- Split View
-
Views
-
Cite
Cite
Theo G.M.F. Gorgels, Xiaofeng Hu, George L. Scheffer, Allard C. van der Wal, Johan Toonstra, Paulus T.V.M. de Jong, Toin H. van Kuppevelt, Christiaan N. Levelt, Anneke de Wolf, Willem J.P. Loves, Rik J. Scheper, Ron Peek, Arthur A.B. Bergen, Disruption of Abcc6 in the mouse: novel insight in the pathogenesis of pseudoxanthoma elasticum, Human Molecular Genetics, Volume 14, Issue 13, 1 July 2005, Pages 1763–1773, https://doi.org/10.1093/hmg/ddi183
- Share Icon Share
Abstract
Pseudoxanthoma elasticum (PXE) is a heritable disorder of connective tissue, affecting mainly skin, eye and the cardiovascular system. PXE is characterized by dystrophic mineralization of elastic fibres. The condition is caused by loss of function mutations in ABCC6. We generated Abcc6 deficient mice (Abcc6−/−) by conventional gene targeting. As shown by light and electron microscopy Abcc6−/− mice spontaneously developed calcification of elastic fibres in blood vessel walls and in Bruch's membrane in the eye. No clear abnormalities were seen in the dermal extracellular matrix. Calcification of blood vessels was most prominent in small arteries in the cortex of the kidney, but in old mice, it occurred also in other organs and in the aorta and vena cava. Newly developed monoclonal antibodies against mouse Abcc6 localized the protein to the basolateral membranes of hepatocytes and the basal membrane in renal proximal tubules, but failed to show the protein at the pathogenic sites. Abcc6−/− mice developed a 25% reduction in plasma HDL cholesterol and an increase in plasma creatinine levels, which may be due to impaired kidney function. No changes in serum mineral balance were found. We conclude that the phenotype of the Abcc6−/− mouse shares calcification of elastic fibres with human PXE pathology, which makes this model a useful tool to further investigate the aetiology of PXE. Our data support the hypothesis that PXE is in fact a systemic disease.
INTRODUCTION
Pseudoxanthoma elasticum (PXE) is a heritable disorder of the connective tissue with multiple systemic manifestations. Clinical expression is variable and most frequently involves skin, eye and blood vessels (1). Skin abnormalities consist of yellowish papules and plaques on the neck. Flexural areas, such as axils and groin, develop redundant folds of lax skin (2). Examination of the fundus of the PXE eye reveals angioid streaks representing breaks in Bruch's membrane, which may lead to choroidal neovascularization, haemorrhages and, consequently, loss of visual acuity (3). Cardiovascular complications include hypertension, arterial insufficiencies in the extremities and gastrointestinal haemorrhages (3). Histopathological analysis of the affected tissues reveals calcification and fragmentation of elastic fibres as well as abnormalities in collagen fibrils (3,4). PXE symptoms usually appear in the second or third decade of life, although we have seen patients with skin changes from 4 years on. The disease is progressive and as yet incurable.
The prevalence of PXE is at least one in 25 000 (5,6). The vast majority of patients are sporadic cases. In familial cases, the disease almost exclusively segregates in an autosomal recessive fashion (6). Recently, PXE was found to be caused by loss of function mutations in the ABCC6 gene (7–9). Furthermore, we found that a frequent (founder) mutation in ABCC6, R1141X (10), may be associated with a strong increase in the prevalence of premature coronary artery disease (5).
The gene ABCC6 encodes a protein of 1503 amino acids, which contains three membrane-spanning domains and two ATP-binding cassettes (ABC). On the basis of amino acid alignments, ABCC6 has been assigned to the subfamily of multidrug resistance proteins (MRP) (11,12). Members of the MRP family are known to transport drugs, toxic substances or lipids across the cell membrane, which may lead to clinical multidrug resistance of cancer cells. The physiological substrate of ABCC6 is not yet known. However, Ilias et al. (13) found that glutathione conjugates, including leukotrien-C4 (LTC4) and N-ethylmaleimide S-glutathione (NEM-GS), are actively transported by human ABCC6. In three ABCC6 mutant forms associated with PXE, loss of ATP-dependent transport activity was observed (13).
Expression of ABCC6 is high in liver and kidney (8,14). The protein is located at the basolateral side of hepatocytes (15). This location suggests that ABCC6 exports substances out of the liver cell into the blood stream. Interestingly, ABCC6 mRNA expression is low in tissues usually affected by the disease: skin, eye and blood vessels (8). This finding has led to the hypothesis that PXE is a systemic disease (16). On the other hand, the protein can be detected in cultured fibroblasts, which could support a local origin of PXE pathology (10,17).
Mutation analysis of ABCC6 (10,18) and the in vitro transport studies with PXE-associated ABCC6 mutants (13) suggest that PXE is primarily caused by deficiency in transport activity of this ABC transporter. Yet, the exact role of ABCC6 in the aetiology of PXE is still unclear. To investigate the biological function of ABCC6, we constructed a genetically modified mouse in which Abcc6 was disrupted. Analysis of Abcc6 deficient mice showed calcification of elastic fibres in Bruch's membrane in the eye and in blood vessel walls. In addition, alterations in plasma HDL cholesterol and plasma creatinine, but not in mineral balance, were found. Consequently, the Abcc6 deficient mouse appears to be a suitable animal model for further studies of PXE.
