Abstract
Acute intermittent porphyria (AIP) is an autosomal dominant disorder characterized by insufficient porphobilinogen deaminase (PBGD) activity. When hepatic heme synthesis is induced, porphobilinogen (PBG) and 5-aminolevulinic acid (ALA) accumulate, which causes clinical symptoms such as abdominal pain, neuropathy, and psychiatric disturbances. Our aim was to investigate if hepatocyte transplantation can prevent or minimize the metabolic alterations in an AIP mouse model. We transplanted wild-type hepatocytes into PBGD-deficient mice and induced heme synthesis with phenobarbital. ALA and PBG concentrations in plasma were monitored, and the gene transcriptions of hepatic enzymes ALAS1, PBGD, and CYP2A5 were analyzed. Results were compared with controls and correlated to the percentage of engrafted hepatocytes. The accumulation of ALA and PBG was reduced by approximately 50% after the second hepatocyte transplantation. We detected no difference in mRNA levels of PBGD, ALAS1, or CYP2A5. Engraftment corresponding to 2.7% of the total hepatocyte mass was achieved following two hepatocyte transplantations. A lack of precursor production in less than 3% of the hepatocytes resulted in a 50% reduction in plasma precursor concentrations. This disproportional finding suggests that ALA and PBG produced in PBGD-deficient hepatocytes crossed cellular membranes and was metabolized by transplanted cells. The lack of effect on enzyme mRNA levels suggests that no significant efflux of heme from normal to PBGD-deficient hepatocytes takes place. Further studies are needed to establish the minimal number of engrafted hepatocytes needed to completely correct the metabolic abnormality in AIP and whether amelioration of the metabolic defect by partial restoration of PBGD enzyme activity translates into a clinical effect in human AIP.
Keywords
Introduction
Acute intermittent porphyria (AIP) is an autosomal dominant disorder, which is characterized by insufficient hydroxymethylbilane synthase activity, also known as porphobilinogen deaminase (PBGD). Different endogenous or exogenous factors can cause hepatic PBGD to become overloaded by its substrate porphobilinogen (PBG). Both PBG and 5-aminolevulinic acid (ALA), the substrate of the previous enzyme, are subsequently found in high concentrations in plasma and urine (2). These presumably neurotoxic metabolites are associated with clinical symptoms such as abdominal pain, neuropathy, and psychiatric disturbances. Treatment includes intravenous hemin infusion, which acts by repressing the hepatic heme biosynthesis (5) by negative feedback mechanisms.
In some patients, attacks are frequent, severe, potentially life threatening, and may lead to long-term complications such as neuropathy, renal impairment, chronic hypertension, hepatocellular carcinoma, and a very poor quality of life (3,6,7). For this subset of AIP patients, orthotopic liver transplantation (OLT) has emerged as a curative treatment option (9,22). The natural course and mortality risk in severe AIP is difficult to predict, and some patients spontaneously improve after years of frequent attacks (1,6). Since OLT is a radical treatment with significant short- and long-term risks, appropriate indications are difficult to determine.
Hepatocyte transplantation is a less invasive and reversible procedure that has shown promising results when used as a treatment for several inherited metabolic diseases in which the liver has a main causative role for phenotypic expression (13,19). The beneficial effect is achieved by partially restoring the activity of a deficient enzyme. In symptomatic AIP, the predominant problem is accumulation of toxic metabolites. This may be important for the feasibility of hepatocyte transplantation. In order to be of clinical benefit, transplanted non-AIP hepatocytes corresponding to less than 10% of the total liver mass cannot be expected to stop metabolite production and must presumably normalize or significantly reduce elevated circulating concentrations of ALA and PBG by metabolizing surplus metabolites from the recipients' cells.
The aim of this study was to explore whether hepatocyte transplantation would ameliorate or even correct the metabolic abnormality in a mouse model of AIP. We investigated this by transplanting male wild-type hepatocytes into PBGD-deficient female mice and induced heme synthesis with phenobarbital, a well-known porphyrogenic drug. We monitored metabolite concentrations in plasma and induction of relevant hepatic enzymes: 5-aminolevulinate synthase (ALAS1), porphobilinogen deaminase (PBGD), and cytochrome P450 family 2 subfamily A polypeptide 5 (CYP2A5). Results were compared with controls and correlated to the percentage of engrafted hepatocytes.
