Human Molecular Genetics

Human Molecular Genetics Advance Access originally published online on March 17, 2004

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 13/9/905    most recent ddh112v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (26) Request Permissions Google Scholar Articles by Schmidt, A. Articles by Isbrandt, D. Search for Related Content PubMed PubMed Citation Articles by Schmidt, A. Articles by Isbrandt, D. Social Bookmarking          What's this?

Human Molecular Genetics, 2004, Vol. 13, No. 9 905-921
DOI: 10.1093/hmg/ddh112

Severely altered guanidino compound levels, disturbed body weight homeostasis and impaired fertility in a mouse model of guanidinoacetate N-methyltransferase (GAMT) deficiency

Andreas Schmidt¹, Bart Marescau², Ernest A. Boehm³, W. Klaas Jan Renema⁴, Ruben Peco¹, Anib Das^5,, Robert Steinfeld^5,, Sharon Chan³, Julie Wallis³, Michail Davidoff⁶, Kurt Ullrich⁵, Ralph Waldschütz¹, Arend Heerschap⁴, Peter P. De Deyn², Stefan Neubauer³ and Dirk Isbrandt^1,*

¹Center for Molecular Neurobiology Hamburg (ZMNH), Institute for Neural Signal Transduction, Hamburg, Germany, ²Laboratory of Neurochemistry and Behaviour, Born-Bunge Foundation, University of Antwerp, Belgium, ³Department of Cardiovascular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK, ⁴Department of Radiology, UMC Nijmegen, Nijmegen, The Netherlands and ⁵Department of Pediatrics and ⁶Department of Anatomy, University Hospital Hamburg, Hamburg, Germany

Received December 9, 2003; Accepted March 9, 2004

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
We generated a knockout mouse model for guanidinoacetate N-methyltransferase (GAMT) deficiency (MIM 601240), the first discovered human creatine deficiency syndrome, by gene targeting in embryonic stem cells. Disruption of the open reading frame of the murine GAMT gene in the first exon resulted in the elimination of 210 of the 237 amino acids present in mGAMT. The creation of an mGAMT null allele was verified at the genetic, RNA and protein levels. GAMT knockout mice have markedly increased guanidinoacetate (GAA) and reduced creatine and creatinine levels in brain, serum and urine, which are key findings in human GAMT patients. In vivo ³¹P magnetic resonance spectroscopy showed high levels of PGAA and reduced levels of creatine phosphate in heart, skeletal muscle and brain. These biochemical alterations were comparable to those found in human GAMT patients and can be attributed to the very similar GAMT expression patterns found by us in human and mouse tissues. We provide evidence that GAMT deficiency in mice causes biochemical adaptations in brain and skeletal muscle. It is associated with increased neonatal mortality, muscular hypotonia, decreased male fertility and a non-leptin-mediated life-long reduction in body weight due to reduced body fat mass. Therefore, GAMT knockout mice are a valuable creatine deficiency model for studying the effects of high-energy phosphate depletion in brain, heart, skeletal muscle and other organs.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Creatine, phosphocreatine and the creatine kinase reaction are long known components of the high-energy phosphate metabolism of cells and tissues with high or rapidly changing energy demand (1). The recent identification of creatine deficiency syndromes, which are caused by defects in creatine biosynthesis (2,3) or transport (4), highlights the importance of intact creatine metabolism for psychomotor development and cognitive function in humans.

The biosynthesis of creatine involves the enzymes arginine : glycine amidinotransferase (AGAT) and guanidinoacetate N-methyltransferase (GAMT, EC 2.1.1.2) (Fig. 1). AGAT forms guanidinoacetate (GAA) and ornithine in a rate-limiting reaction from arginine and glycine. Subsequently, GAMT catalyzes the formation of creatine through methyl group transfer from S-adenosylmethionine to GAA (5). In cells depending on exogenous creatine synthesis, creatine is actively taken up by the creatine transporter (1). Creatine is a central compound in the energy metabolism of cells in tissues with a highly fluctuating energy demand, such as heart, skeletal muscle and brain. Reversible phosphorylation of creatine to phosphocreatine by the five known creatine kinase isoenzymes provides a temporal and spatial energy buffering system in these tissues. Due to non-enzymatic conversion of creatine to creatinine, which is finally excreted in urine, the creatine body pool must be maintained by nutritional intake and de novo synthesis. The latter is believed to be mainly localized to liver, kidney and pancreas (1).

View larger version (16K): [in this window] [in a new window]   Figure 1. Biosynthesis of creatine and disorders of creatine metabolism. Creatine is synthesized in two steps. First, AGAT catalyzes the formation of GAA, the immediate precursor of creatine. Subsequently, creatine is formed by methyl group transfer from S-adenosyl-L-methionine to GAA catalyzed by GAMT. Creatine is utilized within the synthesizing cell or actively taken up via the creatine transporter (CRTR) by cells dependent on exogenous creatine synthesis. Creatine is a substrate of the different creatine kinase (CK) isoforms and can be phosphorylated to creatine phosphate. The three hitherto identified disorders of creatine metabolism comprise AGAT deficiency, GAMT deficiency and the CRTR defect.

 
GAMT deficiency, the first discovered creatine deficiency syndrome (MIM 601240), is an autosomal recessively inherited disorder of creatine biosynthesis (2,6,7). The disease usually manifests as developmental delay or arrest during the first months of life. Neurological symptoms are heterogeneous and may include muscular hypotonia and weakness, poor head control, extrapyramidal movement disorders, epilepsy and autistic or self-aggressive behavior in older patients (2,6–10). The diagnosis is based on the detection of excessive amounts of guanidinoacetate (GAA), the precursor of Cr, in body fluids such as urine, serum and cerebrospinal fluid (CSF), which is a pathognomonic feature of GAMT deficiency (7). Using MR spectroscopy, GAMT deficiency-related changes such as the absence of creatine/phosphocreatine and the accumulation of significant amounts of GAA can be detected in the brains of patients (6).

The pathophysiology of GAMT deficiency may involve both deficiency of high-energy phosphates in neurons and neurotoxic and/or neuromodulatory action of GAA, a partial agonist at GABA_A receptors (11). Single case studies of GAMT-deficient patients indicate that oral creatine supplementation leads to some improvement in clinical symptoms but is not sufficient to cure the disease (7). The same applies to a more complex dietary approach, which was more successful in the treatment of therapy-resistant epilepsy (12). Since systematic patient studies on the pathophysiology of GAMT deficiency cannot be conducted due to the low incidence of the disease, we decided to generate a knockout mouse model by targeted disruption of the murine GAMT gene in embryonic stem cells.

Here we show that GAMT expression profiles are very similar in humans and mice and much more widespread than previously thought. GAMT knockout mice exhibit key biochemical changes in body fluids, brain and muscle that are typical of human GAMT deficiency. Furthermore, we show that GAMT deficiency in mice is associated with increased neonatal mortality, life-long reduction in body weight due to reduced body fat content, muscular hypotonia and decreased male fertility. Therefore, GAMT knockout mice provide a valuable model for studying the effects of high-energy phosphate depletion in brain, heart, skeletal muscle and other organs.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Expression profiles of human and mouse GAMT
Comparable expression patterns of GAMT in human and mouse tissues are essential for the generation of a mouse model that is aimed at studying human GAMT deficiency. We performed a detailed expression analysis of human (h) and mouse (m) GAMT using molecular biological methods that are more sensitive than biochemical assays, which were previously used to localize GAMT activity mainly to liver and pancreas (1). We investigated the expression profiles of hGAMT and mGAMT mRNA and protein by means of northern and western blots containing samples from human and mouse tissues (Figs 2 and 3). Northern blot analysis with a ³²P-radiolabeled cDNA probe of hGAMT (13) demonstrated that hGAMT mRNA was present in high amounts in skeletal muscle, liver, heart, kidney and in smaller amounts in brain (Fig. 2Ai). In human brain (Fig. 2Aii–iii), the highest hGAMT mRNA levels were found in cerebellum, cerebral cortex, medulla, caudate nucleus and thalamus. Lower amounts were detected in corpus callosum, amygdala, putamen, hippocampus, substantia nigra, subthalamic nucleus as well as in occipital, frontal and temporal lobes. The estimated hGAMT transcript size is about 1.2 kb and thus corresponds well with our previous observations in human liver total RNA (2). No alternative transcripts were detected. Our results are in line with biochemical data on GAMT activity in liver and kidney. However, the high expression found in skeletal muscle and heart and the ubiquitous expression in brain were unexpected, because creatine synthesis (and hence GAMT expression) was previously mainly attributed to liver, kidney and pancreas (1). In order to confirm our results and to analyze tissues other than those used for the northern blots, we hybridized the hGAMT cDNA probe to a human multiple tissue mRNA dot blot (Fig. 2B). Again, high expression of hGAMT was detected in skeletal muscle, cardiac tissues (especially ventricles), liver, pancreas and kidney. Lower levels were found in all brain areas, adrenal, salivary, thyroid and mammary glands, trachea, testis and ovary. In the intestinal tract and lymphatic tissue, only low or background expression levels were observed. Furthermore, hGAMT transcripts were detected in fetal tissues such as liver, heart, kidney and in small amounts in brain.

