Circulating Factor Rejuvenates the Hearts of Older Mice


Two researchers associated with the Harvard Stem Cell Institute, one of whom is a practicing cardiologist at Brigham and Women’s Hospital, and a cell biologist have identified a protein in the blood of mice and humans decreases during aging and may prove to be the first effective treatment for the form of age-related heart failure that affects millions of Americans.

Growth Differentiation Factor 11 or GDF11 circulates throughout the bloodstream of humans and mice. Injections of GDF11 into old mice that have developed thickened heart walls in a manner similar to aging humans, the hearts were reduced in size and thickness, and resembled the healthy hearts of younger mice.

Even more important than the implications for the treatment of diastolic heart failure, the finding by Richard T. Lee, from Harvard Medical School and Amy Wagers, who is a professor in Harvard’s Department of Stem Cell and Regenerative Medicine, might rewrite our understanding of aging. These findings were published in the prestigious, international journal Cell.

“The most common form of heart failure [in the elderly] is actually a form that’s not caused by heart attacks but is very much related to the heart aging,” said Lee, who, like Wagers, is a principal faculty member at HSCI.

“In this study, we were able to show that a protein that circulates in the blood is related to this aging process, and if we gave older mice this protein, we could reverse the heart aging in a very short period of time,” Lee said. “We are very excited about it because it opens a new window on the most common form of heart failure.”
He added, “This is the coolest thing I’ve ever been a part of.”

Doug Melton, HSCI co-director, called the discovery “huge. It’s going to change the way we think about aging.”

Lee, who practices medicine as a cardiologist, noted that he has approximately 300 patients, and of these “I think I have about 20 who are suffering from this type of heart failure, which we sometimes call diastolic heart failure,” said Lee. “They come into the hospital, have a lot of fluid taken off, then they’ll go home. Then they come back again. It’s really frustrating because we don’t have any drugs to treat this. We need to work as hard as we can to figure out if this discovery can be turned into a treatment for heart failure in our aging patients.”

The Lee and Wagers labs would like to move GDF-11 toward clinical trials. Lee predicts that they might be able to begin these studies in four to five years. However, they need to determine the other tissues that are affected by GDF11.

Wagers was a postdoctoral research fellow at Stanford University, where she learned how to work with the “parabiotic” mouse system. Parabiosis refers to two animals that share a common circulatory system. Wagers used this system to link the circulatory system of young mice with that of older mice. When she did that, she and her colleagues discovered that soluble factors in the blood of young animals have a rejuvenating effect on various tissues in older animals. In particular, the spinal cord and the musculature of the older animals showed marked improvements.

“As we age, there are many changes that occur in different parts of the body,” Wagers said, “and those changes are often associated with a decline in the function of our bodies. One of the interests of my laboratory is in understanding why this happens and whether it is an inevitable consequence of aging, or if it might be reversible.

“In this study, we compared young and old animals and identified a substance in the blood that is present at high levels when you’re young and lower levels when you’re old. We further found that when we supplemented the low levels of this substance that were present in old animals to the levels normally seen in youth, this could have a dramatic effect on the heart.

“It’s been observed for many, many years that when aging occurs it affects multiple body systems sort of in a semi-synchronous way,” Wagers said, “and this suggests that there may be some common signal that drives the body’s response to getting older. We hypothesized that this common signal might be a substance that was traveling in the bloodstream, because the bloodstream accesses organs throughout the body.”

“I think Amy and I started thinking about something like this almost five years ago,” said Lee, who added that he and Wagers were brought together by HSCI. “Without the Harvard Stem Cell Institute, this never would have happened,” he said.

Lee and Wagers conducted their first experiment about four years ago, and the results were startling, Lee said. “A fellow named Francesco Loffredo was examining the hearts of the aging mice. He came to me and said, ‘You don’t have to analyze it; you can see it with the naked eye.’ I couldn’t believe that, and I said ‘Go back, analyze it, and do it blinded.’ Then I looked at the hearts, and I could see he was correct,” Lee recalled.

“When we started these experiments, I actually was thinking that there would not be a response,” Wagers said. “We had been using similar kinds of approaches in other tissues, regenerative tissues, tissues that we know have the capacity to heal themselves after they’ve been injured. But the heart is not well known for doing that, and so I was quite convinced that there would be no response. When I saw the dramatic difference in heart size that was very apparent after this exposure of an old animal to young blood, it was very clear that we had to figure out what was going on,” she said.

“The blood is full of all kinds of things,” the biologist said, “and trying to narrow down what might be the responsible factor was going to be a big challenge. I think that’s where the collaboration was so wonderful; in that we could take advantage of the expertise in both of our laboratories to really home in on what might be the responsible substance.”

