William Check, PhD
March 2016—At the annual meeting of the Association for Molecular Pathology in November 2015, one plenary session was called “Exciting Times for Translational Research in Molecular Hematology.” In accord with the title, Margaret Goodell, PhD, gave an exciting talk about how hematopoietic stem cells are regulated in mice. While Dr. Goodell’s basic research was impressive, what was most remarkable was how it meshed with and anticipated research in human hematopoietic malignancies from other laboratories.
Dr. Goodell, the Vivian L. Smith chair of regenerative medicine at Baylor College of Medicine, described finding two genes that largely control hematopoietic stem cell renewal in mice. When human leukemias and lymphomas were probed for prevalent oncogenes, these same two genes turned up. It may be possible to use the mouse model to shed light on the process of oncogenesis in humans, work that is going on now in Dr. Goodell’s laboratory.
In a related AMP webinar presentation, in January, Benjamin Ebert, MD, PhD, an associate professor of medicine at Harvard Medical School, talked about clonal hematopoiesis of indeterminate potential, or CHIP, which is a premalignant human condition consistent with Dr. Goodell’s discoveries. Several laboratories have found that human hematopoietic stem cells become increasingly clonal with age, and that this clonality is regulated by the same two genes that govern mouse hematopoiesis and are often found mutated in human hematopoietic cancers. Dr. Ebert, a member of the Harvard Stem Cell Institute, raised the possibility that clonal hematopoiesis of indeterminate potential is a premalignant state that can be identified on the way to myeloid malignancies.
Dr. Goodell
In a CAP TODAY interview, Dr. Goodell said her laboratory had arrived at findings that are similar to those of Dr. Ebert but from a different direction. “We started with the genes, while Dr. Ebert started with the disease and found the genes,” said Dr. Goodell, who is also director of the Stem Cells and Regenerative Medicine Center. Asked how far mouse work could be extrapolated to humans, Dr. Goodell said, “We don’t know, but if I were a betting person, I would put my money on the overall understanding we are gleaning from the mouse.”
Dr. Goodell said she “stumbled” on the two genes that have been found to be so important in hematopoietic stem cells (HSC) in both mice and humans. Her work has focused on regulation in murine HSC. She asked, What controls a mouse stem cell to differentiate or self-renew? “We did a fishing experiment,” she said. Using stem cells from wild-type mice, they did microarray expression profiling and picked out a gene called DNA methylation transferase, or Dnmt3a. The DNMT3A protein methylates cytosine to methylcytosine, which is important for regulating gene expression.
Loss of DNA methyltransferase activity in embryonic stem cells obstructs differentiation (Mayle A, et al. Blood. 2015;125:629–638). To determine its role in somatic stem cells, Dr. Goodell’s laboratory did serial transplantation of hematopoietic stem cells from the knockout mouse. They found a great increase in peripheral blood cells derived from the knockout stem cells. Knockout-derived stem cells went from 250 cells (the number transferred) to around 70,000, an increase Dr. Goodell calls “remarkable.” Normally a mouse has about 5,000 bone marrow stem cells. The number found when knockout stem cells were transplanted was extraordinary, she says, concluding that with the knockout HSC you get ramped-up self-renewal rather than differentiation (Challen GA, et al. Nat Genet. 2011;
44:23–31).
Further work identified other enzymes involved in regulation of differentiation; DNMT3B was also found to contribute (Challen GA, et al. Cell Stem Cell. 2014;15:350–364). Other laboratories studied the ten-eleven translocation (TET) proteins. This family of enzymes converts 5-methylcytosine to 5-hydroxylmethylcytosine, thereby regulating expression of developmental genes. “We are currently studying what happens when you knock out both of them,” she says.
About five years ago there were no genetic data on mutations in DNMT3A in humans with hematopoietic malignancies. However, within a few years several groups reported that DNMT3A is mutated in about 20 percent of cases of acute myelogenous leukemia. “Now DNMT3A is considered one of the most important tumor suppressors in hematologic malignancies,” Dr. Goodell says. Current data show that DNMT3A and TET are both mutated in a similar range of human hematologic malignancies and are sometimes co-mutated (Yang L, et al. Nat Rev Cancer. 2015;15:152–165).