View larger version (92K): [in this window] [in a new window]   Figure 2. Expression profiles of GAMT mRNA and protein in human tissues. (A) Multiple tissue and brain northern blots (Clontech, Palo Alto, CA, USA). Poly (A+) mRNA of adult human tissues (i) and specific human brain areas (ii, iii), as indicated at the top of each lane, was hybridized with an (-32P)-labeled cDNA probe specific for hGAMT. Size markers (in kb) are indicated on the left side. (B) Multiple tissue northern dot blot (Clontech) containing poly (A+) mRNA of adult human tissues was hybridized as described in (A). The tissue source of the mRNA preparation is given in the right panel. (C) Western blot with selected human tissue lysates (Clontech) probed with affinity-purified polyclonal anti-hGAMT antibodies. A size marker (in kDa) is given on the left. Bands with lower molecular weight observed in brain, skeletal muscle and liver most likely represent degradation products because the samples originate from post-mortem tissues. PBL, peripheral blood lymphocytes.

 

View larger version (121K): [in this window] [in a new window]   Figure 3. Expression profiles of GAMT mRNA and protein in mouse tissues. (A) (i) Multiple tissue northern blot containing 10 µg total RNA of adult mouse tissues, as indicated at the top of each lane, was hybridized with an (-32P)-labeled cDNA probe specific for mGAMT. The marker on the left indicates the localization of 18S RNA. (ii) Multiple tissue western blot with lysates of the same tissues that were used in (Ai) was probed with affinity-purified polyclonal anti-mGAMT antibodies. Size marker (in kD) is given on the left. (B) Western blots with total embryo (Bi) or postnatal (Bii) mouse tissues, as indicated on the left, were probed as described in (Aii). (C) Immunoperoxidase staining by the affinity-purified anti-mGAMT antibodies in coronal sections through adult mouse brain reveals glial expression of mGAMT. Coronal section from the corpus callosum of wild-type (Ci) or GAMT–/– animals (Cii). Since no immunoperoxidase staining was visible in (Cii), the image was taken in phase contrast to reveal tissue structures. At higher magnification, distinct immunolabeling is visible in oligodendrocytes in corpus callosum (Ciii) and striatum (Civ, v) or in astrocytic cells in cerebral cortex (Cvi). Magnifications: left column 10x, middle column 20x, right column 40x.

 
Next, we investigated whether the high mRNA levels in liver, kidney, skeletal muscle and brain would result in detectable hGAMT protein amounts. We used commercially available protein preparations from these human tissues and our polyclonal anti-hGAMT antiserum raised against recombinant hGAMT purified from Escherichia coli (see Materials and Methods). Human GAMT immunoreactivity was detected in all four tissue samples, with highest amounts in liver and skeletal muscle (Fig. 2C), whereas no hGAMT immunoreactivity could be identified in the intestinal tract, a tissue in which hGAMT-specific RNA was absent as well.

The expression profile of mGAMT in adult mouse tissues proved to be very similar to that found in samples of human origin. Using a ³²P-radiolabeled mGAMT cDNA probe covering the entire reading frame, strong mGAMT-specific signals were detected in total RNA from liver, kidney, skeletal muscle, brain, testis and ovary (Fig. 3Ai). Lower intensities were observed in spleen and heart, and only faint signals were detected in stomach, intestine and lung. As in human brain, comparable amounts of GAMT transcripts were found in all brain areas analyzed, including cortex, hypothalamus, striatum and cerebellum.

Western blot analysis using our polyclonal anti-mGAMT antibodies (see Materials and Methods) and the lysates of the tissues that had been employed for northern blot analysis showed mGAMT-specific immunoreactivity in all tissues containing mGAMT-specific transcripts, although the observed intensities varied (Fig. 3Aii). In mouse embryos, mGAMT immunoreactivity was present at all stages investigated (E7–E17) and tended to increase with development (Fig. 3Bi). Furthermore, mGAMT protein was detected in postnatal (P1–P21) brain, liver and skeletal muscle (Fig. 3Bii).

The ubiquitous expression pattern of GAMT transcripts and protein in brain prompted us to investigate the GAMT distribution in brain at higher resolution. In coronal sections from adult mouse brain, mGAMT immunoreactivity was widespread but mainly restricted to glial cells, such as oligodendrocytes in corpus callosum (Fig. 3Ci and iii) and striatum (Fig. 3Civ and v), or to astrocytes in cortex (Fig. 3Cvi). In contrast, immunoreactivity could be detected neither in brain sections from wild-type mice incubated with blocked antibodies (not shown) nor in sections from GAMT^–/– mice (Fig. 3Cii).

In conclusion, the largely overlapping GAMT mRNA and protein expression patterns in human and mouse tissues clearly suggest that a mouse model is suitable for the study of GAMT deficiency. The generation of GAMT knockout mice is described below.

Targeted disruption of the mGAMT gene
In order to construct a targeting vector for disruption of the mGAMT gene, we screened a 129 SV/J mouse genomic library and isolated a single clone containing the complete mGAMT open reading frame (ORF). Restriction and sequence analysis of the clone revealed that the mGAMT ORF is encoded in six exons (Fig. 4A), which is in agreement with previously published data (14). The strategy for targeted disruption of the mGAMT ORF in ES cells is shown in Figure 4A. Insertion of a neomycin selection cassette in the MluI site in exon 1 of the mGAMT gene caused a frame shift and premature stop codon 15 codons downstream of the insertion site and resulted in the elimination of 210 of the 237 amino acids present in wild-type mGAMT. Homologous recombination in neomycin-resistant embryonic stem cells was identified by Southern blot analysis of genomic DNA using 5' and 3' external probes (Fig. 4Bi). The respective embryonic stem cells were microinjected into C57Bl/J6 blastocyst stage embryos. Two of three chimeric mice that were mated gave rise to germ-line transmission of the disrupted allele, the offspring of which was randomly intercrossed to obtain mGAMT^+/+, mGAMT^+/– and mGAMT^–/– progeny. Genotypes of the offspring were determined by genomic Southern blotting (Fig. 4Bii) and/or PCR with appropriate primer pairs (Fig. 4Biii).

View larger version (25K): [in this window] [in a new window]   Figure 4. Targeting of the mGAMT gene locus by homologous recombination in ES cells. (A) Schematic representation of mGAMT gene locus (top), targeting vector (middle) and mGAMT gene locus after the homologous recombination event (bottom). A fragment containing the complete mGAMT ORF with six exons was isolated from a SV129 genomic library. Insertion of a neomycin selection cassette (neo) into the MluI site resulted in disruption of the gene in exon 1 and in insertion of a premature stop codon (stop). Lines below top and bottom panels give the locations of 5' and 3' external probes used for Southern blot genotyping. PCR primer sites are indicated with open arrows. (B) Southern blots of DraI or HindIII-digested genomic DNA were performed to detect homologous recombination at the mGAMT gene locus in ES cells (Bi) and the presence of knockout alleles in mice (Bii). Blots were hybridized with (-32P)-labeled 5' and 3' external DNA probes homologous to sequences outside the genomic fragment present in the targeting construct. Presence of the knockout allele caused distinct 1.5-kb size shifts of the respective restriction fragment, namely a reduction for the DraI digest (Bi, left) and an increase for the HindIII digest (Bi, right and ii). PCR analyses with internal primers P1, P2 and P3 (see A) were used for routine genotyping of genomic DNA isolated from mouse ear biopsies (Biii).

 
To demonstrate the creation of a GAMT null allele, we performed northern and western blot experiments with liver preparations obtained from GAMT-deficient and control animals. A northern blot with total liver RNA was hybridized with a ³²P-radiolabeled mGAMT cDNA probe (see above) and confirmed the absence of GAMT-specific mRNA, because no signal could be detected for GAMT^–/– animals (Fig. 5Ai). Subsequently, the blot was hybridized with a mouse glyceraldehyde-3-phosphate dehydrogenase (mGAPDH) probe in order to verify the integrity of the RNA and blotting quality (Fig. 5Aii). In addition, western blots with liver lysates were probed with our polyclonal anti-mGAMT antiserum and showed that mGAMT-specific immunoreactivity was completely absent (Fig. 5Bi) in samples from GAMT^–/– animals, although equal amounts of protein had been loaded (Fig. 5Bii). Taken together, these experiments provide evidence for the successful generation of an mGAMT null allele at the genetic, RNA and protein levels.

