Anal Bioanal Chem. 2015 Oct;407(25):7747-56. doi: 10.1007/s00216-015-8940-7.

Vibrational spectroscopic analysis of peripheral blood plasma of patients with Alzheimer’s disease.

 

Carmona P1, Molina M2, López-Tobar E3, Toledano A4.
  • 1Instituto de Estructura de la Materia, CSIC, Serrano 121, 28006, Madrid, Spain. p.carmona@iem.cfmac.csic.es.
  • 2Facultad de Óptica y Optometría, Universidad Complutense, Avda Arcos de Jalón 118, 28037, Madrid, Spain.
  • 3Instituto de Estructura de la Materia, CSIC, Serrano 121, 28006, Madrid, Spain.
  • 4Instituto Cajal, CSIC, Avda. Doctor Arce 37, 28002, Madrid, Spain.

 

Abstract

Using Raman and infrared spectroscopy, we monitored spectral changes occurring in the blood plasma of patients with Alzheimer’s disease (AD) in relation to healthy controls. The protein secondary structure as reflected by amide I band involves β-sheet enrichment, which may be attributable to Aβ peptide formation and to increasing proportion of the globulins that are β-sheet rich. Likewise, the behavior of the infrared 1200-1000-cm(-1) region and the Raman 980-910- and 450-400-cm(-1) regions can be explained in terms of the said plasma composition change. Further, the 744-cm(-1) Raman band from healthy control plasma shows frequency upshifting in the course of AD, which may be generated by the platelets collected in blood plasma. Linear discrimination analysis and receiver operating characteristic (ROC) analysis have been used to distinguish between patients with AD and age-matched healthy controls with a diagnostic accuracy of about 94 %. Graphical Abstract ROC assessment of sample classifications by Raman spectroscopy (dashed line) and linear combination of infrared and Raman spectroscopies (solid line).

KEYWORDS: Alzheimer’s disease; Blood plasma; Infrared spectroscopy; Raman spectroscopy

PMID: 26255297

 

Supplement:

Alzheimer’s disease (AD) is the most common dementia and approximately one in eight people over 65 years old are at risk. AD is an age-related and insidious-onset neurodegenerative disease. Diagnosis of AD is time consuming and requires a combination of psychological testing, imaging and exclusion of other neurological disorders. In light of these facts, the identification of peripheral biomarkers of the disease process leading to an effective diagnostic test for AD would be valuable for monitoring the efficacy of disease interventions during clinical trials. Currently, cerebrospinal fluid (CSF) has provided the most promising source of validated AD biomarkers (1). However, compared with CSF, blood analysis has advantages as an approach to population-based disease screening because it is simpler and less invasive. Because the brain controls many body functions via the release of signaling proteins, and because central and peripheral immune and inflammatory mechanisms are increasingly implicated in Alzheimer’s (2) and related diseases (3), we hypothesized that the pathological processes leading to Alzheimer’s would cause characteristic changes in the concentrations of proteins in the blood, generating a detectable disease-specific molecular phenotype. Moreover, about 500 ml of human CSF is absorbed into the blood daily (4), making blood plasma a suitable source of neurodegenerative disease biomarkers. In this connection, infrared and Raman spectroscopies are physical techniques that may potentially play a critical role in this area of diagnosis. These spectroscopic approaches are very convenient because no reagent and only small amounts of samples are required. Moreover, these spectroscopic methods are not time consuming and measure the variations in the molecular structure and composition of biological materials.

Figure1

Figure 1. Bar diagram showing means±SD of area percentages of protein b-structure for patients with AD (mild, moderate and severe) and age-matched healthy controls, as determined by infrared spectroscopy.

 

We asked whether blood plasma of patients with AD could be distinguished from that of healthy controls having normal cognitive functions, and in the same way we also performed infrared and Raman analyses in terms of AD-induced biochemical changes. With this aim, in our study we considered 50 patients affected by AD at different levels of severity (mild, moderate and severe), and 14 matched-age healthy controls.

