Not that simple. Dmitry recently joined Wistar and you might discuss your question with him. Perhaps your best approach would be to use culture supernatant(s) from a tumor cell line.

the induced suppression in MDSC depends upon the tumor mileau which contains many cytokines which are required for their accumulation and induction of suppression like IL6, PGE2, VEGF,GMCSF,etc. To differentiate bone marrow into MDSC combination of cytokines at different concentration (optimum) will work. Use of culture supernatant from tumor cell is a bit risky you do not know which cytokines are secreted by the cells.

Interesting question for which I suspect you wont get a straight answer. There are several of them out there. For example, Highhill et al (Blood 2010) used BM cells culturing with GM-CSF and IL-13. In there case suppression was mediated by arginase (likely due to IL-13). Zhu et al (Khoury's group) are using CFA-induced Gr-1+ splenocytes cultured for a few hours with LPS/IFNg (i believe??). You can look up a couple of their papers in JI. Lutz's group reported that LPS/IFNg inhibits DC differentiation. Marigo et al (Bronte's group) published a method using BM cells cultured with GM-CSF+IL-6 (Immunity 2010). Inflammation-induced MDSC have not been well characterized and studied, so I suspect that without tumor factors (or without culturing cells in media from tumor cells) these in vitro produced MDSCs are really not the same ones that you will see in tumor-bearing mice.

Dmitry Ostanin provided a cogent comment that addresses your question in part. As you know many growth factors can expand MDSCs. They include but are not limited to G-CSF, M-CSF, GM-CSF, Flt3L, SCF, VEGF etc. The utility of these growth factors in vitro is primarily associated with hematopoiesis i.e. M-CSF induces monocytes/macrophages; G-CSF, granulocytes; Flt3L DCs; etc. These processes can be further manipulated by the addition of additional factors such as the co-culture of GM-CSF and IL-4 or IL-13 that results in DCs rather than macrophages . There are other cytokines/TFs associated with MDSC expansion that includes a mechanism that prevents their terminal differentiation/maturation. These factors are obscure/unknown and a focus of research for several laboratories. Thus, MDSCs are definitively described as cells with suppressive activity, are hypodense, possess a limited ability to mature and with a plastic phenotype. Further, MDSCs appear to be activated via LTs, PGs, LPS, IFN-g etc providing additional phenotypic diversity. The reason to suggest the use of tumor culture supernates is that we may not know the cytokines, growth factors etc that expand/activate MDSCs. Look forward to other comments on this interesting topic.

James Talmadge is absolutely right about the obscurity of factors that promote MDSC expansion and accumulation. Several papers showed that tumor cells produce some of the factors that are thought to contribute to MDSC expansion and activation (one of Dmitry Gabrilovich's recent reviews addresses it nicely). I am curious if the nature of these tumor-derived factors is being studied? Obviously culturing of BM cells in the presence of tumor-derived medium (and obviously hematopoietic factors to keep these cells alive and proliferating) will not mimic the complexity of the in vivo system, but may be a good start. What do others think?

Thanks a lot to everybody, I found all your comments very helpful. I just started to look at MDSCs a couple of weeks ago

To James: I read with interests your article with Dr. Gabrilovich in Nat. Rev Cancer, and of course I asked him some suggestions. One of his staff scientist is helping me with these experiments and he is introducing me in the MDSCs world. Studying the literature I found this article from Bronte (Immunity, 2010, the one Dmitry was suggesting) in which they use GM-CSF and IL-6.

After a couple of experiments to improve my skills with BM flushings and RBC lysis, I was able to obtain a cell population positive for GR1 and CD11b. I will do some biochemical studies on them, but as everyone of you and Dr. Gabrilovich's scientist told me, probably these cells do not have any suppressive activity, so maybe using tumor supernatant would be better, even if the milieu cannot be precisely defined.

In conclusion, what I can do is to try different protocol and see how they behave.

Thanks a lot again for your help, I will keep the post updated as soon as I will have results.

Bye,

Michele

actually, we did compare fresh BM cells and cultured and even fresh BM cells have low level of suppressive activity but culturing enhanced it drastically. We were testing WT CD4 T cells stimulated with antiCD3/CD28, so not an Ag-specific method but works for us.

