Statistics from Altmetric.com
The clinical need
Liver disease is an increasing clinical burden, causing over 10 000 deaths last year in the UK alone (http://www.britishlivertrust.org.uk). Liver insufficiency describes the clinical situation in which cumulative (chronic) or one-off massive (acute) insults exceed the liver's normal physiological capacity to functionally regenerate. Untreated, liver insufficiency invariably leads to death. The current gold standard of care in this setting is whole organ transplantation. Due to the increasing burden of disease within the population, however, the number of patients requiring transplantation far exceeds the number of available donor organs. As a result, many patients with liver insufficiency die prematurely.1
Liver cell therapy-intra vs. extra hepatic
The liver is composed of several cell types (endothelial cells, stellate cells, biliary ductal cells, Kupffer cells and natural killer cells) which together provide a supportive niche for the principle cell type—the hepatocyte.2 Treatment of liver disease by hepatocyte replacement therefore appears a logical alternative to whole organ transplant (figure 1i). Along these lines, intrahepatic hepatocyte transplantation (I-HTx) has repeatedly proven efficacious in small animal models of numerous liver diseases. Results from human applications have unfortunately proven less convincing however with the best, possibly only, positive results observed in paediatric patients suffering from hepatocyte driven inherited metabolic disorders.3 This suggests targeting the surrounding niche may be as important as replacing diseased hepatocytes themselves if one is to treat the majority of liver diseases.4 Such efforts to manipulate the complex nature of the surrounding niche will no doubt be far from trivial. So an alternative, more tangible approach that targets more specific clinical problem could be to use hepatocytes in extra-hepatic anatomical sites to ‘bridge’ patients with liver insufficiency to transplant or self repair (E-HTx).5 Some preliminary clinical experiences suggest a therapeutic effect for alginate-bead encapsulated hepatocytes delivered into the peritoneum of paediatric patients with acute liver failure can be achieved in this manner for example (Dhawan et al, unpublished data). These observations are highly encouraging but are still undergoing further validation. Key to expanding upon this success is an understanding of whether E-HTx is effective by a direct (transplanted cells working as an extra-hepatic organ) or indirect (transplanted cells secreting factors capable of augmenting native liver recovery) mechanism.6 Understanding the mechanism more precisely may in turn allow greater numbers of patients to be more effectively treated by E-HTX based approaches.
A stem cell based approach for the future
Most liver specialists would agree that development of both intrahepatic and extra-hepatic HTx programmes are potentially very attractive but ultimately limited by shortage of donor hepatocytes. Similarly, investigating the mechanistic basis underlying E-HTx, though necessary, is similarly stunted by scarcity of reproducibly matched primary tissue. Alternative, more abundant sources of hepatocytes, such as those derived from stem cells, could address both these challenges. The key questions for the field are (1) how do we generate those cells? and (2) how will we know if we have made the right cell?
Though stem cells are proposed to exist in all tissues, evidence for their existence in adult human livers is controversial.7 To date, no one has successfully captured bona fide liver stem cells capable of clinically relevant expansion making this starting population, however ideal, currently unviable.
On the other hand, the existence of embryonic stem cells (ESCs) is not in doubt.8 They are derived from the inner cell mass of preimplantation embryos and can be retransplanted into the same locus as a functional demonstration of their defining properties. Furthermore, ESCs are pluripotent—that is, they are highly expandable and able to form any cell type of the body, unlike tissue/organ stem cells which are more restricted in their differentiation and expansion potential. The derivation of ESCs is limited in turn by the paucity of available donor embryos and certain ethical constraints. For these reasons, the authors do not see an easy path to clinical application using ESCs. In 2006, however, the stem cell field was presented with an exciting new alternative. Shinya Yamanka, a stem cell biologist from Japan (originally trained as an orthopaedic surgeon), discovered a second type of pluripotent stem cell derived not from embryos but from adult cells.9 He called it the induced pluripotent stem cell (iPSC). This cell type can actually be generated from any adult cell type of the human body but has been most often derived from skin fibroblasts because of their practical ease of procurement. The technique to make iPSCs is simple, involving transient overexpression of just a few critical transcription factors. In this way, the starting cell is epigenetically reprogrammed back into an ESC-like state. Due to their unlimited capacity for expansion, if the correct cocktail to drive iPSCs into hepatocytes can be identified, it offers the potential for generating limitless numbers. On top of addressing the hepatocyte supply problem, this new technology also offers the possibility of immune-suppression-free transplantation since the starting material is isolated from the intended transplant recipient.10 The technique of reprogramming has proven highly reproducible across hundreds of laboratories since its discovery adding further weight to the possibility of downstream clinical applications soon becoming a reality.