Materials and Methods
Animals
Male C57BL/6 10-16-week-old mice served as donors for hepatocytes and female PBGD-deficient C57BL-based mice (the “Vincent” mouse; Urs A. Mayers laboratory, University of Basel, Basel, Switzerland) were used as recipients. PBGD-deficient mice were generated by gene targeting and exhibit typical biochemical characteristics of human AIP after treatment with porphyrogenic drugs (16). The mice were 14-16 months old and weighed 28-30 g.
The local ethics committee approved all animal procedures used in this study. The animals were treated in accordance with the Swedish regulations and laws for care and use of laboratory animals.
Experimental Design
The study protocol is shown in Figure 1. The AIP mice were divided into a group receiving hepatocyte transplantation (HTx group) and a control group (Control). All mice underwent partial hepatectomy (45-50%) in order to enable enhanced repopulation by donor hepatocytes (18). One animal did not survive surgery. Hepatectomy was followed by intrasplenic hepatocyte transplantation in the HTx group and intrasplenic injection of hepatocyte-free minimum essential medium (MEM) (Invitrogen AB; Life Technologies, Stockholm, Sweden) in the controls.
Schematic of the study protocol. PBG/ALA, porphobilinogen/5-aminolevulinic acid.
On days 11-14 after surgery, we simulated AIP attacks by repeated intraperitoneal administrations of phenobarbital (Sigma-Aldrich, Stockholm, Sweden) in daily increasing doses (75, 80, 85 and 90 mg/kg of body weight) for 4 consecutive days [modified from Lindberg et al. (16)].
Five mice in the HTx group (1HTx group) and four mice in the control group were sacrificed after the first course of phenobarbital treatment. The remaining HTx mice received a second intrasplenic hepatocyte infusion 3 days after the first course of phenobarbital treatment (2HTx, n = 4), and control mice received a second intrasplenic infusion of cell-free medium (n = 3). These mice received a second phenobarbital course on days 11-14 after the second intrasplenic infusion. After the second phenobarbital course, the mice were sacrificed.
All surgeries were performed under anesthesia with 0.5-1.5% isoflurane (Fluovac Unit, IMS, Cheshire, UK). Surgeries were performed between 10:00 a.m. and 2:00 p.m. to control for diurnal variation. Directly after surgery, animals had free access to water and standard chow.
Hepatectomy
All animals underwent 45-50% hepatectomy. The left, the left median, and the right median lobes of the liver were removed by the modified method of Higgins and Anderson (12) through an upper abdominal transverse incision. The two layers of the abdominal wall were closed separately by a running suture with 5-0 Prolene (Ethicon, San Angelo, TX, USA), and mice were kept under a heating lamp until they recovered from anesthesia.
Hepatocyte Isolation
An in situ two-step collagenase method was used for isolating hepatocytes from the six C57BL/6 donor mice. After skin sterilization, the abdomen was opened via an upper abdominal transverse incision, exposing the portal vein. A 22-gauge intravenous cannula (BD Venflon, Helsingborg, Sweden) was inserted into the portal vein and secured using 6-0 silk ties (AgnTho's Ab, Lidingö, Sweden). The cannula was connected to the perfusate tube, and the perfusion flow rate was controlled by a roller pump (MP-GE, ISMATEC SA, Switzerland) for 15-18 min. The infrahepatic inferior vena cava was incised to allow efflux. The liver was first perfused with calcium and magnesium-free Hank's balanced salt solution (Invitrogen AB; Life Technologies), which contained 10 nM ethylenediaminetetraacetic acid (EDTA) (Becton Dickinson AB, Stockholm, Sweden) and 10 nM N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid (HEPES; BioWhittaker, Lonza, Allendale, NJ, USA) at a flow of 5 ml/min for 10 min. The liver was then perfused with MEM containing 0.5 g/L collagenase P (Roche Diagnostics GmbH, Mannheim, Germany) at a flow of 5 ml/min for 15-18 min. All perfusates were pretreated at 37°C under aeration with room air.