RESULTS
Generation of Abcc6−/− mice
The targeting strategy for generating mice lacking Abcc6 is shown in Figure 1A. We chose to delete the coding sequence of the first nucleotide-binding fold (NBF1), because PXE causing mutations occur in this region (18,19). In addition, in vitro studies showed that NBFs are essential for the function of ABC proteins in general (12) and ABCC6 in particular (12,13). Finally, this strategy has been previously successful in creating functional knock-out mice for other ABCC transporters, such as MRP1 (20).
The largest part of Abcc6 NBF1, including the Walker A and B and signature C motifs, is encoded in exons 16, 17 and 18. In the targeting construct, the genomic fragment containing these exons was replaced by a selectable hygromycin resistance marker gene (Fig. 1). After electroporation of the construct into ES cells, hygromycin-resistant clones were screened for homologous recombination. Out of 96 ES cell clones examined, only nine clones showed correct integration of the construct at the targeted site as indicated by Southern analysis using probes located external to the construct (Fig. 1) and by PCR using primers located in the undeleted part of exon 18 and in the hygromycin resistance gene (data not shown). ES cells of correctly targeted clones were injected into C57Bl/6 blastocysts, resulting in chimeric mice that transmitted the targeted allele, as detected with Southern blot analysis and PCR. Heterozygous offspring were mated with C57Bl/6 mice, for backcrossing to C57Bl/6, and were interbred to generate Abcc6+/+, Abcc6+/− and Abcc6−/− mice. Both Abcc6+/− and Abcc6−/− mice were viable and fertile. Inheritance of the mutated allele showed no deviation from the expected Mendelian ratio.
Confirmation of disruption of Abcc6
We analyzed mRNA and protein expression of the targeted locus. Abcc6 mRNA expression was first studied by RT–PCR on Abcc6−/− mouse liver cDNA with three different primers sets with amplicons located up- and down-stream of the targeted area and within the targeted area. Using cDNA from Abcc6−/− mouse liver, no product was amplified with the primer set located within the deleted area, but the amplicons up- and down-stream of the deletion were present. This indicates that an aberrant Abcc6 transcript is formed, lacking the targeted NBF1 region. Next, we sequenced a 2 kb RT–PCR product between exons 15 and 19, spanning the targeted area. The latter showed that the open reading frame, in fact any putative reading frame, was interrupted by several stop codons through introduction of the anti-sense hygromycin cassette in the transcript.
To study Abcc6 protein expression, we raised rat monoclonal antibodies (Mabs) against a fusion protein containing amino acids 843–1000 of mouse Abcc6. Two Mabs (M6II-24 and M6II-68) were used in western blots with protein fractions derived from liver and kidney of wild-type and Abcc6−/− mice (Fig. 2). In the wild-type fractions, both Mabs stained a protein band at 165 kDa, which corresponds to the predicted size of full-length mouse Abcc6. Abcc6 protein was not detected in the protein fractions of Abcc6−/− mice (Fig. 2).
Phenotype of Abcc6−/− mice
To find out whether Abcc6 deficient mice spontaneously develop pathology, we kept Abcc6−/− and wild-type mice under standard conditions in our stables and checked them regularly. The oldest Abcc6−/− mice analyzed were 22 months old and no gross abnormalities or a clear difference in mortality with wild-type littermates was seen. Mice of various ages were sacrificed for histological examination.
Abcc6−/− mice spontaneously developed calcification of blood vessels. This phenomenon was most prominent in the cortex of the kidney (Fig. 3). Here, calcification of blood vessels first appeared ∼6 months and progressed with age (Table 1). Both male and female Abcc6−/− mice were affected. In medium-sized arteries, mineral deposits were noticed within the vessel wall (Figs 3 and 4A and E). Elastica staining in adjacent sections revealed that calcium was deposited along the elastic fibres in the vessel wall (Fig. 4B and D). Smaller blood vessels and capillaries were extensively calcified with deposits protruding into the lumen (Figs 3 and 4F).
In older Abcc6−/− mice, blood vessel calcification was also found in other organs. In 17–19-month-old Abcc6−/− mice, calcium deposits were seen in small blood vessels in many tissues examined, such as adipose tissue at various topographic locations, skin, submandibular gland and the tongue. In addition, calcium deposition was observed in the elastic lamellae of the aorta, in the elastic fibres of the vena cava (Fig. 5A and B) and in the internal elastic lamina of coronary arteries. Examination of these tissues in age-matched wild-type mice revealed no calcifications.
As skin and eye are usually the first organs to show clinical signs in patients, we examined these structures extensively in Abcc6−/− mice. The mice had no skin abnormalities that were visible to the naked eye. Light microscopic analysis of the skin of the neck, arm pit, groin and back showed some sparse calcified blood vessels, but calcium deposition at dermal elastic fibres was not observed. Next, we employed electron microscopy on the skin of the groin of 18-month-old three Abcc6−/− and three wild-type mice. No clear differences were seen in the dermal elastic fibres, although these fibres in Abcc6−/− mice sometimes had a mottled appearance (data not shown).
In the eyes of old Abcc6−/− mice, calcification of Bruch's membrane was demonstrated with Alizarin red S staining (Fig. 5D). With electron microscopy, we observed accumulation of electron dense material in the lamina elastica of Bruch's membrane of Abcc6−/− mice (Fig. 6). In addition, abnormal fibres were seen having an electron lucent core and a very electron dense cortex. These fibres often branched and ran crisscross in the lamina elastica, the inner collagenous zone and, to lesser extent, in the outer collagenous zone of Bruch's membrane. These abnormal structures were found in all three of the 18-month-old Abcc6−/− mice examined, but were not observed in the age-matched wild-type mice (n=3).
