Fatal Propionic Acidemia in Mice Lacking Propionyl-CoA Carboxylase and Its Rescue by Postnatal, Liver-specific Supplementation via a Transgene*

Propionic acidemia (PA) is an inborn error of metabolism caused by the genetic deficiency of propionyl-CoA carboxylase (PCC). By disrupting the α-subunit gene of PCC, we created a mouse model of PA (PCCA−/−), which died in 24–36 h after birth due to accelerated ketoacidosis. A postnatal, liver-specific PCC expression via a transgene in a far lower level than that in wild-type liver, allowed PCCA−/− mice to survive the newborn and early infant periods, preventing a lethal fit of ketoacidosis (SAP+PCCA−/− mice). Interestingly, SAP+PCCA−/− mice, in which the transgene expression increased after the late infant period, continued to grow normally while mice harboring a persistent low level of PCC died in the late infant period due to severe ketoacidosis, clearly suggesting the requirement of increased PCC supplementation in proportion to the animal growth. Based on these results, we propose a two-step strategy to achieve an efficient PA prevention in human patients: a partial PCC supplementation in the liver during the newborn and early infant periods, followed by a larger amount of supplementation in the late infant period.

Propionic acidemia (PA) is the most frequent autosomal recessive disorder of organic acid metabolism in humans. It is caused by a deficiency of propionyl-CoA carboxylase (PCC), a ubiquitously expressed, heteropolymeric mitochondrial enzyme involved primarily in the catabolism of branched-chain amino acids and fatty acids, which harbor odd-numbered chain lengths (1). A variety of mutations in the genes of either subunit (PCCA, PCCB) abolish or reduce the PCC function, resulting in a considerable clinical heterogeneity; some patients die in early childhood due to rapid and furious acceleration of the ketoacidosis in proportion to the increase of protein intake (2,3). The primary treatment of PA has been dietary restriction of the precursor amino acids of propionyl-CoA. Despite improvement of the dietary therapy, however, its long-term efficacy still appears controversial, and the overall outcome of severe forms of PA often remains disappointing (1,4,5). As in other genetic disorders, supplementation of PCC activity, perhaps primarily via gene therapy, may efficiently temper even the severe forms of PA (6 -9). It seemed important, then, to generate an animal model of PA to test possible therapeutic strategies. We therefore created a strain of mice lacking PCC activity by disrupting the ␣-subunit gene (PCCA gene) via homologous recombination in embryonic stem cells. To address the potentiality of a tissue-specific gene supplementation as a PA therapy, we also generated a transgenic mouse strain expressing PCCA exclusively in the liver and then crossbred these two types of mice. By analyzing these mice, we determined the potential therapeutic efficacy of the liver-specific PCC supplementation and propose a new mode of treatment to achieve an efficient PA prevention in human patients.

EXPERIMENTAL PROCEDURES
Generation of PCCA Ϫ/Ϫ Mice-A mouse genomic DNA clone containing exons of the PCCA gene was isolated by screening a genomic DNA library derived from the 129/sv strain. The targeting vector was prepared using the 12-kb EcoRI-HindIII genomic DNA fragment, pMC-1-neo-poly(A) (Stratagene) plasmid and the pBluescript plasmid (Stratagene). This construct was designed to delete a 2-kb EcoRI-XbaI fragment from PCCA, encompassing exons 1-4 as schematized in Fig. 1A. The linearized targeting vector was electroporated into the E14.1 embryonic stem line as we previously described (10). G418-resistant clones were isolated and screened by polymerase chain reaction and Southern blotting. Two of 146 Gly-418-resistant clones had undergone the desired homologous recombination. Positive clones were injected into C57BL/6 (B6) blastocysts, and male chimeric offspring were mated to B6 females. Mice carrying the mutation in the heterozygous state (PCCAϩ/Ϫ) were intercrossed to produce homozygote mutants (PCCA Ϫ/Ϫ ).
