Recent epidemiological studies have identified plasma triglyceride as a risk factor for atherogenesis. We have generated a mouse transgenic line that carries a recessive mutation designated lpd (lipid defect). Homozygous lpd mice develop as runts and die by age 10-15 days with striking liver pathology characterized by the presence of numerous lipid-containing vacuoles and extensive accumulation of triglycerides. Cloning of the mutant insertion locus and the wild-type lpd locus have revealed a duplication of host genomic sequences at the site of integration. Mapping of the lpd locus with the Jackson Laboratory BSS interspecific backcross panel of (C57BL/6JEi×SPRET/Ei) F1 ×SPRET/Ei placed the lpd locus to the distal part of chromosome 16. These observations suggest that the transgene disrupts a putative gene at the lpd locus and that lpd is a novel locus related to triglyceride metabolism. The lpd mutant mice may serve as models for human disorders of fatty livers or hypertriglyceridemia.
While plasma cholesterol has been well recognized as an important cardiovascular risk factor, it is only recently that plasma triglyceride has been identified and emphasized as a risk factor for atherogenesis (1,2). The sources of plasma triglycerides include dietary triglycerides and endogenous triglycerides synthesized in the liver. Dietary triglycerides are packaged into triglyceride-rich lipoprotein particles called chylomicrons in the intestine. The main protein component of chylomicrons is apolipoprotein-B-48 (ApoB-48), with smaller quantities of apolipoprotein C-peptides, ApoC-I, ApoC-II and ApoC-III. Chylomicrons are secreted into the intestinal lymphatics, enter the blood stream through the thoracic duct, and are then transported to the peripheral tissues where the core triglycerides are hydrolysed by lipoprotein lipase (LPL). ApoC-I and ApoC-II are required for activity of LPL. In the endogenous pathway, triglycerides and cholesterol are synthesized in liver cells and are incorporated into very low density lipoproteins (VLDLs) which are secreted into the circulation. VLDLs are transported to the tissues where triglycerides are hydrolysed by LPL. The resulting VLDL remnants are called intermediate density lipoproteins (IDLs). Some IDLs are cleared directly by the liver through the Apo-E and LPL receptors. The remaining IDLs are hydrolysed by another lipase, the hepatic lipase (HL), and all the triglycerides are hydrolysed and replaced with cholesterol ester. In this process, all the apoproteins, except ApoB-100, are lost and the remnant particles are transformed into cholesterol-ester-rich low density lipoproteins (LDLs), which constitute 60-70% of plasma cholesterol in humans and function in the delivery of cholesterol to tissues for the synthesis of membranes and steroid hormones.
Most of the known genes involved in lipid metabolism have been cloned and mapped, leading to a better understanding of lipid metabolism and its abnormalities (3,4). Transgenic technologies are valuable tools for dissecting the functions of these genes by overexpressing human mini-genes in mice. Several human apoprotein genes, including ApoA-I, ApoC-II and ApoC-III, have been introduced into transgenic mice. ApoA-I comprises 70% of high density lipoproteins (HDLs). In transgenic mice that overexpressed human ApoA-I, increased plasma levels of HDL-C and cholesterol were observed (5). In humans, the genetic defect of ApoC-II, the cofactor of LPL, results in hypertriglyceridemia. However, overexpression of human ApoC-II in transgenic mice also resulted in hypertriglyceridemia, suggesting a more complex role for ApoC-II in the metabolism of plasma triglycerides (6). Transgenic mice that carried human ApoC-III were generated and developed hypertriglyceridemia. Further analysis suggested that the ApoC-III mice had an impaired clearance of ApoB-48 remnants (7).
In addition to functional studies of lipid-related genes in mice that overexpress these transgenes, it is also possible to generate mice with mutations in novel genes related to lipid metabolism as the result of the random insertion of foreign DNA into a gene. Microinjection of DNA into fertilized eggs of mice yields such insertional mutations at a frequency of ~7% of the transgenic lines (8). The phenotypes associated with these mutations include embryonic lethality at various stages of development (9-11), limb deformities (12,13), germ line deficiencies (14,15), and neurological defects (16-18). The advantage of analyzing a mouse insertional mutation is that the insertional locus is genetically tagged by the transgene which provides a unique marker to clone the locus and to identify the affected gene. We describe here the generation and characterization of a mouse transgenic line Z-14 which carries an insertional mutation designated lpd (lipid defect). Phenotypic and genetic analyses reveal that lpd is a novel locus on mouse chromosome 16 related to triglyceride metabolism.
