Stem Cell Transfer and the Uterus: Title and subTitle BreakThe Egg Teaches the Chicken

Early humans associated the generative life force with the "woman-mother" in the form of large-breasted, round-bellied figures of veneration. One of the oldest Paleolithic female goddess statues, the Venus of Willendorf, dating from 22 000 BC, portrays a fecund, likely pregnant, woman.¹ The Ebers Papyrus (c 1550 BC), one of the earliest Egyptian anatomic records, clearly describes the vagina and uterus.¹ Given that both the early Egyptians and Greeks based their anatomic descriptions on animal dissections—human dissection was forbidden—the external genitalia, vagina, and cervix were faithfully described. However, descriptions of the uterus remained highly imaginative.

The uterus was understood to be important for childbirth but its position was uncertain and primarily dictated by fantasy. For example, the Pythagoreans believed the uterus was bifid, whereas Aristotle thought it was made up of 7 cells: 3 to the left, 3 to the right, and 1 on top.² The followers of Hippocrates believed the uterus could wander throughout the body, a belief strengthened by the observation of uterine prolapse and that its displacement resulted in the symptom of hysteria—a word applied originally to afflictions of the womb.²

The Roman physician Soranus of Ephesus (AD 98-138) apparently was the first to provide an accurate description of the uterus, as a result of cadaver dissections. Soranus accurately described the position of the uterus and its relationship to the bladder and also distinguished between the muscular myometrium and inner, lush endometrium, which he clearly understood was related to fetal implantation and menstruation.³ In the 16th century, again as the result of cadaver dissections, the anatomy of the pregnant uterus, fetus, and membranes was finally accurately and realistically rendered by Leonardo da Vinci.⁴ Thus, by 1513, both physicians and artists understood the anatomy and cyclical, generative function of the uterus. For the next 500 years, common wisdom viewed the uterus as an incubator supporting the gestation and birth of progeny, with the uteroplacental unit allowing the exchange of nutrients and gases.

During the past decade, several studies have provided data that have revolutionized this view by demonstrating that the uteroplacental unit can serve as a conduit for fetal cells to enter and persist in the mother's body, as well as allow maternal cells to become incorporated into offspring.^{5 - 7} The reports by Khosrotehrani et al⁸ and Taylor⁹ in this issue of JAMA further reveal that the uterus is, respectively, a dynamic organ permeable to fetal stem cells capable of transdifferentiation and an end organ in which adult bone marrow stem cells may find a home and differentiate. Thus, the uterus can now be thought of as a bidirectional conduit for stem cell transfer.

Various animal and in vitro models of end organ tissue regeneration by adult bone marrow stem cells, including pluripotent hematopoietic stem cells (HSCs), have been proposed and demonstrated.^{10 - 12} However, organ regeneration by bone marrow stem cells in humans has as yet been confirmed only in the liver and brain.^{13 - 15} The report by Taylor provides new evidence of endometrial regeneration by bone marrow stem cells in endometrial tissues from 4 women who received single-HLA antigen mismatched bone marrow transplant for leukemia.⁹ Donor-derived endometrial stromal and epithelial cells were identified by reverse-transcription polymerase chain reaction and immunostaining of the mismatched HLA antigen, with donor-derived endometrial cells composing 0.3% to 50% of cells in the biopsied tissues.⁹ The endometrial biopsy with the highest proportion of donor-derived cells was from a woman who underwent dual-agent chemotherapy and total body irradiation, thereby raising the question of whether the degree of donor-derived stem cell regeneration is related to the extent of organ damage.

These findings are important because they suggest a biological explanation for the clinical observation of endometrial regeneration after both endometrial ablation and treatment for Asherman syndrome. In addition, endometrial regeneration by bone marrow stem cells may have therapeutic implications in fertility treatment for cancer survivors who have received radiation or chemotherapy. However, the most important therapeutic potential of this work likely rests on the possibility that some forms of endometriosis may arise from differentiation of bone marrow stem cells into endometrial glands and stroma at ectopic sites, perhaps in response to certain microenvironmental cues. Since endometriosis is a disease that affects an estimated 5 million women of reproductive age in the United States and can result in debilitating pain, multiple operations, and infertility,¹⁶ testing this hypothesis in animal bone marrow transplantation models is a priority.

