Is it a good approach to get information about which genes are expressed? Or should I use another type of experiment design etc. loop design to get better results?
Expression microarray experiments with two color labelling, Agilent human 8x60k with Spike In control. Compare RNA from untreated cells labeled with Cy3 and RNA from treated cells labeled with Cy5 on one microarray (8 biological replication).
Which genes are differentially expressed after treating? This is main question to know pathways of cytotoxicity of two substances, single and both.
I have isolated RNA from 18 wells (6 wells per plate) – one the same passage - , each RNA from well to separate probe – this is control untreated tumor cell line, which will be label with Cy3.
RNA (will be label with Cy5) from 18 wells marked from A to R from three another plates with treated cells. One the same passage.
From first plate I isolated RNA from 6 wells treated with one substance, each RNA from well to separate probe. Each probe with RNA marked as A, B, C, D, E, F. From second plate I isolated RNA from 6 wells treated with another substance, each RNA from well to separate probe. Each probe with RNA marked as G, H, I, J, K, L. From third plate I isolated RNA from 6 wells treated with both substances, each RNA from well to separate probe. Each probe with RNA marked as M, N, O, P, Q, R.
So I have control untreated labelled with Cy5 RNA marked from 1 to 18 and three treated groups of RNA: A, B, C, D, E, F (first); G, H, I, J, K, L (second); M, N, O, P, Q, R (third).
Next I pool the RNA from three of the wells and randomize it as below to obtained 8 probe (because I have 8 microarray per slide):
see .docx file
As you have 18 samples, Agilent array (SurePrint G3 Human GE Microarray, 8x60K) array would be economic compared to Affymetrix.
I do not suggest to use the data from pooled samples.
If I understand correctly, you have 3 groups (control, treatment A and treatment B) with 6 biological replicates per group.
Then, you do not need to pool the samples as these 6 biological replicates are enough to apply statistical test like one-way anova (if all groups are independent).
So, if you DID NOT yet perform the labeling of the samples, I suggest you NOT TO use two-color labeling.
Agilent has One-Color labeling kit (One-Color Microarray-Based Gene
Expression Analysis, Low Input Quick Amp Labeling, Cat# 5190-2305).
Here you can label each sample with Cy3 which can be used for hybridization.
I used this kit with 50ng of total RNA and worked well.
For 18 samples, you just need 3 arrays.
Then extract the raw data using feature extraction software. Review the QC Report to identify any outliers due to technical variance.
Analyze the data and apply statistical test to get significantly regulated transcripts in the group of interest. For example: treatment A vs. control, treatment B vs. control or treatment A vs. treatment B.
I have enclosed the protocol to this answer.
RNA-spike in kit for one color should also be ordered separately (RNA Spike-In Kit, One-Color Agilent p/n 5188-5282).
Refer Page#12 in the manual to check the required kits/reagents to use this kit.
Spike-ins serve as positive controls and help you to monitor the process of labeling efficiency.
Good luck...
http://www.genomics.agilent.com/en/Gene-Expression-Amplification-Labeling-Reagents/Low-Input-Quick-Amp-Labeling-Kits/?cid=AG-PT-102
Thank you for answer,
I have also fourth group where I have RNA isolated from treating cells with both substances (treatment A and B). Use RNA One-Color Spike In Kit as you suggested is difficult to achieve now, especially instead Two-Color Spike In Kit. So I would need four slides instead three if I want use RNA One-Color Spike In Kit to compere control vs. treatement A, control vs. treatement B, control vs. treatement A and B, treatemnt A vs. treatement B.
- Badges
- Science topic
- Similar topics
- Psychology
- Mental Processes
- Learning