Generation of SAP ϩ PCCA Ϫ/Ϫ Mice-A cDNA clone containing the whole mouse PCCA coding sequence was isolated by screening a mouse spleen cDNA library (CLONTECH) using a human PCCA cDNA fragment as a probe. The 2.8-kb mouse PCCA cDNA fragment was subcloned into the EcoRI site of the pSG-1 expression vector, which contains human SAP promoter and rabbit ␤-globin non-coding exon/intron (11). The complete plasmid (SAP-PCCA) was digested with HindIII and XhoI, and the transgene fragment no longer containing vector sequence was purified by Geneclean II kit (Bio101, Vista, CA). DNA was microinjected into fertilized eggs of (B6xDBA/2)F1 mice. Resulting founders were screened for transgene by polymerase chain reaction and Southern blotting. Three independent founders carried the transgene. All three mice were crossbred with PCCA ϩ/Ϫ mice. Progenies harboring Transgene ϩ PCCA ϩ/Ϫ genotype were intercrossed to obtain SAP-Transgene ϩ PCCA Ϫ/Ϫ (SAP ϩ PCCA Ϫ/Ϫ ) mice.
PCC Activity Assay-PCC activity in the liver tissues was determined by fixation of [ 14 C]bicarbonate as described elsewhere (12).
Histology-Liver and kidney specimens were fixed in 10% formalin. 10 m-thick sections were cut from paraffin-embedded samples and stained by hematoxylin-eosin or periodic acid-Schiff (PAS). For fat detection, frozen liver sections, which had been embedded in Tissue-Tek OCT compound, were cut at 10-m thickness and then stained with oil red O (Sigma) as described elsewhere (13).

RESULTS
Lethal Ketoacidosis in PCC-deficient Mice-We created PCCdeficient mice by disrupting the gene encoding the ␣-subunit of PCC protein (PCCA). Null mutation of the PCCA gene was generated by replacing an ϳ2-kb fragment of the PCCA gene, which contains exons encoding the N-terminal sequence of PCCA protein, with the neomycin resistance gene (neo r ), via homologous recombination in embryonic stem cells (Fig. 1A). Although the PCCA gene is expressed ubiquitously from the early embryonic stage throughout the whole development, 2 PCCA-null (PCCA Ϫ/Ϫ ) mice were born at the expected Mendelian frequency, and were indistinguishable from wild-type (PCCA ϩ/ϩ ) littermate mice at birth. This is compatible with human PA patients, most of whom have no clinical complaints at birth (1). During embryogenesis, the toxic propionic acid derived from propionyl-CoA that has accumulated within cells might be purified via maternal circulation. The abolished expression of PCCA in PCCA Ϫ/Ϫ mice was confirmed by Northern blot analysis of the liver RNA (Fig. 1B), and by the absence of PCC enzyme activity in the liver tissue (1590 Ϯ 18 in PCCA ϩ/ϩ versus 15 Ϯ 17 in PCCA Ϫ/Ϫ , mice; n ϭ 3 for each, pmol/min/mg) (12).
Twenty-four h after birth, PCCA Ϫ/Ϫ mice showed remarkably high concentrations of serum propionyl-carnitine, urinary 3-OH-propionate (3-HP) and urinary methylcitrate, all of which increase in human PA patients as the result of abnormal metabolism of the accumulated propionyl-CoA, and thus are useful for PA diagnosis (1,15) (Table I). Interestingly, the serum propionyl-carnitine concentration was significantly high in PCCA Ϫ/Ϫ mice even at birth ( Table I), suggesting that the maternal circulation appears not to be able to completely overcome the accumulation of the catabolites of propionyl-CoA. Probably due to the intake of protein from the milk, as well as to the lack of the maternal circulation, the concentrations of propionyl-carnitine, methylcitrate, and 3-HP were strikingly increased in PCCA Ϫ/Ϫ mice within 10 h after birth ( Table I). The ketonuria, which indicates the presence of acidosis caused by the catabolism of the accumulated propionyl-CoA (1,16), was detected 10 h after birth and was rapidly accelerated in PCCA Ϫ/Ϫ mice ( Table I).
As a consequence of progressed acidosis, PCCA Ϫ/Ϫ mice no longer took milk after 10 -12 h, although both PCCA Ϫ/Ϫ and PCCA ϩ/ϩ mice started sucking milk right after birth. This decrease of intake is also often observed in human PA infants as a primary symptom of the disease (1). Twenty-four h after birth, PCCA Ϫ/Ϫ mice were apparently dehydrated (dry skin, poor urination), poor in movements, and possessed little or no milk in their stomachs, whereas PCCA ϩ/ϩ littermate mice were healthy and their stomachs were filled with milk. A vicious circle by the acceleration of dehydration and the progression of acidosis resulted in death of all PCCA Ϫ/Ϫ mice in 24 -36 h after birth.