Try activating them. We use classic macrophage activating agents and the suppressive activity dramatically increases.

I followed the protocol described in Marigo's Immunity paper (Bronte's group) by culturing 2.5x105 BM cells in 10 ml complete RPMI with GM-CSF (40ng/ml) and IL-6 (40ng/ml) for 4-5 days. The cells are very suppressive, however they do not solely express Gr-1. In those cultures, more than 90% are CD11b, with Ly6ChiLy6G- being around 40% and Ly6CintLy6G+ being around 20%. What is confusing to me however, is the relevance of using such cells in different experiments, since different reviewers will hold different opinions of whether it is OK to use in vitro cultured cells or not. All the best.

Dose the MDSC differentiate from bone marrow cells, which only stimulated with high-dose GM-CSF about 5days?

You can differentiate different flavors of MDSCs using a wide number of growth factors.  I am primarily discussing in vivo differentiation but should work in vitro but perhaps with differing activity/outcomes.  Cytokines that can do this, but are not limited to these include: G-CSF, GM-CSF, M-CSF, SCF, Flt3L......... 

Hi all, i am too interested in MDSCs and i wanted to look MDSCs in peripheral blood. Can anyone suggest how to isolate MDSC from blood

Hi Michele,

Did you figure out if MDSC would attach to the plate, or do are they in suspension when generated from Bone Marrow? I collected all cells (floating and attached) and the percentage of CD11b+ Gr-1+ cells was about 40% (it is reported to be 60% in literature).

Thanks!

Mehdi

MDSCs don't stick to plastic, unless you have differentiated them to macrophages. What did you culture them in and for how long? Please remember that MDSC are hypodense and you can use this characteristic to enrich them.

Thank you for your response James. I cultured bone marrow cells in RPMI supplemented with 10mM HEPES, 20uM b-2mercaptoethanol, 20ng/mL GM-CSF and 50% (V/V) cancer cell CM. The entire purpose of the experiement was to test differentiating agents that would potentially favor APC rather than MDSC differentiation. The reason I collected adherent as well as suspension cells is to check how the macrophage, DC and MDSC populations would change across treatment groups. I also did qPCR for arg-1, Cox2 and NOS2, and the group that had more macrophages and less MDSCs had higher levels of arg-1. Is it possible that these macrophages become M2-like, and that's where arg-1 is coming from?

Thanks!

Mehdi

The conditioned media is a black box of cytokines. The Arg-1, as you suggest could be M2 Macs and or mMDSCs or more likely both. To push towards MDSCs I think the use of GM-CSF and IL6 would be better. If you want Macs M-CSF would be the cytokine of choice. If you want DCs FLT3-L would be the growth factor of choice.

One thought that came to mind would be to collect non-adherent and adherent cells and compare the flow profile. One other point that might help is that MDSCs are hypodense. So if you do a F/H gradient you will find that the G-MDSCs are in with the lymphocytes. I would also do a cytospin as well as flow cytometry. You might find those results interesting.

With BM if you want DCs as your APCs you would culture in GM-CSF and IL4. The IL4, or you can use IL-13 prevents differentiation to macrophages.

Hope these comments help.

James E Talmadge Could you recommend a protocol to induce MDSC from human PBMC? Does human MDSC stick to plastic?Thanks!

Lots of different questions.

Could you recommend a protocol to induce MDSC from human PBMC?

From a normal, healthy individual no. You need hematopoietic progenitor cells. Thus, BM, mobilized PB stem cells or blood from a patient with advanced cancer, preferably with multiple cycles of chemotherapy. Note, mobilized PB and BM from normal individuals will have MDSCs, as MDSCs and hematopoietic progenitor cells are largely the same thing. You can induce MDSC proliferation with CSF-G or CSF-GM and IL6 and Flt3L seems to work as well. Activation to prolong survival and bioactivity helps. Standard myeloid activating ligands would do this.

As an aside, G-MDSCs, which have a cytologic appearance of PMNs are hypodense and so travel with lymphs following ficoll hypaque separation.

Does human MDSC stick to plastic? Interesting question. G-MDSCs no. M-MDSCs depending on maturation may stick to glass. iMDSCs will not stick.

Hope these superficial responses help.

Thanks!

James E Talmadge Thank you! Very useful recommendation.