Limits...
Whose Gene Is It Anyway? The Effect of Preparation Purity on Neutrophil Transcriptome Studies.

Thomas HB, Moots RJ, Edwards SW, Wright HL - PLoS ONE (2015)

Bottom Line: The level of contamination was assessed by morphology and flow cytometry.RNA-seq analysis identified only 25 genes that were significantly differentially-expressed between Polymorphprep and negatively-selected neutrophils across all three treatment groups (untreated, GM-CSF, TNFα).The expression levels of 34 cytokines/chemokines both before and after GM-CSF or TNFα treatment were not significantly different between neutrophil isolation methods and therefore not affected by contributions from non-neutrophil cell types.

View Article: PubMed Central - PubMed

Affiliation: Institute of Integrative Biology, University of Liverpool, Liverpool, United Kingdom.

ABSTRACT
Protocols for the isolation of neutrophils from whole blood often result in neutrophil preparations containing low numbers (~5%) of contaminating leukocytes, and it is possible that these contaminating cells contribute to highly sensitive assays that measure neutrophil gene expression (e.g. qPCR). We investigated the contribution of contaminating leukocytes on the transcriptome profile of human neutrophils following stimulation with inflammatory cytokines (GM-CSF, TNFα), using RNA-Seq. Neutrophils were isolated using Polymorphprep or the StemCell untouched neutrophil isolation kit (negative selection of "highly pure" neutrophils). The level of contamination was assessed by morphology and flow cytometry. The major source of contamination in Polymorphprep neutrophil preparations was from eosinophils and was highly donor dependent. Contaminating cells were largely, but not completely, absent in neutrophil suspensions prepared using negative selection, but the overall yield of neutrophils was decreased by around 50%. RNA-seq analysis identified only 25 genes that were significantly differentially-expressed between Polymorphprep and negatively-selected neutrophils across all three treatment groups (untreated, GM-CSF, TNFα). The expression levels of 34 cytokines/chemokines both before and after GM-CSF or TNFα treatment were not significantly different between neutrophil isolation methods and therefore not affected by contributions from non-neutrophil cell types. This work demonstrates that low numbers (<5%) of contaminating leukocytes in neutrophil preparations contribute very little to the overall gene expression profile of cytokine-stimulated neutrophils, and that protocols for the isolation of highly pure neutrophils result in significantly lower yields of cells which may hinder investigations where large numbers of cells are required or where volumes of blood are limited.

No MeSH data available.


Venn diagram showing differentially expressed genes between neutrophil samples prepared by either Polymorphprep™ (poly) or magnetic beads (bead).Comparisons performed by Cufflinks using treatment specific paired-samples from two biological replicates. All genes displayed were significantly differentially expressed due to a higher RPKM in Polymorphprep prepared samples. Significance was calculated by Cuffdiff and adjusted for 5% false discovery rate by Benjamini-Hochberg correction for multiple-testing.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4581699&req=5

pone.0138982.g005: Venn diagram showing differentially expressed genes between neutrophil samples prepared by either Polymorphprep™ (poly) or magnetic beads (bead).Comparisons performed by Cufflinks using treatment specific paired-samples from two biological replicates. All genes displayed were significantly differentially expressed due to a higher RPKM in Polymorphprep prepared samples. Significance was calculated by Cuffdiff and adjusted for 5% false discovery rate by Benjamini-Hochberg correction for multiple-testing.

Mentions: We performed differential expression (DE) analysis on the RNA-seq data using Cufflinks[20]. The transcriptomes from Donor 1 and Donor 2 were input as n = 2 biological replicates to compare DE in Polymorphprep and negatively selected neutrophils under the three experimental conditions (untreated, GM-CSF or TNFα for 1h). This analysis identified only 25 genes whose expression was significantly DE across all samples, representing just 0.1% of annotated genes. Of these, 23 genes were significantly DE in the untreated samples, and 9 genes were significantly DE in all three treatment pairings (Fig 5 and S1 Table). Only 2 genes in the treatment group samples (1 each in the paired GM-CSF-treated and TNFα-treated incubations) were significantly DE. In GM-CSF-treated neutrophils, ITGB7 was expressed at higher levels in Polymorphprep isolated neutrophils (RPKM 1.55 compared to RPKM <0.3 with negative selection), and in TNFα-treated neutrophils, IDO1 was expressed at higher levels in Polymorphprep isolated neutrophils (RPKM = 6.64 compared to RPKM<0.3 with negative selection). The greatest discrepancy in expression levels between the two isolation methods was for CLC, which is expressed at high levels in eosinophils. Genes encoding the α- and β- subunits of haemoglobin (HBA1, HBA2 and HBB) were also detected in the Polymorphprep samples.