In the histology of tissues at 24 h after birth, PCCA ϩ/ϩ mice harbored significant amounts of a PAS-positive product in the liver ( Fig. 2A), which was identified as glycogen by a glycogenspecific tissue staining method using ethanol-fixed liver specimens (Fig. 2C). In contrast, PCCA Ϫ/Ϫ mice showed no or very little glycogen in the liver (Fig. 2B). Interestingly, PCCA Ϫ/Ϫ mice did possess glycogen in the liver like PCCA ϩ/ϩ mice at the very late gestation, E18.5 (Fig. 2, D and E). Therefore, this disappearance of glycogen in PCCA Ϫ/Ϫ mice appears to be due to a drastic consumption. Serum concentration of either glucose or insulin was comparable in PCCA ϩ/ϩ and PCCA Ϫ/Ϫ mice (data not shown), clearly indicating that there was no defect in the glucose metabolism pathway that may cause a complemental consumption of glycogen. In the kidney of PCCA Ϫ/Ϫ mice, the collecting ducts were enlarged with marked incorporation of hyaline droplet casts, which implicated an acceleration of reabsorption (Fig. 2, F and G). All of these pathological changes in the liver and kidney of PCCA Ϫ/Ϫ mice might be primarily caused by the dehydration and malnutrition due to no intake of milk after 10 -12 h of age, since similar phenotypes were also observed in PCCA ϩ/ϩ mice that had undergone dietary treatment for 16 h at the newborn period ( not shown), implying that the deposition appeared to be rapidly generated in proportion to the progression of acidosis after birth. Also in human, a few cases of similar fat deposition in the liver of infant severe-form PA patients have been also reported (17,18). Accelerated dehydration and malnutrition in addition to accumulation of propionyl-CoA might disturb the general oxidation of fatty acids in the liver, perhaps resulting in their accumulation and deposition in the liver tissue of PCCA Ϫ/Ϫ mice (17)(18)(19)(20)(21)(22)(23). There was, however, no difference in RNA expression of various genes related to either synthesis or oxidation of fatty acids, e.g. SREBP-1 (a transcription factor), fatty acid synthase, glucose 6-phosphate dehydrogenas, acyl-CoA oxidase, and carnitine palmitoyl CoA transferase in PCCA ϩ/ϩ and PCCA Ϫ/Ϫ mice as assessed by Northern blotting (data not shown). Further study will be required to clarify the precise mechanism of the fat deposition in the liver. This fat deposition in the liver does not seem to provoke hepatic dysfunction because serum concentrations of various parameters for hepatic damage, such as GOT, GPT, ␥-GTP and bilirubin, were within the normal range in PCCA Ϫ/Ϫ mice and comparable with those in PCCA ϩ/ϩ mice (data not shown). Overall, the PCCA Ϫ/Ϫ mouse properly represents the severe form of PA patients and thus is a useful animal model to test therapies for the disease.
Prevention of Lethal Ketoacidosis in Early Infant PCCA Ϫ/Ϫ Mice by a Weak, Liver-specific PCCA Expression via Transgene-By using PCCA Ϫ/Ϫ mice, we attempted to address the potential efficacy of PCC gene supplementation as an alternative strategy that might prevent PA more efficiently. In general, gene transfer via the DNA vaccination or the virus-based vectors has been studied extensively (6 -9, 24 -26). It appears, however, impossible to directly test either strategy in PCCA Ϫ/Ϫ mice because unlike human patients, PCCA Ϫ/Ϫ mice die very shortly (in 24 -36 h) after birth, and that is not sufficient time for transferred genes to yield efficient expression. Thus, we designed a PCCA transgenic mouse system, in which the transgene expression pattern is similar to that in postnatal gene transfer therapies i.e. 1) the transgene will not express during the embryonic stages, 2) the transgene will express in a tissue specific fashion, and 3) the supplementation of PCC activity per a tissue by the transgene will be within an achievable level by gene therapies in general according to the previous reports of various other genes (27)(28)(29)(30). To satisfy these conditions, we generated transgenic mice expressing PCCA under the regulation of the serum amyloid P component (SAP) gene promoter based on the following characteristics of SAP promoter (11). First, the SAP promoter regulates the target gene to express exclusively in the liver (11,31,32). The liver appears ideal as a recipient organ for PCC gene transfer because the liver tissue is one of the primary organs that metabolize the odd-numbered amino acids and fatty acids. Second, it is known that a transgene under the SAP promoter regulation starts its expression only after birth, preserving a low expression level for the first few days and increases its expression thereafter (31,32). Hence, SAP-PCCA transgenic mouse on a PCCA Ϫ/Ϫ background (SAP ϩ PCCA Ϫ/Ϫ ) appears to enable one to address the therapeutic efficacy of supplementation of a low amount of PCC activity in the liver during the newborn and the early infant periods.