Taylor's findings will also serve to motivate reproductive biologists as well as stem cell scientists to definitively address whether bone marrow stem cells turn into endometrial cells by fusing with preexisting, differentiated endometrial cells, ie, cell fusion. Alternatively, bone marrow stem cells with developmental plasticity may differentiate into endometrial cells by transdifferentiation, which involves a complete reprogramming of gene expression to effect specialized cellular functions. Both cell fusion and stem cell transdifferentiation can occur, at least in the liver in animal models, and the operative mechanism appears to depend on the type of stem cells studied and microenvironmental cues of tissue injury.^{17 - 18} It is especially encouraging that evidence for neuron regeneration by transdifferentiation of donor-derived bone marrow stem cells in humans has also been reported.¹⁵ Thus, whereas therapeutic brain regeneration requires migration and transdifferentiation of stem cells into neurons, potential endometriosis treatment would require inhibition of these processes. Both processes share the same need to understand relevant subpopulations of HSCs and the mechanisms regulating their mobilization, homing, and differentiation.

Just as bone marrow transplantation permits the recipient to become a chimera with all the attendant potential benefits and risks, pregnancy permits at least some, and possibly most or all, mothers to become chimeras of their children. Previous work by Bianchi et al⁵ and others^{6 - 7} has demonstrated that pregnancy-associated progenitor cells (PAPCs) or fetal cells of lymphocytic or leukocytic phenotype can persist in the maternal circulation and maternal tissues. The potential immunologic functions have causally implicated PAPCs in maternal autoimmune diseases with a proposed graft-vs-host response mechanism. The complex roles of these PAPCs continue to unfold as Khosrotehrani et al report strong morphologic and immunohistochemical evidence for the plasticity of these fetal cells to differentiate into specialized cell types, including hepatocytes as well as epithelial cells in thyroid, cervix, intestine, and gallbladder.⁸ Furthermore, the presence of single X and Y chromosomes demonstrated by fluorescence in situ hybridization strongly suggests that these cells undergo transdifferentiation rather than cell fusion. Thus, their plasticity, multilineage differentiation potential, and their previously described CD34⁺ phenotype qualify PAPCs as multipotent stem cells of fetal origin that contribute to maternal tissue formation through their transplacental journey into the mother's body.

The most interesting finding by Khosrotehrani et al is that hepatocytes of fetal stem cell origin were identified in liver tissues of one woman with liver injury and another woman following hepatic transplantation.⁸ Although the functional contribution of these fetal-derived hepatocytes remains to be determined, the possibility that newly implanted or persistent fetal stem cells may promote tissue regeneration in maternal disease states is novel and exciting. Originally only viewed as possible culprits of maternal autoimmune disease, PAPCs have now emerged as fetal stem cells with potential therapeutic relevance not only for the mothers who harbor them, but possibly also for first-degree family members or even unrelated individuals.

This observation of maternal organ chimerism raises a multitude of questions. Are all mothers chimeras, and of both daughters and sons? Do organ "embedded" PAPCs play a role in pregnancy-related diseases such as preeclampsia? Could these cells be stimulated to fight maternal disease? If their therapeutic potential is realized, and their isolation from women with current or past pregnancies is feasible, the use of PAPCs in stem cell therapy has significant advantages in addition to a high degree of HLA matches with the biological parents and siblings. For example, the type of long-term planning and expense associated with umbilical cord blood banking would not be required. With further understanding of the physiologic controls that alternately suppress and stimulate PAPC transfer, homing, and differentiation, the time may soon come when the prenatal child heals the mother and perhaps in the far distant future becomes the ultimate health insurance for the whole family.