Lee explained, “We thought it was interesting right away, and we repeated it right away. But we had to show that this was not a blood pressure effect, that the young mice didn’t just cause the old mice to have lower blood pressure. We had to build a custom device to measure blood pressures off their tails. It took a year to do the analysis to show that it was not a blood pressure effect.

“After about 2½ years we were convinced, and said, ‘we really have to identify this factor.’ It took about six months to find something, and another year to be convinced that it was real,” Lee said. “We looked at lipids; we looked at metabolites. Then we set up a collaboration with a startup company in Colorado, called SomaLogic that had an interesting technology for analyzing factors in blood. And by working closely with SomaLogic, we found the likely factor.”

They discovered that at least one of the factors responsible for heart rejuvenation was GDF-11, “a member of a very important family of proteins called TGF-beta proteins, for transforming growth factor. There are around 35 members of the family,” Lee said. “Some have been very well studied, and this is one that is relatively obscure.”

The work was supported in part by HSCI, the National Institutes of Health, and the American Heart Association.

Direct Conversion of Skin Cells into Neural Precursor Cells


Cell reprogramming involves the use of genetic engineering techniques to push cells into a new cell type WITHOUT passing those cells through the embryonic stage. Several different studies have shown that transferring particular genes into specific cell types or removing distinct genes from them can drive them to become other cell types. There are several published examples of transdifferentiation:
1) In 1989, Weintraub and colleagues overexpressed a gene called MyoD in cultured fibroblasts to convert them into muscle cells. Unfortunately, this conversion was incomplete and required continuous expression of MyoD (Weintraub H et al., Proc. Natl. Acad. Sci. USA 1989;86:5434-8).
2) Tachibana and colleagues overexpressed a gene called MITF to transdifferentiate fibroblasts into pigment-synthesizing melanocytes (Tachibana et al., Nature Genetics 1996;14:50-4).
3) Xie and others overexpressed genes that encode two transcription factors (C/EBP and PU.1) in B cells, T cells, and fibroblasts into transdifferentiated them into cells that looked like macrophages (Xie et al., Cell 2004;117:663-76).
4) Deletion of a gene called Pax5 can transdifferentiate antibody-secreting B lymphocytes into common lymphoid progenitors, macrophages and antigen-presenting T cells (Cobaleda C, Jochum W, and Busslinger M. Nature 2007;449:473-7).
5) Doug Melton’s laboratory at Harvard University transferred a specific combination of three transcription factor genes (Ngn3, which is also known as Neurog3, Pdx1 and Mafa), into pancreatic exocrine cells (those cells that produce and secrete digestive enzymes).  This reprogrammed the cells into insulin-secreting beta cells (Qiao Zhou et a., In vivo reprogramming of adult pancreatic exocrine cells to β-cells. Nature 2008;455, 627-632).
6) Deletion of a gene that encodes a transcription factor called Foxl2 converts granulosa and thecal cells (found in the ovary) into Sertoli and Leydig cells, which are found in the testes (Uhlenhaut et al., Cell 2009;139:1130-42).
7) Thomas Vierbuchen and colleagues in the laboratory of Marius Wernig at Sanford University School of Medicine used a combination of three genes (Asc1, Brn2 and Myt1l) to convert fibroblasts into functional neurons (Vierbuchen et al., Nature 2010;463:1035-42).
8) In 2010, Ieda and co-workers in the laboratory of Deepak Srivastava have used ectopic expression of three genes (GATA4, MEF2C, and TBX5) to directly convert heart-based fibroblasts into heart muscle cells. These reprogrammed cells did not require expression of the introduced transgenes (Ieda et al., Cell 2010;142:375-86).

A recent study has extended these results even further. In an earlier study, Marius Wernig’s lab at Stanford University School of Medicine showed that skin fibroblasts can be transdifferentiated into functional neurons. Wernig’s lab has followed up in a paper that was published online on Jan. 30, 2012 in the Proceedings of the National Academy of Sciences.

In this study, Wernig’s lab used mouse skin cells and directly transdifferentiated them into the three main parts of the nervous system. These transdifferentiation experiments show that pluripotency (a term that describes the ability of stem cells to become nearly any cell in the body) is NOT necessary for a cell to transform from one cell type to another. Together, these results raise the possibility that embryonic stem cell research and induced pluripotency could be superseded by a more direct way of generating specific types of cells for therapy or research.

In the new study, Wernig and his colleagues converted fibroblasts in to neural precursor cells (NPCs). NPCs have the capacity to differentiate into neurons, but they can also become the two other main cell types in the nervous system: astrocytes and oligodendrocytes. In addition to their greater versatility, newly derived NPCs offer another advantage over neurons because they can be cultivated to large numbers in the laboratory — a feature critical for their long-term usefulness in transplantation or drug screening..