Three lines of SAP-PCCA transgenic mice were generated and crossbred to PCCA Ϫ/Ϫ background to obtain SAP ϩ PCCA Ϫ/Ϫ mice. As expected, the transgene expression was restricted in the liver in all of three transgenic mouse lines as assessed by Northern blotting (data not shown). Two lines of SAP ϩ PCCA Ϫ/Ϫ mice were analyzed together with PCCA Ϫ/Ϫ and PCCA ϩ/ϩ littermate mice (both two SAP ϩ PCCA Ϫ/Ϫ lines revealed identical phenotypes). At birth, all of these types of mice were indistinguishable and started sucking milk. After 24 h, most of PCCA Ϫ/Ϫ mice revealed severe ketonuria, and died as described above. At the same time point, however, SAP ϩ PCCA Ϫ/Ϫ mice were healthy and TABLE I Rapid increase of abnormal metabolites in PCCA Ϫ/Ϫ mice Blood and urine (10 -20 l) were collected from either PCCA Ϫ/Ϫ (Ϫ/Ϫ) or PCCA ϩ/ϩ (ϩ/ϩ) mice at various ages on a filter paper (Guthrie card). Each number is the average of the results of 4 -6 mice. Urine 3-HP and MC were analyzed by the selected ion monitoring (SIM) method (36). Urine ketone was assessed by a test-tape (Lab-sticks, Bayer Co. Ltd). Ϫ/Ϫ mice at 24 h after birth showed the maximum level (ϩϩϩϩ (14) of the PCCA ϩ/ϩ sections. Liver sections of both types of mouse embryos at E18.5 were also PAS-stained and analyzed (D, E). PAS (ϩ) products are observed also in the PCCA Ϫ/Ϫ embryo liver. F, PAS-staining of the kidney sections of PCCA ϩ/ϩ and G, PCCA Ϫ/Ϫ . Enlarged collecting ducts harboring the hyaline droplets in PCCA Ϫ/Ϫ kidney is indicated by arrows. PAS (ϩ) layer is also observed on the interior surface of the collecting duct epithelium both in PCCA Ϫ/Ϫ and PCCA ϩ/ϩ , which is due to normal reabsorbing function by the collecting ducts. H-K, wild-type (PCCA ϩ/ϩ ) newborns had undergone a dietary treatment for 16 h after birth (fed with phosphate-buffered saline and kept in an incubator at 32°C with humidity), and then their liver and kidney were analyzed by PAS-staining (H, J) together with samples from mice that had been fed by maternal milk as a normal control (I, K). After the dietary treatment the PAS (ϩ) products in the liver were almost disappeared (H), and the collecting ducts in the kidney were enlarged and revealed incorporation of hyaline droplets (J, indicated by arrows). L, M, analysis for fat deposition (oil red O staining) of the liver from PCCA ϩ/ϩ or PCCA Ϫ/Ϫ at 24 h after birth. Red spots indicate fat (lipid) droplets. Magnifications: ϫ400.