 

The differences related to protein b-sheet content were interesting. Figure 1 shows that significant differences exist between AD patients and healthy controls. However, nonsignificant differences were found between any pair drawn from the three AD levels studied here. Ab amyloid peptides as well as a-2-macroglobulin and a-1-antitrypsin, which have b-sheet-rich polypeptide backbones (5, 6), have been reported as biomarkers of AD (7, 8). The same can be said for IgG protein as major fraction of g-globulin (9,10). Therefore, the increasing of protein b-sheet content as determined by vibrational spectroscopy in the 1700-1600 cm-1 region (amide I band) may be explained in terms of contributions of the above protein substances.

 

Another spectral profile which is sensitive to AD is located in the 1200-1000 cm-1 region of plasma infrared spectrum (Figure 2). Generation of highly reactive secondary products of lipid peroxidation may contribute to cellular damage and therefore measurement of these products has been commonly used to assess oxidative stress/injury. Hydroxyl compounds (for instance, isoprostanes, hydroperoxides) are products of lipid peroxidation and have been reported as markers of oxidative damage in AD (11). Therefore, the changes in the 1200-1000 cm-1 spectral profile may be attributable to C-O stretching vibrations of hydroxyl compounds stemming from lipid peroxidation. In addition, these spectral changes in the course of AD, which we have used for AD diagnosis, may also be due to contributions of C-O stretching bands from carbohydrate moieties in plasma globulins.

 

Figure2

Figure 2. Scheme of the AD diagnostics procedure including interactions of either the laser beam or the infrared beam with plasma molecules and the resulting Raman spectrum or infrared spectrum and subsequent ROC curves.

 

Another interesting feature is related to the origin of the frequency upshifting of the 744 cm-1 Raman band in the course of AD, which can be attributable to changes in platelets from AD patients because this frequency upshifting is not visible in spectra from either serum or platelet-poor plasma. In fact, for many years platelets were accepted as peripheral model to study the pathophysiology of AD because platelets display the enzymatic activities to generate amyloid-ß peptides. In addition, platelets are also considered to be a biomarker for early diagnosis of AD (12).

 

With the aim of validating the significance of the above spectral parameters as potential AD biomarkers, we carried out discriminant analysis and subsequent receiver-operating characteristic analysis (ROC curve, Figure 2) of the discriminant function by assuming that the levels of these parameters follow normal distributions in two groups (13). An experimental measure of diagnostics test accuracy is provided by the area under the curve (AUC). Figure 2 reveals that the linear combination of Raman and infrared spectral parameters displayed the highest AUC value of 0.94 (solid line) when compared with the Raman parameters alone (0.87) (dotted line). This indicates that, within a binary classification scheme provided by the discriminant function, a randomly selected positive case has a 94% probability of obtaining a higher score than a randomly selected negative case. Therefore, the AUC is an experimental measure of test accuracy, and on the basis of linear discrimination analysis and subsequent ROC curve of combined Raman and infrared spectral parameters, the discrimination accuracy is about 94%. In addition, the discriminant function alone of these combined parameters provided a similar percentage of cross-validated cases correctly classified. On the other hand, ROC analysis of discriminant function shows also that the combination of the said spectral parameters leads to improved sensitivity and can identify those patients who missed the diagnosis of AD by using infrared and Raman techniques separately. A classification model based on artificial neural networks (ANN) has been carried out for differential diagnosis of AD by using Raman spectra of blood serum (14). Although its overall accuracy (95%) was very similar to the best discrimination accuracy described above, our model is expected to be more robust because the number of AD cases and healthy controls studied here are significantly higher (50 patients affected by AD and 14 matched-age healthy controls), and because we have combined two vibrational spectroscopic techniques. Furthermore, unlike the said ANN work, spectral changes have been described here in terms of particular plasma substances whose relative contents can vary in the course of AD. Nevertheless, it will be necessary to analyse more samples including those from other conditions with differential diagnosis purposes.