Whose Gene Is It Anyway? The Effect of Preparation Purity on Neutrophil Transcriptome Studies.

Thomas HB, Moots RJ, Edwards SW, Wright HL - PLoS ONE (2015)

Venn diagram showing differentially expressed genes between neutrophil samples prepared by either Polymorphprep™ (poly) or magnetic beads (bead).Comparisons performed by Cufflinks using treatment specific paired-samples from two biological replicates. All genes displayed were significantly differentially expressed due to a higher RPKM in Polymorphprep prepared samples. Significance was calculated by Cuffdiff and adjusted for 5% false discovery rate by Benjamini-Hochberg correction for multiple-testing.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4581699&req=5

pone.0138982.g005: Venn diagram showing differentially expressed genes between neutrophil samples prepared by either Polymorphprep™ (poly) or magnetic beads (bead).Comparisons performed by Cufflinks using treatment specific paired-samples from two biological replicates. All genes displayed were significantly differentially expressed due to a higher RPKM in Polymorphprep prepared samples. Significance was calculated by Cuffdiff and adjusted for 5% false discovery rate by Benjamini-Hochberg correction for multiple-testing.
Mentions: We performed differential expression (DE) analysis on the RNA-seq data using Cufflinks[20]. The transcriptomes from Donor 1 and Donor 2 were input as n = 2 biological replicates to compare DE in Polymorphprep and negatively selected neutrophils under the three experimental conditions (untreated, GM-CSF or TNFα for 1h). This analysis identified only 25 genes whose expression was significantly DE across all samples, representing just 0.1% of annotated genes. Of these, 23 genes were significantly DE in the untreated samples, and 9 genes were significantly DE in all three treatment pairings (Fig 5 and S1 Table). Only 2 genes in the treatment group samples (1 each in the paired GM-CSF-treated and TNFα-treated incubations) were significantly DE. In GM-CSF-treated neutrophils, ITGB7 was expressed at higher levels in Polymorphprep isolated neutrophils (RPKM 1.55 compared to RPKM <0.3 with negative selection), and in TNFα-treated neutrophils, IDO1 was expressed at higher levels in Polymorphprep isolated neutrophils (RPKM = 6.64 compared to RPKM<0.3 with negative selection). The greatest discrepancy in expression levels between the two isolation methods was for CLC, which is expressed at high levels in eosinophils. Genes encoding the α- and β- subunits of haemoglobin (HBA1, HBA2 and HBB) were also detected in the Polymorphprep samples.

Bottom Line: The level of contamination was assessed by morphology and flow cytometry.RNA-seq analysis identified only 25 genes that were significantly differentially-expressed between Polymorphprep and negatively-selected neutrophils across all three treatment groups (untreated, GM-CSF, TNFα).The expression levels of 34 cytokines/chemokines both before and after GM-CSF or TNFα treatment were not significantly different between neutrophil isolation methods and therefore not affected by contributions from non-neutrophil cell types.

View Article: PubMed Central - PubMed

Affiliation: Institute of Integrative Biology, University of Liverpool, Liverpool, United Kingdom.

ABSTRACT
Protocols for the isolation of neutrophils from whole blood often result in neutrophil preparations containing low numbers (~5%) of contaminating leukocytes, and it is possible that these contaminating cells contribute to highly sensitive assays that measure neutrophil gene expression (e.g. qPCR). We investigated the contribution of contaminating leukocytes on the transcriptome profile of human neutrophils following stimulation with inflammatory cytokines (GM-CSF, TNFα), using RNA-Seq. Neutrophils were isolated using Polymorphprep or the StemCell untouched neutrophil isolation kit (negative selection of "highly pure" neutrophils). The level of contamination was assessed by morphology and flow cytometry. The major source of contamination in Polymorphprep neutrophil preparations was from eosinophils and was highly donor dependent. Contaminating cells were largely, but not completely, absent in neutrophil suspensions prepared using negative selection, but the overall yield of neutrophils was decreased by around 50%. RNA-seq analysis identified only 25 genes that were significantly differentially-expressed between Polymorphprep and negatively-selected neutrophils across all three treatment groups (untreated, GM-CSF, TNFα). The expression levels of 34 cytokines/chemokines both before and after GM-CSF or TNFα treatment were not significantly different between neutrophil isolation methods and therefore not affected by contributions from non-neutrophil cell types. This work demonstrates that low numbers (<5%) of contaminating leukocytes in neutrophil preparations contribute very little to the overall gene expression profile of cytokine-stimulated neutrophils, and that protocols for the isolation of highly pure neutrophils result in significantly lower yields of cells which may hinder investigations where large numbers of cells are required or where volumes of blood are limited.

No MeSH data available.