Limits...
Huntington's disease biomarker progression profile identified by transcriptome sequencing in peripheral blood.

Mastrokolias A, Ariyurek Y, Goeman JJ, van Duijn E, Roos RA, van der Mast RC, van Ommen GB, den Dunnen JT, 't Hoen PA, van Roon-Mom WM - Eur. J. Hum. Genet. (2015)

Bottom Line: Relative to previous studies, this yielded novel genes and confirmed previously identified genes, such as H2AFY, an overlap in results that has proven difficult in the past.Using a highly parallelized microfluidics array chip (Fluidigm), we validated 12 of the top 20 significant genes in our discovery cohort and 7 in a second independent cohort.Finally we suggest a first empiric formula predicting total motor score from the expression levels of our biomarker panel.

View Article: PubMed Central - PubMed

Affiliation: Department of Human Genetics, Leiden University Medical Center, Leiden, The Netherlands.

ABSTRACT
With several therapeutic approaches in development for Huntington's disease, there is a need for easily accessible biomarkers to monitor disease progression and therapy response. We performed next-generation sequencing-based transcriptome analysis of total RNA from peripheral blood of 91 mutation carriers (27 presymptomatic and, 64 symptomatic) and 33 controls. Transcriptome analysis by DeepSAGE identified 167 genes significantly associated with clinical total motor score in Huntington's disease patients. Relative to previous studies, this yielded novel genes and confirmed previously identified genes, such as H2AFY, an overlap in results that has proven difficult in the past. Pathway analysis showed enrichment of genes of the immune system and target genes of miRNAs, which are downregulated in Huntington's disease models. Using a highly parallelized microfluidics array chip (Fluidigm), we validated 12 of the top 20 significant genes in our discovery cohort and 7 in a second independent cohort. The five genes (PROK2, ZNF238, AQP9, CYSTM1 and ANXA3) that were validated independently in both cohorts present a candidate biomarker panel for stage determination and therapeutic readout in Huntington's disease. Finally we suggest a first empiric formula predicting total motor score from the expression levels of our biomarker panel. Our data support the view that peripheral blood is a useful source to identify biomarkers for Huntington's disease and monitor disease progression in future clinical trials.

No MeSH data available.


Related in: MedlinePlus

DeepSAGE gene expression levels for the best three TMS predictor genes as these were reported from the LASSO algorithm prediction analysis and across controls, presymptomatics and different HD symptomatic stages.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4592077&req=5

fig4: DeepSAGE gene expression levels for the best three TMS predictor genes as these were reported from the LASSO algorithm prediction analysis and across controls, presymptomatics and different HD symptomatic stages.

Mentions: To evaluate which panel of genes would optimally predict TMS, we fitted a linear regression model with a lasso penalty using the Fluidigm expression data, age and gender as predictors and TMS as the response. The gene expression values of three genes (AQP9, ANXA3 and ARL4C), together with age and gender, were the best predictors of TMS. The last gene (ARL4C) was non-significantly downregulated in HD blood and specifically served the purpose of enlarging the ‘biomarker chip' set towards tolerance for smaller individual gene changes, providing additional informativeness. The results of the cross-validated prediction analysis can be seen in Figure 3. The prediction model performed better for earlier disease stages (Stage I, II), while it was less accurate for later stages (Stage III-V) and especially for patients with a motor score of 50 points and over. Only one patient was assigned a predicted TMS >50 points (patient no.29). This patient was the oldest HD carrier (>70 years). We also observed that for one patient the blood-based signature indicated a higher predicted motor score compared with the clinical motor score. This could be explained by the fact that this patient had a much lower TFC score (TFC=4) compared with other patients with similar motor score, indicative of a more advanced disease stage. Finally, the control sample with the highest clinical motor score (control no.4) was our oldest control sample (69 years) and also received a higher predicted score. When we plotted the DeepSAGE gene expression levels of these three genes across the controls, the presymptomatic carriers and the different HD TFC-based disease stages, we could confirm that for ANXA3 and AQP9 there was an increase in gene expression even in the presymptomatic stage. For ARL4C, contrary to ANXA3 and AQP9 there was a decrease in gene expression, the expression changes were more prominent in the more advanced disease stages and hence provided complementary information to the other two genes (Figure 4). On the basis of this analysis, we formulated the following TMS predictive equation to measure the disease stage based on gene expression of the three genes:


Huntington's disease biomarker progression profile identified by transcriptome sequencing in peripheral blood.