We have previously defined the function of sequences in the [gamma]F-crystallin promoter by directing the expression of the Escherichia coli lacZ gene to the mouse lens in transgenic mice. Most of the transgenic lines established expressed the lacZ transgene and showed a narrow or wide staining pattern in the lens nucleus (19). Surprisingly, the progeny from one transgenic line, Z-14, when bred to homozygosity, showed runted phenotype (Fig. 1a), and the mice died ~10 days after birth. The lack of lacZ blue staining indicated that the transgene was not expressed in the lens of the Z-14 transgenic mice. When lpd/+ male mice were mated with lpd/+ females, litters with the expected number of offspring were born. The genotypes of these litters were typed by Southern blot analysis with a transgene probe that included the [gamma]F-crystallin promoter. The DNAs which only showed the endogenous [gamma]F-crystallin band were typed as wild-type. The mice whose DNAs contained both endogenous and exogenous [gamma]F-crystallin bands were typed as hemizygous or homozygous animals which could be further distinguished by the intensity of transgene band (data not shown). From (lpd/+) × (lpd/+) matings, 121 pups were genotypically and phenotypically typed; the numbers of wild-type (+/+), heterozygotes (lpd/+) and homozygotes (lpd/lpd) were scored as 29, 57 and 26 (1:1.9:0.9), implying that the lpd is a recessive mutation. Since the mutant phenotype was tightly linked to the lacZ transgene, and the expression of the lacZ gene in other transgenic lines did not give rise to the mutant phenotype (19), we assume that the insertion of the transgene caused the mutation, disrupting a putative gene in the genome.
At birth, the homozygous mutant lpd mice were indistinguishable from their littermates. About 3 days later, the homozygous animals showed obviously retarded growth. Fur growth was also retarded (Fig. 1a). At ~5-10 days after birth, the homozygotes usually developed a generalized tremor and unsteady gait. Almost all of the mutant runts died between 10 and 15 days of age. After systematically examining the histology of the main organs, we observed striking liver pathology in homozygous mutant mice. The livers of the homozygotes were highly vacuolated (Fig. 1c), and the vacuoles could be stained by Oil-Red-O, which is specific for fat (data not shown). In severe cases, only ~20% normal hepatic structure was left. Further biochemical analysis of neutral lipids extracted from the livers of the homozygous animals demonstrated an elevated level of triglycerides, whereas cholesterol levels appeared to be normal (Fig. 2). Consistent with the liver pathology, the homozygous mice showed a 2- to 3-fold increase of plasma triglycerides, while the plasma cholesterol levels appeared to be normal.
When introduced by microinjection, exogenous DNA can form tandem repeats which can complicate the cloning of flanking sequences (20-22). Therefore, we first constructed a physical map of the lpd locus. Restriction mapping of the lpd insertion locus by Southern blot analysis with probes derived from different regions of the injected DNA construct revealed that the transgene integrant spanned ~5 kb and corresponded to a partial tandem duplication of the injected 3.9 kb [gamma]F-lacZ-SV40 construct (data not shown).
To map the lpd locus, we next analyzed the segregation of restriction fragment length polymorphisms (RFLPs) in the Jackson Laboratory BSS panel of 94 N2 animals from the cross (C57BL/6JEi × SPRET/Ei) F1 × SPRET/Ei (Jackson BBS) (23). Probes derived from each flanking side of the transgene were tested for informative RFLPs. Probe P3 derived from the 5'-flanking genomic clone 3A (Fig. 3) detected a 6.8 kb band and a 4.8 kb band on NcoI digests of C57BL/6JEi and SPRET/Ei DNA, respectively. The H06 probe, derived from the 3'-flanking side, detected a RFLP of 6.7 and 5.9 kb on NcoI digests of C57BL/6JEi and SPRET/Ei DNA, respectively. These two probes were then used in a Southern blot analysis of a DNA panel from 94 N2 progeny of the interspecific backcross (data not shown). The allele segregation patterns obtained from this panel were entered into the mapping database of the Jackson Laboratory, and linkage was analyzed by comparison with the ~2500 other loci previously typed in this cross.