View larger version (35K): [in this window] [in a new window]   Figure 5. Northern and western blot analysis of mGAMT–/–, mGAMT+/– and mGAMT+/+ mice. (A) (i) A northern blot containing total liver RNA isolated from mice with the genotypes indicated at the top of each lane was hybridized with an (-32P)-labeled cDNA probe specific for mGAMT. (ii) Subsequent to the detection of mGAMT-specific RNA, the northern blot was hybridized with an (-32P)-labeled cDNA probe specific for mouse GAPDH. Size markers on the left indicate the positions of 28S and 18S RNAs. (B) (i) A western blot with liver lysates obtained from the same mice that were used in (Ai) was probed with affinity-purified polyclonal anti-mGAMT antibodies. Size markers (in kDa) are given on the left. (ii) Coomassie blue-stained SDS polyacrylamide gel loaded with 20 µg of the same liver homogenates used for the western blot in (Bi).

 
Analysis of guanidino compounds
The biochemical hallmarks of human GAMT deficiency are reduced levels of creatine (Cr) and creatinine (Crn) and the accumulation of GAA in serum, urine and cerebrospinal fluid (CSF) (7). We therefore investigated the levels of these guanidino compounds in serum, brain homogenates and 24 h urine samples from GAMT-deficient, heterozygous and wild-type animals of both sexes (Table 1). No significant differences were found in the concentrations of these substances in the three types of samples between wild-type and heterozygous mice. In clear contrast, marked and characteristic changes were observed in homozygous GAMT-deficient mice. Creatine deficiency in GAMT knockout animals was present in urine, serum and brain samples (Fig. 6 and Table 1). Mean creatine concentrations were decreased 14- to 122-fold. Likewise, the concentrations of creatinine, the breakdown product of creatine, were reduced 7- to 27-fold (Fig. 6 and Table 1). In contrast, GAA, key metabolite that accumulates due to the metabolic block in GAMT deficiency, was increased on average 10-fold in urine, 52-fold in serum and 142-fold in brain (Fig. 6 and Table 1). In addition, further abnormalities were observed in the composition of guanidino compounds in brain homogenates, such as, for example, a 13-fold increase in guanidinosuccinic acid (GSA) and a 6-fold increase in ß-guanidinopropionic acid (GPA) and -guanidinobutyric acid (GBA) [controls versus GAMT^–/– (in nmol/g tissue, P<0.001 for all comparisons): GSA 0.18±0.02 versus 2.35±0.21, GPA 0.17±0.01 versus 1.08±0.08, GBA 1.66±0.07 versus 10.7±0.9].

View this table: [in this window] [in a new window]   Table 1. Guanidino compounds in urine, serum and brain of GAMT–/– and control mice

 

View larger version (25K): [in this window] [in a new window]   Figure 6. Relative changes of diagnostically relevant guanidino compounds in GAMT–/– mice. The guanidino compounds creatine (Cr), creatinine (Crn) and GAA used in the clinical diagnosis of GAMT deficiency were determined biochemically in brain, serum and 24 h urine of GAMT–/– and control mice. Increases or decreases in the respective compound levels are expressed as mean fold change (±range) in knockout mice in comparison to the levels found in controls. Please note the logarithmic scale. The corresponding absolute concentrations and statistical comparisons are given in Table 1.

 
In summary, the biochemical analyses of brain tissue, serum and urine of GAMT knockout mice show severe creatine deficiency and accumulation of GAA. These findings closely resemble those previously observed for GAMT patients and thus prove the success of our knockout strategy at the metabolic level.

In vivo ¹H and ³¹P MRS analyses of brain and skeletal muscle
The first GAMT patient was identified by detection of a reduced Cr/CrP signal on ¹H MRS (6). Since then, ¹H or ³¹P MRS has been used to confirm the diagnosis and to monitor the therapeutic success of creatine substitution therapy in GAMT-deficient patients. Therefore, we investigated the extent to which biochemical changes in brain homogenates, serum and urine are reflected in in vivo ¹H or ³¹P MRS of GAMT^–/– mice. Localized ¹H MRS in muscle and brain showed detectable but markedly reduced Cr/CrP resonance signals in comparison to controls (Fig. 7A, B, E and F). In contrast to human studies, a GAA ¹H MRS resonance signal was not observed in vivo. The exact reasons for these findings are unclear. However, we detected a GAA singlet in GAMT^–/– mice using high resolution in vitro studies of intact muscles (15), which might point to a methodological problem. In addition to reduced CrP levels, localized ³¹P MRS of GAMT^–/– brain or muscle revealed another signal near the CrP resonance. This signal corresponded to guanidinoacetate phosphate (PGAA, Fig. 7D and H) and was not detected in control spectra (Fig. 7C and G). Interestingly, the relative amounts of PGAA compared to CrP in muscle were much higher than those observed in brain of GAMT^–/– mice. Furthermore, phosphorylation of GAA indicated that it might serve as a substrate in creatine kinase reactions. These in vivo MRS measurements confirmed creatine deficiency and PGAA accumulation in brain and skeletal muscle of GAMT^–/– mice. A detailed quantitative analysis of the ¹H and ³¹P MRS data and of experimental results suggesting that GAA can be reversibly phosphorylated was performed by us in a separate MRS study (15).

View larger version (24K): [in this window] [in a new window]   Figure 7. Localized in vivo 1H and 31P MRS in brain and skeletal muscle of GAMT–/– and control mice. Representative localized 1H and 31P MRS analyses of brain (A–D) and hindleg muscle (E–H) obtained from control or GAMT–/– (–/–) mice. The left column shows localized 1H MR spectra, whereas the right column shows localized 31P MR spectra. Peak assignments: Pi, inorganic phosphate; Cr, creatine; CrP, creatine phosphate; PGAA, guanidinoacetate phosphate. The resonances of nucleotide phosphates (e.g. ATP) are labeled with , ß and . Vertical scaling is arbitrary. The circles in (D) and (H) show a higher magnification of the double peak observed in knockout animals.

 
Respiratory chain activities in brain and skeletal muscle
As creatine levels were markedly reduced in brain and skeletal muscle of GAMT^–/– mice and marked changes in in vivo ATP concentrations were not present, we hypothesized that adaptive changes in oxidative phosphorylation might be present in both tissues. We therefore measured respiratory chain activities in muscle and brain homogenates. While complex I–IV activities in GAMT^–/– and GAMT^+/+ mice were comparable for both tissues, complex V (ATP synthase) activities were increased by 68% in brain and by 80% in skeletal muscle of GAMT^–/– mice (Table 2). When normalized to citrate synthase, a mitochondrial marker enzyme, ATP synthase activities were still significantly higher in brains from GAMT^–/– mice (Table 2). Due to the observed (albeit not statistically significant) increase in citrate synthase activity in GAMT^–/– muscles, normalized ATP synthase activities did not differ between knockout and control animals (Table 2). Active regulation of ATP synthase, i.e. down-regulation in response to mitochondrial uncoupling or anoxia and up-regulation in response to increased calcium concentration in the incubation medium, was normal in both brain and muscle from GAMT^–/– and control mice (data not shown). These findings indicate adaptive changes in the activity of ATP synthase in mitochondria of GAMT-deficient muscle and brain tissue. Furthermore, higher citrate synthase activities point to an increase in mitochondrial mass in knockout muscle.

View this table: [in this window] [in a new window]   Table 2. Respiratory chain activities in brain and skeletal muscle of GAMT-deficient and control mice

 
Structural adaptive rearrangements in response to CK deficiency were previously reported in a study on combined mitochondrial and cytosolic CK deficiencies. These changes included an increase in the number of mitochondria in skeletal muscle (16). We investigated whether the biochemical alterations present in GAMT^–/– muscle, such as the lack of Cr/CrP, accumulation of GAA/PGAA, and altered ATP synthase/citrate synthase activities, may also lead to morphological alterations. Using electron microscopy, we analyzed sections from gastrocnemius muscle of three male GAMT^–/– and three male control animals but did not observe any striking difference (Fig. 8). However, when using a grip strength meter for analysis of the maximum grip force of male GAMT^–/– mice, we observed impaired muscular performance [GAMT^–/– versus control: 724.7±21.6 mN versus 857.1±32.4 mN (n=20); P=0.005]. Furthermore, when lifted by the tail to be placed in a new cage and held in a vertical position for a few seconds, GAMT^–/– mice regularly showed reduced body tension. This symptom was prominent and allowed the identification of GAMT-deficient animals by the caretaker, practically without genotyping, as early as 4 to 5 weeks after birth. Both muscular hypotonia of the trunk musculature and attenuated grip force, albeit difficult to quantify, point to a muscle phenotype with biochemical adaptation, but without marked morphological changes.

View larger version (92K): [in this window] [in a new window]   Figure 8. Electron microscopy of skeletal muscle sections from control and GAMT–/– mice. Electron micrographs of longitudinal sections from gastrocnemius muscle from a control (A) or a GAMT-deficient (B) mouse. Magnification: 12 000x.

 
Viability, genotype frequencies and fertility of GAMT knockout mice
In general, GAMT^–/– mice were viable and required no specific precautions for survival and growth. However, analysis of mGAMT^–/– genotype frequencies from intercrosses of heterozygous mutant mice revealed a too small number of GAMT^–/– offspring, which resulted in a deviation from the Mendelian inheritance pattern (GAMT^+/+: 24.9%; GAMT^+/–: 60.6%; GAMT^–/–: 14.5%; n=531). This decrease might be due to an increased neonatal mortality in GAMT^–/– mice, because we occasionally found that newborn mice with the GAMT^–/– genotype (n=18) died after birth. The number of perinatally deceased animals was probably much higher than this number, as parental mice often eat the dead offspring. For this reason, these animals could neither be genotyped nor registered. We do not know the exact cause of death, but all these mice apparently had difficulty suckling because their stomachs contained no milk.

