71.

Wang

, Du

, Yang

, Chen

, Tian

, Shang

, Liu

, Yao

, Wang

(2014) Raman spectroscopy detection of platelet for Alzheimer’s disease with predictive probabilities. Laser Physics 24, 085702–085708.

72.

Sajid

, Elhaddaoui

, Turrell

(1997) Fourier transform vibrational spectroscopic analysis of human cerebral tissue. J Raman Spectrosc 28, 165–169.

73.

Chen

, Shen

, Zhao

, Baek

S-J

, Yuan

, Hu

(2009) Raman signature from brain hippocampus could aid Alzheimer’s disease diagnosis. Appl Opt 48, 4743–4748.

74.

Sajid

, Elhaddaoui

, Turrell

(1997) Investigation of the binding of Congo red to amyloid in Alzheimer’s diseased tissue. J Mol Struct 408–409, 181–184.

75.

De Grauw

, Sijtsema

, Otto

, Greve

(1997) Axial resolution of confocal Raman microscopes: Gaussian beam theory and practice. J Microsc 188, 273–279.

76.

Michael

, Otto

, Lenferink

, Gelpi

, Montenegro

, Rosandić

, Tresserra

, Barraquer

, Vrensen

GFJM

(2014) Absence of amyloid-beta in lenses of Alzheimer patients: A confocal Raman microspectroscopic study. Exp Eye Res 119, 44–53.

77.

Schie

, Krafft

, Popp

(2015) Applications of coherent Raman scattering microscopies to clinical and biological studies. Analyst 140, 3897–3909.

78.

Enejder

, Kiskis

, Fink

, Nyberg

, Thyr

, Li

J-Y

(2014) CARS microscopy of Alzheimer’s diseased brain tissue. In SPIE BiOS International Society for Optics and Photonics, pp. 89480U-89480U-89487.

79.

Kiskis

, Fink

, Nyberg

, Thyr

, Li

J-Y

, Enejder

(2015) Plaque-associated lipids in Alzheimer’s diseased brain tissue visualized by nonlinear microscopy. Sci Rep 5, 13489–13498.

80.

Lee

, Kim

, Song

, Oh

, Ko

(2015) Label-free imaging and quantitative chemical analysis of Alzheimer’s disease brain samples with multimodal multiphoton nonlinear optical microspectroscopy. J Biomed Opt 20, 56013–56020.

81.

Stevens

, Iping Petterson

, Day

JCC

, Stone

(2016) Developing fibre optic Raman probes for applications in clinical spectroscopy. Chem Soc Rev 45, 1919–1934.