 

Importance of the study:

This work shows how important is the use of combined infrared and Raman spectroscopies as potential tools for diagnosis of AD, from three points of view: firstly, because no reagent and only small amounts of samples are required; secondly, these techniques are non-invasive and not time consuming; and thirdly, the high sensitivity and specificity percentages obtained through the use of the above procedure.

 

References:

  1. Roe CM, Fagan AM, Williams MM, Ghoshal N, Aeschleman M, Grant EA, …, Morris JC (2011) Improving CSF biomarker accuracy in predicting prevalent and incident Alzheimer disease. Neurology 76:501-510.
  2. Wyss-Coray T (2006) Inflammation in Alzheimer disease: driving force, bystander or beneficial response? Nat Med 12:1005-1015.
  3. Steinman L (2004) Elaborate interactions between the immune and nervous systems. Nat Immunol 5:575-581.
  4. Hye A, Lynham S, Thambisetty M, Causevic M, Campbell J, Byers HL, …, Lovestone S (2006) Proteome-based plasma biomarkers for Alzheimer’s disease. Brain 129:3042-3050.
  5. Marrero A, Duquerroy S, Trapani S, Goulas T, Guevara T, Andersen GR, Navaza J, Sottrup-Jensen L, Gomis-Rüth FX (2012) The crystal structure of human a2-macroglobulin reveals a unique molecular cage. Angew Chem Int Ed 51:3340-3344.
  6. Whisstock JC, Skinner R, Carrell RW, Lesk AM (2000) Conformational changes in serpins: I. The native and cleaved conformations of a1-antitrypsin. J Mol Biol 296:685-699.
  7. Hye A, Lynham S, Thambisetty M, Causevic M, Campbell J, Byers HL, Hooper C, Rijsdijk F, Tabrizi SJ, Banner S, Shaw CE, Foy C, Poppe M, Archer N, Hamilton G, Powell J, Brown RG, Sham P, Ward M, Lovestone S (2006) Proteome-based plasma biomarkers for Alzheimer’s disease. Brain 129:3042-3050.
  8. Liao PC, Yu L, Kuo CC, Lin C, Kuo YM (2007) Proteomics analysis of plasma for potential biomarkers in the diagnosis of Alzheimer’s disease. Proteomics Clin Appl 1:506-512.
  9. Girardi E, Holdom MD, Davies AM, Sutton BJ, Beavil AJ (2009) The crystal structure of rabbit IgG-Fc. Biochem J 417:77-83.
  10. Koehler NKU, Stransky E, Shing M, Gaertner S, Meyer M, Schreitmüller B,…, Richartz-Salzburger E (2013) Altered serum IgG levels to a-synuclein in dementia with Lewy bodies and Alzheimer’s disease. Plos One. Doi: 10.1371/journal.pone.0064649.
  11. Song F, Poljak A, Smythe GA, Sachdev P (2009) Plasma biomarkers for mild cognitive impairment and Alzheimer’s disease. Brain Res Rev 61:69-80.
  12. Gowert NS, Donner L, Chatterjee M, Eisele YS, Towhid ST, Münzer P,…, Elvers M (2014) Blood platelets in the progression of Alzheimer’s disease. Plos One 9:e90523.
  13. Carmona P, Molina M, López-Tobar E, Toledano A (2015) Vibrational spectroscopic analysis of peripheral blood plasma of patients with Alzheimer’s disease. Anal Bioanal Chem 407:7747-7756.
  14. Ryzhikova E, Kazakov O, Halámková L, Celmins D, Malone P, Molho E,…, Lednev IK (2014) Raman spectroscopy of blood serum for Alzheimer’s disease diagnostics: specificity relative to other types of dementia. J Biophotonics 8:1-13.

 

Acknowledgements

This work was supported by the Spanish Ministerio de Ciencia e Innovación (Project CTQ2009-09538/BQU) and Ministerio de Economía y Competitividad (Project IPT-2012- 0769-010000).

 

Multiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier SchönmannMultiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier Schönmann