Analysis of the RFLP distribution patterns shows that both the P3 and H06 probes are closely linked to a number of genetic loci on the distal part of chromosome 16, including Htr1f, Pmv14 (polytropic murine virus 14) and Pmv16 (24) (Fig. 4). Among the 94 N2 progenies, 93 showed the same RFLP distribution pattern on Southern blots probed with P3 and H06. However, a recombination event between P3 and H06 sequences was scored on N2 progeny #15. Based on the frequency of genetic recombination, the calculated genetic distance between these two probes was 1.06 cM (Fig. 4). However, the H06 probe mapped at the same chromosomal location as the D16Bir11 and Pmv14 loci with no recombination being detected. The P3 probe mapped to the same position as D16Mit141. Apod, another gene involved in lipid transportation and metabolism, maps several centimorgans proximal to the lpd locus in this same cross thus ruling out identity between these two loci. The inferred order of the nearby markers and calculated genetic distances is: proximal-D16Mit5-(1.06 cM ± 1.06 SE)-Htr1f, D16Bir7-(2.13 cM ± 1.49 SE)-D16Bir10-(4.25 cM ± 2.08 SE)-P3 5'-probe, D16Mit141-(1.06 cM ± 1.06 SE)-H06 3'-probe, Pmv14, D16Bir11-(1.06 cM ± 1.06 SE)-Gabpa-(1.06 cM ± 1.06 SE)-D16Mit117-(3.19 cM ± 1.81 SE)-D16Mit6-distal (Fig. 4).
To determine the structure of the lpd locus in the transgenic Z-14 mouse strain, we undertook a Southern blot analysis of genomic DNA from these mice using probes flanking the insertion site. First, probes directly flanking the transgene were used in Southern blot analyses of wild-type, hemizygous and homozygous tail DNAs. The 5'-flanking P01 probe detected a 7 kb BamHI fragment in wild-type CD1 mouse DNA (Fig. 5A, lane 1). In hemizygous (lpd/+) mouse DNA, the P01 probe detected a 7 kb BamHI fragment from wild-type DNA as well as the 8.9 kb band corresponding to the lpd insertion locus. In homozygous (lpd/lpd) mutant DNA, it was expected that the P01 would detect only the 8.9 kb band representing the post-integration locus. However, P01 detected the 7 kb `wild-type' band as well as the 8.9 kb band (Fig. 5A, lane 3). Similar results were observed following BglII digestion, in that both `wild-type' and transgene-specific bands were detected in homozygous mutant DNA (Fig. 5A). In the 3'-flanking side, H06 detected a 19 kb BamHI fragment in wild-type mouse DNA. In homozygous mutant DNA, H06 detected the 8.9 kb fragment, but also the 19 kb `wild-type' band (Fig. 5B). These extra `wild-type' sequences in the mutated lpd locus could only be explained as the result of gene rearrangement. Secondly, we have shown that the two flanking wild-type genomic clones 3A and D3 did not overlap (Fig. 3). Instead, they were mapped 1.1 cM apart by the genetic analysis of an interspecific backcross panel, implying that these sequences are well separated in the mouse genome. The data suggest that the lpd locus underwent either a large deletion or insertion of host genomic sequences during transgene integration. To test whether the lpd locus carried a deletion, probes P2, P3 and RX (Fig. 3) were used in a Southern blot analysis of homozygous mutant DNA. If these sequences were deleted in the lpd mutant locus, the P2, P3 and RX probes would not detect signals in DNAs from the lpd homozygotes. However, all the P2, P3 and RX probes detect bands in Southern blots with the homozygous mutant DNAs (data not shown), suggesting that at least at this level of resolution, the genomic sequences adjacent to the insertional site were not deleted.
