MSC Transplantation Reduces Bone Loss via Epigenetic Regulation of Notch Signaling in Lupus

Mesenchymal stem cells from bone marrow, fat, and other tissues have been used in many clinical trials, experiments, and treatment regimens. While these cells are not magic bullets, they do have the ability to suppress unwanted inflammation, differentiate into bone, cartilage, tendon, smooth muscle, and fat, and can release a variety of healing molecules that help organs from hearts to kidneys heal themselves.

Mesenchymal stem cell transplantation (MSCT) is the main means by which mesenchymal stem cells are delivered to patients for therapeutic purposes. However, the precise mechanisms that underlie the success of these cells are not fully understood. In a paper by from the University Of Pennsylvania School Of Dental Medicine published in the journal Cell Metabolism, MSCT were able to re-establish the bone marrow function in MRL/lpr mice. The MRL/lpr mouse is a genetic model of a generalized autoimmune disease sharing many features and organ pathology with systemic lupus erythematosus (SLE). Such mice show bone loss and poor bone deposition, a condition known as “osteopenia.” Because mesenchymal stem cells are usually the cells in bone marrow that differentiate into osteoblasts (which make bone) a condition like osteopenia results from defective mesenchymal stem cell function.

In this paper, Shi and his coworkers and collaborators showed that the lack of the Fas protein in the mesenchymal stem cells from MRL/lpr mice prevents them from releasing a regulatory molecule called “miR-29b.” This regulatory molecule, mir-29b, is a small RNA molecule known as a microRNA. MicroRNAs regulate the expression of other genes, and the failure to release miR-29b increases the intracellular levels of miR-29b. This build-up in the levels of miR-29b causes the downregulation of an enzyme called “DNA methyltransferase 1” or Dnmt1. This is not surprising, since this is precisely what microRNAs do – they regulate genes. Dnmt1 attaches methyl groups (CH3 molecules) to the promoter or control regions of genes.

Decrease in the levels of Dnmt1 causes hypomethylation of the Notch1 promoter. When promoters are heavily methylated, genes are poorly expressed. When very methyl groups are attached to the promoters, then the gene has a greater chance of being highly expressed. Robust expression of the Notch1 genes activates Notch signaling. Increased Notch signaling leads to impaired bone production, since differentiation into bone-making cells requires mesenchymal stem cells to down-regulate Notch signaling.

When normal mesenchymal stem cells are transplanted into the bone marrow of MRL/lpr mice, they release small vesicles called exosomes that transfer the Fas protein to recipient MRL/lpr bone marrow mesenchymal stem cells. The presence of the Fas protein reduces intracellular levels of miR-29b, and this increases Dnmt1-mediated methylation of the Notch1 promoter. This decreases the expression of Notch1 and improves MRL/lpr BMMSC function.


These findings elucidate the means by which MSCT rescues MRL/lpr BMMSC function. Since MRL/lpr mice are a model system for lupus, it suggests that donor mesenchymal stem cell transplantation into lupus patients provides Fas protein to the defective, native mesenchymal stem cells, thereby regulating the miR-29b/Dnmt1/Notch epigenetic cascade that increases differentiation of mesenchymal stem cells into osteoblasts and bone deposition rates.


Mesenchymal Stem Cells Derived from Induced Pluripotent Stem Cells are Epigenetically Rejuvenated

Earlier this year, Miltalipov and his research group published a paper in Nature that compared the genetic integrity of embryonic stem cells made from embryos, to induced pluripotent stem cells and embryonic stem cells made from cloned embryos.  All three sets of stem cells seemed to have comparable numbers of mutations, but the induced pluripotent stem cells had “epigenetic changes” that were not found in either stem cell line from cloned or non-cloned embryos.

Genetic characteristics have to do with the sequence of the DNA molecules that make up the genome of an organism.  Epigenetic characteristics have nothing to do with the sequence of DNA, but instead are the result of small chemicals that are attached to the DNA molecule.  These small chemical tags affect gene expression patterns.  Every cell has a specific epigenetic signature.