Successful perfusion was considered complete when blanching of the entire organ was observed. The collagenase perfusion was stopped when the perfusate oozed from the surface of the liver, the capsula separated completely from the surface, and the blanching disappeared.
Before removing the collagenase-perfused liver, 3 ml of 4°C MEM was infused intraportally. The resected liver was moved to a dish with MEM on ice and agitated gently after the liver capsule was incised. The released cell suspension was first filtered through a two-layer sterilized gauze (Selefatrade, Danderyd, Sweden), then centrifuged at 50 × g for 2 min and washed with MEM three times to yield homogenous hepatocytes. The cell suspension was then filtered through a 100-μm nylon mesh (BD Falcon, Stockholm, Sweden) before the last centrifugation. The cell viability was determined by 4% trypan blue (Sigma-Aldrich) exclusion in a hemocytometer (Burker, Marienfeld, Germany). After storage on ice for 3 h, hepatocytes with >85% viability were used for transplantation.
Hepatocyte Transplantation
In the HTx groups, 2 × 106 hepatocytes in 200 μl MEM were slowly injected into the lower pole of the spleen with a 29-gauge needle (B. Braun, Melsungen AB, Germany). Hemostasis was secured with a ligature encircling the injection site. In the control group, 200 μl hepatocyte-free MEM was injected in the same manner. The same procedures were used for the first and the second hepatocyte transplantation.
Sample Collection
Blood samples were drawn from the tail vein after each phenobarbital administration. The blood was collected in MiniCollect tubes (Greiner Bio-One, Longwood, FL, USA) and centrifuged at 3,000 × g for 10 min. The plasma was stored at −80°C until analysis. One day after the last phenobarbital treatment, mice were sacrificed by an anesthesia overdose. The liver was then perfused in situ with cold isotonic saline, and the whole liver, spleen, and right quadriceps muscle were harvested, snap-frozen in liquid nitrogen, and stored at −80°C until preparation.
Determination of Plasma PBG and ALA Concentrations
PBG and ALA in plasma were quantified by using liquid chromatography coupled with tandem mass spectrometry (Quattro-LC, Micromass, UK) after derivati-zation with phenylisothiocyanate (PITC) as previously described (10). The automated electrospray ionization system was controlled by a computer (DEC 266 IPC) with MassLynx (Micromass, Waters, Milford, MA, USA) 3.4 and Janeiro II (Flux Instruments, Reinach, Switzerland) software.
Quantitative Real-Time PCR
All primers (Table 1) were designed using the Primer Express 2.0 software (Applied Biosystems, Foster City, CA, USA). Real-time PCR was performed using a Prism 7000 sequence detection system (SDS; Applied Biosystems). One microliter of DNA or cDNA template was analyzed in triplicate 25-μl reaction volumes using Platinum SYBR Green qPCR SuperMix (Invitrogen, Paisley, Scotland) according to the manufacturer's instructions. For each sample, the threshold cycle (Ct) value for each target transcript was normalized to that of the gene encoding 18S by calculating the DCt according to the formula ΔCt = Cttarget - Ct18s. These DCt values were used for the statistical analyses.
Sequences of Primers Used in the Quantitative Real-Time PCR and Gen-Bank Accession Numbers
Sry, sex-determining region Y; PBGD, porphobilinogen deaminase; ALAS1, 5-aminolevulinate synthase; CYP2A5, cytochrome P450 family 2 subfamily A polypeptide 5.
Genomic DNA Extraction
Genomic hepatocyte DNA was purified using the Easy-DNA® Kit (Invitrogen) according to the manufacturer's instructions. In order to avoid the sampling error phenomenon from a single biopsy, the whole liver or spleen was homogenized in 4 ml or 1 ml PBS, respectively, after which 200 μl of the homogenate was taken for further genomic DNA purification.
In order to assess hepatocyte engraftment efficiency, we measured the expression of sex-determining region Y (Sry) gene with real-time PCR in hepatocyte genomic DNA. We established a standard curve by mixing DNA extracted from male and female hepatocytes in serial ratios (1:100,000, 1:10,000, 1:1,000, 1:100, and 1:10) and used the Ct values obtained from the Sry assay.
RNA Extraction and cDNA Synthesis
Total RNA was extracted from 350 μl of liver homogenate using silica gel-based spin column (RNeasy Kit, Qiagen, Hilden, Germany) according to the manufacturer's instructions. Two micrograms RNA was incubated with 1 unit of DNase I (Invitrogen) at 37°C for 30 min to remove residual DNA followed by inactivation at 65°C for 10 min. Twenty microliters first-strand cDNA was generated by using Moloney murine leukemia virus reverse transcriptase (SuperScript II, Invitrogen) and an oligo (dT) primer in combination with the SMART II oligonucleotide (SMART PCR cDNA synthesis kit, Invitrogen) according to the manufacturer's instructions.
Motor Function and Muscle Histopathology
As a crude measure of motor function, we measured the hind leg stride length by paw prints on paper covering a walkway (8.0 cm × 29.5 cm). The test was done twice on all animals before they were sacrificed and on five C57BL/6 (control) mice. All measurements were included in calculated group means.
Frozen sections of the right quadriceps muscle were stained with hematoxylin/eosin (Sigma-Aldrich) and viewed under a light microscope (Zeiss Axiovert 40C; Zeiss, Tegeluddsvägen, Sweden).
Statistical Analysis
We used PRISM software (GraphPad, San Diego, CA, USA) for data analysis. Two-sample t tests were used for differences between group means. Data were expressed as mean ± SD and a value of p < 0.05 was considered significant.
Results
Plasma PBG and ALA Concentrations
After hepatectomy and intrasplenic infusion of hepatocytes (HTx) or cell-free medium (control), plasma concentrations of PBG and ALA increased in both groups. ALA tended to decrease faster in the HTx group (Fig. 2B), and PBG concentrations remained elevated only in the control group (Fig. 2A).
The effect of partial hepatectomy on plasma concentrations of PBG (A) and ALA (B) in a subset of the transplanted and nontransplanted PBGD-deficient mice. On day 0, the animals underwent partial hepatectomy followed by intrasplenic hepatocyte transplantation (2 × 106) or by intrasplenic minimum essential medium (MEM) injection (control) (*significant difference between groups at p < 0.05).
Ten days after hepatectomy and the first intrasplenic infusion (1HTx), all mice were administered daily with increasing doses of phenobarbital. Plasma PBG and ALA concentrations are shown in Figure 3. The concentration of both metabolites increased gradually in both groups with a trend toward lower concentrations in the transplanted group. The PBG concentration was significantly lower in the transplanted animals after the first two phenobarbital doses (Fig. 3A).
Effect of the first hepatocyte transplantation on plasma concentrations of PBG (A) and ALA (B) in transplanted and nontransplanted PBGD-deficient mice. Intraperitoneal phenobarbital administration was performed daily with increasing doses (75, 80, 85, and 90 mg/kg body weight) for 4 consecutive days (*significant difference between groups at p < 0.05). PBGD, porphobilinogen deaminase.
Ten days after the second intrasplenic infusion of hepatocytes (2HTx) or cell-free medium, all mice were once again treated with phenobarbital (Fig. 4). In the control animals, PBG and ALA increased significantly on days 3 and 4 of phenobarbital administration. After the fourth dose, the PBG (Fig. 4A) and ALA (Fig. 4B) concentrations in the controls were twice that of the transplanted animals.
Effect of the second hepatocyte transplantation on plasma concentrations of PBG (A) and ALA (B) in transplanted and nontransplanted PBGD-deficient mice. Intraperitoneal phenobarbital administration was performed daily with increasing doses (75, 80, 85, and 90 mg/kg body weight) for four consecutive days (*significant difference between groups at p < 0.05).
