SLIDES & TRANSCRIPTS
October 30-31, 2000

Biology of Hematopoiesis and Apoptosis

Alan List, MD
University of Arizona

but what I will try to do is offer an overview of our current understanding. In the last ten years we have learned a lot about the biology of this disease. Much of this research derives from serminal observations of an impairment in progenitor growth which affects not only the committed progenitors, but also the long-term initiating cells. There is limited progenitor maturation capacity which is influenced in part by the hostile environment in which these progenitors reside. There is deficient stromal support as evidenced by the inability of myelodysplastic stroma to support the growth of normal hematopoietic progenitors and excessive production of inflammatory or pro-apoptotic cytokines.

TOP

A major focus of investigation has been to identify what principal abnormality(s) contributes a causality to the ineffective hematopoiesis in this disease. Certainly apoptosis, as a manifestation of ineffective blood cell production, has emerged as a potentially reliable index of abortive hematopoiesis in MDS in the last 5 years. Dr. Raz's group and many others, have shown that no matter what the assay employed, i.e. measurement of nucleosome generation or phosphatidylserine exposure, the magnitude of medullary apoptosis inversely correlates with leukemia burden.

TOP

What evidence supports a caused relation between medullary apoptosis index and ineffective hematopoiesis? The latter has proven difficult, however, in addition to the relationship to blast percentage as mentioned, an inverse relationship to erythroid burst recovery and the percent of white blood cell count offers correlative supports. Peter Greenberg's laboratory has shown that c-myc: BCL2 oncoprotein ratio in the progenitor and non-progenitor compartment correlates directly with apoptotic index. Clonogenic studies indicate that impaired erythroid progenitor growth correlates with caspase-3 activation within the erythroid progenitor compartment. The most convincing evidence derives from clinical investigations demonstrating a reduction in apoptotic index in those patients who respond to treatment with erythropoietin and GCSF.

TOP

What abnormalities which govern cell survival are responsible for the abortive progenitor growth? Much of the data are simply observational and how they relate to the apoptosis initiation is not clear. The most promising leads have been reported in the last few years. Our Japanese colleagues have shown that a truncated EPO receptor, which is unable to transmit the stimulatory signal, is detected in the majority of patients with MDS. EPO receptor legation is associated with decreased DNA binding of STAT5 and decreased GATA-1 activation in erythroid progenitors. However, the latter does not correlate with impaired in vitro erythroid progenitor growth. Sustained expression of the full length CD34 is demonstrated in bone marrow mononuclear cells, which in transfection studies suppresses myeloid differentiation. The high apoptotic index in low grade MDS occurs in a setting of high S-phase fraction, which Dr. Raza's group has termed, signal antonomy. This observation is important, and offers insight into the disease pathobiology. If we view the apoptotic response simply as a growth factor signal withdrawal, we should expect accompanying growth arrest. However, in MDS we see a high S-phase fraction that parallels apoptotic index, something that our solid tumor colleagues recognize as anoikys. Dislocation of adherent cells from counter-receptor adhesion molecules triggers anoikys, a model of programmed death that merit investigation in MDS. Excessive telomere shortening is also demonstratable that which correlates with anemia severity and abnormal karyotype. Although these findings offer a biologic profile of some intrinsic abnormalities in MDS, the relation to apoptosis initiation is not proven.

TOP

The most convincing evidence to date _______ evidence for activation of the fas ligand system. This is particularly true for the anemia and ineffective erythropoiesis in MDS. Erythroid bursts from a normal individual display low density expression of the fas receptor (CD95) and fas ligand, under dominant negative regulation by Epo. In this manner, stimulation with Epo triggers down regulation of fas receptor, creating resistance to fas ligand induced apoptosis to promote effective hematopoiesis. What we see in MDS is very different.

TOP

. There is high density display of both fas receptor and fas ligand that is resistnat Epo modulation permitting _______ or paracrine activation of the apoptotic pathway. Increased expression of fas and fas ligand is seen in all hematopoietic precursors but it is highest in the erythroid progenitors. Decreased expression of the fas associated phosphatase 1, a negative regulator of the fas death signal, is demonstrable, suggesting impaired suppression of fas signaling. Increased fas display directly correlates impaired growth of CFUE with decrease in blast percentage. Also, apoptotic bone marrow cells display highest fas density, supporting a pathogenic role in apoptosis. Similarly, c-myc:bcl-2 ratio directly correlates with increased fas and fas ligand display. The fas receptor density and the percentage of fas positive cells correlate with anemia and red cell transfusion requirements. Indeed, in vitro neutralization of fas ligand using soluble fas receptor, promotes the outgrowth of committed progenitors. Such data provides convincing evidence for activation of the fas(F)-fas ligand (FL) system, particularly in relation to ineffective erythropoiesis.