During development, the cells that will form our eggs and sperm in our bodies, the “primordial germ cells,” begin their lives in the outer layer of the embryo.  During the third week of life, these primordial germ cells or PGCs move like amoebas and wander into the yolk sac wall and collect near the exit of a sac called the “allantois.”  The PGCs are outside the embryo at this time or extraembryonal.  Incidentallyyolk sac is a terrible name for this structure, since it does not produce yolk proteins.  Therefore other textbooks have renamed it the “primary umbilical vesicle,” which is a bit of a mouthful, but it probably better than “yolk sac.”


1 - Primordial germ cells 2 - Allantois 3 - Rectum 4 - Ectoderm 5 - Foregut 6 - Primordial Heart 7 - Secondary yolk sac 8 - Endoderm 9 - Mesoderm 10 - Amniotic cavity
1 – Primordial germ cells
2 – Allantois
3 – Rectum
4 – Ectoderm
5 – Foregut
6 – Primordial Heart
7 – Secondary yolk sac
8 – Endoderm
9 – Mesoderm
10 – Amniotic cavity

The embryo around this time undergoes a bending process as a result of its growth and the head bends toward the tail (known as the cranio-caudal curvature) and then the sides of the embryo fold downwards and eventually fuse (lateral folding).  This bending of the embryo allows the PGCs to wander back into the embryo again between the fourth and sixth week.  The PGCs move along the yolk sac wall to the vitelline and into the wall of the rectum.  After crossing the dorsal mesentery (which holds the developing intestines in place) they colonize the gonadal or genital ridge (which is the developing gonad). During their journey, and while in the gonadal ridge, the PGCs divide many times.

1 - Rectum 2 - Vitelline 3 - Allantois 4 - Nephrogenic cord (pink) 5 - Gonadal ridge (green) 6 - Primordial germ cells (red dots) 7 - Heart prominence
1 – Rectum
2 – Vitelline
3 – Allantois
4 – Nephrogenic cord (pink)
5 – Gonadal ridge (green)
6 – Primordial germ cells (red dots)
7 – Heart prominence

When the PGCs move into the developing gonad, the chemical tags on their DNA are completely removed (rather famous paper – Lee, et al., Development 129, 1807–1817 (2002).  This epigenetic erasure proceeds in order for the PGCs to develop into gametes and then received a gamete-specific set of epigenetic modifications.  These epigenetic modifications also extend to the proteins that package the DNA into chromosomes – proteins called histones.  Specific modifications of histone proteins and DNA lead to gamete-specific expression of genes.  Once fertilization occurs, and the embryological program is initiated, tissue-specific epigenetic modifications are conveyed onto the DNA and histones of particular cell populations.

This is a long-winded explanation, but because many cancer cells have abnormal epigenetic modifications, these epigenetic abnormalities in induced pluripotent stem cells (iPSCs) have been taken with some degree of seriousness.  Although, there is little evidence to date that links the cancer-causing capabilities of iPSCs with specific epigenetic modifications, although it certainly affects the ability of these cells to differentiate into various cell types.

A paper has just come from the laboratory of Wolfgang Wagner from the Aachen University Medical School, in Aachen, Germany that derived iPSCs from mesenchymal stem cells from human bone marrow, and then in a cool one-step procedure, differentiated these cells into mesenchymal stem cells (MSCs).  These  iPS-MSCs looked the same, and acted the same in cell culture as the parent MSCs, and had the same gene expression profiles as primary MSCs.  However, all age-related and tissue-specific epigenetic patterns had been erased by the reprogramming process.  This means that all the tissue-specific, senescence-associated, and age-related epigenetic patterns were erased during reprogramming.  Another feature of these iPS-MSCs is that they lacked but the ability to down-regulate the immune response, which is a major feature of MSCs.

Thus, this paper by the Wagner lab shows that MSCs derived from iPSCs are rejuvenated by the reprogramming process.  Also, the donor-specific epigenetic features are maintained, which was also discovered by Shao and others last year.  This suggests that epigenetic abnormalities are not an inherent property of the derivation of iPSCs, and that this feature is not an intractable characteristic of iPSCs derivation and may not prevent these cells from being successfully and safely used in the clinic.  However, this might be a cell type-specific phenomenon.  Also, the loss of the immune system regulatory capabilities of these iPS-MSCs is troubling and this requires further work.


