Should hematopoietic scheme be revised? Dr. Jerimiah Bell interview

by Alexey Bersenev on June 16, 2008 · 4 comments

in interviews, under discussion

Post to Twitter

To get closer to the topic of discussion I’d like to give a short introduction.

It’s known that B-, T-, and NK cells arise from common lymphoid progenitor (CLP), while the rest blood cells (monocytes/macrophages, granulocytes, mast cells, dendritic cells, megakariocytes and erythrocytes) – arise from common lymphoid progenitor (CMP). Both CMP and CLP arise from one parent hematopoietic stem cell (HSC). This scheme (classical or canonical) was well-established in 1997-2000 by Irving Weissman lab and has been commonly accepted.

canonical scheme of hematopoietic tree:
(picture credit)

Let’s get close to the definition of “lineage commitment” through an example of T-cell development and CLP. For T-cell development, it was considered that CLPs migrate from the bone marrow to the thymus and give rise to mature T-cell types. This concept was recently challenged in the review by Avinash Bhandoola who reported that:

– several of the T-potent subsets in bone marrow and circulation in blood may contribute to the intrathymic pool;
– nearly all steps of commitment toward the T cell lineage, including the loss of myeloid, erythroid, and B potential, can occur extrathymically;
– many hematopoietic progenitors are capable of T lineage development if exposed to Notch signaling within an otherwise lymphopoietic environment;
– Unlike T- and B- potent Earliest Thymic Progenitor (ETPs) many others ETPs retain NK, DC and also demonstrate some myeloid potential.

Scheme of progenitor-successor relationships in T-cell development:

These conclusions based on these latest findings indicate that a simplistic canonical hematopoietic division tree to the lymphoid and myeloid branches could be challenged. The last 2 papers, appeared in one of April’s issue of Nature, from USA and Japan show that a previously well-recognized distinction between two main hematopoietic branches — lymphoid and myeloid — needs to be reconsidered.

Today I’m interviewing Jerimah Bell – the first author of the paper: The earliest thymic progenitors for T cells possess myeloid lineage potential. Dr. Bell is the postdoctoral researcher in Avinash Bhandoola lab at University of Pennsylvania (USA).

1. You show for the first time that Earliest Thymic Progenitors (ETPs), which reside in adult mouse thymus, possesses not only lymphoid but also myeloid potential. So, what kind of myeloid cells arise from ETPs and why does thymus need them? Is this process occur in the steady-state organ or is it created under special conditions?

JB: As we show in the paper, ETPs have the ability to generate granulocytes, macrophages, and dendritic cells, in addition to T cells, in culture. In vivo, we focused on the granulocytic potential of ETPs because granulocytes can be unambiguously identified as myeloid cells. In the accompanying paper in the same issue, Wada et al showed that macrophages also develop from early T cell progenitors in vivo. In addition, Ken Shortman has previously demonstrated dendritic cell potential from these progenitors in vivo.

The granulocytes that we observed appear to be present at steady-state, as they are readily detectable, albeit at very low frequency, in unmanipulated animals. We are currently investigating the necessity, if any, of thymic granulocytes. Our current speculation is that they may have some role in tolerance to granulocytic antigens.

2. If the thymus has many progenitors, which one of them is capable of giving rise to many mature blood cell types?, Are they involved in general hematopoiesis or are they still restricted to T-cell development?

JB: There are several distinct progenitor populations present within the thymus, each increasingly more restricted towards the T cell lineage as they mature until they reach a fully T cell-committed stage. The earliest defined progenitors, ETPs, have T cell, dendritic cell, NK cell, and myeloid cell potentials, but limited B cell potential. As alluded to above, we are still unsure whether there is a reason for T cell progenitors to generate these alternative lineages.

3. Can you speculate that “lympoid-restricted” or CLP cells have myeloid potential not only in the thymus but also in bone marrow?

JB: Others have suggested that this may be the case, though I think further investigation and more complete analysis is required.

4. Can you explain in simple terms the advantages of methods (such as single cell clonal assay and lineage-tracing in vivo system) used in your experiments compared to others?

JB: In particular, the single cell assays that we performed allowed us to answer questions that could not be previously. As we cited, others had suggested that early thymic progenitors as a population had myeloid potential, but it was argued that within this population there may be myeloid-biased or myeloid-committed precursors which were responsible for those observations. By examining the lineage potentials of single cells we were able to show that the majority of single ETPs possessed both myeloid and T cell potentials.

From the lineage-tracing experiments we were able to show that a cell that appeared to be programmed for a particular lineage had actually reverted towards an alternative lineage in vivo.

5. We know that many progenitors from bone marrow circulate in the blood, but only some of them enter the thymus and make mature T-cells. Do we know something about regulation of this process? Any signal or existence of a specific thymic niche?