Expression of Enzyme mRNA
After 4 consecutive days of phenobarbital treatment, the expression of transcripts encoding ALAS1, PBGD, and CYP2A5 was analyzed by quantitative RT-PCR. No significant differences were seen between the control group and the transplanted groups (Table 2).
Expression of Enzyme mRNA in Liver Homogenates After 4 Consecutive Days of Phenobarbital Administration
Mice receiving intrasplenic infusions of cell-free medium, compared to mice receiving one (1HTx) or two (2HTx) hepatocyte transplantations. No significant differences were seen.
Hepatocyte Engraftment
The amount of transplanted hepatocytes was based on the sex mismatch between donors (male) and recipients (female). The percentage of transplanted cells accounted for 0.8 ± 0.3% of the recipient liver cell mass after the first cell infusion and for 1.6 ± 0.4% after the second. Only minor numbers of transplanted hepatocytes were found in the spleen (data not shown).
Motor Function
The right stride length was significantly shorter in the PBGD-deficient mice compared to the wild-type C57BL/6 donor mice (Fig. 5). The stride length did not differ between controls or transplanted animals undergoing one or two hepatocyte transplantations. Histological examination of the hind leg quadriceps muscles revealed typical signs including both atrophic and hypertrophic muscle fibers in all PBGD-deficient mice (results not shown). No improvements or changes were seen during the 3 weeks of the study.
Mean right hind leg stride length by paw prints on paper. C57BL/6 (WT, n = 5) compared to PBGD-deficient mice that had undergone intrasplenic infusion of cell-free medium (Control, n = 7), one (1HTx, n = 5) or two hepatocyte transplantations (2HTx, n = 4).
Discussion
Liver transplantation is now an established curative treatment option in severe AIP (9,22), but hepatocyte transplantation has not been explored in this disease. While risks and donor shortage are inherent obstacles for the expanded use of OLT in AIP, hepatocyte transplantation has potential advantages in being less invasive, less expensive, reversible, and possible to repeat. Hepatocyte transplantation is a promising therapy in metabolic liver diseases, both in experimental and clinical situations, in which hepatocyte engraftment and function has been demonstrated for up to 3.5 years (13,19,21).
In many inherited metabolic liver diseases, partial restoration of 5-10% of the enzyme activity is sufficient. This has been shown in, for example, factor VIII deficiency (8), Crigler Najjar syndrome (11), and citrullinemia (17). In other diseases, like primary hyperoxaluria, restoration of enzyme activity is not enough; the entire liver has to be replaced to stop oxalate overproduction (4).
In AIP, accumulation of toxic metabolites gives rise to the symptoms. Whether partial restoration of PBGD enzyme activity by hepatocyte transplantation would improve the clinical condition is unknown.
The introduction of a limited number of hepatocytes with normal enzyme function into a PBGD-deficient liver might help metabolize surplus ALA and PBG from native hepatocytes. These metabolites easily cross cellular membranes, whereby adjacent transplanted cells may metabolize precursors produced and accumulated in some hepatocytes. Conceivably, reduction of metabolite accumulation both by partial enzyme restoration and by metabolism in transplanted cells would be required.
In this study, we used knockout mice with a targeted disruption of PBGD to investigate if hepatocyte transplantation can prevent or minimize the metabolic severity of acute AIP attacks. The PBGD-deficient mouse is a good model for studying the pathogenesis and mechanisms involved in acute intermittent porphyria (16). This mouse model has 30% residual PBGD activity. We treated the mice with phenobarbital to challenge the heme pathway further. Phenobarbital has pleiotropic effects and is known to induce massive synthesis of P450 isoenzymes (15,23), which leads to a drain of cellular heme, which, in turn, induces ALAS1 via reduced feedback inhibition. Phenobarbital can also directly induce ALAS1 (14). The resulting amplification of the heme biosynthesis pathway turnover leads to accelerated accumulation of ALA and PBG (16). Although the mechanisms whereby these porphyrin precursors produce neurotoxic symptoms have not been fully elucidated, clinical manifestations are consistently related to their accumulation. A reduced precursor accumulation is thus a good marker for treatment affect.