During breeding of the GAMT mice, we often noticed that no litters were born when male GAMT^–/– mice had been mated to wild-type or heterozygous mutant females. In scheduled breeding experiments, we obtained no offspring from plug-positive females (GAMT^+/+ or GAMT^+/–) in 16 of 24 breeding experiments. For comparison, breeding pairs with GAMT^+/– or GAMT^+/+ genotypes produced offspring in 20 of 20 cases. When GAMT^–/– males were fertile, litter sizes showed a tendency to be smaller, although the difference was not statistically significant [control: 9.9±0.9; GAMT^–/–: 7.9±0.9 (n=20); P=0.14]. These data, which indicate impaired fertility in GAMT-deficient males, prompted us to investigate testis morphology. Semi-thin sections from adult male testes of control and knockout mice (n=3) revealed striking differences (Fig. 9). We observed severely attenuated spermatogenesis in knockout mice at the level of spermatid development; that is, elongated spermatids were hardly present and mature spermatozoa were almost absent in the lumen of seminiferous tubules (Fig. 9B and C). Furthermore, the structure of the seminiferous tubules was highly unordered and showed a large number of resorption holes and a larger lumen (Fig. 9B–D). Multinucleated giant cells were frequently observed and indicate increased phagocytic activity. These changes were not present in control tubules from wild-type mice, which were processed in parallel to exclude fixation and processing artifacts. In some sections from GAMT^–/–mice, however, intact spermatogenesis was observed (Fig. 9D), which might explain why some GAMT-deficient males were—at least temporarily—fertile.

View larger version (139K): [in this window] [in a new window]   Figure 9. Reduced fertility in male GAMT–/– mice is associated with altered testis morphology. Representative illustrations of testis sections showing seminiferous tubules from control (A) and GAMT–/– mice (B–D). Spermatogenesis is severely affected in GAMT-deficient mice, because elongated spermatids or spermatozoa are almost absent in large areas of the seminiferous tubules. In every section of the GAMT–/– testis tissue, several multinucleated giant cells were seen in seminiferous tubules (white arrows in B–D). In some regions, though, a few (B) or numerous elongated spermatids and spermatozoa (D) were observed (black arrows in B, D). Magnifications: A, B: 20x; C, D: 40x.

 
Body weight development and body composition
Female and male GAMT^–/– animals consistently weighed less than control littermates (Fig. 10A, B and F). The weight difference was noticeable as early as during the first weeks of life and increased with time (Fig. 10A and B). The differences were more prominent in females, in particular in adult animals aged 6 months or older (Fig. 10D). Although GAMT^–/– mice appeared smaller, no significant length differences were found in adult male (data not shown) or female mice (Fig. 10E). These findings suggest that the reduced weight of GAMT^–/– mice cannot be simply attributed to length differences. Therefore, we investigated the body composition of female GAMT^–/– (n=11) and female control mice (n=10) aged between 6 and 10 months (Fig. 10G and H). The absolute lean mass was not different between GAMT^–/– and control female mice. However, we found a reduction in total body fat mass in GAMT^–/– mice to levels below 30% of those observed in controls (Fig. 10G), which corresponds to a decrease in relative fat mass by about 50% (Fig. 10H). Furthermore, the absolute water content was decreased by 20%. When normalized to the body weight, it showed a relative increase by 20% (Fig. 10G and H). Thus, the markedly reduced body weight of female GAMT^–/– mice can be mainly attributed to alterations in the body composition, in particular to reduced fat mass.

View larger version (43K): [in this window] [in a new window]   Figure 10. Reduced body weight and altered body composition in GAMT–/– mice. (A) Weight development of female control and GAMT–/– mice during a period of 6 months (control: n=11; GAMT–/–: n=11). (B) Weight development of male control and GAMT–/– mice during a period of 6 months (control: n=17; GAMT–/–: n=14). (C) Example of two female littermates housed together for 6 months in a mixed genotype group with ad libitum food supply. (D) Normalized weight development of female and male GAMT–/– mice. At each time point, the weight of GAMT–/– mice was normalized to the weight of control littermates of the same sex, which were housed in the same cage [numbers of mice as given in (A) and (B)]. Body length (E), body weight (F) and absolute (G) or relative (H) quantification of body composition in female control and GAMT–/– mice (control: n=10; GAMT–/–: n=11) aged 6–10 months. Significance levels: *P<0.05, **P<0.01, ***P<0.001 (A, B: two-way repeated measurements ANOVA with Tukey's HSD post test; E–H: two-tailed heteroscedastic t-test). Absence of error bars indicates errors smaller than symbol size.

 
Since body weight regulation and fat content are under tight hormonal control, we wondered whether the changes in GAMT-deficient animals might be associated with altered hormone levels. In particular, we suspected that leptin might be altered since it is the gene product of the ob gene and a key metabolic hormone regulating food intake, energy expenditure and body weight. However, serum leptin levels in female GAMT^–/– did not differ significantly from those of female controls but showed a tendency to be slightly reduced (Table 3). Consequently, relative food intake (that is, intake normalized to body weight) in GAMT^–/– mice was only marginally increased (Table 3), pointing to an intact leptin-signaling pathway. Likewise, the levels of the two key metabolic hormones insulin and adiponectin remained unchanged (Table 3). Furthermore, home cage activity levels were not different between the two groups of mice (Table 3). Hence, disturbed body weight regulation and altered body composition cannot be explained by reduced food intake, altered cage activity or pathological changes in the secretion of the three metabolic hormones leptin, adiponectin or insulin, but rather may be due to metabolic alterations in the synthesis or metabolism of storage fat.

View this table: [in this window] [in a new window]   Table 3. Cage activity, food assimilation and hormone levels of female GAMT–/– and control mice

 
Guanidino compounds and contractile function in the hearts of GAMT^–/– mice
Besides metabolic or hormonal causes of underweight, chronically impaired cardiac function could also play a role. As cardiac muscle is an excitable tissue with changing energy demand that expresses high amounts of creatine transporter protein and, like skeletal muscle, contains high creatine levels, it is believed to depend on a functional Cr/CrP high-energy phosphate-buffering system (17). We therefore investigated the levels of Cr, CrP and PGAA, as well as baseline cardiac function in isolated perfused hearts from adult control (older than 8 months, n=9) and GAMT knockout animals (n=11). Total creatine levels, measured by HPLC, were 21.5 ±1.7 nmol/mg protein in knockout heart muscle and thus represented about 27% of the control creatine content (77.4±10.6 nmol/mg protein, P<0.001). ³¹P-MR spectroscopy measurements of cardiac high-energy phosphate levels revealed that GAMT-deficient mice showed a 67% reduction in phosphocreatine/ATP ratios and a 68% decrease in phosphocreatine concentrations. ATP levels remained at control levels (control: 10.0±0.3 mM; GAMT^–/–: 10.2±0.8 mM). Inorganic phosphate resonances were too small to be reliably detected. Hearts from GAMT knockout mice showed the same additional ³¹P resonance at the right shoulder of the phosphocreatine resonance that was also observed in brain and skeletal muscle and corresponds to PGAA. The amplitude of this resonance signal (5.51±0.36 mM) was similar to that of phosphocreatine (5.32±0.47 mM). As with brain and skeletal muscle ³¹P-MR spectroscopy, this resonance was not observed in control hearts. Furthermore, analysis of the citrate synthase activity revealed no changes in response to the disruption of the GAMT gene in knockout mice (data not shown).

The GAMT^–/– mice used for the study of cardiac function had a 31% lower body weight and a 17% lower heart weight. This resulted in an increase in the heart weight/body weight ratio by 15%. Isolated buffer-perfused hearts of GAMT knockout mice showed normal isovolumetric contractile function and both left ventricular developed pressure (LVDP; control: 132±4 mmHg; GAMT^–/–: 135±7 mmHg) and heart rate (control: 388±16 bpm; GAMT^–/–: 414±7 bpm) were similar for controls and GAMT^–/– hearts. Moreover, when we tested the inotropic reserve by adding 20 nM isoproterenol to the perfusate, GAMT^–/– and control mice showed a similar increase in their LVDPs (control: 171±5 mmHg; GAMT^–/–: 167±4 mmHg).

Thus, severely reduced total creatine and phosphocreatine levels, unchanged ATP levels, accumulation of PGAA, as well as normal left ventricular function and contractile reserve were observed in GAMT-deficient hearts. The data do not suggest a cardiac contribution to impaired weight development.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
In the present study, we generated a knockout mouse model of GAMT deficiency through targeted disruption of the mouse GAMT gene in embryonic stem cells. Deletion of more than 90% of the GAMT ORF due to a reading frame shift in exon 1 resulted in undetectable mGAMT transcript and protein levels and caused characteristic biochemical changes in serum, urine, skeletal and cardiac muscle, and brain of GAMT^–/– mice. Creatine levels decreased to levels below 1–27% of those found in control littermates. More importantly, accumulation of GAA, the key diagnostic finding in human GAMT deficiency, was detected in tissues and body fluids of GAMT^–/– mice. The highest increase in GAA concentrations, more than 100-fold, was seen in brain. In serum and urine, the two most frequently used body fluids for GAMT deficiency screening, there was a 10- to 50-fold increase in GAA levels. The extent of GAA accumulation was remarkably comparable between mice and humans. In GAMT patients, the GAA levels were more than 200-fold elevated in CSF, about 20-fold in serum and about 5- to 10-fold in urine (18). The relatively high residual creatine detected in GAMT^–/– mice is surprising and does not originate from the creatine-free food. Since the animals used in this study were kept in mixed-genotype groups, coprophagia is a likely source of dietary creatine and may account for the variation in creatine and creatinine levels. Hence, ‘metabolic genotyping’ merely based on these compounds provided no absolute diagnostic certainty—another parallel to the human disease (7,19). In contrast, the determination of GAA in serum and urine represented a reliable method for the identification of knockout mice, which was further improved when we calculated the ratio of GAA to the sum of creatine and creatinine. This ratio increased on average 5400-fold in brain, 1400-fold in serum and 13 500-fold in urine. When we compared the knockout animal with the lowest and the control animal with the highest ratio, the increase was still 50-fold in urine, 200-fold in serum and about 800-fold in brain. Thus, our results and data on human GAMT deficiency (20) suggest that this parameter is useful and reliable in the diagnostic screening of GAMT patients.

In addition to the metabolic alterations found in vitro, localized in vivo ¹H MRS of brain and skeletal or cardiac muscle from GAMT^–/– mice revealed markedly reduced Cr/CrP resonances to levels below 27% of normal. In addition, we detected accumulation and phosphorylation of GAA using in vivo ³¹P MRS. Interestingly, the PGAA/CrP ratios in cardiac muscle and brain were close to 1, whereas in skeletal muscle this ratio was about 3.5. Since CrP levels were comparable, this difference may be due to either active GAA uptake via the creatine transporter or due to endogenous GAA synthesis, or both. In skeletal muscle high levels of creatine transporter are expressed that may also transport GAA (1), especially in our mice, in which low serum creatine levels and increased GAA levels were found. Alternatively, skeletal muscle is likely to synthesize its own creatine (1) because both AGAT (D. Isbrandt, unpublished data) and GAMT RNAs are present. Furthermore, the detection of substantial CrP levels in skeletal muscle of creatine transporter-deficient individuals by in vivo ³¹P MRS clearly points to significant amounts of creatine synthesis in human muscle (21). In GAMT deficiency, the accumulated GAA may become a substrate of creatine kinases, leading to the formation of PGAA. It is interesting to note that PGAA, like phosphocreatine, may act in vivo as energy buffer, as it can be reversibly de-phosphorylated during ischemia-reperfusion of hindleg muscle, and may thus provide high-energy phosphates to stabilize ATP levels (15).

The most obvious symptom of GAMT deficiency in mice was a marked reduction in body weight throughout the life of the animals. When we systematically followed the weight development of control and knockout animals, we found that, despite comparable birth weight, GAMT^–/– mice gained less weight than their control littermates. The differences increased with age and were more pronounced in females than in males. As we did not find significant body length differences, we analyzed the body composition of female GAMT^–/– mice. Quantitative analyses revealed a marked reduction in absolute fat mass to one-third and in relative fat mass to one-half of control levels. In contrast, the absolute lean weight was not different. The physiological control of energy balance and body weight regulation involves the hypothalamus, which integrates several peripheral signals, such as leptin and insulin levels, in order to increase or decrease energy intake (22,23). Leptin signaling in female GAMT^–/– mice appeared to be normal because slightly lower leptin levels were paralleled by moderately increased food intake in GAMT^–/– mice. The levels of insulin, the key hormone for glucose metabolism, were also comparable. Likewise, no change was found in the levels of adiponectin, another secretory protein predominantly released from adipocytes (24) that regulates glucose utilization and fatty acid oxidation (25). Furthermore, the reduction in body fat could not be explained by increased locomotor activity or chronic stress due to impaired cardiac function. Together, our data suggest that GAMT—through synthesis of creatine—is not critical to the systemic signal pathways regulating appetite. Rather, they point to cellular changes in either the regulation of fat synthesis in liver or fat metabolism in white adipose tissue, or both.

GAMT^–/– mice surviving the perinatal period were viable and had no neurological symptoms such as ataxia or seizures, which were observed in some GAMT patients (7). This is particularly surprising in view of the accumulation of GAA in brain, as this is thought to contribute essentially to the neurological symptoms in human GAMT deficiency (11). Favorable biochemical adaptations, energy buffering by PGAA or residual CrP, or both may prevent serious neurological symptoms in adult GAMT-deficient mice. Alternatively, GAA levels in the extracellular space of mouse brain might be heterogeneous and possibly also lower than in the CSF of humans. A thorough behavioral assessment of adult mice with a homogenous genetic background will provide insights into the learning and memory performance of GAMT-deficient mice in the future.

The expression profiles of human and mouse GAMT in our study, along with previously published reports (26,27), point to a role of creatine synthesis in the reproductive tract, in particular in males. The severely reduced fertility of our male GAMT knockout mice and pronounced morphological changes in seminiferous tubules resulting in disturbed spermatogenesis provide evidence that intact creatine synthesis plays an important role in male reproductive function. In mouse and rat testis, GAMT expression was found in Sertoli cells, the type of cell to which spermatids attach for nourishment during spermatogenesis (26,27). Most likely, attenuation of Sertoli cell function interfered with maturation of spermatids, which only rarely reached the elongated stage or developed to mature spermatozoa. Instead, the presence of polynucleated giant cells indicated increased phagocytic activity, most likely for removal of degenerated spermatids. As already proposed for muscle and brain tissues, residual CrP or high-energy phosphate buffering by PGAA may explain the normal maturation of spermatids in some males producing offspring. Although GAMT mRNA and protein was present in comparable levels in testis and ovary, we did not observe a similar reduction in the fertility of female GAMT^–/– mice, which suggests gender-specific differences in the importance of creatine for the maturation of oocytes versus sperm. As the number of sperm cells to be produced in testis is several orders of magnitude higher than the number of oocytes, which develop during each estrus cycle, it is most likely that the different and more challenging metabolic activity in testis accounts for the male-specific reduction in fertility.

In summary, we generated a knockout mouse model of GAMT deficiency that replicates many of the biochemical hallmarks of the human disease. Furthermore, this model shows that intact creatine synthesis is essential for body weight homeostasis and reproduction.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 
Northern blot analysis
Human multiple tissue northern blots, brain northern blots and human multiple tissue arrays were purchased from Clontech (Palo Alto, CA, USA). Northern blots containing mouse total RNAs were prepared, hybridized and washed using standard methods. We used human or mouse [³²P]-labeled GAMT cDNA probes that covered the respective GAMT reading frame.

Western blot and immunohistochemistry
Western blots.
We generated polyclonal antibodies in rabbit against recombinant N-terminally, six-histidine (His)-tagged mGAMT or hGAMT proteins purified from E. coli using standard procedures according to the manufacturer (Qiagen, Hilden, Germany). The antibodies used in this study were affinity purified against the respective antigen immobilized on NHS-activated Sepharose (Amersham, Freiburg, Germany). The specificity of the antibodies was first verified by western blot experiments using antigen-blocked and unblocked antibodies (not shown). Upon availability of GAMT^–/– mice, specificity of the antibodies was reconfirmed (Figs 3Ci–ii and 5B).

Human tissue extracts for western blot analysis were purchased from Clontech. Mouse tissue extracts from the same brain regions and tissues (the same as used for RNA preparation) were prepared as follows: tissue pieces of up to 100 mg were homogenized in lysis buffer (50 mM Tris pH 7.4, 150 mM NaCl, 0.5% Triton X-100) supplemented with proteinase inhibitor cocktail (Sigma, Taufkirchen, Germany) at 4°C and incubated on ice for 15 min. Subsequently, insoluble material was removed by centrifugation at 10 000g. The protein content of the extracts was determined using the Bradford method. The cell extracts were separated by 12.5% SDS polyacrylamide gel electrophoresis (40 µg protein per lane) and transferred by electroblotting to nitrocellulose membranes (Protran, Schleicher and Schüll, Germany) according to standard protocols. The blot was blocked with phosphate-buffered saline (PBS) containing 5% milk powder and incubated with rabbit anti-mGAMT or anti-hGAMT antibodies (1 : 1000 in blocking solution). The blot was then washed twice in PBS containing 0.05% Tween 20, incubated with a horseradish peroxidase-conjugated anti-rabbit IgG (1 : 10 000 in blocking solution; Dianova, Hamburg, Germany) and washed as before. Signals were visualized using the enhanced chemiluminescence (ECL) technique and BioMax MR X-ray films (Amersham, Freiburg, Germany).

