| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
1 Children’s Hospital, University Hospital, Carl Gustav Carus, Technical University Dresden, Dresden, Germany
2 Institute of Transfusion Medicine, and Immunology, German Red Cross Blood Service, University of Heidelberg, Faculty of Clinical Medicine, Mannheim, Germany
aAddress correspondence to this author at: Institute of Transfusion Medicine and Immunology, German Red Cross Blood Service, University of Heidelberg, Faculty of Clinical Medicine, Mannheim, Germany. Fax 49-621-3706851; e-mail p.bugert{at}blutspende.de.
To the Editor:
Growing interest in isolating residual amounts of RNA from blood platelets (PLTs) to perform multiple genetic studies, such as RNA profiling (1)(2) and real-time PCR (3), calls for a protocol that can be applied reliably to blood samples with limited PLT numbers. Problems with limited blood volumes occur, in particular, in pediatric patients for whom blood samples often cannot exceed 3–5 mL. A high yield of purified platelets is also mandatory in thrombocytopenic patients. We and our coworkers recently published a validation of PLT RNA amplification based on the SMART (switch mechanism at the 5' end of RNA templates) technology (4). We described the processing of 40 mL of citrated whole-blood samples, with a final mean (SD) loss of 72 (3)% of PLTs after leukocyte depletion from the platelet-rich plasma (PRP). To make PLT RNA profiling feasible for studies on pediatric and thrombocytopenic patients, we optimized PRP preparation and evaluated the quality of isolated RNA by microarray hybridization analysis.
The optimized protocol for PRP preparation has the following steps: (a) determination of PLT and leukocyte counts on an automated hematology analyzer (CellDyn 3700; Abbott); (b) centrifugation of the blood sample for 20 min at 100g; (c) filtration of the upper nine-tenths of the PRP through Purecell PL leukocyte removal filters (Pall); (d) washing of the filter with 3 mL of Tyrode’s HEPES buffer and gentle blow out of residual liquid; (e) determination of PLT counts on an automated hematology analyzer and of leukocyte counts by manual cell counting on a light microscope.
To evaluate the optimized procedure, we obtained citrated blood samples of low volume (
3 mL) from 12 healthy volunteers with initial mean (SD) total PLT numbers (per 3 mL) of 7.6 (1.3) x 108 (range, 5.8 x 108–10.7 x 108) and leukocyte numbers of 1.4 (0.2) x 107 (1.0 x 107–1.8 x 107). After PRP preparation the final PLT number was 2.8 (0.5) x 108 (2.3–3.7 x 108) corresponding to a PLT recovery of 36.8 (3.9)%. In 9 of the 12 samples, residual leukocytes counts were 215–490, corresponding to a ratio of 1 leukocyte in 1.0 x 106 PLTs.
Thrombocytopenic blood samples were produced from 8 healthy volunteers by centrifugation of a 10-mL citrated blood sample for 20 min at 150g, transfer of 75% of the PRP and centrifugation for 10 min at 2000g to obtain platelet-poor plasma, and pooling of the platelet-poor plasma with the corresponding blood sample. Mean (SD) PLT counts in these samples were 90.8 (8.1) x 103/µL, which gave total PLT numbers of 8.3 (0.6) x 108 (7.2 x 108–9.0 x 108) and leukocyte numbers of 5.0 (0.7) x 107 (4.1 x 107–5.9 x 107). Using the optimized PRP preparation, we could achieve a PLT recovery of 38.6 (2.9)% with final PLT numbers of 3.2 (0.4) x 108. In 6 of the 8 samples, leukocyte counting revealed 360–816 leukocytes, corresponding to a ratio of 1 leukocyte in 0.8 x 106 PLTs. Both PLT recovery and residual leukocyte contamination were significantly higher (P <0.05) in thrombocytopenic blood samples than in low-volume blood samples with normal PLT counts. Because PLTs are believed to contain 100 000 times less RNA than leukocytes, the given leukocyte:PLT ratios have to be recalculated for the RNA concentration. Taking this into account, the mean (SD) PLT RNA purity was 93.1 (5.6)% (range, 84.3%–100%) for the low-volume blood samples and 88.2 (8.5)% (76.6%–100%) for the thrombocytopenic samples.
To further evaluate the procedure, we processed 3 samples each of the low-volume and the thrombocytopenic samples for RNA isolation, SMART-based RNA amplification, and microarray analysis of 9850 genes as described previously (4). Both types of samples showed RNA profiles comparable to those of previous studies and to each other, with correlation coefficients 
0.9. Comparison of PLT RNA profiles from a thrombocytopenic sample with no leukocyte contamination and a sample with 200 leukocytes in 108 PLTs revealed a correlation of 0.896 (Fig. 1
). Thus, low amounts of leukocyte RNA contamination, as achieved by the given protocol, do not interfere with the microarray-based PLT RNA profile. According to the described protocol, the investigation of PLT RNA can be performed on patient blood samples with initial total PLT numbers of 8 x 108, which is reasonable for clinical studies on pediatric and thrombocytopenic patients. Even lower PLT numbers would be possible because we used only one-half of the extracted RNA for microarray analysis in this study to keep backup material.
|
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