One model to explain the data described above involves DNA duplication at the lpd insertion site (Fig. 6). This model suggests that the transgene integrant is flanked by the duplicated genomic sequences in the same orientation. The genomic clone 3T contains the transgene as well as the J2 and J3 junctions. The wild-type genomic clones 3A and D3 contain the junctions J1 and J4, respectively. This model explains why the P01 and H06 probes detected wild-type bands in the lpd homozygous mutant DNAs because the J1 and J4 sequences are present in the lpd mutant locus, due to the duplication (Fig. 6). This model further proposes that the P2 and RX probes, derived from the 3A and D3 clones, respectively, are not included in the region of duplication, whereas the P4 and H06 probes are duplicated in the lpd locus (Fig. 6b).
We have described a recessive mouse mutation, lpd, generated after microinjection of DNA into the pronucleus of CD1 zygotes. Homozygous mice showed retarded growth and developed fatty livers and hypertriglyceridemia. The phenotype and molecular analyses revealed that the transgene has disrupted a gene on chromosome 16 which is involved in triglyceride metabolism. As the homozygous mutant lpd mice were born at the expected frequency and appeared phenotypically normal at birth, the lpd locus does not appear to be necessary for embryogenesis. These data suggest that this locus is required soon after birth, and that the loss of its function gives rise to triglyceride accumulation. The accumulation of triglycerides in the liver is first observed inside the hepatic cells as small fat droplets. As the disorder progresses, the droplets grow and fuse together to form large vacuoles, and finally the hepatic cells lyse. In some mutant mice, the liver impairment was so severe that only ~20% of the hepatic architecture was left, implying that the death of the mutant mice might be due to liver failure.
In mice, two naturally occurring mutations associated with defects in triglyceride metabolism have been described. Combined lipase deficiency (cld) is a recessive mutation which results in deficiency of activities of both lipoprotein lipase and hepatic triglyceride lipase (26). Mice homozygous for the cld locus develop lethal hyperchylomicronemia within 2 days after birth as a consequence of nursing. Plasma triglyceride values in affected mice are up to 100 times higher than the normal littermates. The cld locus maps to chromosome 17 and thus appears genetically distinct from the LPL and HL loci. Though the lpd mutant mice also developed triglyceridemia, the plasma triglyceride level showed only about a 2- to 3-fold increase above the levels of their normal littermates. The major impairment in the lpd mutant mice was in the liver, and the homozygotes live longer than the cld mice.
The mouse fatty liver dystrophy (fld) mutation is another recessive mutation which was identified in neonatal mice by their enlarged and fatty liver (27). Like lpd, the fld mutation is an autosomal recessive mutation characterized by neonatal hypertriglyceridemia and a fatty liver. The growth of homozygous fld mutants is retarded and they develop an unsteady gait. However, the hypertriglyceridemia in the fld mice occurs only in the suckling period. Many fld homozygotes die between the 19th and 35th postnatal days. Surviving females are usually fertile. The histological examination of the livers from 12 day old homozygous fld mutant mice reveals lipid droplets within the hepatocytes, but this lipid accumulation is transient and does not affect the hepatic architecture. The fld locus has recently been mapped to mouse chromosome 12 (28).
Almost all known genes involved in triglyceride metabolism have been mapped in humans, and most of them have been mapped in mouse (3,4). The genes for ApoE, ApoC-I, and ApoC-II are tightly clustered on human chromosome 19, and the genes for ApoA-I, Apo-IV and ApoC-III are clustered together on human chromosome 11. The genes for ApoB and ApoD are located on human chromosomes 2 and 3, and on mouse chromosomes 12 and 16, respectively. Genes for LPL, HL and LIPA are related by sequence, but they mapped to separate mouse chromosomes 8, 9 and 19. The interspecific backcross mapping placed the lpd locus to the distal part of mouse chromosome 16 near the Pmv14 locus, the region with syntenic homologue to human chromosome locations 21q21-q22 and 3p11 (Fig. 4). On human chromosomes 21 and 3, the only gene involved in lipid metabolism is APOD. This gene has been mapped to human chromosome 3q27-qter, and mapped to the proximal to middle region of mouse chromosome 16. Based on its distinct chromosomal location, we conclude that the lpd locus is a novel locus different from all of the previously identified genes involved in lipid metabolism.