Immunocytochemistry for light microscopy.
Mice were deeply anesthetized with sodium pentobarbital (80 mg/kg i.p.) and brains were perfused via the ascending aorta with 4% paraformaldehyde and 0.1% glutaraldehyde in phosphate buffer pH 7.4. Brains were dissected and shock frozen after 48 h of pretreatment in 30% sucrose. Cryostat sections (25 µm) were cut and incubated free floating in 1% NaBH₄ for 15 min. After blocking of unspecific binding sites in 10% normal goat serum and 0.5% BSA in PBS and permeabilization in an ethanol series of 10, 20, 40, 20, 10% in PBS, sections were incubated overnight with affinity-purified polyclonal antibodies against mGAMT at 4°C. For control sections, binding of the antibodies was blocked by addition of recombinant mGAMT proteins (0.5 mg/ml) to the antibody solutions. Furthermore, specificity of the antibodies was verified by the absence of immunoreactivity on sections of GAMT^–/– animals. Detection of the bound mGAMT antibodies was performed with a secondary biotinylated anti-rabbit antibody followed by the peroxidase-coupled Avidin-Biotin-Complex and diaminobenzidine reaction (Vector/Camon Vectastain Elite Kit, Vector Labs, Peterborough, UK). Sections were mounted, coverslipped and examined by light microscopy.

Cloning of the mGAMT gene
We screened a 129/SvJ mouse genomic library (Stratagene, La Jolla, CA, USA) using the mGAMT cDNA as a probe and obtained one positive clone. Complete insert sequencing showed that this clone contained the complete mGAMT open reading frame. The 5.2 kb HindIII/BglII restriction fragment isolated by partial digestions of the phage DNA, which contained all six exons, was subcloned into pBluescriptKS (Stratagene). The 5' end of the fragment was shortened by 500 bp, yielding pBKS-mGAMT. The region deleted from the 5' end of the genomic clone was subcloned to pBluescript KS and used as 5' external probe in Southern blot experiments (see below).

Gene targeting
pBKS-mGAMT DNA was cut with MluI, filled with Klenow polymerase and ligated with a blunt-ended restriction fragment containing the neomycin resistance gene (neo) under the control of the phosphoglycerate kinase (PGK) promoter yielding the targeting vector pmGAMT-KO (Fig. 4A). The linearized targeting vector was electroporated into 129 (R1) embryonic stem (ES) cells, which were subjected to selection by geneticin (G418, Invitrogen, Karlsruhe, Germany). Southern blotting was performed on 280 resistant ES cell clones, 10 of which were positive for the targeting event. Genomic ES cell DNA was digested with Dra, separated on 0.8% agarose gels, transferred to nylon membranes and hybridized with the 3' external probe derived from DNA distal to the pBKS-mGAMT BglII site (Fig. 4A). The wild-type 6.8-kb DraI restriction fragment was 1.5 kb shorter (5.3 kb) when homologous recombination occurred (Fig. 4Bi, left). Homologous recombination at the 5' end was verified by Southern blot analysis of HindIII-digested genomic ES cell DNA as described above using the 5' external probe. In the case of homologous recombination, the wild-type 4.3-kb HindIII fragment was 1.5 kb longer (5.8 kb) (Fig. 4Bi, right). One of the positive ES clones (#167) was expanded and microinjected into C57BL/6J mouse blastocysts, which were then transferred into pseudopregnant CBA/C57BL/6J females. Two of three chimeric mice that were mated gave rise to germ-line transmission of the disrupted allele. Males and females with different genotypes from different litters were randomly intercrossed to obtain mGAMT^+/+, mGAMT^+/– and mGAMT^–/– progeny. Genomic DNA from mouse-tail or ear biopsies was screened either by Southern analysis or multiplex polymerase chain reaction (PCR) following standard protocols (Fig. 4Bii and iii). The binding sites of the primers used for PCR detection of the wild-type or mutant alleles (P1–P3) are indicated in Figure 4A. The studies described below were performed on mice belonging to generations F3 to F8.

Animal groups and care
If not otherwise stated, the control group (control) consisted of wild-type (GAMT^+/+) and heterozygous (GAMT^+/–) littermates. The knockout group consisted of homozygous GAMT-deficient mice (GAMT^–/–).

Mixed genotype groups of not more than five animals were housed in standard mouse cages under conventional laboratory conditions (12/12 h dark–light cycle, constant temperature, constant humidity, and food and water ad libitum). According to the manufacturer, the food did not contain animal ingredients (rodent maintenance chow R/M-H; ssniff, Soest, Germany) and should therefore be free of creatine. Our own analysis of random food samples showed only trace amounts of creatine close to the detection limit (for details of the method used, please see below). Generation, care and use of the animals were in accordance with institutional guidelines. All experimental procedures were in accordance with the German Law for the Protection of Experimental Animals and complied with the regulations of the National Institutes of Health and those of the Society for Neuroscience (USA). In addition, all experiments were approved by the respective local animal ethics committees.

Determination of guanidino compounds
The concentration of the guanidino compounds was determined using a Biotronic LC 5001 (Biotronik, Maintal, Germany) amino acid analyzer adapted for guanidino compound determination. The guanidino compounds were separated via a cation exchange column using sodium citrate buffers and were detected with the fluorescence ninhydrin method, which is described in detail elsewhere (28).

In vivo ¹H and ³¹P magnetic resonance spectroscopy (MRS) of brain and skeletal muscle
The MR experiments have been described in detail elsewhere (15) and are briefly summarized here. During the in vivo magnetic resonance (MR) experiments, all animals were anesthetized using 1.2% isoflurane in a gas mixture of 50% O₂ and 50% N₂O delivered through a facemask. Breathing frequency was monitored and a warm waterbed (37°C) was used to keep the animals at body temperature. In vivo MR was performed on a 7.0 T horizontal bore magnet. Metabolic profiles of brain and hind leg were assessed in vivo both by ¹H and ³¹P MRS. For the ¹H MR studies on the mouse brain, localized spectroscopy was performed of a 3x3x3 mm³ (27 µl) voxel using a STEAM sequence (TE=10 ms, TM=15 ms) in 10 GAMT^–/– and seven control animals. The voxel was positioned in the center of the brain guided by gradient echo MR images. Phosphorous compounds were measured in the brain of seven GAMT^–/– and seven control animals. Two ³¹P surface coils working in quadrature mode were used together with an ISIS pulse sequence to measure a spectrum from an image-guided voxel of 6.5x6.5x4.5 mm³ (volume 190 µl).

Proton measurements on the mouse hind leg were done using an Alderman–Grant type of coil. The STEAM sequence used was the same as that employed for ¹H brain measurements (voxel size 1.8x1.8x3.4 mm³). Seven GAMT^–/– and seven control animals were measured. ³¹P spectra of the hind leg were obtained by a pulse acquire experiment in seven GAMT-deficient animals and seven control animals.

Cardiac physiology and metabolism
Isolated heart function.
After anesthetizing the mice with an intraperitoneal injection of sodium pentobarbitone (240 mg/kg body weight), their hearts were rapidly removed after thoracotomy (29) and retrogradely perfused with Krebs–Henseleit buffer via the aorta at a constant perfusion pressure of 80 mmHg. Isovolumic function was measured by means of an LV balloon inserted via the mitral valve. The experimental setup was the same as previously described (17).

³¹P-MRS.
The hearts were placed in a 10-mm diameter MR tube and perfused as described above. ³¹P-MR spectra were obtained using a 400-MHz wide bore magnet (Oxford Instruments, Inc., Oxford, UK) and a Varian spectrometer. Each spectrum was acquired in 4 min and consisted of 112 summed transients of 60° pulses with a 1.97-second interpulse delay. Each resonance was fitted to a Lorentzian lineshape using the NMR1 line-fitting software program (Tripos, St Louis, USA). Total creatine in cardiac muscle was measured by HPLC (30). Citrate synthase activities were measured with spectrophotometry as previously described (30).

Morphology
Muscle.
To analyze muscle tissue, mice were perfused transcardially with 4% paraformaldehyde (PFA) and 2% glutaraldehyde (GA) in PBS. Tissues were postfixed in 1% OsO₄, dehydrated and embedded in Epon. Ultrathin sections were stained with uranylacetate and lead citrate and examined with a Zeiss EM902.

Testis.
Whole mouse testes were fixed in 3.5% glutaraldehyde in 0.05 M sodium phosphate buffer pH 7.1–7.4 for 3 h. Small pieces of the testes were then postfixed in 1% OsO₄, dehydrated in ascending alcoholic series and finally embedded in Epon 812. Semi-thin free-floating sections (1 µm thick) were stained with toluidine blue/pyronine G and covered with Caedax (Merck, Darmstadt, Germany).