We transplanted female PBGD-deficient mice with male wild-type hepatocytes after performing partial hepa-tectomy without pretreatment. This mimics our current protocol for clinical hepatocyte transplantation with one exception; in the clinical situation, the hepatocytes are delivered directly into the right portal vein. Higher levels of engraftment may have been achieved if pretreatment with a plant alkaloid in combination with 70-90% hepa-tectomy had been used (18), but this would have been a clinically less relevant model. Hepatectomy resulted in initial accumulation of ALA and PBG in plasma. This confirms clinical observations, which indicate that surgery can induce acute AIP attacks. Changes in the plasma concentration of ALA and PBG early after hepatocyte transplantation suggested a negative effect of intrasplenic cell infusion per se, which may be related to impaired hepatic microcirculation (24).
In order to increase the number of engrafted hepatocytes, we transplanted the mice twice. The results suggest that no significant metabolic improvement was achieved after the first transplantation but that the metabolic abnormalities were ameliorated after the second transplantation. Upon phenobarbital treatment, plasma concentrations of ALA and PBG were reduced by approximately 50% after the second hepatocyte transplantation.
The biochemical response and the gene transcription of ALAS1, PBGD, and cytochrome P450 resulting from phenobarbital challenge was compared with and without preceding hepatocyte transplantation. ALAS1 is the rate-limiting enzyme in hepatic heme biosynthesis, and its level is regulated by intracellular heme concentration through a negative feedback mechanism.
To determine the effect on P450 activity, we focused on the 2A5 isoenzyme. Phenobarbital causes a reduction of P450 2A5 gene transcription, and previous studies have suggested that P450 2A5 is a sensitive indicator of altered heme status in the mouse liver (14). We detected no difference in mRNA levels of PBGD, ALAS1, or CYP2A5 between transplanted and control mice (Table 2).
The level of engraftment was analyzed after sacrificing transplanted mice. After the second transplantation, we found 1.6% normal hepatocytes, which correspond to 2.7% of the total hepatocyte mass. A lack of precursor production in less than 3% of the hepatocytes thus resulted in a 50% reduction in plasma precursor concentrations. This disproportional finding suggests that ALA and PBG produced and accumulated in PBGD-deficient hepatocytes, indeed crossed cellular membranes, and was metabolized by transplanted cells. By comparison, in a study on liver-directed gene therapy, a 2.2-fold PBGD overexpression in 10-15% of hepatocytes was sufficient to correct the metabolic alterations associated with phenobarbital induction in the AIP mouse model (20).
Replacement of a small portion of the endogenous hepatocyte mass is obviously not enough to correct the metabolic abnormality induced by phenobarbital, but our results suggest proof of concept. Whether the 50% reduction in precursor production translates into a meaningful clinical effect remains to be seen. Human AIP displays a large interindividual variability in phenotypic expression. Less than 50% of individuals with the AIP genotype ever have acute symptoms (6). While some are asymptomatic but display high urinary excretion of ALA and PBG, others have moderate excretion but recurrent severe symptoms (6). Amelioration of the metabolic defect in the latter group may be enough to make them asymptomatic. Possibly, a limited reduction of metabolite accumulation below a hypothetical threshold above which acute attacks are possible would be enough.
The limited size of this study calls for cautious interpretation. Further studies on larger numbers of mice and higher engraftment rates are needed to explore the minimal number of engrafted hepatocytes needed to make a clinical difference.
Hepatocyte transplantation may become a viable treatment option in human AIP if engraftment of hepatocytes corresponding to less than 10% of the total liver mass can be shown to prevent or reduce the severity or frequency of acute AIP attacks. One may then envision a protocol with repeated, perhaps annual, hepatocyte transplantations until spontaneous improvement occurs or until better treatment options become available.
Footnotes
Acknowledgments
The authors thank Christer Möller and Zymenex for help with financing. The authors declare no conflicts of interest.