The switch from skin cells to NPCs occurred with high efficiency and only took about three weeks after the addition of just three transcription factors. Wernig’s research group used a different combination of three transcription factors than those used to generate mature neurons (Brn2, Sox2 and FoxG1) than was used to generate mature neurons. This combination of transcription factors drove the fibroblasts to transdifferentiate into “tripotent” NPCs that have the ability to form neurons and astrocytes but also into oligodendrocyte. The finding implies that it may one day be possible to generate a variety of neural-system cells for transplantation that would perfectly match a human patient.

The lab’s previous success with transdifferentiation experiments led Wernig to wonder if his lab could convert skin-based fibroblasts into the more-versatile NPCs. To do so, Wernig’s research group infected embryonic mouse skin cells — a commonly used laboratory cell line — with a virus that encoded 11 transcription factors known to be expressed at high levels in NPCs. Just over three weeks later, about 10 percent of the cells began to look and act like NPCs.

They then winnowed down the original panel of 11 transcription factors to just three that still converted fibroblasts to NPCs. Three of these genes (Brn2, Sox2 and FoxG1; in contrast, the conversion of skin cells directly to functional neurons requires the transcription factors Brn2, Ascl1 and Myt1l.) drove fibroblasts to differentiate into NPCs that were “tripotential” – that is, the NPCs could differentiate into not just neurons and astrocytes, but also oligodendrocytes, which make myelin that insulates nerve fibers and allows them to effectively transmit nerve impulses. Wernig’s lab workers dubbed the newly converted population “induced neural precursor cells,” or iNPCs.

In vitro experiments showed that the astrocytes, neurons and oligodendrocytes made from iNPCs expressed the same genes and morphologically resembled that they resembled astrocytes, neurons and oligodendrocytes found in living organisms. However, Wernig’s lab wanted to know how iNPCs would react when transplanted into an animal. Therefore, they injected them into the brains of newborn laboratory mice that were bred to lack the ability to myelinate neurons. After 10 weeks, they found that the injected cells had differentiated into oligodendroytes and had begun to coat the animals’ neurons with myelin.

Marius Wernig, MD, assistant professor of pathology and a member of Stanford’s Institute for Stem Cell Biology and Regenerative Medicine, said: “We are thrilled about the prospects for potential medical use of these cells. We’ve shown the cells can integrate into a mouse brain and produce a missing protein important for the conduction of electrical signal by the neurons. This is important because the mouse model we used mimics that of a human genetic brain disease. However, more work needs to be done to generate similar cells from human skin cells and assess their safety and efficacy.”

Pediatric cardiologist Deepak Srivastava, MD, who was not involved in these studies noted, “Dr. Wernig’s demonstration that fibroblasts can be converted into functional nerve cells opens the door to consider new ways to regenerate damaged neurons using cells surrounding the area of injury. It also suggests that we may be able to transdifferentiate cells into other cell types.” Srivastava is the director of cardiovascular research at the Gladstone Institutes at the University of California-San Francisco. In 2010, Srivastava’s lab transdifferentiated mouse heart fibroblasts into beating heart muscle cells.

The first author of this article, Ernesto Lujan, added: “Direct conversion has a number of advantages. It occurs with relatively high efficiency and it generates a fairly homogenous population of cells. In contrast, cells derived from iPS cells must be carefully screened to eliminate any remaining pluripotent cells or cells that can differentiate into different lineages.” Pluripotent cells can cause cancers when transplanted into animals or humans.

“Not only do these cells appear functional in the laboratory, they also seem to be able to integrate appropriately in an in vivo animal model,” said Lujan.

Wernig’s group is now working to replicate the work with skin-based fibroblasts from adult mice and humans, but Lujan emphasized that more research is needed before any human transplantation experiments could be conducted. Until that time, the ability to quickly and efficiently generate NPCs that can be grown in the laboratory to mass quantities and maintained over time will be valuable in disease and drug-targeting studies.

“In addition to direct therapeutic application, these cells may be very useful to study human diseases in a laboratory dish or even following transplantation into a developing rodent brain,” said Wernig.

Induced Pluripotent Stem Cells


Embryonic stem cells might provide the means to heal a variety of physical ailments. However the problem with embryonic stem cells is not necessarily in their use, but in their derivation. In order to make embryonic stem cell lines, human embryos are destroyed.

The following video shows Alice Chen from Doug Melton’s laboratory at Harvard University destroying embryos to make embryonic stem cells:  http://www.jove.com/index/details.stp?ID=574.

Now that federal funding is available to not only work with existing embryonic stem cell lines but to MAKE new lines, there is nothing to stop researchers from thawing and (I’m sorry to be so blunt) killing human embryos. Can we have our “cake and eat it too?” Can we have the benefits of embryonic stem cells and not destroy embryos? Perhaps we can.