Making Better Induced Pluripotent Stem Cells

On July 2nd of this year, a paper appeared in the journal Nature that performed complete genomic analyses of embryonic stem cells derived from embryos or cloned embryos, and induced pluripotent stem cells (iPSCs), which are made from reprogrammed adult cells.  They found that both embryonic stem cells made from cloned embryos and iPSCs derived from the same types of adult cells contained comparable numbers of newly introduced mutations.  However, when it came to the epigenetic modification of the genome (the small chemical tags attached to specific bases of DNA that gives the cell hints as to which genes to turn off), the epigenetic pattern of the embryonic stem cells made from cloned embryos more closely resembled that from embryonic stem cells.  The iPSCs still had some similarities with the adult cells from which they were derived whereas the embryonic stem cells made from cloned embryos were more completely reprogrammed.  From this the authors claimed that making embryonic stem cells by means of cloning is ideal for cell replacement therapies.

There is a big problem with this conclusion:  This was tried in animals and it did not work because of immunological rejection of the products from the stem cells.  For more information on this, see my book, The Stem Cell Epistles, chapter 18.

Despite this “bad news” for iPSCs, two recent papers have actually provided some good news for stem cells that can heal without destroying embryos.  The first paper comes from Timothy Nelson’s laboratory at the Mayo Clinic in Rochester, Minnesota.  Differentiation of iPSCs is, in some cases, rather efficient and the isolation procedures fail to effectively isolate the differentiated cells from potentially tumor-causing cells.  However, in other cases, the differentiation is inefficient and the isolation procedures are also rather poor, which leaves a large enough population of undifferentiated tumor-causing cells.

Nelson’ group has discovered that treating iPSCs and their derivatives with anti-cancer drugs like etoposide (a topoisomerase II inhibitor for those who are interested) increases engraftment efficiency and decreases the incidence of tumors.  My only problem with Nelson’s paper is that he and his colleagues used lentiviral vectors to make their iPSCs.  These vectors tend to produce iPSCs that are rather good at causing tumors.  I would have rather that he tried making iPSCs with other methods that do not leave permanent transgenes in the cells.  Nelson and his group transplanted their iPSC-derived cells into the hearts of mice where they could use high-resolution imaging to determine the number of cells that integrated into the heart and the presence of cell masses that were indicative of tumors.  None of the ectoposide-treated cell transplants caused tumors whereas 4 of the 5 transplants not treated with ectoposide caused tumors.  This paper appeared in Stem Cells and Development.

The second “good news” paper for iPSCs comes from Junji Takeda at the University of Osaka and Ken Igawa from the Tokyo Medical and Dental University, Japan.   In their paper from Stem Cells Translational Medicine, the Japanese groups collaborated to make iPSCs from skin based fibroblasts and then differentiate them into skin cells (keratinocytes).  However, they made the iPSCs in two different ways.  The first protocol utilized the piggyBac transposon system to make iPSCs.  The piggyBac system comes from moths, but it is highly active in mammalian cells.  It can deliver the genes to the cells, but the segment of DNA is then easily excised from the host cells without causing any mutations.  This system, therefore, will generate iPSCs that do not have any transgenes in them.  The second protocol used a system based on cytomegalovirus that leaves the transgenes in the cells but gradually inactivates their expression.

When these two types of iPSCs were compared, they seems to be essentially identical when grown in culture.  Thus in the pluripotent state, the cells were equivalent for the most part.  But once the iPSC lines were differentiated into skin cells, the transgene-free iPSCs formed skin cells that looked, behaved and had the same gene expression profile as normal human skin cells.  The transgene-containing iPSCs differentiated into skin cells, but they did not look quite like skin cells, did not have the same gene expression profile as normal human skin cells, and did not behave like normal human skin cells.

The moral of this story is that not all iPSC lines are created equally and the way you derive them is as important as the cell type from which they were derived.  Also, even incomplete differentiation does not need to be an obstacle for iPSCs, since the cancer-causing cells can be removed by means of specific drugs.  Finally, not all that glitters is gold.  Cloned embryos may give you stem cells that look more like embryonic stem cells, but so what.  These might still suffer from many of the same set backs.  Add to that the ethical problems with getting women to give up their eggs for research and cures (see Jennifer Lahl’s movie Eggsploitation for more disturbing information about that), and you have a losing combination.