Immunolocalization of Abcc6
Abcc6 protein expression and localization was studied using the anti-Abcc6 Mabs on frozen sections of several tissues of wild-type and Abcc6−/− mice. In the liver of wild-type mice, clear Abcc6 staining was observed in the basolateral membranes of the hepatocytes. In the kidney of wild-type mice, the Mabs localized the protein to the basal membranes of the proximal tubules. Negative control stainings, using either an irrelevant, isotype matched antibody or frozen tissue sections from Abcc6−/− mice, showed no staining (Fig. 7).
As PXE patients develop pathology in skin, eye and blood vessels, we examined these structures for the presence of Abcc6 protein in wild-type and Abcc6−/− mice. No immunoreactivity for Abcc6 was seen in blood vessels of the kidney, which, as mentioned earlier, can develop pathology in the Abcc6−/− mouse. Eye and skin also did not show Abcc6 immunoreactivity (Fig. 7). In addition, we examined tissues of newborn mice to see whether Abcc6 protein was developmentally expressed at these pathogenic sites: no Abcc6 protein was detected in skin, eye and renal blood vessels in 0- and 3-day-old wild-type and Abcc6−/− mice (data not shown).
Analysis of blood plasma
In view of a putative systemic origin of PXE (16), we analyzed several potentially relevant constituents of blood plasma of 11-month-old Abcc6−/− mice and healthy, age-matched control mice. Interestingly, no significant changes were found in plasma mineral levels (Table 2). However, statistically significant differences were found in HDL-cholesterol levels, which were reduced in Abcc6−/− mice (P=0.01), and in creatinine levels, which were elevated in the knock out (P=0.03), when compared with wild-type mice. To confirm and extend these findings, we analyzed plasma lipid spectrum and plasma levels of creatinine and urea of 2.5- and 8-month-old mice, which were fasted 7–9 h prior to collection of blood (Table 3). In 8-month-old mice, the significant changes in HDL cholesterol and in creatinine levels were reproduced. In young animals, however, no difference with wild-type mice was found. Plasma urea levels did not differ at these ages.
DISCUSSION
When ABCC6 was identified in the year 2000 as the gene causing PXE, this came as a surprise since the relation between this ABC transporter and elastic fibre calcification was not readily apparent (7–9). We now disrupted Abcc6 in the mouse to further investigate this relationship. The most important finding is that Abcc6−/− mice spontaneously developed calcification and elastic fibre abnormalities in Bruch's membrane and blood vessel walls. This phenotype resembles the ophthalmic and cardiovascular pathologies of PXE (3) and provides us with a suitable animal model for further investigation of the aetiology of PXE.
Phenotype of Abcc6−/− mice partly resembles PXE pathology
Progressive calcification and fragmentation of elastic fibres in skin, eye and blood vessels are characteristics for human PXE (2,3,21). The phenotype of the Abcc6−/− mouse partly resembles the pathology of PXE patients: Abcc6−/− mice spontaneously developed calcification and elastic fibre abnormalities in blood vessels and Bruch's membrane in the eye, whereas no clear changes were seen in the extracellular matrix of the skin.
In the Abcc6−/− mouse, calcification apparently affected primarily small arteries, with small arteries in the cortex of the kidney being most vulnerable. Calcium deposits could be detected at the elastic component in the vessel wall. In old Abcc6−/− mice, large blood vessels were also affected, as exemplified by calcification of the elastic lamellae of the aorta and elastic fibres of the vena cava.
These features correspond well to the pathology in PXE patients. Intermittent claudication, which is due to mineral deposits in arteries in the legs, is a common cardiovascular symptom of PXE (2,3). In addition, calcification of arteries in internal organs of PXE patients has been demonstrated by various techniques such as radiography and echography. In view of the kidney pathology in the Abcc6−/− mouse, it is of interest that 25% of PXE patients develop renovascular hypertension and echographic opacities due to calcifications of arteries in kidneys, spleen and pancreas (3,22–24). Finally, there are histopathologic data of blood vessel calcification in various organs of PXE patients. Recently, a large number of tissues from two PXE patients were examined ultrastructurally and alterations of the elastic component were seen in almost all small- and medium-sized vessels in all organs. In addition, fragmentation and mineralization of elastic fibres were observed in the aorta and vena cava (4).
In Bruch's membrane of old Abcc6−/− mice, calcification was found by employing light microscopy and Alizarin red S staining. At the ultrastructural level, we observed electron dense material in the lamina elastica of Bruch's membrane. In addition, abnormal fibres with an electron dense cortex were present in Bruch's membrane. Similar fibres have been described as calcified elastic fibres in ageing human Bruch's membrane (25). As we did not see these abnormalities in age-matched control mice, this may resemble the calcification process of Bruch's membrane in PXE patients (3,4), eventually leading to breaks in Bruch's membrane.
The skin of the Abcc6−/− mouse appears to be less involved in the disease than is generally the case in PXE patients. PXE patients frequently have extensive mineralization of elastic fibres in the dermis, typically in the neck and flexural areas (2). In contrast, in the Abcc6−/− mouse, von Kossa staining did not reveal calcification of elastic fibres in these skin areas, whereas in the same tissue, calcification of blood vessels had already occurred. The reason for this difference between man and mice is not clear. However, in PXE patients, the relative involvement of eye, skin and cardiovascular system varies considerably, even between family members having the same mutations (3,18). This suggests that environmental factors or other genetic factors may have an impact on the expression of the disease (26), which may shift the balance towards a cardiovascular and eye phenotype in the mouse.