Stem cells—reprogramming and conversion to hepatocytes (r-Heps)
Reprogramming donor cells to a pluripotent state is now relatively easy. What is significantly more challenging on the other hand is their conversion into hepatocytes. The strategy employed for hepatocyte conversion by most groups so far has been to try and replicate, using standard two-dimensional (2D) tissue culture techniques in vitro, the sequential steps of mammalian development observed in utero.11 In a second, related but more recent approach, several groups have reported a more direct strategy of conversion by reprogramming donor cells into hepatic or endoderm precursors before enforcing the final steps of differentiation in the same manner as the first wave of studies.12 ,13 Flying more direct is no doubt faster and intuitively sounds more sensible; however, such studies still need to be replicated. Of perhaps greater interest is that some of the phenotypic characterisations reported in this second wave of studies suggested the more direct method could yield better ‘quality’ cells. At this stage having the right quality of cell is paramount to translational advancement of the field (discussed below).
Assuming reprogramming techniques continue to be further optimised, we envisage billions of hepatocytes will soon be available in clinical grade conditions from such experiments relatively cheaply and quickly (figure 1iv). The next challenge will be to implement quality control measures, that is, how do we know we have the right cell coming off the production line? Though several key characteristics of primary hepatocytes are indeed recapitulated by reprogrammed hepatocytes (r-Heps), justified concerns remain about the ‘maturity’ and ‘safety’ of these cells. The field seems to be in agreement that r-Heps are ‘immature’—that is, they are essentially ‘correct’ but stunted with respect to complete recitation of their developmental programme. To achieve full maturity may therefore require current blocks to be overcome through further manipulation of cell intrinsic, cell extrinsic or both types of signalling pathways. Many investigators are currently focusing their efforts on these investigations to try and make ‘mature’ hepatocytes in vitro with some success.14 ,15 It is important to concurrently consider alternative explanations and solutions, however. The inference that r-Heps are ‘immature’ for example is deduced from the fact that maturity has so far almost universally been defined as the functional readout pertaining to two essential criteria (i) in vitro drug metabolism and (ii) in vivo reconstitution of specific rodent models of liver injury (FRG/u-PA/TK-NOG). In vitro generated cells are therefore considered ‘mature’ if they perform in a manner analogous to that observed with primary hepatocytes. Human primary hepatocytes are in turn isolated from livers using standard techniques first described over 50 years ago and largely unchanged since.16 What has not been investigated thoroughly however is whether the processes used for isolating primary hepatocytes select out a particular subpopulation of cells.17 ,18 If this were true, it could be that r-Heps currently available through in vitro reprogramming methods are indeed of the ‘right’ quality now but functionally different to isolated primary-Heps. Clarifying this is essential. What is even more important is working out which type of hepatocyte would achieve therapeutic efficacy in which human disease context. For this, we either need to increase the breadth of preclinical (animal) models available for assessment or take what we have now and press forward with human trials. While pursuing continued investigations and improvements in cell production will no doubt eventually yield benefit, it is our opinion that the time has now come to start testing efficacy of currently available cells in man. However, we must bear in mind an important caveat to this pursuit—that the genetic/epigenetic stability of such cells still remains unknown.19 Therefore, to use reprogrammed cells in humans where their anatomical distribution is unselected would we feel be too dangerous at this time. Instead, we would propose initial studies where cells can be safely trapped within removable devices to ensure cells cannot spread into the body and form tumours. This requirement naturally lends itself to the extra-hepatic as opposed to intrahepatic therapeutic approach. It also pushes us to consider how the multi-dimensional, multi-cellular arrangement of hepatocytes within such structures might affect their functionality and whether the next iterations of liver cell therapies will need to encompass such biological designs in order to achieve maximal functionality.