Determination of respiratory chain enzyme activities
Respiratory chain enzyme activities, including active regulation of the mitochondrial ATP synthase (complex V), were determined spectrophotometrically after sonication as described in detail in a previous study (31).

Analysis of body composition
Carcasses were weighed and oven dried at 60°C for at least 2 weeks until weight was constant. Total body water was calculated as the difference between the weights before and after drying. The carcass was then chloroform-extracted using a Soxhlet apparatus. The extracted carcass was dried and weighed to calculate fat mass and lean mass.

Determination of metabolic hormones
Leptin, insulin and adiponectin levels in serum were measured by ELISA assays according to the protocols provided by the manufacturers (leptin, insulin: Crystal Chemistry, Downers Grove, IL, USA; adiponectin: BioCat, Heidelberg, Germany).

Grip strength measurements
Maximum grip force of 20 male GAMT^–/– and 20 control mice aged between 8 and 10 weeks was measured using a grip strength meter (TSE-Systems, Bad Homburg, Germany). Within each group, the mean grip force of each mouse was calculated from 15 appropriate trials.

Cage activity
The home cage activity was monitored for 3 consecutive days with an infrared motion detector with a sampling frequency of 1 Hz and a bin size of 4 min (INFRA-E-MOTION GmbH, Hamburg, Germany). Activity levels were calculated as the sum of activity values obtained during daylight and dark phases.

Statistical analyses
If not otherwise stated, data are given as mean ± SEM. The following statistical tests were applied: heteroscedastic two-tailed t-test (Excel X, Microsoft, Redmond, WA, USA), or ANOVA for repeated measurements with Tukey's honest significant differences (HSD) post test (Statistica 5, Statsoft, Tulsa, OK, USA).

	ACKNOWLEDGEMENTS

 
We thank O. Pongs, ZMNH, for continued support throughout the study. Furthermore, we are indebted to H. Voss (ZMNH Animal Facility, Hamburg) for expert animal care and assistance with body weight measurements; M. Klingenspohr (Department of Biology, University of Marburg, Germany) for help with analysis of body composition; M. Schweizer, S. Fehr and S. Siegel (Morphology group, ZMNH) for help with immunohistochemistry and electron microscopy; A. Elizabeth Sang for help with mouse heart perfusion; and Kathrin Sauter, Mina Monfared and Stefan Schillemeit for expert technical assistance. We thank B. Wieringa (Department of Cell Biology, University of Nijmegen, The Netherlands) for helpful discussions of our MRS results. The study was supported by the Deutsche Forschungsgemeinschaft (SFB545, project A3 to D.I. and K.U.), by the Born–Bunge Foundation, University of Antwerp and Fund for Scientific Research, Flanders (FWO-Vlaanderen, grant G.0027.97 to P.P.D.D. and B.M.); by the British Heart Foundation (Program Grant 2000008 to S.N.); and by the Netherlands Organization for Scientific Research (NWO-ZONMW to B. Wieringa, W.R. and A.H.).

	FOOTNOTES

 
^* To whom correspondence should be addressed at: Zentrum für Molekulare Neurobiologie Hamburg, Institut für Neurale Signalverarbeitung, Universität Hamburg, Martinistrasse 52, 20246 Hamburg, Germany. Tel: +49 40428036650; Fax: +49 40428036643; Email: isbrandt{at}uni-hamburg.de

Present address: Department of Pediatrics, University of Hanover, Germany.

Present address: Department of Pediatrics, University of Göttingen, Germany.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS REFERENCES

 

1. Wyss, M. and Kaddurah-Daouk, R. (2000) Creatine and creatinine metabolism. Physiol. Rev., 80, 1107–1213.[Abstract/Free Full Text]

2. Stöckler, S., Isbrandt, D., Hanefeld, F., Schmidt, B. and von Figura, K. (1996) Guanidinoacetate methyltransferase deficiency: the first inborn error of creatine metabolism in man. Am. J. Hum. Genet., 58, 914–922.[Web of Science][Medline]

3. Item, C.B., Stockler-Ipsiroglu, S., Stromberger, C., Muhl, A., Alessandri, M.G., Bianchi, M.C., Tosetti, M., Fornai, F. and Cioni, G. (2001) Arginine : glycine amidinotransferase deficiency: the third inborn error of creatine metabolism in humans. Am. J. Hum. Genet., 69, 1127–1133.[CrossRef][Web of Science][Medline]

4. Salomons, G.S., van Dooren, S.J., Verhoeven, N.M., Cecil, K.M., Ball, W.S., Degrauw, T.J. and Jakobs, C. (2001) X-linked creatine-transporter gene (SLC6A8) defect: a new creatine-deficiency syndrome. Am. J. Hum. Genet., 68, 1497–1500.[CrossRef][Web of Science][Medline]

5. Walker, J.B. (1979) Creatine: biosynthesis, regulation, and function. Adv. Enzymol. Relat. Areas Mol. Biol., 50, 177–242.[Medline]

6. Stöckler, S., Holzbach, U., Hanefeld, F., Marquardt, I., Helms, G., Requart, M., Hanicke, W. and Frahm, J. (1994) Creatine deficiency in the brain: a new, treatable inborn error of metabolism. Pediatr. Res., 36, 409–413.[Web of Science][Medline]

7. von Figura, K., Hanefeld, F., Isbrandt, D. and Stöckler-Ipsiroglu, S. (2001) Guanidinoacetate methyltransferase deficiency. In Scriver, C.R., Beaudet, A.L., Sly, W.S. and Valle, D. (eds), The Metabolic and Molecular Bases of Inherited Disease., McGraw-Hill, New York, Vol. II, pp. 1897–1908.

8. Ganesan, V., Johnson, A., Connelly, A., Eckhardt, S. and Surtees, R.A. (1997) Guanidinoacetate methyltransferase deficiency: new clinical features. Pediatr. Neurol., 17, 155–157.[CrossRef][Web of Science][Medline]

9. Schulze, A., Hess, T., Wevers, R., Mayatepek, E., Bachert, P., Marescau, B., Knopp, M.V., De Deyn, P.P., Bremer, H.J. and Rating, D. (1997) Creatine deficiency syndrome caused by guanidinoacetate methyltransferase deficiency: diagnostic tools for a new inborn error of metabolism [see comments]. J. Pediatr., 131, 626–631.[CrossRef][Web of Science][Medline]

10. van der Knaap, M.S., Verhoeven, N.M., Maaswinkel-Mooij, P., Pouwels, P.J., Onkenhout, W., Peeters, E.A., Stockler-Ipsiroglu, S. and Jakobs, C. (2000) Mental retardation and behavioral problems as presenting signs of a creatine synthesis defect. Ann. Neurol., 47, 540–543.[CrossRef][Web of Science][Medline]

11. Neu, A., Neuhoff, H., Trube, G., Fehr, S., Ullrich, K., Roeper, J. and Isbrandt, D. (2002) Activation of GABA(A) receptors by guanidinoacetate: a novel pathophysiological mechanism. Neurobiol. Dis., 11, 298–307.[CrossRef][Web of Science][Medline]

12. Schulze, A., Ebinger, F., Rating, D. and Mayatepek, E. (2001) Improving treatment of guanidinoacetate methyltransferase deficiency: reduction of guanidinoacetic acid in body fluids by arginine restriction and ornithine supplementation. Mol. Genet. Metab., 74, 413–419.[CrossRef][Web of Science][Medline]

13. Isbrandt, D. and von Figura, K. (1995) Cloning and sequence analysis of human guanidinoacetate N-methyltransferase cDNA. Biochim. Biophys. Acta., 1264, 265–267.[Medline]

14. Jenne, D.E., Olsen, A.S. and Zimmer, M. (1997) The human guanidinoacetate methyltransferase (GAMT) gene maps to a syntenic region on 19p13.3, homologous to band C of mouse chromosome 10, but GAMT is not mutated in jittery mice. Biochem. Biophys. Res. Commun., 238, 723–727.[Medline]

15. Renema, W.K., Schmidt, A., Van Asten, J.J.A., Oerlemans, F., Ullrich, K., Wieringa, B., Isbrandt, D. and Heerschap, A. (2003) MR Spectroscopy of muscle and brain in guanidinoacetate methyltransferase (GAMT)-deficient mice: validation of an animal model to study creatine deficiency. Magn. Reson. Med., 50, 936–943.[CrossRef][Web of Science][Medline]

16. Steeghs, K., Oerlemans, F., de Haan, A., Heerschap, A., Verdoodt, L., de Bie, M., Ruitenbeek, W., Benders, A., Jost, C., van Deursen, J. et al. (1998) Cytoarchitectural and metabolic adaptations in muscles with mitochondrial and cytosolic creatine kinase deficiencies. Mol. Cell. Biochem., 184, 183–194.[CrossRef][Web of Science][Medline]

17. Spindler, M., Niebler, R., Remkes, H., Horn, M., Lanz, T. and Neubauer, S. (2002) Mitochondrial creatine kinase is critically necessary for normal myocardial high-energy phosphate metabolism. Am. J. Physiol. Heart Circ. Physiol., 283, H680–687.[Abstract/Free Full Text]