Localization of Abcc6
An important unresolved issue is whether PXE is a systemic disease (16). The alternative is that it is caused by local dysfunction of the protein within the connective tissue itself, e.g. in fibroblasts (4,17). When examining ABCC6 expression in human tissues, only low levels of ABCC6 mRNA were found in tissues frequently affected by PXE (8) and no ABCC6 protein was detected by immunohistochemistry in skin and eye (15). In contrast, ABCC6 mRNA and protein were highly expressed in human liver, which is unaffected by PXE (8,14). Beck et al. (27) examined mouse tissues with polyclonal antibodies and found a more widespread Abcc6 protein expression in multiple tissues including liver, kidney, eye, skin and blood vessels. In the present study, we used polyclonal and monoclonal antibodies to localize Abcc6, with Abcc6−/− mouse tissues as negative control. In the mouse, we found essentially the same localization as previously reported for the human tissues (15): Abcc6 was localized to the basolateral membranes of hepatocytes and basal membranes of the proximal tubules of the kidney. Furthermore, examining tissue of adult as well as newborn mice, we could not detect Abcc6 protein at pathogenic sites such as renal blood vessels, the skin or Bruch's membrane in the eye. Therefore, our data do not support a local but a systemic aetiology of PXE in the mouse.
Analysis of blood plasma
The high expression of Abcc6 in liver and kidney (8,14,28), its putative function as transporter and its localization to basolateral membranes of hepatocytes (15) are ideal for a potential systemic origin of PXE by changes in blood composition. Analysis of blood plasma resulted in three important findings.
First, the plasma concentration of minerals, such as calcium, was not changed in the Abcc6−/− mouse. Therefore, no support was obtained for a systemic disturbance of the electrolyte balance as systemic cause of PXE. This is particularly relevant as several studies suggested that mineral intake and balance may influence the PXE disease state (24,29,30).
Secondly, creatinine levels were significantly increased in older Abcc6−/− mice (≥8 months), which points toward impaired kidney function (31). This is an interesting finding, since, in patients with chronic kidney disease and renal failure, increased calcification as well as thickening and elastic changes of the blood vessel wall have been reported (32). These changes may be similar to those found in the carotid artery of PXE patients (33). In this way, kidney dysfunction could not only be a consequence of generalized vascular disease in PXE but also a contributor.
Thirdly, plasma HDL cholesterol was reduced by 25% in Abcc6−/− mice of ≥8 months. This is remarkable because low HDL-cholesterol levels are associated with vascular pathology. In the literature, there is also support for the involvement of abnormal serum lipid levels in PXE. In a family with a proven apoa1/CIII deficiency of the Detroit type, some striking clinical similarities with PXE have been reported (34). Cross-sectional studies of PXE patients and controls suggested depressed plasma HDL cholesterol (2) and hypertriglyceridemia in PXE (2,29). Recently, a DNA sequence polymorphism in ABCC6 was found associated with altered plasma HDL cholesterol and triglyceride levels (35).
The observed HDL reduction in Abcc6−/− mice of ≥8 months may be the result of the impairment of kidney function. In patients with chronic renal failure, hepatic Apo A-I synthesis decreases and HDL levels fall, whereas plasma levels of triglycerides rise (36,37). Alternatively, it is tempting to speculate that Abcc6, like many other ABC transporters (38), can be directly involved in lipid transport or metabolism. Although it is well known in humans that lipid distribution and composition changes considerably during development and ageing (39), very little is known about an age-dependent regulation of Abcc6 expression (40), which could be a cause of reduction of HDL in older mice.
In summary, we observed three, probably related, pathological events in the ageing Abcc6−/− mouse. At young age (2.5 months), calcification of renal vessel walls (and other tissues) was absent and renal function was normal as evidenced by normal creatinine, urea and cholesterol plasma levels. In 8–19-month-old mice, progressive calcification of blood vessels and other tissues occurred, serum creatinine levels rose, suggesting renal failure, and HDL-cholesterol levels fell. It is currently not clear what exact mechanism(s) underlie this potential chain of events.
Combining our data on Abcc6 protein localization and blood plasma values with those of the literature, we suggest that alterations in plasma (lipid) composition may be involved in the development of clinical features of PXE. In this respect, it is important to test lipid substrates in in vitro Abcc6 transport studies. In addition, further studies of plasma (lipid) composition in PXE patients with defined molecular lesions in ABCC6 are warranted, whereas the Abcc6−/− mouse seems to be a good model for dietary studies.
MATERIALS AND METHODS
Generation of Abcc6−/− mice
Isogenic mouse Abcc6 genomic DNA was obtained by screening a 129/Ola mouse bacteriophage artificial chromosome (BAC) library (Incite) with mouse Abcc6 cDNA probes. A BAC containing the Abcc6 gene was isolated (BAC no. A-962B4) and a 14.5 kb HindIII restriction fragment was subcloned into pGEM7 (Promega, Madison, WI, USA). This fragment spans the genomic region of exons 15–22 and encompasses the coding sequence of the NBF1 (roughly exons 16–18).
A gene-targeting construct was designed in which most of the NBF1 coding sequence was replaced by a marker gene to select for hygromycin resistance (Fig. 1). The targeting vector was constructed on the basis of pBluescript II KS (+/−) phagemid with a hygromycin resistance gene driven by the mouse phosphoglycerate kinase promoter (PGK-hygro vector, kindly provided by Dr J. Wijnholds). A fragment of 4.1 kb surrounding exon 19 and containing part of exon 18 was cut out of genomic Abcc6 with Eco47III and BssHII and was inserted into the PGK-hygro vector, to serve as the 3′ homologous arm of the targeting construct. Next, a 2.5 kb fragment surrounding exon 15 was isolated after digestion with PstI and XbaI and ligated to the PGK-hygro cassette as the 5′ arm. Sequencing of the final targeting construct showed that exons 16, 17 and part of 18 were replaced by the PGK–hygromycin sequence, which was oriented in antisense direction. Correct targeting would delete the coding sequence for amino acids 647–782 of Abcc6 protein. This eliminates the largest part of NBF1 including the Walker A and B and the signature C motifs.