As discussed, current protocols for hepatocyte conversion, focusing on uni-cellular differentiation, most likely produce somewhat heterogeneous populations of cells with suboptimal levels of functional activity. A key missing concept in these past approaches has been the idea of inter-cellular communications, spatiotemporally regulated throughout organogenesis. The spatial regulation occurs in 3D and encompasses critical interactions with supporting cell types such as mesenchymal and vascular cells. The precise timing of these cellular interactions on the other hand adds another dimension of complexity. Cumulatively, this demonstrates natural liver development is acquired through 4D multicellular communications. There is currently no stem cell culture system that allows for such sequential exposure to developmental cues within a 4D context.
To address this challenge, a novel approach was recently described to produce ‘organ buds’ in culture by recapitulating early organogenic multi-cellular interactions. Specified human iPSC-derived hepatic cells are self-condensed to form a 3D organoid by coculturing with supporting stromal cell populations—mesenchymal and endothelial progenitors.20 This condensed tissue formed after multiple progenitors were allowed to interact in a 4D manner, and subsequently self-organised into a liver bud-like structure (or miniature liver). Impressively, following transplantation of the bud in vivo, vascularisation stimulated hepatocyte maturation in the absence of any specific environmental niche (such as severe liver damage). The matured liver constructs were then able to considerably improve survival of mice with chemically induced liver failure following mesenteric transplantation. The iPSC-derived liver bud approach therefore seems to offer real therapeutic potential as an extra-hepatic liver support system. Although unanswered questions still remain such as whether a direct or indirect mechanism underlies the observed therapeutic effect, this new bud therapy appears to be a readily translatable approach for eventually treating liver insufficiency (figure 1v). It also further underlies the hypothesis that complex biological interactions between key cellular players may need to be reconstructed in order to achieve the most meaningful therapeutic outcomes.
Hepatocytes as part of a whole organ
While ‘single’, ‘multiple’ or ‘bud’ cell therapies could certainly make a huge impact for a significant number of people suffering with liver disease, they will not be able to treat the majority of clinical conditions for which a whole organ replacement is required. To address that challenge, we would have to realise the ‘holy grail’ of stem cell biology—synthesis of an entirely new organ in the laboratory. A number of reports recently suggested laboratory engineered scaffolds reseeded with stem cell-derived products may be a profitable avenue of research in this regard.21 Replicating this approach on the scale and in the complexity needed for humans has proven too difficult so far and given the previously mentioned challenges of generating functional hepatocytes in vitro does not appear an easily surmountable hurdle.
In the interim, a hybrid approach, using large animal ‘living scaffolds’ to nurture conversion of hepatocytes from iPSCs, could provide a more effective solution since organs such as the liver and pancreas in these large animals (sheep and pig) share common physiology with their human counterparts.22 To do this, one would have to find a way of seeding human iPSCs/liver buds into a suitable animal recipient long enough to allow satisfactory 4D organogenesis. Along these lines, it was recently shown that compensating a genetically engineered organ deficient animal with donor stem cells at the preimplantation embryo stage could be used to generate functional pancreas in both rodents23 and pigs.24 What is not clear at this point is whether donor cells taken from higher order species (primate and human) could similarly recapitulate organ reconstitution. If such cross species complementation of the conceptus is possible, we could exploit this remarkable quirk of developmental biology to grow human livers in pigs or sheep by the thousand. Indeed, if proven feasible, we could one day be talking of housing animal ‘organ bioreactors’ in hospital basements as a solution to our shortage of donor organs (figure 1vi). Perhaps this notion is a little beyond scientific reality for now, but even if the conceptus complementation approach does not lead to an instantly useable end product, the strategy will undoubtedly prove useful as a supplement to ongoing efforts for generating appropriately functional stem cell-derived therapies in the laboratory.
Contributors All authors made substantial contributions to the conception and design of the work, and approved of the final version.
Competing interests None.
Provenance and peer review Commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.