Jump to content
Powered by

Christian Buske and his research into leukaemia stem cells

A clinical accolade for a model: the gene signature of leukaemia stem cells allows predictions to be made on the course of disease in adults suffering from acute myeloid leukaemia (AML). A study carried out by an international team of researchers shows that AML contains cells with stem cell properties (leukaemia stem cells, LSC). The study also found that patients who strongly expressed the LSC signature had a much shorter survival time than patients with a low LSC expression.

The findings of an international team of researchers provide evidence for the clinical relevance of the leukaemia stem cell (LSC) concept. “We found that LSCs are not only significant in experimental models. We also found that the LSC gene expression profile can be used to accurately predict patient survival,” said Christian Buske commenting on the importance of the paper that has recently been published in the renowned journal “Nature Medicine”.

Ulm researchers focused on the clinical part of the study

Prof. Christian Buske, Director of the Institute of Experimental Tumour Research at the University of Ulm. © University of Ulm

Christian Buske, Director of the Institute of Experimental Tumour Research, co-authored the study which involved biologists and clinicians from other parts of the world. Buske, a specialist in AML pathogenesis, was really the person who made the publication possible, as it was he who linked the LSC gene signature with patient samples by making the connection between the LSC signature and testing whether it correlated with patients’ response to therapy.

Buske, 46, believes that it is too early to say whether the study will pave the way to cancer stem cell-based therapies. However, he is hoping that the international team’s finding will provide clinicians with a tool to identify high-risk ALM patients and put in place more aggressive, though more promising therapies.
Buske is convinced that the study was able to show that the cancer stem cell model is more than just an artefact. The model was established around 20 years ago and it proposes that acute leukaemias, and AML in particular, are “hierarchically organised”. Animal experiments involving immunocompromised mice have shown that not all injected human AML cancer cells are able to induce cancer. When the AML sample was divided into different fractions, the researchers found “harmless” leukaemia as well as cancer-inducing cells (i.e. cancer stem cells).

More than an artefact of xenograft models

Buske is well aware that there are some good reasons to reject the xenograft mouse model, according to which some solid tumours and leukaemias are organised as cellular hierarchies sustained by cancer stem cells (CSCs). He himself believes that the model is “suboptimal”. On the other hand, he knows that mice only select the fittest cells and that these cells have the potential to induce tumour development. However, he is hesitant to completely refute the objections to the hierarchy theory, even though he has shown that mouse leukaemias are organised as cellular hierarchies. He believes that the “Nature” paper provides evidence to support the hierarchical organisation of AML according to the CSC model and that LSCs are clinically relevant and not just artefacts of experimental xenograft models.

If LSCs do indeed induce leukaemia, then it would be safe to assume that the genes expressed by these cells are indicative of a bad prognosis. Buske proposes that AML patient samples that express many LSC genes are indicative of a particularly bad outcome. The international group of researchers have been able to confirm this hypothesis: they investigated the specific properties of CSCs in leukaemia by defining the gene expression signatures of LSCs and haematopoietic stem cells (HSCs) from functionally validated, sorted fractions of 16 human AML samples, and found that the LSC gene expression profile could be used as prognostic marker for the survival of AML patients.

Gene signature provides highly significant results

The study found that around 50% of all AML patients with the LSC-specific gene signature died within a year of diagnosis whereas patients without the LSC-specific gene signature had a median survival time of 1500 days. “That’s a big difference,” said Buske explaining that the difference makes the results both representative and reliable because - although only 16 primary human AML samples were used – the genetic signature was tested in hundreds of AML patients.

Leukaemia stem cells (LSC) cannot be defined phenotypically, but only functionally. Buske’s Canadian partners provided evidence using an extremely complex approach in mice that the stem cells under investigation were actually LSCs. Two researchers (Becker, Jordan), who were not part of the study, confirmed the study’s major scientific value, by confirming that only functionally validated LSCs were used.

Difficult search for differences

Haematopoietic stem cells are very similar to leukemia stem cells. © University Hospital Ulm

Buske is sure that it will be difficult, or even impossible, to identify a cancer stem cell marker because cancer stem cells originate from normal stem cells with which they share many properties. Nevertheless, many research groups around the world, including Buske’s, are working on the identification of differences between leukaemia stem cells and normal haematopoietic stem cells (which must not be destroyed by anti-cancer therapies), and this has proved to be a rather difficult task. No cancer stem cell-specific therapy is currently available. Although there are approaches that use interferon alpha to block cancer stem cells involved in myeloid leukaemia by trying to lure them out of their niche and reactivate their cell cycle in order to then destroy them with chemotherapeutics, Buske believes that these approaches are not cancer stem cell-specific.

Another aspect that makes the translation of scientific results into clinical application even more difficult is that leukaemia stem cells and normal haematopoietic stem cells (HSCs) are fairly similar in many aspects including their gene expression profile. It has been found that LSCs have the core set of biological functions common to all stem cells, including haematopoietic stem cells.

Researchers find it very difficult to identify key differences between leukaemia stem cells and healthy haematopoietic stem cells. Buske calls the process in which LSCs use and copy the core properties of healthy haematopoietic stem cells for producing harmful cell material “hijacking”.

Healthy haematopoietic stem cells are also of clinical relevance

The Nature paper has also shown that the gene expression profile of normal haematopoietic stem cells is of prognostic value. Buske regards this as evidence for the similarity between cancer stem cell and normal stem cell signatures. Some genetic differences between cancer stem cells and healthy haematopoietic stem cells are known, but future research will need to show whether these differences can be used as pharmaceutical targets.

Detailed insights into the biology of cancer stem cells

The retrospective study that was the basis of the Nature paper will now be followed by a prospective study in which tumour tissue will be removed from AML patients before they undergo chemotherapy. The gene signature of AML tissue will then be compared with that of leukaemia stem cells and its suitability for predicting the clinical outcome of AML patients will be assessed. From a biological perspective, future research will need to focus on the characterisation of LSCs, as information is currently only available on their gene expression pattern. The differently expressed LSC-specific and HSC-specific genes could then be selected and investigated for their role in LSC biology. Buske explains that between 50 and 60 candidate genes have been identified and are now being tested to see whether they are “druggable”, i.e. can be used as therapeutic targets. Whether these experiments will at some stage produce positive results, and make it possible to develop cancer stem cell-associated or even cancer stem cell-specific therapies is as yet unknown.

As both LSCs and HSCs have canonical stem cell functions, Buske and his research partners have generated HSC gene expression profiles to find out whether human LSCs share molecular mechanisms and gene expression programmes with HSCs. The search for common properties (rather than differences) is due to the frustrating finding that leukaemia stem cells are genetically highly complex, as Buske’s colleagues have recently found out. To put it another way: “Every tumour is different.”

Buske’s team of 25 is now looking for answers to fundamental biological questions such as: how plastic are leukaemia cells? What are the molecular properties of leukaemia stem cells in comparison to leukaemia cells that are unable to induce AML? One project that is currently closer to clinical application is Buske’s attempt to indirectly block leukaemia-inducing oncogenes (by way of the signal transduction pathways involved). He has already been able to switch off these oncogenic transcription factors in animal models and is now working with a pharmaceutical company to develop antibody therapies for AML.


Eppert, K.; Takenaka, K. et al.: Stem cell gene expression programs influence clinical outcome in human leukemia, in: Nature Medicine, online 28.08.20011, doi: 10.1038/nm.2415

Becker, M.; Jordan, Craig: Leukemia Stemness Signatures Step forward to the Clinic, in: Cell Stem Cell 9, 2. Sept. 2011, S. 185f., doi: 10.1016/j.stem.2011.08.006.

Gupta, P.; Fillmore, C. et. al.: Stochastic State of Transitions Give Rise to Phenotypic Equilibrium in Population of Cancer Cell, Cell, 146, 4, 19.08.2011, doi: 10.1016/j.cell.2011.07.026.

Website address: https://www.gesundheitsindustrie-bw.de/en/article/news/christian-buske-and-his-research-into-leukaemia-stem-cells