In 2001, Masako Tada reported the fusion of embryonic stem cells with a connective tissue cell called a fibroblast. This fusion reprograms the fibroblasts so that they behave like embryonic stem cells (Current Biology 11, no. 9 (2001): 1553–8). This suggests that something within embryonic stem cells can redirect the machinery of somatic cells to become more like that of embryonic stem cells. In 2006 Kazutoshi Takahashi and Shinya Yamanaka were able to generate embryonic stem cell lines by introducing four specific genes into mouse skin fibroblasts. These “induced pluripotent stem cells” (iPSCs) shared many of the properties of embryonic stem cells derived from embryos, but when transplanted into mouse embryos, they were not able to participate in the formation of an adult mouse (Cell 126, no. 4 (2006): 663–76). This experiment showed that it is possible to convert adult cells into something that resembles an embryonic stem cell. Could we push adult cells further? In 2007, three different research groups used retroviruses to transfer four different genes (Oct3/4, Sox2, c-Myc and Klf4) into mouse skin fibroblasts and completely transformed them into cells that had all the features and behaviors of embryonic stem cells (Cell Stem Cell 1, no. 1 (2007): 55–70; Nature 448 (2007): 313–7; Nature 448 (2007): 318–24.).

These experiments drew a great deal of excitement, but there were several safety concerns that had to be addressed before iPSCs could be used in human clinical trials.  Scientists used engineered retroviruses to introduce genes into adult cells in order to reprogram them into iPSCs (Current Topics in Microbiology and Immunology 261 (2002): 31-52).  Retroviruses insert a DNA copy of their genome into the chromosomes of the host cell they have infected.  If that viral DNA inserts into a gene, it can disrupt it and cause a mutation.  This can have dire consequences (see Folia Biologia 46 (2000): 226-32; Science 302 (2003): 415-9).  Fortunately this is not an intractable problem.  The conversion of adult cells into iPSCs only requires the transient expression of the inserted genes.  Secondly, scientists have created retroviruses that self-inactivate after their initial insertion (Journal of Virology 72 (1998): 8150-7; Virology 261, (1999).  One laboratory has also discovered a way to make iPSCs with a virus that does not insert into host cell chromosomes (Science 322 (2008): 945-9).  Other researchers have designed ingenious ways to move the necessary genes into adult cells without using viruses (Science 322 (2008): 949-53).  Both procedures avoid the dangers associated with the use of retroviruses.

A second concern involves the genes used to convert re-program adult cells into iPSCs.  One of these genes, c-Myc, is found in multiple copies in human and animal tumors.  Thus increasing the number of copies of the c-Myc gene might predispose such cells to form tumors (Recent Patents on Anticancer Drug Discovery 1 (2006): 305-26; Seminars in Cancer Biology 16 (2006): 318-30). Indeed, the increased ability of iPSCs made by Yamanaka to cause tumors in laboratory animals underscore this concern (Hepatology 46, no 3 (2009): 1049-9).  Several groups, however, have succeeded in making iPSCs from adult cells without the use of the c-Myc gene (Science 321 (2008): 699­-702; Nature Biotechnology 26 (2008): 101-6; Science 318 (2007): 1917–20), although the conversion is much less efficient.  Additionally, several groups have established that particular chemicals, in combination with the addition of a subset of the four genes originally used, can effectively transform particular cells into iPSCs (Cell Stem Cell 2 (2008): 525-8).   Thus the larger safety concerns facing iPSCs have been largely solved.

Finally, patient-specific iPSCs have been made in several labs, even though they have not been used in clinical trials to date.  Here is a short list of some of the diseases for which patient-specific iPSCs have been made:

Amylotrophic Lateral SclerosisScience 321 (2008): 1218­21.

Spinal Muscular AtrophyNature 457 (2009): 277­81.

Parkinson’ DiseaseCell 136, no. 5 (2009): 964­77.

Adenosine deaminase deficiency-related severe combined immunodeficiency – Cell 134, no. 5 (2008): 877­86.

Shwachman-Bodian-Diamond syndrome – Cell 134, no. 5 (2008): 877­86.

Gaucher disease – Cell 134, no. 5 (2008): 877­86.

Duchenne and Becker muscular dystrophy – Cell 134, no. 5 (2008): 877­86.

Huntington disease – Cell 134, no. 5 (2008): 877­86.

Juvenile-onset type 1 diabetes mellitus – Cell 134, no. 5 (2008): 877­86.

Down syndrome – Cell 134, no. 5 (2008): 877­86.

Lesch-Nyhan syndromeCell 134, no. 5 (2008): 877­86.

Thus iPSCs represent an exciting, embryo-free alternative to embryonic stem cells that provide essentially all of the opportunities for regenerative medicine without destroying embryos.