Mastrokolias A, Ariyurek Y, Goeman JJ, van Duijn E, Roos RA, van der Mast RC, van Ommen GB, den Dunnen JT, 't Hoen PA, van Roon-Mom WM - Eur. J. Hum. Genet. (2015)

DeepSAGE gene expression levels for the best three TMS predictor genes as these were reported from the LASSO algorithm prediction analysis and across controls, presymptomatics and different HD symptomatic stages.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4592077&req=5

fig4: DeepSAGE gene expression levels for the best three TMS predictor genes as these were reported from the LASSO algorithm prediction analysis and across controls, presymptomatics and different HD symptomatic stages.
Mentions: To evaluate which panel of genes would optimally predict TMS, we fitted a linear regression model with a lasso penalty using the Fluidigm expression data, age and gender as predictors and TMS as the response. The gene expression values of three genes (AQP9, ANXA3 and ARL4C), together with age and gender, were the best predictors of TMS. The last gene (ARL4C) was non-significantly downregulated in HD blood and specifically served the purpose of enlarging the ‘biomarker chip' set towards tolerance for smaller individual gene changes, providing additional informativeness. The results of the cross-validated prediction analysis can be seen in Figure 3. The prediction model performed better for earlier disease stages (Stage I, II), while it was less accurate for later stages (Stage III-V) and especially for patients with a motor score of 50 points and over. Only one patient was assigned a predicted TMS >50 points (patient no.29). This patient was the oldest HD carrier (>70 years). We also observed that for one patient the blood-based signature indicated a higher predicted motor score compared with the clinical motor score. This could be explained by the fact that this patient had a much lower TFC score (TFC=4) compared with other patients with similar motor score, indicative of a more advanced disease stage. Finally, the control sample with the highest clinical motor score (control no.4) was our oldest control sample (69 years) and also received a higher predicted score. When we plotted the DeepSAGE gene expression levels of these three genes across the controls, the presymptomatic carriers and the different HD TFC-based disease stages, we could confirm that for ANXA3 and AQP9 there was an increase in gene expression even in the presymptomatic stage. For ARL4C, contrary to ANXA3 and AQP9 there was a decrease in gene expression, the expression changes were more prominent in the more advanced disease stages and hence provided complementary information to the other two genes (Figure 4). On the basis of this analysis, we formulated the following TMS predictive equation to measure the disease stage based on gene expression of the three genes:

Bottom Line: Relative to previous studies, this yielded novel genes and confirmed previously identified genes, such as H2AFY, an overlap in results that has proven difficult in the past.Using a highly parallelized microfluidics array chip (Fluidigm), we validated 12 of the top 20 significant genes in our discovery cohort and 7 in a second independent cohort.Finally we suggest a first empiric formula predicting total motor score from the expression levels of our biomarker panel.

View Article: PubMed Central - PubMed

Affiliation: Department of Human Genetics, Leiden University Medical Center, Leiden, The Netherlands.

ABSTRACT
With several therapeutic approaches in development for Huntington's disease, there is a need for easily accessible biomarkers to monitor disease progression and therapy response. We performed next-generation sequencing-based transcriptome analysis of total RNA from peripheral blood of 91 mutation carriers (27 presymptomatic and, 64 symptomatic) and 33 controls. Transcriptome analysis by DeepSAGE identified 167 genes significantly associated with clinical total motor score in Huntington's disease patients. Relative to previous studies, this yielded novel genes and confirmed previously identified genes, such as H2AFY, an overlap in results that has proven difficult in the past. Pathway analysis showed enrichment of genes of the immune system and target genes of miRNAs, which are downregulated in Huntington's disease models. Using a highly parallelized microfluidics array chip (Fluidigm), we validated 12 of the top 20 significant genes in our discovery cohort and 7 in a second independent cohort. The five genes (PROK2, ZNF238, AQP9, CYSTM1 and ANXA3) that were validated independently in both cohorts present a candidate biomarker panel for stage determination and therapeutic readout in Huntington's disease. Finally we suggest a first empiric formula predicting total motor score from the expression levels of our biomarker panel. Our data support the view that peripheral blood is a useful source to identify biomarkers for Huntington's disease and monitor disease progression in future clinical trials.

No MeSH data available.


Related in: MedlinePlus