Conversely, knockout mice in which hematopoietic stem cells carry mutated Dnmt3a genes can recapitulate the range of human DNMT3A-associated diseases. This creates the opportunity to use Dnmt3a knockout mice to model specific malignancies found in patients. Dr. Goodell is now working on the interactions between Dnmt3a and Tet, as well as other genes, to understand how they affect methylation, gene expression, and cell renewal (Jeong M, et al. Nat Genet. 2014;46:17–23).
Dr. Ebert introduced his topic in the AMP webinar with an analogy to colon cancer, asking, “Are there polyps of the bone marrow?”
“We know,” he continued, “that in myelodysplastic syndrome or AML there are commonly mutations in three or more driver genes leading to full-blown malignancy.” Logically, many people must have an expanded population of cells that are not yet malignant with fewer mutations. Indeed, past studies found that some healthy older women have clonal hematopoiesis, and some mutations were detected in these clones.
Dr. Ebert
Dr. Ebert’s group sought to find out whether there are point mutations that lead to clonal hematopoiesis in the general population. To do this they examined whole exome sequencing data from the peripheral blood of more than 17,000 people; the data were acquired for studies investigating the genetic basis of diabetes or cardiovascular disease.
They found that the incidence of clonal hematopoiesis of indeterminate potential was “surprisingly high” in the elderly, Dr. Ebert said. By age 70, approximately 10 percent of individuals had clonal hematopoiesis—one dominant clone that makes up between 20 percent and 80 percent of their peripheral blood and that contains a mutation with one of the known driver genes (Jaiswal S, et al. N Engl J Med. 2014;371:2488–2498). Dr. Ebert called this “likely a conservative estimate.”
Corresponding with Dr. Goodell’s work, this analysis showed that the DNMT3A gene was most frequently mutated, by a large margin. In second place were TET2 mutations. Giving a nod to Dr. Goodell’s work, Dr. Ebert said, “In mice these mutations cause clonal expansion in hematopoietic stem cells without symptoms of disease.”
Among people with CHIP, a large fraction of cells in the blood bear the same mutation—a mean of 24 percent of blood peripheral cells, or, as Dr. Ebert said, “a sizable clone.” This is different from the discovery of rare BCR/ABL clones in some individuals that may be transient. An additional sample was available for analysis from a later time. All variants persisted during the subsequent four to eight years, and in a few cases a second mutation was acquired during follow-up.
As one would predict, the risk of a hematologic malignancy was increased in subjects with clonal hematopoiesis. They had about a 10-fold risk over 150 months. For those with larger clones, the risk was increased 50-fold. Dr. Ebert said that CHIP can be seen as a premalignant state much like monoclonal gammopathy of unknown significance, in that the odds of progression to overt neoplasia are 0.5 percent to one percent per year.
Clonal hematopoiesis was associated with significantly decreased survival; it increased overall mortality about 40 percent. This was not due to a higher risk of cancer but to an increased probability of developing coronary heart disease and ischemic stroke—“a surprising finding,” Dr. Ebert said.
Similar results from a number of other groups at roughly the same time verified the findings of Dr. Ebert and his collaborators.
One clinical implication of the findings, he said: “We need to make sure that individuals coming in for rule-out MDS aren’t misdiagnosed by finding DNMT3A mutations.”
A key question is, When do we say that CHIP has progressed to MDS? “Our challenge now,” Dr. Ebert said, “is that we don’t have any consensus for diagnostic criteria for drawing cutoffs based on mutant allele fractions and number of mutations.”
When, he asked, would you do a CHIP screen? With no intervention to alter outcomes, he said, “there is no rationale for wide-scale screening.”
Dr. Goodell calls the finding of CHIP “transformative.”
“To me, as a stem cell biologist who has worked on HSC for 25 years, I am astounded by the implications of the fact that when we are old we might end up relying on only a handful of different stem cells.”
If these findings are borne out, she says, “Pathologists will end up testing for this. It may be related to the aging phenomenon. Right now we are 10 years away from that.”n
William Check is a writer in Ft. Lauderdale, Fla.
[hr]