There Is No “Stem Cell Mess”

Abstract

Mesenchymal stem cells (MSCs) can be used as therapeutic agents or as starting cells for tissue-engineered skeletal tissues such as cartilage. In vitro tissue engineering requires that optimal conditions be arranged for their tissue-specific differentiation. MSCs naturally affect the regenerative and immunomodulatory activities at in vivo sites of injury. The latter functionality stimulated the name change to medicinal signaling cells and the important distinction that MSCs are not stem cells. This commentary emphasizes these differences.

Impact Statement

The impact should encourage continued research and clinical trials using mesenchymal stem cells.

I first used the name mesenchymal stem cells (MSCs) for bone marrow cells that we isolated and culture expanded.¹ The fact that we could drive these cells in culture into a number of mesenchymal phenotypes not only seemed to justify the use of the term “stem cells,” but more importantly it stimulated us to start using these cells to tissue engineer autologous skeletal tissues that could be used to replace or augment injured or diseased tissue. This stimulated me and my colleagues to start Osiris Therapeutics, Inc. in 1992 as a bio-orthopedic tissue engineering company. Indeed, to this day, my laboratory is still actively involved in trying to use MSCs to tissue engineer articular cartilage.

I write this short position article in response to the Sipp et al. comment ² since the MSC technology and my research were mentioned several times. I fully agree with their introductory thesis that clinics and practitioners should stop marketing “stem cell and regenerative medicine therapies” especially using MSC logics and nomenclature. Indeed, in 2010, I asked that we in the cell therapy industry change the name of MSCs that I invented in my 1991 article¹ to medicinal signaling cells³ because in the body, most MSCs do not differentiate and they only do this in mass in a petri dish or in culture where those in vitro conditions^4–7 induce huge mesenchymal plasticity. To further emphasize the derepressive nature of cell culture conditions, others have published that MSCs can be converted in culture to neuronal phenotypes,⁸ which certainly never happens in the body.

The extremes of Sipp et al.'s proposal involving “omics approaches” are completely unacceptable to the experimentalist who work with MSCs since they want us to “omics” almost every cell in the MSC preparations to ensure that we have a “pure” and rigorously characterized preparation. This suggestion completely misses the key issue that current MSC preparations, even the very impure and variable stromal vascular fraction from adipose tissue, have been reported to have substantial therapeutic potency in both humans and large animals.^9,10 The objective of the multitude of publications on MSCs is to better characterize these cells and test their therapeutic potential (as is the case in any subdiscipline, some publications are more rigorous than others, but nonetheless contributory to the MSC information base). It is clear that all of these therapeutic uses are to utilize the paracrine activities of MSCs not their ability to differentiate in culture or to be used in a tissue-engineered context.

One must understand that the initial Food and Drug Administration (FDA)-monitored high-level clinical trials using any new technology and especially using MSCs have not failed, quite the opposite. These trials, although missing in their statistical endpoints, provided detailed information with regard to the strong paracrine therapeutic potential of MSCs. Currently, one could say that the 500 patient graft-versus-host disease (GvHD) trial conducted by Osiris Therapeutics failed to make its endpoints.^11–14 Careful analysis of their data, however, reveals that there was an overaggressive recruitment of GvHD patients with an abundance of skin GvHD patients whose positive response to steroid treatment pushed the control group out of the normal placebo range. Indeed, the liver GvHD and GI GvHD patients exhibited very high and impressive therapeutic efficacy effects of the administered MSCs. In this regard, Mesoblast, which now owns the Osiris technology and data, has completed a 60-patient single-site phase 3 trial whose results should help Mesoblast gain FDA approval MSCs for the GvHD indications of that trial.

My view is to not throw away the last 25 years of MSC data (many thousands of publications) and clinical experience as suggested by Sipp et al. because I called these unique cells stem cells without equating them with embryonic or tissue-specific stem cells. The error was to assume that what I could do in culture in terms of differentiating these cells to bone, fat, cartilage, muscle, hematopoietic support cells, etc. was what happened in the body; MSCs are highly plastic multipotent stem cells in culture but that is not what is observed in the body.¹⁵ The fact that we are still trying to use the differentiation potential of MSCs into skeletal phenotypes for in vitro tissue engineering purposes only indicates the high degree of difficulty of such a use. In contrast, we now have rigorously organized high-level clinical trials that use these potent paracrine cells that function at injury and disease sites; such clinical trials¹⁶ were unimagined in the late 1980s when this study began in my laboratory. In simplistic terms, luckily we did not throw out the baby or the bath water since both are quite useful for medicinal purposes.

My plea is that this nonsense of characterizing every cell in every MSC preparation will get us nowhere simply because most cell preparations have become culture adapted and the resulting cells reflect neither their previous innate in vivo tissue characteristics nor their in vivo therapeutic functionality when reintroduced into patients. Relative to the MSCs in culture, we are very carefully analyzing the transcript complexity of the in vitro differentiation of hMSCs into the chondrogenic phenotype and publishing our insights from this analysis that clearly identifies the temporal activities of these differentiating cells.^17,18 Furthermore, although some want this pellet culture to be a model for articular cartilage differentiation, the cartilage formed in pellets is molecularly different from that of in vivo articular cartilage.¹⁸

Human marrow MSCs and their in vivo and in vitro characteristics including their functional heterogeneity are the subject of a recent poster for Nature Protocols.¹⁹ We and others continue to analyze MSCs by rigorous and sensible criteria without all of the negativity espoused by Sipp et al.² We are not using false marketing, nor false claims, as are some practitioners or clinics, nor are we exhausting our meager resources to omics every MSC preparation. This request is as useful as other suggestions to clone the efficacious cell in the MSC preparation to ensure good results in clinical trials. MSCs are not stem cells; that said, I would not now retract any of my publications because the data are real, they are reproducible and they are contributory to the professional acquisition of scientific information and biologic insight into these therapeutically active cells with profound immunomodulatory and trophic paracrine capacity. The mess is that there are too many suggestions of how to provide “rigorous” science with no clue as to how to do it. If MSCs can stop GvHD, help protect hearts from the negative effects of a heart attack,²⁰ and can help people with low-back pain (Mesoblast Phase 2 clinical trial), my view is to use them. We scientists will eventually figure out how they do this and companies and their staffs will make them work better, more inexpensively and we will together design better isolation, cultivation, and patient presentation protocols. Do not mess with MSCs but use them wisely. An overedited version of this thesis has been published as a Correspondence.²¹

Disclosure Statement

No competing financial interests exist.