JB: Both the egress from the bone marrow and the entry into the thymus appear to be regulated, though both processes are incompletely understood. It is thought that downregulation of retention signals may be responsible for exit from the bone marrow, but how particular cells are selected to leave, I’m not sure we know. Once in the blood, we have a pretty good idea that CCR9 is involved in recruitment of progenitors to the thymus and that PSGL1 is somehow involved in getting those cells into the thymus. Again, how entry exactly is regulated, we don’t fully understand.

6. Another paper, which was published in the same issue of Nature is a great complimentary work to your study and basically “argues against the classic model of haematopoiesis in which the CLP is located at the branch point towards T and B-cell lineages, and strongly suggest that the myeloid-based model is applicable to both fetal and adult haematopoiesis”. Would you agree that we should exclude CLP from the scheme at all and redraw it in terms of common myelo-erythroid progenitors and myelo-lymphoid progenitors?

(credit: Wada et al )

JB: I suppose the answer to this question would depend upon the answer to your previous question – whether or not CLPs possess myeloid potential. If so-called CLPs have myeloid potential then they are actually myelo-lymphoid progenitors rather than lymphoid-restricted progenitors. If instead they lack myeloid potential, then they might still be important progenitors in the bone marrow for B cell production. Further, if found to lack myeloid potential, CLPs might prove to be important T cell progenitors, though it may be that they enter into the T cell pathway independent of ETPs

7. Why is it important to study a relationship among different progenitors within the blood-forming system? To me it seems that scheme is pretty flexible and can be revised from year to the year and from lab to the lab. Don’t you think that the vast majority of hematopoietic progenitors are very plastic and it’s very hard to define commitment? In other way Thomas Graf asked in his commentary whether earliest lineage decisions occur in a sequential fashion or at random?

JB: I think it’s important to understand the lineage relationships among different progenitors in order to ask the questions concerning what drives precursors towards a particular lineage. I think we are starting to understand that a given lineage needs to be defined not only by what it is, but also what it is not. In other words, there is a complex transcriptional program that progenitors must go through in order to reach final commitment. For a T cell, we want to understand molecularly what makes it a T cell, but also what prevents the majority of precursors from becoming a myeloid cell. Only after we understand and define the true lineage relationships of the progenitors can we begin to ask such questions. Others have shown that Notch can limit myeloid potential of bone marrow progenitors, but it has been unclear why that might happen. Now, since we understand that ETPs have myeloid potential, which is especially easily revealed in the absence of Notch signals, we may understand why Notch has the ability and the need to limit that myeloid potential.

I don’t think it’s necessarily that difficult to define lineage commitment. Most in the T cell development field would define the DN3 population as T cell committed. It is true that when transduced with transcription factors that the DN3 cells do not normally express that they can be diverted towards the myeloid lineage, as Thomas Graf has shown. I agree that this may reflect “plasticity” but I do not agree that this is the same as lineage potential. I don’t believe that DN3 cells in the thymus are ever likely to develop into myeloid cells, B cells, NK cells, ect, and so are regarded as lineage committed. We need to be very careful in the way that we use terms like “plasticity” and “lineage potential” and make certain it is clear what we are describing.

8. I’m just curious how long it took for you from an idea and initial hypothesis to a paper in the Nature journal?

JB: I worked on the project for about 1.5 years from beginning to publication. I have to be certain to qualify this, however. We must remember that this is an old question, as we pointed out in the paper. We are certainly not the first to ask if early T cell progenitors have myeloid potential. There were two keys to the success of the project. First, in our in vitro cultures, we were able to look early before the myeloid cell disappeared, which to my knowledge had not been done. Second, we managed to use a lot of wonderful tools like the OP9-DL culture system and the H2S-VEX and lysozyme GFP mice that were recently developed. So we took new tools and applied them to an old question.

Dr. Bell I’d like to thank you for interview.

Interviewer: Alexey Bersenev

{ 4 comments… read them below or add one }

oracy June 18, 2008 at 1:36 am

Very informative. Well done to both of you.


JWS June 19, 2008 at 3:28 pm

Cool interview: old question + new techniques..


Alex June 20, 2008 at 11:57 pm

I’d agree with Dr. Bell that you can not revised current scheme without same kind of experiments with bone marrow. I just wondering why it is not been done yet.


Lei August 5, 2008 at 8:48 pm

Cell reprogramming is always fascinating. Goodell’s Cell Stem Cell paper in 2007 showed distinct gene expression profile between HSC and different lineages, so-called lineage fingerprint. This group further demonstrated that ZFP105 and ETS2 can biase HSC to NK and MNC after BMT, respectively. While inducible stem cell research found OCT4 promoter cannot be demethylated at DNA level after induction thus hampers it to achieve its full expression since OCT4 is not supposed to reactivated after differentiation. It seems we’ll screw ourselves up before we can screw the cells up.


Leave a Comment

Previous post:

Next post: