The Effect of Recombinant Human Alpha-1,2-Fucosyltransferase and Alpha-Galactosidase A on the Reduction of Alpha-Gal Expression in the Liver of Transgenic Pigs

2020. The effect of recombinant human alpha-1,2-fucosyltransferase and alpha-galactosidase A on the reduction of alpha-gal expression in the liver of transgenic pigs. Genetically modified pigs lacking Gala1 3Gal and other immunogenic carbohydrates are considered as the most promising, alternative source of various tissues and organs for human transplantation. Here, we tested the hypothesis that combining the expression of human al,2-fucosyltransferase (h FUT2) and a-galactosidase A (h GLA) genes would allow for the removal of this specific carbohydrate in porcine transgenic livers. We investigated the expression profile of human al,2-fucosyltransferase and a-galactosidase A proteins and the amount of Gala1 3Gal antigen in the liver of single transgenic h FUT2 (n=5), h GLA (n=5), and double transgenic hFUT2xhGLA (n=5) pigs. Both human proteins, al,2-fucosyltransferase and a-galactosidase A, were abundantly expressed in the liver tissue in respective transgenic lines as was revealed by confocal microscopy and Western blotting. The level of Gala1 3Gal epitope evaluated by lectin histochemistry and lectin blotting was significantly lower (p<0.05) in all genetically modified livers than that in the control non-transgenic porcine livers. Importantly, the double transgenic line expressed a significantly lower (p<0.05), but still detectable level of this antigen, compared to both single transgenic pigs, as shown by lectin blotting. Histological evaluation of the liver samples stained with haematoxylin and eosin showed no morphological evidence of hepatic abnormalities in all transgenic pigs. Our study indicates that the simultaneous expression of two protective transgenes hFUT2xhGLA indeed improves the removal of the Gala1 3Gal epitope in porcine liver. However, this modification alone is not sufficient enough for complete elimination of this antigen from porcine liver tissue.

growth at a relatively low cost. In addition, porcine endogenous retroviruses could be effectively inacti vated by CRISPR-Cas9 to prevent their transmission to human recipients (NlU et al. 2017).
Unfortunately, the phylogenetic distance between pigs and humans entails a complex immune response leading to hyperacute rejection as well as acute hum o ral and cellular rejection o f transplanted pig organs (COOPER et al. 2015(COOPER et al. , 2016LU et al. 2020). Hyper acute rejection is an immediate reaction o f the human immune system to the major carbohydrate xenoantigen Gala I '3Gal. which is abundantly present on the surface o f porcine cells, in particular endothelial cells. The formation ofthe Galo.1 -SCial epitope is driven by the enzyme al,3-galactosyltransferase, encoded by the GGTA1 gene (SANDRIN & M c K f .N7.te 1994). GGTA1 is functional in most mammals, including pigs, but not in humans and Old World apes (G a l i l i etal. 1988;LU etal. 2020). In turn, primates produce natu ral xenoreactive antibodies, o f which about 90% rec ognise the Galo.1 -3Gal epitope (SANDRIN etal. 1993;M e M o r r o w et al. 1997).
A promising approach to overcome hyperacute rejection is generating homozygous al,3-galactosyltransferase gene-knockout (GTKO) pigs using zinc finger nucle ase (ZFN) technology, transcription activator-like ef fector (TALE) nucleases and modifications o f the CRISPR/Cas system (LA I etal. 2002;PIE R SO N 2009;LADOWSKI et al. 2019). Another strategy to remove the Gala. I -3Gal epitope from porcine cells involves the combined transgenic expression o f recombinant human enzymes al,2-fucosyltransferase (rhal,2-FT ) and a-galactosidase A (rha-Gal A) (OSM AN et al. 1997). al,2-fucosyltransferase is encoded by \\FUT2 and in humans occurs in the cis compartment o f the Golgi apparatus. This human enzyme acts earlier than porcine endogenous al,3-G T , which is present in the trans compartment. This strategy is based on the com petition o f these two enzymes acting on the same sub strate during oligosaccharide processing in transgenic cells. As an oligosaccharide moves through the Golgi apparatus from the cis to the trans com partm ent, it is first flicosylated by rhal,2-FT. Hence, this oligo saccharide cannot accept the terminal galactose resi due in the subsequent reaction catalysed by al,3-G T (H A R TEL-Sc h e n k et al. 1991). In turn, the recombi nant human a-galactosidase A, encoded by the h GLA gene, is responsible for the cleavage o f term inal D-galactose residues (LU O et al. 1999). The expres sion o f rha.l,2-FT or rha-Gal A alone does not, how ever, allow for the complete elimination o f the G alal^3G al epitope from pig cells. Thus, it was sug gested that the co-expression o f both rhal,2-FT and rha-Gal A would be more efficient method for this epitope reduction (LU O et al. 1999;ZEYLAND et al. 2014). Indeed, ZEYLAND etal. (2014) reported forthe first time the successfi.il production of double transgenic pigs that expressed human a l , 2 -fucosyltransferase and a-galactosidase and showed a considerable re duction o f the a-Gal antigen level on the surface o f skin fibroblasts. However, the effect o f transgenic modification may vary among different cell types, due to their specific glycosylation pattem, and so far the impact o f hFUT2 and hGLA transgenes on carbo hydrate antigen Gala. I oG al expression has not been investigated in any solid organ. Liver is the largest in ternal organ in the body and exhibits remarkable re generative capacity. The glycosylation o f liver cells and secreted proteins play an essential role in regulat ing various metabolic and immune functions, per formed by parenchymal hepatocytes, sinusoidal endothelial cells, hepatic stellate cells, resident macrophages (including Kupffer cells), and various lymphocytes.
In this context, in the present study we hypothesised that the combined actions o f human a 1 ,2fucosyltransferase and a-galactosidase A enzymes in double transgenic \lFUT2x\iGLA pigs are more effi cient in eliminating the Gala. I oCial epitope in por cine liver than a single expression o f ci ther \\FUT2 or h GLA transgenes. Fresh tissue specimens were taken immediately after slaughtering and the opening o f the abdominal cavity from 15 transgenic pigs (n=5 for each transgenic vari ant) and from 8 non-genetically modified Polish Large White pigs. All liver samples were frozen in liquid nitrogen for further analyses. The transgenic pigs were designed to show the expression o f re com binant hum an al,2-fucosyltransferase (hFUT2), a-galactosidase (hGLA) and hFUT2*hGLA (LIPIŃSKI et al. 2010;ZEYLAND et al. 2014). The transgenic pigs and non-transgenic control pigs were healthy and normal in terms o f reproductive capability. They were 12-to 18-months old and weighed 150-200 kg.

Histological analysis
The frozen liver samples were sectioned at 6 pm in Leica CM 1850 UV cryostat (Leica, Biosystems, Nussloch, Germany) and collected onto poly-Llysine coated microscopic slides. Cryosections were fixed with 4% paraformaldehyde in PBS for 10 min, washed in PBS three times for 5 min and then stained with haematoxylin (Shandon, ThennoFisher Seien-tific, Waltham, MA, USA) for 3 min. Subsequently, slides were washed 1 0 min in tap water and were then stained with eosin (ThermoFisher Scientific, Waltham, MA, USA) for 2 min. After dehydration in a graded series o f ethanol (70%-100%), followed by two changes of xylene, sections were mounted with Consul-Mount™ (Shandon™, ThennoFisher Scientific, Waltham, MA, USA) and coverslipped. Finally, sections were ana lysed using aNikon Optihot-2 bright field microscope (Nikon, Tokyo, Japan) equipped with aN ikon Digital Camera DXM 1200F (Nikon, Tokyo, Japan) and 1 0 x, 2 0 x, and 40x objective lenses.

Immunofluorescence staining
The porcine liver samples were cryosectioned at 6 pm and collected onto poly-U-lysine coated m icro scopic slides. Sections were fixed with a 4% parafor maldehyde solution in PBS for 10 min, washed in PBS and blocked in 5% Normal Goat Serum/PBST (Phosphate buffer saline with 0.1 % v/v Triton X-100, Bioshop Inc., Burlington, Canada) for 45 min. Then sections were incubated with the primary antibodies against: human al,2-fucosyltransferase (Rabbit poly clonal antibodies, ab 198712, Abeam) diluted 1:150 in PBST; human a-galactosidase (Rabbit polyclonal an tibodies, PA5-27349; ThennoFisher Scientific, Waltham, MA, USA) diluted 1:200, overnight at +4°C in a humidified chamber. In the next step, sections were washed several times in PBST and treated with appropriate secondary antibodies labelled with Cy3 (Goat anti Mouse or Goat anti Rabbit, Jackson Immuno Research) diluted 1:600 in PBST, for 1 h at room temperature. After final washes, sections were mounted in Fluoroshield with DAPI mounting m e dium (F6057, Sigma-Aldrich, St. Uouis, MO, USA) and coverslipped. Fluorescently labelled sections were examined by Olympus FV1200 Confocal Mi croscope (Olympus, Tokyo, Japan). Relative intensi ties o f fluorescence were quantified in each, randomly chosen region o f interest (ROI) using Im-ageJ version 1.46r software (National Institutes of Health, Bethesda, MD, USA) in a greyscale of 256 levels (GAJDA et al. 2011;ROM EK et al. 2017). For each genetically modified and control pig, three sections were sampled by 70 ROI.

Lectin histochemistry
The localization and semi quantitative comparison o f G alal^3G al epitope expression in the liver o f \iFUT2, hGLA, and hFUT2A\GLA was assessed by using a specific lectin from Griffonia simplicifolia (GS-IB4) conjugated with Alexa Fluor 647 fluores cent dye (132450, Molecular Probes, Invitrogen , ThennoFisher Scientific, W altham, MA, USA). Lectin histochemistry was carried out on 6 pm cryostat sections, which were fixed with 4% paraformal dehyde in PBS, washed in PBS and then blocked in 1% BSA (Bovine Serum Albumin, Bioshop Inc., Bur lington, Canada) in PBST for 1 h. Samples were washed three times in PBS and treated with lectin GS-IB4 diluted 1:200 in DPBS, at +4°C overnight in a dark humidified chamber. After final washes, sec tions were mounted in Fluoroshield with DAPI and coverslipped. Fluorescently labelled sections were examined as described in the paragraph Immunofluo rescence staining.

Total protein isolation and Western/Lectin blot analyses
Total protein was extracted from frozen tissue sam ples by using a radioimmunoprecipitation assay lysis buffer (RI P A buffer, Thermo Fisher Scientific, Waltham, MA, USA) containing 1% o f proteinase in hibitor cocktail (RIPA+PI; Bioshop Inc., Burlington, Canada). Liver samples were cut into small, approxi mately 2 mm pieces and homogenized in 300 pi RIPA+PL Tissue lysates were centrifuged at 14 000 x g for 15 min at +4°C and supernatants collected. Protein concentration was determined with microassay DC™ Protein Assay (Bio-Rad Laboratories, Hercules, CA, USA) using bovine serum albumin as a standard. Pro tein samples were stored at minus 80°C for further analyses.
Protein samples were diluted in 2 x Laemmli Sample Buffer (Bio-Rad Laboratories, Hercules, CA, USA) with ß-mercaptoethanol (BME; Sigma-Aldrich, St. Louis, MO, USA) and denatured at 100°C for 5 mm. Then, proteins were separated in SDS-PAGE using 5% stacking and 10% resolving gels. Molecular weights were estimated with reference to standard proteins (Precision Plus Dual Color Protein Standard, Bio-Rad Laboratories, Hercules, CA, USA). For immunoblotting, lectin blotting proteins were electro transferred onto PVDF membrane (Immobilon, Merck, Demistadt, Germany) at a constant amperage o f 250 mA for 120 min.
For immunoblotting, membranes were blocked for 1 h in 5% non-fat milk in TBST (Tris buffer saline with 0.1% v/v Tween20, Bioshop Inc., Burlington, Canada) and after several washes in TBST, incubated overnight at +4°C with the following primary anti bodies (the same as for immunofluorescent labelling) against: human a l , 2 -fucosyltransferase diluted 1 : 1 0 0 0 in TBST and human a-galactosidase diluted 1:1000 in TBST. Then, membranes were washed several times in TBST and incubated with appropriate secondary antibodies horse radish peroxidase (HRP) conjugated (Goat anti-Rabbit or Goat anti-Mouse ThennoFisher Scientific, Waltham, MA, US A) at a dilution o f 1:6000 in TBST for 1 h at room temperature. The ß-actin was used as a reference protein (ab8224 Abeam, Mouse monoclonal antibodies anti-ß-actin) diluted 1 : 2 0 0 0 in TBST For lectin blotting, membranes were blocked for 30 min in 1% BSA in TBST. After several washes in DPBS followed by TBS, membranes were incubated overnight at +4°C with lectin GS-IB4 labelled with HRP (L5391, Sigma-Aldrich, St. Louis, MO, USA) diluted 1:2000 in DPBS. Finally membranes were washed in TBS.
For both immunoblotting and lectin blotting, pro tein bands were detected by chemiluminescence us ing Clarity™ Western ECL Blotting Substrate (Bio-Rad Laboratories, Hercules, CA, USA) and visualized with the ChemiDocTM XRS+ Imaging System (Bio-Rad Laboratories, Hercules, CA, USA). Protein bands were quantified using Image Lab™ 2.0 Software (Bio-Rad Laboratories, Hercules, CA, USA) by measurement o f their relative optical densities. Following protein detection in Western blotting, mem branes were stripped and reprobed with anti-ß-actin antibody, which was used as a reference protein (ab8224 Abeam, Mouse monoclonal antibodies anti-ß-actin) diluted 1:2000 in TBST.

Statistical analysis
For each genetically modified and non-modified group o f pigs and for all analyses, three repeats were performed. Quantitative data were expressed as the mean ± standard error o f the mean and examined using the Shapiro-Wilk (SW) test for nonnality. Compari sons between the appropriate means were performed by one-way analysis o f variance (ANOVA), followed by the Tukey HSD post hoc test. Statistical signifi cance was marked accordingly: *p<0.05, **p<0.01, ***p<0 .0 0 1 .

Histological analysis
Histological analysis oftransgenic and non-transgenic porcine livers was performed to examine any possible adverse effects o f transgenesis on the morphology of porcine liver. Routine staining with haematoxylin and eosin showed no visible differences between livers from transgenic and non-transgenic pigs in the histo logical organisation o f tissue structures such as a por tal space, bile ducts, hepatic veins, hepatic arteries, and parenchymal cells. Figure 1 shows representative images o f liver tissue sections stained with H&E from single transgenic liFUT2 and h GLA, double trans genic hFUT2xhGLA, and non-transgenic pigs. The analysis showed no histological differences between the control and the transgenic samples.

Immunofluorescence staining
The localization o f recombinant human a. 1,2fiicosyltransferase (rhal,2-FT ) and a-galactosidase A (rha-Gal A) was examined by the immunofluores cence staining o f liver cryostat sections derived from single transgenic liFUT2 and h GLA, double trans genic hFUT2x\\GLA, and non-transgenic pigs. The positive immunofluorescence signal from rhal,2-F T was evenly distributed in the hepatic lobules o f the porcine liver derived from both transgenic lines hFUT2 and hFUT2xhGLA ( Fig. 2a and 2c). In turn, rha-Gal A immunostaining was less homogenous and detected in small clusters o f hepatocytes as well as in the interlobular space in the porcine liver o f transgenic variants hGLA and hFUT2xhGLA ( Fig. 2b and 2d). In the control non-transgenic livers the rhal,2-F T was barely detectable by immunofluorescence (Fig. 2e), while we did not observe any positive signal from re combinant human a-galactosidase (Fig. 2f). Double immunofluorescence labeling experiments demon strated colocalization o f the recombinant human al,2-fucosyltransferase with the Golgi resident 58K protein in the perinuclear region o f hepatocytes origi nated from both transgenic pig models, liFUT2 and \iFUT2xhGLA. Figure 4 shows representative m i crophotographs of immunofluorescence colocaliza tion o f recombinant human-a.l,2 -fucosyltransferase (rhal,2-FT) and the Golgi apparatus (58K Golgi pro tein) in porcine liver cryostat sections derived from single (hFUT2) and double transgenic pigs (hFUT2xhGLA).
The semi-quantitative analysis o f fluorescence in tensity showed statistically significant, inter individual differences in the expression o f recombi nant human a l , 2 -fucosyltransferase between the con trol (CTR nTG) and both transgenic variants, \\FUT2 (p<0.01) and hFUT2x\\GLA (p<0.001) (Fig. 3a). Moreover, significant differences were also detected between \iFUT2xhGLA and hFUT2 pigs (p<0.01). Considering recombinant human a-galactosidase ex pression, we did observe statistically significant differences between the control (CTR nTG) and both transgenic variants, h GLA (p<0.01) and hFUT2xhGLA (p<0.001), but there was no difference between the transgenic animals (Fig. 3b).

Western blot
Liver tissue samples were derived from single (hFUT2, h GLA) and double (hFUT2xhGLA) trans genic pigs as well as from non-transgenic, control pigs (CTRnTG). Western blot analysis of total protein extracts from porcine liver samples showed the pres ence o f recombinant human a l , 2 -fucosyltransferase and a-galactosiadase A proteins in all corresponding transgenic samples (Fig. 5a-c). In the control group (Fig. 5d), we observed a weak positive signal for rhal,2-FT and a barely detectable signal for human rha-Gal A. The signal intensities o f the analysed pro teins were normalized to beta-actin, which was used as a loading control. Quantitative analysis o f the W estern blot showed that the relative expression o f both tested proteins was significantly higher (at least p<0 .0 1 ) in all transgenic variants than that in the con trol group (CTRnTG) (Fig. 5e and 5f). The specificity o f primary antibodies was detenu ined by the presence o f a single band in the liver samples, at the expected molecular weight, on a W estern blot.