continued sucking, thus their stomachs were full with milk like PCCA ϩ/ϩ mice. The ketouria of SAP ϩ PCCA Ϫ/Ϫ mice was significantly milder than that of PCCA Ϫ/Ϫ mice (ϩ ϳ ϩϩ in SAP ϩ PCCA Ϫ/Ϫ versus ϩϩϩϩ in PCCA Ϫ/Ϫ ). After 36 h, in contrast to PCCA Ϫ/Ϫ mice that had all died, all of SAP ϩ PCCA Ϫ/Ϫ littermate mice had survived. SAP ϩ PCCA Ϫ/Ϫ mice continued growing comparatively in size with PCCA ϩ/ϩ mice, and showed no abnormalities in behavior, responses to various stimuli, sleep, and sexuality. Both male and female adult SAP ϩ PCCA Ϫ/Ϫ mice revealed no defect in breeding efficiency. Fig. 3 demonstrates the RNA expression of the transgene and the PCC activity in the liver as well as the concentrations of the serum propionyl-carnitine and the urine 3-HP in SAP ϩ PCCA Ϫ/Ϫ mice at various ages. As expected, no transgene expression, thus also no PCC activity, was detected in the E12 embryonic liver. Transgene was not expressed even at the late embryonic stage, E18.5, which is immediately before the birth (data not shown). Twenty-four h after birth, the PCC activity in the SAP ϩ PCCA Ϫ/Ϫ mouse liver was 10 -20% of that in the PCCA ϩ/ϩ mouse liver, supplemented by a low expression of the transgene (ϳ20% of the level of the endogenous PCCA mRNA expression in PCCA ϩ/ϩ liver; compare with 24 h ϩ/ϩ lane). The concentrations of both propionyl-carnitine and 3-HP in these mice were significantly decreased as compared with those in PCCA Ϫ/Ϫ mice. Thus, supplementation of 10 -20% PCC activity only in the liver was sufficient to prevent the lethal fate of neonatal PCCA Ϫ/Ϫ mice.
This low PCCA transgene expression in SAP ϩ PCCA Ϫ/Ϫ mice was preserved until 6 -8 days of age, resulting in persistence of high propionyl-carnitine concentration and ketonuria, which were in the comparable levels to those at 24 h after birth (data not shown). SAP ϩ PCCA Ϫ/Ϫ mice, however, showed no growth abnormality as compared with PCCA ϩ/ϩ mice, as described above. The transgene expression started to increase after 8 -10 days of age (data not shown), and at 3 weeks of age, SAP ϩ PCCA Ϫ/Ϫ mice gained comparable PCC activity in the liver with PCCA ϩ/ϩ mice (Fig. 3). This increase of PCC activity was followed by further decrease of propionyl-carnitine and 3-HP concentrations. Thereafter, these concentrations in SAP ϩ PCCA Ϫ/Ϫ mice were at plateau levels similar to those at 3 weeks of age.
The glycogen storage in the liver, which seemed to be consumed rapidly in PCCA Ϫ/Ϫ mice in 24 h after birth, was entirely recovered by transgene expression, and was comparable in SAP ϩ PCCA Ϫ/Ϫ and PCCA ϩ/ϩ mice either at 24 h or 3 weeks of age (Fig. 4, A-E). In addition, the pathological change observed in the kidney of PCCA Ϫ/Ϫ mice, the enlargement of collecting ducts with marked incorporation of hyaline droplet casts, was decreased in SAP ϩ PCCA Ϫ/Ϫ mice (Fig. 4, F-H). These results may be explained mainly by the prevention of dehydration and malnutrition in SAP ϩ PCCA Ϫ/Ϫ mice.
Fat deposition in the liver was apparently decreased in SAP ϩ PCCA Ϫ/Ϫ mice at 24 h of age, though the severity was varied in each individual: some mice were entirely free from the fat-deposition whereas others harbored the deposition that was, however, obviously milder than that in PCCA Ϫ/Ϫ mice (data not shown). Interestingly, at 3 weeks of age, most of SAP ϩ PCCA Ϫ/Ϫ mice revealed no fat-deposition in the liver (data not shown). These results clearly demonstrate that the fatty-change of the liver is reversible, and in addition, supplementation in the liver of at least 10 -20% of the level of the endogenous PCC activity is sufficient for decreasing the fatdeposition in PCCA Ϫ/Ϫ mice. The lack of marked dehydration and malnutrition might have also contributed to the prevention of the fat deposition in SAP ϩ PCCA Ϫ/Ϫ mice.