The 12.5 kb targeting fragment was linearized by BssHII and transfected into 129/OLA-derived E14 ES cells by electroporation. Hygromycin resistant clones were picked and expanded. After genotyping, the karyotype of ES cells of correctly targeted clones was checked. ES cells were injected into C57Bl/6 blastocysts to generate chimeric male mice, which were mated with C57Bl/6 females. Offspring heterozygous for the disrupted allele (Abcc6+/−) were mated with C57Bl/6 mice, for backcrossing to C57Bl/6, and were interbred to generate Abcc6+/+, Abcc6+/− and Abcc6−/− mice. Most data of the present study were obtained on mice of backcross number 2 to C57Bl/6. In addition, Abcc6−/− and wild-type mice of backcross number 5 to C57Bl/6 were used in experiments to confirm renal blood vessel calcification in Abcc6−/− mice and in experiments analyzing blood composition of 2.5- and 8-month-old mice. Mice were kept in a 12 h light (<100 lux)/12 h dark cycle and had free access to water and food.
Genotyping
Genotyping was done with Southern blot and multiplex PCR. ES cells and the first generation of mice were genotyped by both Southern blot and PCR. After confirmation of the genotype and correct chromosomal integration, next generations of mice were genotyped only by PCR.
DNA was isolated from ES cells and snippets of the ears, according to standard protocols (41). Multiplex PCR employed the following primers: 18F (5′-TGA-ATC-TTT-CTG-GGG-GCC-AG-3′), 18R (5′-GTA-CCC-TGG-AGC-AAT-CCA-CT-3′) and T2 (5′-ATG-TGG-AAT-GTG-TGC-GAG-GCC-3′). Primers 18F and 18R amplify a 163 bp fragment of exon 18 on the wild-type allele, whereas primers 18R and T2 yield a 246 bp fragment spanning the boundary of PGK-hygro cassette on the targeted allele. For Southern analysis, genomic DNA was digested by EcoRI and KpnI or by EcoRI and BclI (Biolabs), size fractionated on 0.8% agarose gels and transferred to GeneScreen Plus nylon membranes (NEN). Membranes were probed with radiolabelled PCR-amplified fragments of genomic Abcc6. The probes were located external to the construct (Fig. 1).
Generation of antibodies to mouse Abcc6
Antibodies against mouse Abcc6 were generated using the glutathione S-transferase (GST) gene fusion system (Pharmacia). The third transmembrane domain of Abcc6 was amplified by PCR on mouse liver cDNA. This part of the protein is located downstream of the targeted region. Primers (forward: 5′-GAGACCATGGGSGCCCTGGTGGGTCTT-3′; reverse: 5′-GAGAATTCTTGGGGATGCGAGCGTAG-3′) yielded an 1190 bp amplicon, which was cloned into the pGEMTeasy vector (Promega). Digestion with NcoI resulted in a 471 bp fragment corresponding to amino acids 843–1000 of mouse Abcc6. This fragment was cloned into a derivative of pGEX-3X plasmid (Pharmacia). After transformation of Escherichia coli DH5α cells, the 44 kDa Abcc6–GST fusion protein was produced, isolated and purified using glutathione beads. Abcc6–GST fusion protein was injected in rabbits and rats for production of polyclonal and monoclonal antibodies, respectively. Immunization and fusion protocols for monoclonal production were as described earlier (42,43). In brief, a 12-week-old female Wistar rat received ∼200 µg fusion protein per injection. Four booster injections were given. Draining lymph nodes and the spleen of the rat were fused with Sp20 mouse myeloma cells. Supernatants of obtained hybridoma cells containing Mabs were screened on ELISA plates coated with specific fusion protein and, as a control, on plates coated with irrelevant fusion protein.
Antibody binding was detected using HRP-labelled rabbit-anti-rat serum (1:500, Dako, Copenhagen, Denmark) and 5-amino-2-hydroxybenzoic acid (Merck, Darmstadt, Germany) and 0.02% H2O2 as a chromogen. Screening of hybridoma supernatants yielded several positive hybridomas. Two Mabs, M6II-24 and M6II-68, were selected for further analysis. Isostrips (Serotec, Oxford, UK) showed that M6II-24 is of IgG2a subclass and M6II-68 is of IgG2b subclass.
Immunodetection of Abcc6
For Western blot analysis, liver and kidney samples were thawed in lysis buffer, consisting of 50 mm Tris–HCl, pH 7.4, 150 mm NaCl, 1 mm EDTA, 1% Triton X-100, 1 mm PMSF, 1 mm Na3VO4 and protease inhibitor cocktail (Roche). Next, the tissue was pottered, kept on ice for 30 min and cleared by centrifugation at 10 000g for 15 min. Protein concentration was determined by micro-BCA protein assay reagent kit (Pierce). Protein samples (30 µg/lane) were size fractionated by SDS–polyacrylamide (7.5%) gel electrophoresis and transferred onto nitrocellulose membranes by electroblotting. Membranes were probed with anti-Abcc6 Mabs M6II-24 and M6II-68, described earlier, and stained using HRP-labelled donkey-anti-rat secondary antibodies (Jackson Immunoresearch) and a chemiluminescent detection system (Pierce).
