Early Diagnosis of Sarcocystis neurona Infection Using Blood Gene Expression Biomarkers

A genetic biomarker signature provides a specific and sensitive indicator of active and early equine protozoal myeloencephalitis (EPM) in the horse.

1. Introduction

Profiles of gene expression in circulating white blood cells can be used as surrogate markers of disease activity (genomic biomarkers) [1]. These biomarkers can provide information on the stage of disease and prognosis and can therefore also be used in disease monitoring. In many cases, genomic biomarkers can detect evidence of disease and predict prognosis before clear clinical symptoms are evident, as has been shown in human cancer diagnosis [2]. The purpose of this study was to identify blood-based gene expression biomarkers in the early phases of Sarcocystis neurona infection, the most common cause of equine protozoal myeloencephalitis (EPM).

2. Materials and Methods

In this study, EPM was induced in horses using an established model [3]. Twenty horses were subjected to transportation stress, and 18 were experimentally infected with 1.5 million mouse infective units (MIUs) of S. neurona, which had been characterized in interferon-γ knockout mice. This study was performed as a companion study evaluating the use of a pharmaceutical compound for the prevention of clinical signs of EPM.

Blood was collected at 10 different times over a 4-wk period (days 0, 2, 4, 7, 9, 11, 14, 17, 21, 24, and 28) into standard serum and ethylenediamine tetra-acetic acid (EDTA) tubes for hematology and biochemistry and PAXgene tubes for RNA isolation [a]. Neurological examinations of all horses were conducted by two veterinarians on the same days as the blood draws. Confirmation of S. neurona infection and clinical evidence of disease were based on positive immunoblot tests for S. neurona antibodies in serum and cerebrospinal fluid, characteristic clinical neurologic abnormalities, and gross and histological changes at autopsy.

Total RNA was isolated from the PAXgene tubes using the Qiagen PAXgene RNA Isolation Kit. RNA was amplified and labeled according to Affymetrix standard procedures and hybridized to a custom equine gene expression microarray (Equine GeneChip) [b]. Genetraks' Equine GeneChip contains probe sets for ~3100 unique equine genes derived from Genetraks' proprietary gene sequence database. After hybridization, arrays were imaged and analyzed by standard algorithms using MAS5.0[c] and robust multichip analysis (RMA) [4].

3. Results

Twelve animals (60%) developed clinical evidence of EPM, seven animals (35%) did not develop clinical evidence of disease, and one animal was withdrawn from the study because of an unrelated illness. Differences in gene expression between animals with and without clinical evidence of EPM were determined using the empirical Bayes statistical approach of Lonnstedt and Speed [5]. Genes that showed statistically significant expression differences between horses with EPM symptoms and horses without symptoms were tabulated for each day post-infection and formed the basis for a genomic biomarker signature of disease.

Cross-validated classification scores showed a statistically significant difference between animals that did not show clinical evidence of EPM and those with EPM. From day 2 post-challenge onward, 23 genes were found to be highly statistically differentially expressed between animals with clinical EPM disease and animals without symptoms (p < 0.001). Six of the 23 expression biomarkers identified using the GeneChip have been validated using quantitative reverse transcriptase-polymerase chain reaction (RT-PCR).

The EPM gene signature obtained from horses participating in the trial with clinical evidence of EPM was unique. Sensitivity and specificity were tested by comparison with the DNA signature from 600 additional horses that were either clinically normal or that had non-EPM illness. Sensitivity and specificity were tested on each of the sample collection days up to 28 days after infection for evaluation. Sensitivity ranged from 0.92 to 0.75, and specificity ranged from 1.0 to 0.71. The highest sensitivity and specificity was noted on day 24 after infection; the sensitivity was 0.92 and the specificity was 1.0. The weakest performance was on day 4 after infection; sensitivity and specificity were 0.75 and 0.71, respectively. These results represent a very strong performance of this diagnostic test during the acute stage of infection. Test performance in more chronic stages of the disease was not studied.

4. Discussion

The demonstration of a distinct gene expression biomarker signature for clinical EPM disease in equine peripheral blood leukocytes represents a significant advance over current methods for the diagnosis of EPM, an often intractable neurological disease in horses. The biomarker signature represents a specific and sensitive indicator of active and early disease.

Further studies on the applicability of this gene signature in clinical cases and in chronic disease are currently underway.

Footnotes

[a] PreAnalytiX Inc., Valencia, CA 91355.

[b] Developed by Genetraks Holdings (Bethesda, MD) in collaboration with Affymetrix, Inc. (Santa Clara, CA 95051).

[c] Affymetrix, Inc., Santa Clara, CA 95051.