Requirement of a Larger Amount of PCC Supplementation after the Late Infant Period-Interestingly, the third line of SAP ϩ PCCA Ϫ/Ϫ mouse revealed different expression pattern of transgene, probably due to the influence of the transgene integrating site in the chromosome. In this specific SAP ϩ PCCA Ϫ/Ϫ line (SAP lowϩ PCCA Ϫ/Ϫ ), unlike in the other two lines of SAP ϩ PCCA Ϫ/Ϫ mice described above, a low level of transgene expression, thus low PCC activity (10 -20% of that in PCCA ϩ/ϩ ) in the liver, persisted even after 2 weeks of age (data not shown). Although these SAP lowϩ PCCA Ϫ/Ϫ mice survived for the first 2 weeks, which is the newborn and early infant periods, they died before 3 weeks of age, demonstrating severe ketonutia (ϩϩϩ ϳ ϩϩϩϩ) and dehydration. This result strongly implicated that a larger amount of PCC supplementation is required in proportion to the increase of protein intake as mice grow. This result may also explain the diversity of the age in the first onset of the severe ketoacidosis in human PA patients, each of which possesses different PCC activity depending on the type of PCC gene (either PCCA or PCCB) mutation (2,3,33). DISCUSSION In this report, we established a mouse model that resembles newborn PA patients who have undergone PCC gene transfer in a liver-specific fashion. These mice, SAP ϩ PCCA Ϫ/Ϫ , provided important information as summarized bellow. First, a postnatal supplementation of at least 10 -20% of endogenous PCC activity exclusively in the liver prevents the fatal fit of severe ketoacidosis during the newborn and early infant periods. Second, a persistent mild ketoacidosis (revealing ϳ6to 8-fold higher concentration of propionyl-carnitine than that in wild-type mice and a mild ketonuria) appears not to essentially influence infant development. This is reminiscent of the normal embryonic development in PCCA Ϫ/Ϫ mice, in which a certain accumulation of propionic acid appeared to be present during the embryonic period as described above. Lastly, 10 -20% supplementation of PCCA activity efficiently tempers the fat-deposition observed in the liver of PCCA Ϫ/Ϫ mice. Therefore, it is also very likely that in human, the PCC gene transfer into the liver may be effective to prevent the fatal acceleration of ketoacidosis in neonatal PA patients, thus preserving normal growth during the early infant period. In addition, the early death of SAP lowϩ PCCA Ϫ/Ϫ mice, which harbored a persistent low level expression of PCCA transgene even after the late infant stage, clearly implicated that a larger amount of PCC supplementation appears to be required in proportion to the growth of patients.
Based on these results of SAP ϩ PCCA Ϫ/Ϫ and SAP lowϩ PCCA Ϫ/Ϫ mice, we propose a two-step therapy to achieve efficient prevention of PA complaints in human patients: a partial supplementation of PCC activity exclusively in the liver, followed by a high amount of PCC supplementation in the late infant period. Because the "first step" treatment requires partial (at least 10 -20%) and transient (only during the newborn and early infant periods) supplementation, a virus-based gene transfer in the liver (of either PCCA or PCCB gene depending on which gene is defected in each patient) (34) may be a plausible strategy despite the following disadvantages implicated for the virus-based gene transfer. For example, 1) low supplementation efficiency (in a whole tissue) due to the low expression and limited distribution of the virus-infected area, and 2) difficulty of long term persistence of expression due to the instability of the transfected gene or production of antibodies against either the virus or the introduced gene (35). For the "second step" therapy, a more drastic treatment, such as liver-transplantation, which has been attempted so far in two infant PA patients (5), might be required. The second step therapy, however, will be performed in late infant patients who have been well controlled by the first step therapy. This will lead to less severe damage for patients than such a drastic therapy done in the newborn or the early infant periods as the initial treatment. Judging by the observation from SAP ϩ PCCA Ϫ/Ϫ mice, whether the serum propionyl-carnitine concentration is within ϳ6to 8-fold higher than that in the normal population, might be referred to as a criterion to decide the frequency of the first step gene-transfer treatment as well as the opportunity to perform the second step treatment.
It may be worth emphasizing that SAP ϩ PCCA Ϫ/Ϫ mice showed no apparent clinical complaints without any dietary limitation. This suggests that the two-step therapy in human PA patients also might require no strict dietary limitation. It is, however, strongly possible that more efficient therapeutic results may be obtained if dietary limitation therapy is combined.
In summary, PCCA Ϫ/Ϫ mice enabled us to observe the PA progression in a kinetic fashion from various aspects, including histology, which provided us new insights on the pathogenesis contributing to the severe illness. In addition, SAP ϩ PCCA Ϫ/Ϫ mice indicated the potential efficacy of the two-step PCC supplementation therapy for human PA patients to prevent the fatal acceleration of ketoacidosis, which appears to result in essentially no clinical complaints even without a strict dietary limitation. These two types of mice will be useful animal models for further understanding of PA pathogenesis, as well as to test a variety of strategies toward establishment of the thorough therapy of PA.