Supplementary material for this article can be found on the H EPATOLOGY Web site ( ).

jws-hep.21745.fig1.pdf58K Supplemental Fig. 1Expression of human hepatocyte markers in livers of immunodeficient mice after transplantation of extrahepatic stem cells. Representative figures from published studies are shown. (A) Danet et al. (36): transplantation of purified human Lin(-) CD45(+)CD38(-)CD34(+or-)C1qR(p)(+) cells isolated from human umbilical cord blood. Livers from transplanted mice were analyzed 8?10 weeks after transplantation and immunostained for human albumin. Bar: 10 μm. (B) Newsome et al. (25): transplantation of unsorted mononuclear cell preparations of human cord blood and immunostaining for HepPar1. Bar: 20 μm. (C) Wang et al. (37): transplantation of CD34+ cells from human umbilical cord blood and immunostaining for human albumin. Magnification: 100-fold. (D) Kollet et al. (38): transplantation of CD34+ cells isolated from human cord blood. Immunostaining for human albumin. Magnification: 100-fold. (E) Kakinuma et al. (39): transplantation of adherently proliferating cells isolated from human cord blood. Immunostaining for HepPar1. Bar: 10 μm. (F) Von Mach et al. (13): transplantation of nestin-positive islet-derived adherently proliferating cells. Immunostaining for human albumin. Magnification: 400-fold. (G) Weber et al. (40): transplantation of human hepatoblasts into rag2gammac?/? mice. Immunostaining for human glutathione-S-transferase. (H, I) Aurich et al. (28): transplantation of hBM-MSC into the liver of Pfp/Rag2-/- mice by intrasplenic application. Immonostaining for expression of human albumin (H) and HepPar1 (I). Magnification: 400-fold. (J) Sharma et al. (27): transplantation of unsorted human cord blood cells. Immunostaining for albumin. Magnification: 40-fold.
jws-hep.21745.fig2.pdf511K Supplemental Fig. 2Characterization of adherently proliferating cord blood cells by surface marker analysis (A). Morphology of adherently proliferating cord blood cells (B), nestin-positive hepatopancreatic precursor cells (C) and human blood monocyte derived NeoHep cells (D) cultured in flasks. Scale bar in B ? D: 100μm.
jws-hep.21745.fig3.pdf812K Supplemental Fig. 3Adherently proliferating cord blood cells were cultured on slides and immunostained using antibodies specific for human albumin (A), iron deposition with Prussian blue staining (B), glycogen deposition with PAS-staining (C) and endothelial staining using solanumituberosumlectin (D). Prior to transplantation the cells were all negative for the analysed markers (A ? D). The scale bar is 50μm in all figures.
jws-hep.21745.fig4.pdf74K Supplemental Fig. 4Combined immunostaining for human albumin expressing cells andin situhybridization with alu-probes to identify human nuclei after transplantation of primary human hepatocytes into NOD/SCID mice. A DAB-positive human albumin expressing hepatocyte without hemalum counterstaining is shown in A. The same cell shows an alu-positive nucleus (B). An overlay of DAPI-stained nuclei with a brightfield picture of the DAB-positive cell (black in C) shows co-localization. After removing the coverslip (causing slight alterations in tissue structure) the slice was counterstained with Mayer?s hem alum (D). The scale bar in A ? D is 20μm.
jws-hep.21745.fig5.pdf271K Supplemental Fig. 5Amplification profiles of 16 STR loci for a human cord blood cell line (A) and for NOD/SCID mouse liver tissue three weeks after transplantation with cells from the same cell line (B). Identical profiles were obtained proofing that human DNA in A and B originates from the same individual.
jws-hep.21745.fig6.pdf77K Supplemental Fig. 6Immunostaining for human albumin in order to control the specificity of the applied method. To rule out unspecific binding of the secondary antibody, human liver (A), sham-transplanted NOD/SCID mouse liver (B) and serial slices of NOD/SCID mouse livers transplanted with human cord blood cells (C) were compared using either primary and secondary antibodies (A1,B1,C1) or the primary antibody has been omitted (A2,B2,C2; negative controls). Besides incubation with the first antibody all slices have been treated identically. Scalebar: 50μm, except for C1a and C2a: 100μm.
jws-hep.21745.fig7.pdf76K Supplemental Fig. 7Colocalization of a PAS-positive cell cluster (D) withalu-positive nuclei and CM-DiI cell tracker (E). A mucinous adenocarcinoma served as positive control (A). NOD/SCID-mouse liver (B) and human liver (C) show a weakly positive PAS-staining. 21d after transplantation of CM-DiI-labelled cord blood cells PAS-positive cell clusters were observed in the left liver lobe of NOD/SCID mice (D) that correspond to cells withalu-positive human nuclei (E) identified byin situhybridization performed prior to PAS-staining on the same slice (yellow arrows). Besides glycogen andalu-positive nuclei red fluorescence from CM-DiI was observed in the corresponding region (E). The scale bar in A ? E is 50μm.