Immunolocalization of Abcc6 was done on cryosections (4–6 µm). These were air dried overnight and fixed for 7 min in acetone at room temperature. Slides were incubated with primary antibody for 1 h at room temperature. HRP-labelled goat-anti-rat serum (1:100, Santa Cruz, CA, USA) was used as secondary reagent. Colour development was with 0.4 mg/ml amino-ethyl-carbazole (AEC) and 0.02% H2O2 as a chromogen.
Tissue collection and morphological analysis
Mice were sacrificed by CO2/O2 and cervical dislocation. For immunohistochemistry and biochemical analysis, tissues were excised, frozen in liquid N2 and stored at −80°C until use.
For routine histological analysis, organs were excised, fixed in 4% paraformaldehyde in 0.1 m phosphate buffer for at least 15 h, dehydrated, embedded in paraffin and sectioned in 4–7 µm thick slices. We used standard staining procedures of haematoxylin–eosin (HE), von Kossa, Alizarin red S, Elastica von Gieson (EVG) and Elastica staining. Organs examined included liver, kidney, skin (of groin, neck and back), eye, heart, descending aorta, lung, stomach, tongue, lower lip and ear. For orientation, eyes were marked nasally with Alcian blue (5 in 96% ethanol). Ages of the mice were 6, 10, 11, 14, 17 and 19 months. Per genotype (Abcc6−/− and wild-type) and per age group, 3–5 mice were examined.
In addition, 18-month-old mice were fixed and prepared for electron microscopy. They were anesthetized by intraperitoneal injection with Nembutal and perfused through the heart with 1% glutaraldehyde and 1.25% paraformaldehyde in 0.1 m cacodylate buffer, pH 7.4. Eye, skin, aorta and liver were post-fixed in 1% OsO4 in 0.1 m cacodylate buffer, pH 7.4, and embedded in epon for electron microscopic analysis. Ultra-thin sections were stained with lead citrate and uranyl acetate. Other tissues of these mice were embedded in paraffin and examined at the light microscopical level.
For blood analysis, Abcc6−/− and wild-type mice were sacrificed by CO2/O2 and blood was collected in tubes containing lithium heparin. For measurement of the lipid spectrum, mice were fasted for 7–9 h prior to collection of blood.
Statistical analysis
Statistical analysis was carried out using Student's t-test, two-sided.
ACKNOWLEDGEMENTS
The authors wish to thank Dr J. Wijnholds for excellent assistance in the course of creating the knockout mouse, Professor Dr R.S. Reneman, Dr M. Tripp and Dr A. Plomp for helpful discussions, Ing. I. Versteeg and Ing. A.J.N. Schoonderwoerd for technical assistance. Financial support of the Stichting Blindenpenning and the Algemene Nederlandse Vereniging ter Voorkoming van Blindheid is gratefully acknowledged.
Conflict of Interest statement. None declared.
Age (months) . | Wild-type (n) . | Abcc6−/− (n) . |
---|---|---|
2.5 | 0 (5) | 0 (4) |
6 | 0 (4) | 0.9±0.5 (4) |
10 | 0 (3) | 4.6±2.5 (3) |
17 | 0.1±0.2 (5) | 25.2±18.0 (5) |
Age (months) . | Wild-type (n) . | Abcc6−/− (n) . |
---|---|---|
2.5 | 0 (5) | 0 (4) |
6 | 0 (4) | 0.9±0.5 (4) |
10 | 0 (3) | 4.6±2.5 (3) |
17 | 0.1±0.2 (5) | 25.2±18.0 (5) |
n, number of mice. Per mouse, three von Kossa stained kidney sections were analysed and the counts were averaged. Mean±SD.
Age (months) . | Wild-type (n) . | Abcc6−/− (n) . |
---|---|---|
2.5 | 0 (5) | 0 (4) |
6 | 0 (4) | 0.9±0.5 (4) |
10 | 0 (3) | 4.6±2.5 (3) |
17 | 0.1±0.2 (5) | 25.2±18.0 (5) |
Age (months) . | Wild-type (n) . | Abcc6−/− (n) . |
---|---|---|
2.5 | 0 (5) | 0 (4) |
6 | 0 (4) | 0.9±0.5 (4) |
10 | 0 (3) | 4.6±2.5 (3) |
17 | 0.1±0.2 (5) | 25.2±18.0 (5) |
n, number of mice. Per mouse, three von Kossa stained kidney sections were analysed and the counts were averaged. Mean±SD.
. | Wild-type (n=11) . | Abcc6−/− (n=14) . | t-test (P-value) . |
---|---|---|---|
Sodium | 151±2 | 151±3 | 0.72 |
Calcium | 2.66±0.07 | 2.66±0.11 | 0.94 |
Phosphate | 3.03±0.35 | 3.01±0.37 | 0.87 |
Chloride | 116±2 | 115±2 | 0.21 |
Magnesium | 1.13±0.10 | 1.09±0.08 | 0.38 |
. | Wild-type (n=11) . | Abcc6−/− (n=14) . | t-test (P-value) . |
---|---|---|---|
Sodium | 151±2 | 151±3 | 0.72 |
Calcium | 2.66±0.07 | 2.66±0.11 | 0.94 |
Phosphate | 3.03±0.35 | 3.01±0.37 | 0.87 |
Chloride | 116±2 | 115±2 | 0.21 |
Magnesium | 1.13±0.10 | 1.09±0.08 | 0.38 |
Mean±SD in mM/l.