A few mutant loci have been described on mouse chromosome 16. ckr, chakragati, is an insertional mutation with neuronal defects. The ckr locus maps to the middle of chromosome 16 (29). The other two mutant loci, scid (severe combined immunodeficiency) and md (mahoganoid), map near the proximal end of chromosome 16 (30). Based on their chromosomal location and phenotype, these mutations are not likely be allelic to lpd. The mutant that maps near the lpd locus is wv (weaver) mutation (31). The wv mutation arose spontaneously in the C57BL/6JEi strain. Homozygotes are recognizable in the second postnatal week by their small size, instability of gait, weakness and hypotonia. Many die at weaning age, but some survive to adulthood and females may breed. The cerebellum is small and almost devoid of granule cells, which degenerate during the second week. The lpd mutant mice showed a phenotype similar to the wv mice in their small body size, instability of gait and the postnatal lethality. However, the lpd mutant carried severe liver pathology and a triglyceride deficiency that were not detected in the wv mutation. Therefore, the lpd locus appears distinct from the wv locus. Since the lpd insertional locus underwent a large gene rearrangement and showed some sign of neurological defects (instability of gait), mating between the wv and lpd mice should determine whether the wv locus is also affected by the gene rearrangement.
We proposed a model of DNA duplication (Fig. 6) in the lpd locus similar to the MyK-103 locus (15). This model was supported by quantitative Southern blot analysis with genomic probes derived from different regions of the lpd locus. However the duplication in the lpd locus appears to be much larger than in the Myk-103 locus. Based on the cloning analysis of the lpd wild-type locus, the size of the duplication seems to be >26 kb and it was shown to be 1.06 cM based on genetic analysis of a mouse interspecific backcross of (C57BL/6JEi × SPRET/Ei) F1 × SPRET/Ei. However, this map distance is calculated on a single crossover in 94 informative genetic typings.
If the mutant lpd locus underwent a DNA duplication of >100 kb, how such a duplication could result in a recessive mutation may be complicated. The phenotype may result from the increased gene dosage from the duplication. However, since deletions are frequently detected from the previously characterized transgene-mediated insertional mutations, the duplication/insertion may also be coupled with microdeletions. Either the DNA junctions from transgene integration or the junctions coupled with some undetected microdeletion could also disrupt the normal gene function. Since the P3 sequences derived from the 5'-flanking junction of the transgene is evolutionarily conserved and could be within a candidate gene, we suggest that disruption of the lpd locus is responsible for the phenotype. Because the lpd mutation involved an interesting and distinct biochemical defect, identification of the affected gene should yield important information about triglyceride metabolism. It is hoped that the lpd mice may serve as models for human disorders of fatty livers or hypertriglyceridemia.
The DNA used for microinjection was a linear 3.9 kb DNA fragment containing E.coli lacZ coding sequences driven by the [gamma]F-crystallin promoter sequences from -120 to +45, and a 200 bp polyadenylation sequence from SV40. Transgenic mice were generated by microinjection (32) of the 3.9 kb DNA fragment into the pronucleus of fertilized eggs from the CD1 mouse strain. Transgenic hemizygous and homozygous animals were identified by Southern blot analysis. Southern blots containing 5 µg of digested tail DNA were hybridized with a probe derived from the [gamma]F-crystallin promoter. The DNAs from wild-type mice show only the endogenous [gamma]F-crystallin band, whereas the DNAs from the transgenic mice show both the endogenous and the exogenous insertional [gamma]F-crystallin bands. By using the endogenous [gamma]F-crystallin band as a internal control, the intensity of the transgenic insertional [gamma]F-crystallin band could be used to distinguish the homozygous and hemizygous animals.
Tissues from mutant mice were perfused with a fixative solution of 4% (wt/vol) formaldehyde, 5%(wt/vol) acetic acid. Paraffin-embedded tissues were sectioned at 6-8 µm and stained with hematoxylin and eosin. For fat staining, frozen sections were performed and stained with Oil-Red-O. Total lipids were extracted from mouse liver (33) and fractionated into neutral and polar lipids (34). Neutral lipids were further analyzed by thin layer chromatography employing a diethyl ether/acetic acid system (35). Lipids were visualized with 50% sulphuric acid (36). Serum triglyceride levels were determined with the Koda dry-chemistry system. Sera were obtained from littermates of ~10 days of age.