jws-hep.21745.fig8.pdf1903K Supplemental Fig. 8Type 1 cells do not express CYP3A4. Slices were stained with an antibody directed against human CP3A4 that also recognizes mouse CYP3A (A1-2,B1-2,C1-2,D1). Prior to immunostaining for CYP3A the same slices have been stained byin situhybridization usingalu-probes in order to detect human nuclei (A3,B3,C3,D3). Human nuclei appear green and are indicated with white arrows (C3, D3). Human and mouse hepatocytes stained positive using the anti-human-CYP3A4 antibody (yellow arrowheads). The close-up view (A2 and B2) demonstrates that endothelial cells are negative for CYP3A4 (black asterisks). Using this technique thealu-positive transplanted cells appear negative for CYP3A4 (black arrows in C3 and D1) and show a similar staining intensity as the endothlial cells (asteriks in A2 and B2) that can be considered as background. Scalebar in A1-3, B1-3 and C1-3: 50μm, D1 and D3: 20μm.
jws-hep.21745.fig9.pdf73K Supplemental Fig. 9Slices of livers from NOD/SCID-mice transplanted with 750.000 cord blood cells. A1 and A2 are controls prior to immunostaining. B1 and B2 were dewaxed and counterstained with Mayer?s hemalum. Fotographs before and after the counterstainig were taken from the same location. The scale bar is 100μm in A1and A2, as well as 50μm in B1 and B2.
jws-hep.21745.fig10.pdf123K Supplemental Fig. 10Detection of iron containing cells using Prussian blue staining. Slices of human hereditary hemochromatosis liver served as a positive control, dark blue cells indicate accumulation of iron (A). Normal human liver (B) and NOD/SCID mouse liver stained negative for Prussian blue (C). 21d after transplantation of CM-DiI labelled cord blood cells iron containing cells (D) were observed in the same cell cluster where CM-DiI fluorescence (red) andalu-positive nuclei (green, yellow arrows) have been identified prior to Prussian blue-staining (E) on the same slice. Some of thealu-positive nuclei showed a morphology similar to endothelial cells (G and closeup in H). In a subsequent staining for iron these cells were negative (F). The scale bar in A ? H is 50μm.
jws-hep.21745.fig11.pdf100K Supplemental Fig. 11Co-staining of endothelial cells using solanumituberosum(potato) lectin (red fluorescence) andalu-positive human nuclei (green fluorescence). Solanumituberosumlectin stained endothelial cells from human (A) and NOD/SCID mouse liver (B). Endothelial cells from veins are indicated by the green arrows in A and B, whereas sinusoidal lining cells are indicated by white arrows in A and B. 21d after transplantation of cord blood cellsalu-positive nuclei (as indicated by yellow arrows in C1 ? C4) were detected prior to solanumituberosumstaining (C2). Figure C3 shows that the cell with analu-positive nucleus stains negative for solanumituberosumlectin. C1 (fluorescence microscopy) and C4 (confocal microscopy) are digital overlays of C2 and C3. The scale bar in all images is 50μm.
jws-hep.21745.fig12.pdf97K Supplemental Fig. 12Analysis of apoptosis after transplantation of cord blood cells into NOD/SCID mice. Apoptotic cells were identified by the TUNEL assay and visualized by red fluorescence. Human nuclei were visualized by green fluorescence usingalu-probes. Human liver (A1,A2) and sham transplanted mouse liver (B1,B2) served as positive controls after treatment with DNase (A1,B1) or as negative controls without DNase treatment (A2,B2). Prior to DNase treatment both slices were hybridized usingalu-probes to identify human nuclei. As expected only the histological control of the human liver (A2) showed co-localization of DAPI (blue) andalu-probes (green, Cy2) as indicated by turquoise colored nuclei (A2). One day after transplantation of cord blood cells clusters of TUNEL positive cells neighbouring vital cells could be observed (C). Using a larger magnificationalu-positive but TUNEL-negative nuclei could be observed (D1 and close-up view of green fluorescent nuclei in D2, arrows) neighbouring apoptotic cells (arrowheads). 21d after transplantation of cord blood cellsalu-positive and TUNEL-negative nuclei (E1 and close-up view of green fluorescent nuclei in E2, arrows) could be identified. In contrast to the liver 1d after transplantation (D) much lower numbers of TUNEL positive cells were observed in the neighbourhood of thealu-positive nuclei. Red fluorescence located in the cytoplasm (asteriks) is a consequence of CM-DiI labelling and can be clearly differentiated from TUNEL-staining since the latter shows a nuclear and CM-DiI a cytoplasmic localization. Due to degradation of DNA, DAPI staining and hybridization withalu-probes is much weaker in TUNEL-positive nuclei. Scalebar in A1, A2, B1, B2, E1, E2 50μm, in D1 and D2 20μm, in C 200μm.