Lectin histochemistry
The expression profile o f the G alal^3G al epitope was identified in porcine liver tissue sections by label ling with the specific lectin GS-IB4 Alexa Fluor 647 conjugated (Fig. 6 ). The lectin GS-IB4 strongly la belled G alal -3Gal in all the liver tissue sections o f the control group (Fig. 6 a). In contrast, liver samples from the single transgenic pigs hFUT2, h GLA (Fig. 6 b and 6 c) and from the double transgenic pigs hFUT2xh.GLA (Fig. 6 d) displayed a much lower G alal ^3Gal fluorescence intensity compared to the control group.
Semiquantitative analysis o f fluorescence intensity o f the AlexaFluor 647 dye revealed that the G alal -3Gal epitope expression in the liver o f all three transgenic lines was significantly lower as com pared to the control non-transgenic pigs (Fig. 7) at p<0.01 for both hFUT2 and h GLA, and p<0.001 for hFUT2xhGLA. However, no significant differences were detected among the individual groups o f geneti cally modified animals (p>0.05).

Lectin blot analysis
Lectin blot analysis was used to detennine the ex pression profile o f G alal -3Gal at the total protein level using lectin GS-IB4 labelled with HRP. The re sults showed the presence o f the Gahxl >3Gal epitope in all analysed porcine liver samples (Fig. 8 a). Beta actin served as a loading control. The quantitative analysis ofthe G alal -3Gal epitope showed statistical differences between the control group (CTR nTG) and all genetically modified liver samples (Fig. 8 b). The lowest significant expression o f the G alal >3 Gal epitope was observed in samples from double trans genic porcine livers (hFUT2xhGLA) and this was sta tistically different from both, hFUT2 (p<0.05) and h GLA (pcO.Ol) (Fig. 8 b). GALILI e ta l. (1988) dem- onstrated an abundance o f this epitope on mammalian cells including pig cells (1x106-30x106 epitopes/cell). Since in our study lectin blotting was performed on to tal protein samples, we postulate that the large number o f bands positive for G alal >3Gal result from a great quantity o f this epitope in pig liver cells.