. | Wild-type (n=11) . | Abcc6−/− (n=14) . | t-test (P-value) . |
---|---|---|---|
Sodium | 151±2 | 151±3 | 0.72 |
Calcium | 2.66±0.07 | 2.66±0.11 | 0.94 |
Phosphate | 3.03±0.35 | 3.01±0.37 | 0.87 |
Chloride | 116±2 | 115±2 | 0.21 |
Magnesium | 1.13±0.10 | 1.09±0.08 | 0.38 |
. | Wild-type (n=11) . | Abcc6−/− (n=14) . | t-test (P-value) . |
---|---|---|---|
Sodium | 151±2 | 151±3 | 0.72 |
Calcium | 2.66±0.07 | 2.66±0.11 | 0.94 |
Phosphate | 3.03±0.35 | 3.01±0.37 | 0.87 |
Chloride | 116±2 | 115±2 | 0.21 |
Magnesium | 1.13±0.10 | 1.09±0.08 | 0.38 |
Mean±SD in mM/l.
. | 2.5-month-old mice . | 8-month-old mice . | ||||
---|---|---|---|---|---|---|
. | Wild type (n=8) . | Abcc6−/− (n=6) . | t-test (P-value) . | Wild-type (n=8) . | Abcc6−/− (n=8) . | t-test (P-value) . |
Total cholesterol (mm/l) | 2.2±0.6 | 2.4±0.9 | 0.61 | 3.2±0.7 | 2.3±0.7 | 0.01 |
HDL-cholesterol (mm/l) | 2.3±0.5 | 2.4±0.7 | 0.86 | 3.0±0.2 | 2.2±0.7 | 0.01 |
Triglycerides (mm/l) | 0.46±0.18 | 0.42±0.18 | 0.68 | 0.62±0.17 | 0.49±0.145 | 0.13 |
Creatinine (µm/l) | 8.75±1.3 | 8.67±0.5 | 0.87 | 7.8±1.3 | 9.3±0.7 | 0.01 |
Urea (mm/l) | 7.2±1.0 | 6.9±1.3 | 0.64 | 8.1±1.8a | 8.4±0.9a | 0.71a |
. | 2.5-month-old mice . | 8-month-old mice . | ||||
---|---|---|---|---|---|---|
. | Wild type (n=8) . | Abcc6−/− (n=6) . | t-test (P-value) . | Wild-type (n=8) . | Abcc6−/− (n=8) . | t-test (P-value) . |
Total cholesterol (mm/l) | 2.2±0.6 | 2.4±0.9 | 0.61 | 3.2±0.7 | 2.3±0.7 | 0.01 |
HDL-cholesterol (mm/l) | 2.3±0.5 | 2.4±0.7 | 0.86 | 3.0±0.2 | 2.2±0.7 | 0.01 |
Triglycerides (mm/l) | 0.46±0.18 | 0.42±0.18 | 0.68 | 0.62±0.17 | 0.49±0.145 | 0.13 |
Creatinine (µm/l) | 8.75±1.3 | 8.67±0.5 | 0.87 | 7.8±1.3 | 9.3±0.7 | 0.01 |
Urea (mm/l) | 7.2±1.0 | 6.9±1.3 | 0.64 | 8.1±1.8a | 8.4±0.9a | 0.71a |
aThese urea values were determined in five wild-type and six Abcc6−/− mice.
. | 2.5-month-old mice . | 8-month-old mice . | ||||
---|---|---|---|---|---|---|
. | Wild type (n=8) . | Abcc6−/− (n=6) . | t-test (P-value) . | Wild-type (n=8) . | Abcc6−/− (n=8) . | t-test (P-value) . |
Total cholesterol (mm/l) | 2.2±0.6 | 2.4±0.9 | 0.61 | 3.2±0.7 | 2.3±0.7 | 0.01 |
HDL-cholesterol (mm/l) | 2.3±0.5 | 2.4±0.7 | 0.86 | 3.0±0.2 | 2.2±0.7 | 0.01 |
Triglycerides (mm/l) | 0.46±0.18 | 0.42±0.18 | 0.68 | 0.62±0.17 | 0.49±0.145 | 0.13 |
Creatinine (µm/l) | 8.75±1.3 | 8.67±0.5 | 0.87 | 7.8±1.3 | 9.3±0.7 | 0.01 |
Urea (mm/l) | 7.2±1.0 | 6.9±1.3 | 0.64 | 8.1±1.8a | 8.4±0.9a | 0.71a |
. | 2.5-month-old mice . | 8-month-old mice . | ||||
---|---|---|---|---|---|---|
. | Wild type (n=8) . | Abcc6−/− (n=6) . | t-test (P-value) . | Wild-type (n=8) . | Abcc6−/− (n=8) . | t-test (P-value) . |
Total cholesterol (mm/l) | 2.2±0.6 | 2.4±0.9 | 0.61 | 3.2±0.7 | 2.3±0.7 | 0.01 |
HDL-cholesterol (mm/l) | 2.3±0.5 | 2.4±0.7 | 0.86 | 3.0±0.2 | 2.2±0.7 | 0.01 |
Triglycerides (mm/l) | 0.46±0.18 | 0.42±0.18 | 0.68 | 0.62±0.17 | 0.49±0.145 | 0.13 |
Creatinine (µm/l) | 8.75±1.3 | 8.67±0.5 | 0.87 | 7.8±1.3 | 9.3±0.7 | 0.01 |
Urea (mm/l) | 7.2±1.0 | 6.9±1.3 | 0.64 | 8.1±1.8a | 8.4±0.9a | 0.71a |
aThese urea values were determined in five wild-type and six Abcc6−/− mice.