The lpd mutant genomic library was prepared with the [lambda]-dash II/BamHI cloning kit from Stratagene. In brief, the homozygous lpd mutant DNA was first digested with BamHI and size-fractionated on a sucrose gradient. The 8-12 kb fraction was ligated into the [lambda]-dash vector and packaged into recombinant phage. The resulting library was plated on a lacZ- E.coli strain P2-PLK-17 and screened with the lacZ transgene probe. The wild-type genomic [lambda]-dash library was constructed from a partial Sau3A digest of CD1 genomic DNA. This library was kindly provided by J. Rossant's laboratory (Samuel Lunenfeld Research Institute, Toronto).
Standard recombinant DNA techniques, including DNA preparation, and enzymatic manipulations, bacterial transformation, blotting, and hybridization, were performed according to Sambrook et al. (37). All cloned phage inserts were subcloned into the plasmids pGEM3 or pGEM5Z (Promega) prior to analysis. The chromosomal location of the lpd locus was determined by analysis of the Jackson Laboratory BSS panel. This panel has been typed for >2000 genetic loci that are well distributed among all of the chromosomes in the mouse genome (23). Probes were first used in a Southern blot analysis with the C57BL/6JEi and SPRET/Ei genomic DNAs digested with 10 different restriction enzymes to search for informative RFLPs. The informative probes were then used for a Southern blot analysis of the Jackson BSS N2 backcross DNA panel. The distribution pattern of RFLPs in the backcross panel was entered into the database in the Jackson laboratory.
Apo, apoprotein; cld, combined lipase deficiency; ckr, chakragati; dt, destonia musculorum; fld, fatty liver dystrophy; HDL, high density lipoprotein; HE, heterozygotes; HL, hepatic lipase; lpd, lipid defect; HO, homozygotes; LPL, lipoprotein lipase; LDL, low density lipoprotein; lgl, legless; md, mahoganoid; Pmv14, polytropic murine virus 14; RFLP, restriction fragment length polymorphism; scid, severe combined immunodeficiency; VLDL, very low density lipoprotein; WT, wild-type; wv, weaver.
We thank M. Buchwald, P. Sadowski, and A. Bernstein for advice and support during the course of this work. Thanks to L. Rowe for supplying the Jackson Laboratory BSS DNA panel and data analysis. Also thanks to A. Bernstein, C.C. Hui and J. Clark for reading the manuscript. This work was supported by a grant from the Medical Council (MRC) of Canada to M.L.B. M.L.B. was an MRC scientist.
Human Molecular Genetics
Pages
Introduction
Results
Z-14 transgenic line carries a recessive, insertional mutation
Z-14 mutant mice demonstrate striking liver pathology and develop hypertriglyceridemia
Molecular cloning and analysis of the lpd locus
The lpd locus maps to the distal part of chromosome 16
Duplication of the lpd locus
Discussion
Materials And Methods
Generation and identification of transgenic mice
Tissue histology and lipid analysis
Preparation and screening of the genomic library
Recombinant DNA technology and interspecific backcross mapping
Abbreviations
Acknowledgements
References
REFERENCES
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 14 Mar 1998
Copyright© Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  H.-J. Pan, P. Reifsnyder, D. E. Vance, Q. Xiao, and E. H. Leiter Pharmacogenetic Analysis of Rosiglitazone-Induced Hepatosteatosis in New Mouse Models of Type 2 Diabetes Diabetes, June 1, 2005; 54(6): 1854 - 1862. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 X.-Y. Wen, R. A. Hegele, J. Wang, D. Y. Wang, J. Cheung, M. Wilson, M. Yahyapour, Y. Bai, L. Zhuang, J. Skaug, et al. Identification of a novel lipase gene mutated in lpd mice with hypertriglyceridemia and associated with dyslipidemia in humans Hum. Mol. Genet., May 15, 2003; 12(10): 1131 - 1143. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||