18. Stöckler, S., Marescau, B., De Deyn, P.P., Trijbels, J.M. and Hanefeld, F. (1997) Guanidino compounds in guanidinoacetate methyltransferase deficiency, a new inborn error of creatine synthesis. Metabolism., 46, 1189–1193.[CrossRef][Web of Science][Medline]

19. Verhoeven, N.M., Guerand, W.S., Struys, E.A., Bouman, A.A., van der Knaap, M.S. and Jakobs, C. (2000) Plasma creatinine assessment in creatine deficiency: a diagnostic pitfall. J. Inherit. Metab. Dis., 23, 835–840.[Medline]

20. Carducci, C., Birarelli, M., Leuzzi, V., Battini, R., Cioni, G. and Antonozzi, I. (2002) Guanidinoacetate and creatine plus creatinine assessment in physiologic fluids: an effective diagnostic tool for the biochemical diagnosis of arginine : glycine amidinotransferase and guanidinoacetate methyltransferase deficiencies. Clin. Chem., 48, 1772–1778.[Abstract/Free Full Text]

21. deGrauw, T.J., Cecil, K.M., Byars, A.W., Salomons, G.S., Ball, W.S. and Jakobs, C. (2003) The clinical syndrome of creatine transporter deficiency. Mol. Cell. Biochem., 244, 45–48.[CrossRef][Web of Science][Medline]

22. Schwartz, M.W., Woods, S.C., Porte, D., Jr., Seeley, R.J. and Baskin, D.G. (2000) Central nervous system control of food intake. Nature, 404, 661–671.[Medline]

23. Kalra, S.P., Dube, M.G., Pu, S., Xu, B., Horvath, T.L. and Kalra, P.S. (1999) Interacting appetite-regulating pathways in the hypothalamic regulation of body weight. Endocr. Rev., 20, 68–100.[Abstract/Free Full Text]

24. Scherer, P.E., Williams, S., Fogliano, M., Baldini, G. and Lodish, H.F. (1995) A novel serum protein similar to C1q, produced exclusively in adipocytes. J. Biol. Chem., 270, 26746–26749.[Abstract/Free Full Text]

25. Yamauchi, T., Kamon, J., Minokoshi, Y., Ito, Y., Waki, H., Uchida, S., Yamashita, S., Noda, M., Kita, S., Ueki, K. et al. (2002) Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase. Nat. Med., 8, 1288–1295.[CrossRef][Web of Science][Medline]

26. Lee, H., Ogawa, H., Fujioka, M. and Gerton, G.L. (1994) Guanidinoacetate methyltransferase in the mouse: extensive expression in Sertoli cells of testis and in microvilli of caput epididymis. Biol. Reprod., 50, 152–162.[Abstract]

27. Lee, H., Kim, J.H., Chae, Y.J., Ogawa, H., Lee, M.H. and Gerton, G.L. (1998) Creatine synthesis and transport systems in the male rat reproductive tract. Biol. Reprod., 58, 1437–1444.[Abstract/Free Full Text]

28. Marescau, B., Deshmukh, D.R., Kockx, M., Possemiers, I., Qureshi, I.A., Wiechert, P. and De Deyn, P.P. (1992) Guanidino compounds in serum, urine, liver, kidney, and brain of man and some ureotelic animals. Metabolism, 41, 526–532.[CrossRef][Web of Science][Medline]

29. Bia, B.L., Cassidy, P.J., Young, M.E., Rafael, J.A., Leighton, B., Davies, K.E., Radda, G.K. and Clarke, K. (1999) Decreased myocardial nNOS, increased iNOS and abnormal ECGs in mouse models of Duchenne muscular dystrophy. J. Mol. Cell. Cardiol., 31, 1857–1862.[CrossRef][Web of Science][Medline]

30. Boehm, E., Chan, S., Monfared, M., Wallimann, T., Clarke, K. and Neubauer, S. (2003) Creatine transporter activity and content in the rat heart supplemented by and depleted of creatine. Am. J. Physiol. Endocrinol. Metab., 284, E399–406.[Abstract/Free Full Text]

31. Das, A.M., Jolly, R.D. and Kohlschütter, A. (1999) Anomalies of mitochondrial ATP synthase regulation in four different types of neuronal ceroid lipofuscinosis. Mol. Genet. Metab., 66, 349–355.[CrossRef][Web of Science][Medline]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				L. M. J. W. van Driel, M. J. C. Eijkemans, R. de Jonge, J. H. M. de Vries, J. B. J. van Meurs, E. A. P. Steegers, and R. P. M. Steegers-Theunissen Body Mass Index Is an Important Determinant of Methylation Biomarkers in Women of Reproductive Ages J. Nutr., December 1, 2009; 139(12): 2315 - 2321. [Abstract] [Full Text] [PDF]

				M.C. Bianchi Reply: AJNR Am. J. Neuroradiol., February 1, 2008; 29(2): 214 - 214. [Full Text] [PDF]

				H. E. Kan, E. Meeuwissen, J. J. van Asten, A. Veltien, D. Isbrandt, and A. Heerschap Creatine uptake in brain and skeletal muscle of mice lacking guanidinoacetate methyltransferase assessed by magnetic resonance spectroscopy J Appl Physiol, June 1, 2007; 102(6): 2121 - 2127. [Abstract] [Full Text] [PDF]

				M.C. Bianchi, M. Tosetti, R. Battini, V. Leuzzi, M.G. Alessandri', C. Carducci, I. Antonozzi, and G. Cioni Treatment Monitoring of Brain Creatine Deficiency Syndromes: A 1H- and 31P-MR Spectroscopy Study AJNR Am. J. Neuroradiol., March 1, 2007; 28(3): 548 - 554. [Abstract] [Full Text] [PDF]

				J. A. Vitarius, E. Sehayek, and J. L. Breslow Identification of quantitative trait loci affecting body composition in a mouse intercross PNAS, December 26, 2006; 103(52): 19860 - 19865. [Abstract] [Full Text] [PDF]

				D. R. Boverhof, L. D. Burgoon, C. Tashiro, B. Sharratt, B. Chittim, J. R. Harkema, D. L. Mendrick, and T. R. Zacharewski Comparative Toxicogenomic Analysis of the Hepatotoxic Effects of TCDD in Sprague Dawley Rats and C57BL/6 Mice Toxicol. Sci., December 1, 2006; 94(2): 398 - 416. [Abstract] [Full Text] [PDF]

				M. E. Cullen, A. H.Y. Yuen, L. E. Felkin, R. T. Smolenski, J. L. Hall, S. Grindle, L. W. Miller, E. J. Birks, M. H. Yacoub, and P. J.R. Barton Myocardial Expression of the Arginine:Glycine Amidinotransferase Gene Is Elevated in Heart Failure and Normalized After Recovery: Potential Implications for Local Creatine Synthesis Circulation, July 4, 2006; 114(1_suppl): I-16 - I-20. [Abstract] [Full Text] [PDF]

				H. E. Kan, T. E. Buse-Pot, R. Peco, D. Isbrandt, A. Heerschap, and A. de Haan Lower force and impaired performance during high-intensity electrical stimulation in skeletal muscle of GAMT-deficient knockout mice Am J Physiol Cell Physiol, July 1, 2005; 289(1): C113 - C119. [Abstract] [Full Text] [PDF]

				M. ten Hove, C. A. Lygate, A. Fischer, J. E. Schneider, A. E. Sang, K. Hulbert, L. Sebag-Montefiore, H. Watkins, K. Clarke, D. Isbrandt, et al. Reduced Inotropic Reserve and Increased Susceptibility to Cardiac Ischemia/Reperfusion Injury in Phosphocreatine-Deficient Guanidinoacetate-N-Methyltransferase-Knockout Mice Circulation, May 17, 2005; 111(19): 2477 - 2485. [Abstract] [Full Text] [PDF]

				M. Nahrendorf, M. Spindler, K. Hu, L. Bauer, O. Ritter, P. Nordbeck, T. Quaschning, K.-H. Hiller, J. Wallis, G. Ertl, et al. Creatine kinase knockout mice show left ventricular hypertrophy and dilatation, but unaltered remodeling post-myocardial infarction Cardiovasc Res, February 1, 2005; 65(2): 419 - 427. [Abstract] [Full Text] [PDF]

				H. E Kan, W. K. J. Renema, D. Isbrandt, and A. Heerschap Phosphorylated guanidinoacetate partly compensates for the lack of phosphocreatine in skeletal muscle of mice lacking guanidinoacetate methyltransferase J. Physiol., October 1, 2004; 560(1): 219 - 229. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) All Versions of this Article: 13/9/905    most recent ddh112v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (26) Request Permissions Google Scholar Articles by Schmidt, A. Articles by Isbrandt, D. Search for Related Content PubMed PubMed Citation Articles by Schmidt, A. Articles by Isbrandt, D. Social Bookmarking          What's this?

Online ISSN 1460-2083 - Print ISSN 0964-6906

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics