Salivary Biomarkers of Parkinson's disease

This blog post marks the publication of our first paper on ‘salivary biomarkers of Parkinson’s disease’, resulting from a collaboration between the PREDICT-PD group and the saliva research laboratories at King’s College London.

Biomarkers are measurable indicators of the severity or presence of a particular disease. For example, blood cholesterol levels are a useful biomarker for risk of heart attack or stroke. As of yet, there are no such biomarkers for Parkinson’s disease but having some sort measurable biological indicator would be useful to help achieve earlier diagnosis, to monitor the progression of the disease or even to monitor the response to new treatments in clinical trials aimed at slowing the progression of the disease. Such a biomarker could one day be used alongside the online survey tools being developed as part of the PREDICT-PD project to help identify people with a high risk of developing Parkinson’s disease over the next few years.

Saliva may be an ideal fluid in which to look for biomarkers of Parkinson’s disease for two main reasons. First, collection of saliva is relatively easy compared with taking a blood sample (and certainly much less uncomfortable than having a lumbar puncture to collect spinal fluid!) Second, there are good reasons to believe that saliva may actually be affected in Parkinson’s disease. Some patients with Parkinson’s report that they experience dry mouth, excessive saliva and drooling or that their saliva has become particularly sticky and can stain clothes. There is also evidence that the nerves involved in the control of the saliva glands are affected in Parkinson’s disease.

A test for Parkinson's disease based on saliva would be cheaper, easier to perform and much more acceptable to patients than a test requiring a lumbar puncture to collect cererbro-spinal fluid!

With this in mind, we conducted a small pilot study by collecting saliva from 16 patients with Parkinson’s and 22 control participants of a similar age to the patients and then performed a variety of tests to compare the composition of the saliva from the two groups.

We found that there was more protein in the saliva of patients with Parkinson’s disease. Further tests indicated that the source of this additional protein appeared to be the ‘major’ salivary glands and that the additional protein in patient saliva could reflect a dysfunction of the nerves involved in the control of these major salivary glands (which we know from previous studies to be affected in Parkinson’s disease).

We also found that among the patients with Parkinson’s disease, a specific protein in the saliva, called ‘DJ-1’, appeared to correlate with the severity of their disease (specifically, salivary DJ-1 correlated with a Parkinson’s disease severity rating scale called the UPDRS score). This was particularly interesting because DJ-1 is a protein which we already know to be involved in the process that leads to degeneration of neurons in PD. In addition, DJ-1 has previously been shown to be a potential biomarker of Parkinson’s disease in the blood and in the cerebro-spinal fluid (the fluid which bathes the brain and the spinal cord).

(a) The saliva of patients with PD contained more protein than the saliva of patients with controls. (b) The salivary concentration of a protein called DJ-1 correlated with disease severity (measured by UPDRS score)

By finding that interesting difference exist between the saliva of patients with Parkinson’s disease and controls we believe we have strengthened the case that saliva may be an ideal fluid for further biomarker discovery work. Our ambition is that one day a saliva-based test could be used (alongside the other tools being developed as part of the PREDICT-PD project) to help identify people at high risk of – or even help diagnose – Parkinson’s disease.

Elevated salivary protein in Parkinson’s disease and salivary DJ-1 as a potential marker of disease severity

Parkinsonism and Related Disorders 2015 Published online: July 23, 2015 DOI: http://dx.doi.org/10.1016/j.parkreldis.2015.07.021

Joseph M. Masters, Alastair J. Noyce, Thomas T. Warner, Gavin Giovannoni, Gordon B. Proctor

Abstract

Introduction:
There is an urgent need to identify robust biomarkers for Parkinson’s disease (PD). Previous studies have shown changes in composition and secretion of saliva in patients with PD, including an increase in salivary DJ-1 concentration. Autonomic dysfunction is a known feature of PD and could contribute to abnormal saliva gland function.

Methods:

In this pilot cross-sectional study, characterisation of the saliva of 16 patients with PD and 22 age-matched controls was performed. Salivary DJ-1 concentration was measured with quantitative immunoblotting; total protein concentration with a BCA assay and spectrophotometry; amylase with an amylase activity assay; albumin with an ELISA and mucin concentration with periodic-acid Schiff staining of SDS-gels.

Results:

Patient saliva showed an increase in both total protein concentration (8.4 vs 5.0 mg/ml, p=0.0002) and DJ-1 concentration (0.84 vs 0.42 μg/ml, p=0.001), but there was no difference in salivary DJ-1 after adjusting for total protein concentration. In patients, adjusted DJ-1 levels correlated with disease severity measured with the MDS-Unified Parkinson’s Disease Rating Scale (p=0.019). Patient saliva had elevated concentrations of amylase (127 vs 64 units/ml, p=0.0005) and albumin (110 vs 41 μg/ml, p=0.0003) but not mucins.

Conclusions:

This study suggests that the saliva of patients with PD is different in composition to that of healthy age-matched controls, supporting the notion that saliva may be a good candidate for biomarker discovery in PD. The specific differences suggest that major salivary glands and gingival crevice fluid may both be sources of additional DJ-1 and protein in patient saliva.

Masters, J., Noyce, A., Warner, T., Giovannoni, G., & Proctor, G. (2015). Elevated salivary protein in Parkinson’s disease and salivary DJ-1 as a potential marker of disease severity Parkinsonism & Related Disorders DOI: 10.1016/j.parkreldis.2015.07.021