References
Goodman, R.M., Smith, E.W., Paton, D., Bergman, R.A., Siegel, C.L., Ottesen, O.E., Shelley, W.M., Push, A.L. and McKusick, V.A. (
Hu, X., Plomp, A.S., van Soest, S., Wijnholds, J., de Jong, P.T. and Bergen, A.A. (
Gheduzzi, D., Sammarco, R., Quaglino, D., Bercovitch, L., Terry, S., Taylor, W. and Pasquali-Ronchetti, I.P. (
Trip, M.D., Smulders, Y.M., Wegman, J.J., Hu, X., Boer, J.M., ten Brink, J.B., Zwinderman, A.H., Kastelein, J.J., Feskens, E.J. and Bergen, A.A. (
Plomp, A.S., Hu, X., de Jong, P.T. and Bergen, A.A. (
Le Saux, O., Urban, Z., Tschuch, C., Csiszar, K., Bacchelli, B., Quaglino, D., Pasquali-Ronchetti, I., Pope, F.M., Richards, A., Terry, S. et al. (
Bergen, A.A., Plomp, A.S., Schuurman, E.J., Terry, S., Breuning, M., Dauwerse, H., Swart, J., Kool, M., van Soest, S., Baas, F. et al. (
Ringpfeil, F., Lebwohl, M.G., Christiano, A.M. and Uitto, J. (
Hu, X., Peek, R., Plomp, A., ten Brink, J., Scheffer, G., van Soest, S., Leys, A., de Jong, P.T. and Bergen, A.A. (
Homolya, L., Varadi, A. and Sarkadi, B. (
Borst, P. and Elferink, R.O. (
Ilias, A., Urban, Z., Seidl, T.L., Le Saux, O., Sinko, E., Boyd, C.D., Sarkadi, B. and Varadi, A. (
Kool, M., van der Linden M., de Haas, M., Baas, F. and Borst, P. (
Scheffer, G.L., Hu, X., Pijnenborg, A.C., Wijnholds, J., Bergen, A.A. and Scheper, R.J. (
Uitto, J., Pulkkinen, L. and Ringpfeil, F. (
Boraldi, F., Quaglino, D., Croce, M.A., Garcia Fernandez, M.I., Tiozzo, R., Gheduzzi, D., Bacchelli, B. and Pasquali-Ronchetti, I. (
Hu, X., Plomp, A., Wijnholds, J., ten Brink, J., van Soest, S., van den Born, L.I., Leys, A., Peek, R., de Jong, P.T. and Bergen, A.A. (
Le Saux, O., Beck, K., Sachsinger, C., Silvestri, C., Treiber, C., Goring, H.H., Johnson, E.W., de Paepe, A., Pope, F.M., Pasquali-Ronchetti, I. et al. (
Wijnholds, J., Evers, R., van Leusden, M.R., Mol, C.A., Zaman, G.J., Mayer, U., Beijnen, J.H., van der Valk, M., Krimpenfort, P. and Borst, P. (
Bergen, A.A., Plomp, A.S., Gorgels, T.G. and de Jong, P.T. (
Nikko, A.P., Dunningan, M. and Cockerell, C.J. (
Crespi, G., Derchi, L.E. and Saffioti, S. (
Sapadin, A.N., Lebwohl, M.G., Teich, S.A., Phelps, R.G., DiCostanzo, D. and Cohen, S.R. (
Hogan, M.J., Alvarado, J.A. and Weddell, J.E. (
Uitto, J. (
Beck, K., Hayashi, K., Nishiguchi, B., Le Saux, O., Hayashi, M. and Boyd, C.D. (
Langmann, T., Mauerer, R., Zahn, A., Moehle, C., Probst, M., Stremmel, W. and Schmitz, G. (
Renie, W.A., Pyeritz, R.E., Combs, J. and Fine, S.L. (
Martinez-Hernandez, A., Huffer, W.E., Neldner, K., Gordon, S. and Reeve, E.B. (
Johnson, C.A., Levey, A.S., Coresh, J., Levin, A., Lau, J. and Eknoyan, G. (
Amann, K., Tyralla, K., Gross, M.L., Eifert, T., Adamczak, M. and Ritz, E. (
Kornet, L., Bergen, A.A., Hoeks, A.P., Cleutjens, J.P., Oostra, R.J., Daemen, M.J., van Soest, S. and Reneman, R.S. (
Norum, R.A., Lakier, J.B., Goldstein, S., Angel, A., Goldberg, R.B., Block, W.D., Noffze, D.K., Dolphin, P.J., Edelglass, J., Bogorad, D.D. et al. (
Wang, J., Near, S., Young, K., Connelly, P.W. and Hegele, R.A. (
Kaysen, G.A. and Eiserich, J.P. (
Prinsen, B.H., de Sain-van der Velden, M.G., de Koning, E.J., Koomans, H.A., Berger, R. and Rabelink, T.J. (
Stefkova, J., Poledne, R. and Hubacek, J.A. (
Wilson, P.W., Anderson, K.M., Harris, T., Kannel, W.B. and Castelli, W.P. (
Aranyi, T., Ratajewski, M., Bardoczy, V., Pulaski, L., Bors, A., Tordai, A. and Varadi, A. (
Sambrook, J. and Russel, D.W. (
Scheffer, G.L., Maliepaard, M., Pijnenborg, A.C., van Gastelen, M.A., de Jong, M.C., Schroeijers, A.B., van der Kolk, D.M., Allen, J.D., Ross, D.D., van der Valk, P. et al. (
Scheffer, G.L., Kool, M., Heijn, M., de Haas, M., Pijnenborg, A.C., Wijnholds, J., van Helvoort, A., de Jong, M.C., Hooijberg, J.H., Mol, C.A. et al. (