Discussion
In this study we investigated the effects o f the over expression o f human a l , 2 -fucosyltransferase and a-galactosidase A on the amount o f G alal -3 Gal anti gen and hepatic histology in porcine liver from three transgenic lines. As genetically modified pigs were produced to avoid hyperacute rejection (LIPIŃSKI et al. 2010;ZEYLAND et al. 2013ZEYLAND et al. , 2014, we discuss this aspect in relation to a liver model. Our immuno fluorescence and lectin blot analyses revealed a sig nificant reduction o f G alal -3Gal expression in the liver tissue in all three transgenic lines, hFUT2, h GLA, and hFUT2*hGLA, in comparison to nontransgenic pigs. More importantly, the double trans genic animals showed a significantly lower, but still detectable, level o f the Galo. I *3Gal epitope than two other genetically modified pig lines, as was evidenced by lectin blotting. Collectively, our results indicate that the constitutive co-expression o f cooperating enzymes such as recombinant human a l , 2 -fucosyltransferase and a-galactosidase A is more efficient in removing the G alal -3Gal antigen from porcine cells than the individual function o f either h FUT2 or h GLA. How ever, at present, we cannot exclude the possibility that the more effective removal o f the G alal -3Gal epi tope in the double transgenic line is also due to a higher level o f \xFUT2 expression when compared to the single transgenic liFUT2 line (see Fig. 3a).
Therefore, quantification o f transgene copy number and mRNA level o f \\FUT2 and h GLA in the liver should be carried out to verify this conclusion. Unfor tunately, incomplete removal o f the Galo.1 A Gal in double transgenic hFUT2xhGLA pigs indicates that this genetic modification alone is not to be considered for liver xenotransplantation. Hence, our data support emerging evidence that multi-transgenic pig models with a larger number o f deleted, humanized, or added genes are required for the effective elimination o f m a jo r xenoantigen to prolong vascularised xenograft survival (Niem a n n & Petersen 2016). Undoubt edly, GTKO pigs represent the basis for further opti- For each genetically modified and control animal, three sections were sampled by 70 ROI. Data are presented as mean ± SEM and displayed as arbitrary units in exponential notation (xlO3). Statistics: One-way ANOVA and post hoc Tukey F1SD test, values are denoted as **p<0.01 and ***p<0.001. A significantly lower relative expression of the G alal-3G al epitope in all tested samples compared to the control group (CTR nTG) is noted. The quantitative analysis of G alal-3G al epitope relative expression, data are presented as mean ± SEM from three separate ROD analyses of three animals for each variant, and displayed as arbitrary units in exponential notation (xlO3). One-way ANOVA and post hoc Tukey HSD test, values are denoted as: *p<0.05 and **p<0.01. Die relative expression of the G alal-3G al epitope was significantly lower in the liver of all transgenic pigs compared to the control (CTR nTG) non-transgenic animals. Within the transgenic animals, the lowest significant expression of the G alal-3 G al epitope showed liver samples derived from double transgenic (hFUT2xhGLA) pigs. In other samples (hFUT2, hGLA) the expression remained at a similar level with no significant differences. They also showed that both recombinant human a 1,2-fucosyltransferase and a-galactosiadase A may individually reduce the expres sion o f the G alal -3Gal epitope. Our finding that the single expression o f human a 1,2-fucosyltransferase may decrease the amount o f the Gala I *3Gal epitope in transgenic liver is consistent with earlier studies us ing different tissues (SANDRIN et al. 1995; SHARMA e t a l 1996; COSTA etal. 2008; LIPIŃSKI e t a l 2010). A previous study by JlA et al. (2004) proved that a-galactosidase alone lowered epitope G alal >3Gal expression by 78% , w hile the co-expression o f a-galactosidase and a 1,2-fucosyltransferase reduced this epitope almost to zero on the surface o f SV40immortalised aortic porcine endothelial cells. A n other study by ZEYLAND et al. (2014) estimated the G alal -3Gal epitope reduction in porcine ear skin fi broblasts at 60% for a 1,2-fucosyltransferase, 58.9 % for a-galactosidase, and 66.9 % for both a. 1,2fucosyltransferase and a-galactosidase. Here we demonstrated by lectin histochemistry and lectin blot ting that G alal -3Gal epitope expression in porcine liver was decreased by 54% (histochemistry) and 38% (blotting) in the \\FUT2 pigs. 59% and 35% in the Y\GLA animals, and finally by 62% and 47% in the hl'U T2 y\\GLA pigs. Therefore, it is clear that the most effective approach to eliminating this epitope is generating homozygous GTKO pigs lacking the gene for al,3-galactosyltransferase.
We report here the abundant hepatic expression o f recombinant human proteins al,2-fiicosyltransferase and a-galactosidase A in transgenic pigs, indicating an efficient and stable expression o f human transge nes in porcine livers. Our study also revealed the pre dominant, perinuclear localization o f human a 1,2-fucosyltransferase in porcine hepatocytes, as evidenced by the immunofluorescence analysis o f porcine livers derived from single 1\FUT2 and double liFUT2 xhGLA transgenic animals. Moreover, double immunostaining with antibodies against human a. 1,2-fucosyltransferase and Golgi 58K protein pro vide evidence for the colocalization o f these two pro teins and support the view o f a 1,2-fucosyltransferase distribution within the Golgi apparatus (M lLLAND et al. 2001). Indeed, in human cells a l,2fucosyltransferase is localized within the cis compart ment o f the Golgi complex, while in the porcine cells a.l,3-galactosyltransferase is present in the trans compartment (HARTEL-SCHENK et al. 1991). Hence, distinctive distribution o f these two enzymes in the Golgi network, which compete for the same substrate N-acetyllactosamine, allows for the lowering o f the G alal^3G al epitope in the porcine hepatocytes by rhal,2-FT.
Considering possible adverse effects o f the overex pression o f human transgenes on hepatic tissue integ rity and cell morphology, we performed a routine histological analysis o f the livers from the transgenic pigs. We found no histological evidence o f liver tissue abnormalities in the analyzed pigs, which may sug gest that these genetic modifications do not have dele terious effects on the functional morphology o f porcine hepatic tissue.
In conclusion, we have shown that the overexpres sion o f recombinant human a 1,2-fucosyltransferase and a-galactosidase A in single and double transgenic pig models significantly reduces, but does not elimi nate the G alal >3Gal epitope from hepatic tissue. Hence, the efficacy o f this approach for liver m odifi cation in xenotransplantation-related studies is lim ited as even hyperacute rejection remains a hurdle.