TOP

Why F/FL are aberrantly expressed isn't clear, but may derive in part from soluble signals within the environment in which these cells grow, in particular, the increased elaboration of pro-apoptotic cytokines. TNF-alpha, of IL1, and interferon gamma, to name a few, are demonstrated in excess either in the bone marrow plasma directly, in bone marrow trephine biopsies or the peripheral blood serum. Other strand defects may shorten progenitor survival. Increased fas ligand display is evident on macrophages in the bone marrow stroma, associated with accelerated apoptotic death of endothelial cells, and excess MMP generation. Neutralization of MMPs improves progenitor growth in vitro and decreases the apoptotic fraction. John Barrett and Neal Young's investigations indicate that in some patients there hematopoietic inhibitory lymphocytes contribute to in affect of hematopoiesis, adding to a growing constellation of abnormalities within the bone marrow stroma that contributes to a hostile environment.

TOP

In normal erythropoiesis these inflammatory cytokines, particularly TNF in concert with interferon gamma, trigger up-regulation of fas and fas ligand. Excess generation of these cytokines have been casually linked to the same changes demonstrate in MDS.

TOP

The most convincing evidence derives from in vitro studies demonstrating improvement in erythroid progenitor growth with neutralization of TNF. Data from David Bowen shows direct correlation between plasma concentration of TNF and oxidized DNA pyrimidines in the CD34 compartment which inversely correlates with glutathione defense. The reverse is true in normal cells suggesting a pathogenic role for TNF in pyrimidine oxidation.

TOP

Clues as to what for the pro-apoptotic cytokine excess arises as a primary or secondary stimulas have emerged from investiation of bone marrow angiogenesis in _______. Some of the first observations were reported by Purnari and his colleagues showing that micro vessel density MVD is increased in MDS and increases with blast percentage. This is his data showing an increase in the number of hot spots with marrow blast percentage, but in all FAB types of MDS, MVD is elevated.

TOP

Accumulating evidence suggests that angiogenic molecules in MDS contribute to disease pathobiology as well as angiogenic response. We now know that the myeloblasts in MDS, as well as AML, produce and elaborate VEGF and express VEGF receptors, particularly VEGFR-1 or Flt-1. There is also evidence to show that MMPs, as well as the TIMPS, are elaborated by MDS myeloblasts. Very interesting data from M.D. Anderson indicates that the blast VEGF content also has prognostic import in AML as it relates to induction outcome, and in MDS for overall survival.

TOP

A number of angiogenic molecules are recognized, many of which have been evaluated in MDS. A few of those are listed here.If we just look at the peripheral blood compartment, there is an elevation in plasma VEGF. There are five isoforms of VEGF and only two of those are detected by ELISA assays currently available. These include 2 of the three secretory forms that range from 121 to 165 amino acids. There is also an increase in basic fibroblast growth factor, interleukin-8, as well as hepatocyte growth factor in the bone marrow plasma. From own work at Arizona, bone marrow plasma VEGF content is very low. It is below normal levels, which at odds with the excessive amount of VEGF production by the immature myeloid cells. My suspicion is that it relates in part to the isoforms that are produced. The higher molecular weight isoforms such as the 189 and 206 amino acid VEFG isoforms, bind tightly to heparin sulfate and remain cell bound. If I could have the first slide, VEGF is the one angiogenic molecule that has had the most evaluation in myeloid melignancies.

The pattern of VEGF expression within hematopoietic cells in MDS is quite interesting. The abnormal localized immature precursors or so-called "ALIP", shown with myeloperoxidase activity, intensely express VEGF and also express the Flt-1 (VEGFR-1) receptor. VEGF expression as well as Flt-1 expression is limited largely to the immature myeloids as well as monocytic cells with occasional expression in megakaryocytes.

TOP

This is another example of ALIP and you can see that there is intense VEGF expression, but you can see how much this varies. Here is another case, ALIP with excessive VEGF. In all the cases that we have examined in patients with MDS, ALIP always express VEGF and correspondingly display the VEGF receptor.

TOP

At the Arizona Cancer Center, this same pattern is seen in the majority of patients with MDS. Certainly it is more evident as the blast percentage increases, but the Flt-1 receptor is the dominant VEGF receptor expressed in the immature myeloid cells in MDS. KDR, the VEGFR-2 receptor, is seen very infrequently.

TOP

The same pattern is seen in AML. In endothelial cells it is the VEGFR-2 of the KDR receptor which is believed to be responsible for mitogenesis or a proliferative response, whereas the VEGFR-1 or Flt-1 receptor is responsible for permeability. The question then arises as to what signal is Flt-1 transmitting in these malignant myeloid progenitors.

TOP

Our studies indicate that Flt-1 is providing a mitogenic signal. These are KG1 cells stimulated with agonistic antibodies. KG1 expresses both the Flt-1 and the KDR receptors, and you can see, in the presence of agonistic antibodies to either Flt-1 or KDR, a concentration-dependent clonogenic response occurs. In HL60, cells only the Flt-1 receptor is expressed, and a clonogenic response is triggered by the Flt-1 antibodies.

TOP

Unlike endothelial cells, Flt-1 induces a clonogenic response in MDS and AML. It also induces homotypic adhesion as you can see here through beta-1 integrins which may explain the myeloblast coalescers we see in ALIP in patients with MDS.

TOP

What are the potential biologic implications of medullary VEGF in MDS? In addition to promotion of ______ angiogenesis, it probably also contributes to extracellular matrix degradation, increased secretion of other angiogenic molecules such as interleukin-8, and through MMP induction of endothelial cells this can also generate soluble TNF and fas ligand. MMPs cleave cell membrane-bound fas ligand and TNF to create soluble TNF, which can further contribute to the extracellular matrix (ECM) degradation. As a consequence, ECM degradation may contribute to decreased erythroid adherence. VEGF also suppresses primitive progenitor growth and promotes expansion of the mature myeloid cells. Dr. Broxmeyer will probably go over this later. The data I showed you suggests that VEGF promotes myeloblast self-renewal and adhesion, and also impairs dendritic cell maturation and function which is abnormal in patients with MDS.

TOP

I think if I were to refer to another summary cartoon of these biologic features from our understanding now, at least part of this interaction would be shown like this. ALIP cells produce VEGF which can stimulate endothelial cells as well as macrophages to produce inflammatory cytokines, which reinforce fas ligand expression in erythroid cells and, as a consequence, ineffective erythropoiesis.

TOP

Obviously there is greater complexity in MDS and many patients progress to AML. P15 inactivation occurs through promoter hypermethylation. Our Japanese colleagues have shown that as we move closer to RABT and AML, 70 percent or more patients will have P15 inactivation. In addition, down-regulation of the fas receptor occurs promoting fas ligand resistant myeloblasts.

TOP

Of course the one exception to all this, is chronic myelomonocytic leukemia (CMML). What we see in the laboratory in CMML is autonomous growth of myeloid precursors in the absence of cytokine supplementation. It appears to result from GMCSF hypersensitivity. In the vast majority of these adult patients you will see activating point mutations of ras which generates increased ras GTP pools, which contributes to GMCSF hypersensitivity. In juvenile myelomonocytic leukemia it arises more commonly from inactivation of NF1, rather than from ras activation, which hydrolyzes ras GTP creating increased ras GTP pools.

TOP

In the last few slides I want to summarize where we are and what I think are some of the key questions regarding apoptosis and the biology of MDS. The question still arises, does this occur as a genetic program or is it simply the micro environment or a combination. I am sure it derives from both components. What triggers the excess inflammatory cytokine generation? Certainly VEGF may be one candidate, but there are a number of other potential molecules that need to be evaluated. The apoptotic index may be a valuable tool if it is predictive for treatment response with some of the new anti-apoptotic therapies. It is essential that clinical trials that are being performed that apoptotic index and other biologic features be assessed. What are the relevant response biomarkers? In addition to apoptotic index, there are a number of others which I will propose in the next slide. If we successfully reduce apoptosis, do we alter the natural history of the disease? If this is really a valid target for therapeutic intervention, I think the clinical trials will answer this very soon. The impact on progenitor adhesion, and it is contribution to anoikys in the bone marrow of these patients is a fertile area for investigation in MDS that deserves additional study.

TOP

What are some of the potential response biomarkers that can be evaluated? Certainly apoptotic index and the proliferative capacity have been correlated with response to anti-apoptotic therapy. The majority of investigators assess pro-apoptotic cytokines prior to therapy to see if they are predictive of response. Angiogenesis may now be a relevant biomarker to see how it changes with therapy.

TOP

There are a number of opportunities as it relates to translational therapy in MDS. In CMML, clinical investigations of ras farnesyl transferase inhibitors are under way. There remains the possibility one can effect the same pathway through SRC inhibitors with non-receptor protein kinase inhibitors. P15 de-repression through hypomethylation remains a clinical possibility. The data from the CALGB 5-azocytidine trial are encouraging, but we need good biological co-relation. There are a number of VEGF antagonists that are now in clinical trials that deserve evaluation in MDS. As it relates to TNS, there are a number of molecules that can be used to inhibit TNF actions but I suspect TNF alone is not going to be the answer for this disease. There are a number of potential different targets and they may have some benefit but it won't be a silver bullet affecting apoptosis either through bioreplacement or another approach through protease inhibitors. Neal Young's group has shown that even the AIDS drugs protease inhibitors can impact some of the other caspases within hematopoietic progenitors and these drugs and other more potent ones deserve evaluation in myelin dysplasia. (Applause.)

TOP