jws-hep.21745.fig13.pdf136K Supplemental Fig. 13Ki67 immunostaining to identify proliferating cells (red fluorescent nuclei).in situhybridization withalu-probes to identify human nuclei (green fluorescence) was performed prior to immunostaining. Human tonsil served as positive control (overlay A1, DAPI andalu-positive nuclei A2, DAPI and Ki67-positive nuclei A3). Proliferating cells are indicated by yellow arrows. Human liver showedalu-positive nuclei, but was negative for Ki67 (B), NOD/SCID-liver was negative for both (C). 21d after transplantation of CM-DiI labelled (red fluorescent dye) cord blood cells,alu-positive but Ki67-negative nuclei were identified in the transplanted liver (D1 and D2, white arrows). Scalebar in A and D2 is 20μm, in B, C, D1 50μm.
jws-hep.21745.fig14.pdf71K Supplemental Fig. 14Immunohistochemical detection of human albumin (brownish color, A1, A2) andin situhybridization using mouse major satellite probe for identification of murine nuclei (red fluorescence, B1, B2) as well asalu-probes for identification of human nuclei (green fluorescence, C1, C2) 21d after transplantation of 750.000 cord blood cells. Combined staining of the same slices demonstrates that human albumin positive type II-cells have a mouse nucleus (yellow arrows indicate mouse major satellite positive nuclei [B1, B2] in human albumin positive cells [A1, A2]). The same cells were negative after staining withalu-probes (C1, C2). Scale bar: 50μm.
jws-hep.21745.fig15.pdf39K Supplemental Fig. 15 in situhybridization usingalu-probes for identification of human nuclei in mouse liver tissue 2d after transplantation of 750.000 NeoHep cells into the parenchyma of the left liver lobe (A, close-up view in B and C). White arrows indicate nuclei with human specificalu-sequences (green fluorescence) in A ? C. The yellow arrow indicates a chromatin fragment of human origin ( alu-positive) within a murine cell. Original magnification in all pictures is 200-fold, the scale bar in A is 50μm.
jws-hep.21745.fig16.pdf92K Supplemental Fig. 16Adherent proliferating cordblood cells 21d after transplantation into the left liver lobe of immunodeficient NOD/SCID-mice. Cells were labelled with CM-DiI (red fluorescent dye) prior to transplantation. Human nuclei were detected usingalu-probes with a green fluorescent detection system (Cy2), whereas murine nuclei were detected using mouse-major-satellite-probes and a red fluorescent detection system (Cy3). A: Red fluorescence in the cytoplasm of a mouse-major-satellite positive nucleus. The red fluorescence is due to CM-DiI uptake from deteriorating transplanted human stem cells. The close-up view in B shows that there is noalu-positive human nucleus, but a red fluorescent cytoplasm (C) and a mouse-major-satellite positive nucleus (C). Scalebar: 20μm.
jws-hep.21745.fig17.pdf91K Supplemental Fig. 17Liver damage following injecition of cell suspensions directly into the parenchyma of the left liver lobe at several time points of transplantation. In both pictures a pale colored region indicates damaged liver tissue. The slice in A has been stained for glycogen (PAS), whereas the slice in B has been stained for iron (Prussian blue). In C human nuclei are identified using in situ hybridization withalu-probes in a region of disrupted liver tissue. The liver lobe has been removed 5d (A and B) and 2d (C) after transplantation. Scale bar: 100μm (A, B) and 50μm (C).
jws-hep.21745.fig18.pdf115K Supplemental Fig. 18Immunohistochemical staining for cells expressing human albumin 21d after transplantation. Comparison of two different routes of administration. 750.000 adherent cord blood cells were injected either into the spleen (A, B) or directly into the parenchyma of the left liver lobe (C, D). uPA/RAG mice were used that suffer from permanent deterioratiion of hepatocytes. This mouse model gives a selection advantage to wild-type cells. The scale bar in all pictures is 20μm.
jws-hep.21745.tbl1.pdf30K Supplementary Table 1:Transplantation of human stem and precursor cells into livers of experimental animals: summary of published studies
jws-hep.21745.tbl2.pdf11K Supplementary Table 2:Immunohistochemically identified human albumin positive structures 21d after transplantation of 750,000 cord blood cells directly into the left liver lobe of immunodeficient mice
jws-hep.21745.tbl3.pdf7K Supplementary Table 3:STR typing results of the human cell line CCB6 and mouse liver tissue after transplantation with human cell line for 16 forensic STR markers, according to Fig. 3A and 3B.
jws-hep.21745.tbl4.pdf8K Supplementary Table 4:Fraction of human DNA in livers of NOD/SCID mice 4h, 12h, 24h, 5d and 21d after transplantation of 750,000 cord blood cells directly into the left liver lobe. Quantification was performed using a duplex-PCR (21d estimation using human-specific primers only) according to a technique described by Pelz et al. (17)
jws-hep.21745.tbl5.pdf8K Supplementary Table 5:False-positive results of human albumin positive structures 21d after mock-transplantation of cell culture medium, FCS or 0.9% NaCl
jws-hep.21745.cit1.pdf9KSupporting Information file jws-hep.21745.cit1.pdf
jws-hep.21745.mat1.pdf120KSupporting Information file jws-hep.21745.mat1.pdf

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.