MSCs and endothelial cells are overrepresented in BM organoids

[Ningqing Liu]

Hi Abdullah,

We followed you cancer discovery paper to grow a batch of BMO which seems morphologically close to what was shown in your papers. We did also observe high amount of CD34+ and around 5% CD11b cells by FACS. However, when we did RNA-seq on these organoids, it really seems there are only 10% of myeloid lineages compared to your papers. Moreover, I think we have almost no erthyroid cells (CD235a) is almost not expressed. Instead, I think the percentage of MSCs and endothelial cells are very high in our BM ogranoids.

I wonder whether you have experienced the same thing before, and whether you have any suggestions about the potential solution?

Thank you in advance!

Best,

Ningqing

[Abdullah Khan]

Hi Ningqing, 

I'm afraid this isn't a problem I've come across (If anyone else in the TooT has please chip in!) 

I need a bit more information to be able to help I think but here is a first step: 

1. What hiPSC line are you using? Is it one which you have validated for 2d Diffs before or fairly new? 

2. Morphology

3. Is your high amount of CD34+ cells also CD45+ ? If it is a large CD34+ CD45+ fraction then you probably have a lot of HSPCs and are sampling a bit early (what is your time point)? 

4. It is very odd to not have any Ery if you have EPO in there. Might be a cytokine issue? 

If you could share any FACS plots or single cell data and some of your methodology we can try and help. 

Abs

[Ningqing Liu]

Hi Abdullah,

Thank you very much for the quick reply and helping us out for the troubleshooting.

1. We are using a hiPSC which has been validated using 2D diff protocol for myeloid lineages. When we started the organoid protocol, we did pick up the undifferentiated colonies based on the morphology;

2. We use the same antibodies from your Cancer discovery paper. However, we did not get everything in house before we have to stain our organoids last time (so we don't have CD45 or Lin in the last ). We only have CD34, CD11b and CD235 at that moment. I attach the FACS plots here. Although it seems that we have a lot of CD235+ cells, the signal may be the spillover from other channels since the compensation was not optimal;

3. When we will ran bulk RNA-seq on the same organoid batch,  we did not see any expression of GYPA (CD235). The expression of all the myeloid genes, such as ITGAM (CD11b), S100A8/9/12, are at least 5-10 fold lower than the pseudo-bulk of your scRNA data. With the help of one bioinformatician in our department, we also deconvolute the bulk RNA-seq signal using your scRNA-seq data as a reference. It was clear over 90% cells assigned to MSCs and endothelial cells;

4. EPO may be a problem, although it was a fresh vial from StemCell Technologies. For the next time, we will use EPO from another supplier.

It will be great if you can share some of your thoughts. Thank you in advance!

[Abdullah Khan]

Thanks for sharing Ningqing - I have a busy week in lab but will go through your FACS data. My first thought though which I will share quickly is re this: 

1. We are using a hiPSC which has been validated using 2D diff protocol for myeloid lineages. When we started the organoid protocol, we did pick up the undifferentiated colonies based on the morphology;

Do you have differentiation in the background of your hiPSC cultures? We tend to only use cultures which have no differentiation (all colonies are undifferentiated). Can you share some images of your 'base' culture? 

Also I would suggest setting up the myeloid diff your line is validated for in parallel to an organoid differentiation that should sene check that your line can still make haem/myeloid cells. 

I'll try and get back to you by the end of the week re your FACS data! 

BW
Abs