Atherosclerosis and Lipoproteins |
From the Genome Sciences Department (N.B., J.A., J.C., V.A., E.M.R., L.A.P.), Lawrence Berkeley National Laboratory, Berkeley, Calif; the Department of Atherosclerosis (E.B., J.-C.F., J.F.-N.), Pasteur Institute, Lille, France; United States Department of Energy Joint Genome Institute (E.M.R., L.A.P.), Walnut Creek, Calif.
Correspondence to Dr Len A. Pennacchio, Department of Genome Sciences, MS 84-171, One Cyclotron Road, Lawrence Berkeley National Laboratory, Berkeley, CA 94720. E-mail LAPennacchio{at}lbl.gov
| Abstract |
|---|
|
|
|---|
Methods and Results To address this issue, we generated independent lines of mice that either overexpressed ("double transgenic") or completely lacked ("double knockout") both apolipoprotein genes. We report that both "double transgenic" and "double knockout" mice display normal triglyceride concentrations compared with overexpression or deletion of either gene alone. Furthermore, we find that human ApoAV plasma protein levels in the "double transgenic" mice are
500-fold lower than human ApoCIII levels, supporting ApoAV as a potent triglyceride modulator despite its low concentration.
Conclusions Together, these data support that APOA5 and APOC3 independently influence plasma triglyceride concentrations but in an opposing manner.
To address the relationship between the apolipoprotein A5 and C3 genes, we generated independent lines of mice that either overexpressed or completely lacked both genes. We report both lines display normal triglyceride concentrations compared with overexpression or deletion of either gene alone. Together, these data support that APOA5 and APOC3 independently influence plasma triglyceride concentrations but in an opposing manner.
Key Words: apolipoprotein triglyceride transgenic mice
| Introduction |
|---|
|
|
|---|
In addition to APOA5, the neighboring APOC3 gene has also been reported to have a striking effect on human and mouse plasma triglyceride concentrations. However, APOC3 has an opposite impact on triglycerides, with transgenic mice having increased concentrations (+200% to 2000%) and knockout mice having decreased concentrations (30%).12,13 Furthermore, transgenic mice for the neighboring human genes APOA1/C3/A4 (but not APOA5) also have large increases in triglyceride concentration (+500% to 800%), similar to that found in human APOC3 transgenic mice.14 Finally, genetic studies have supported that common polymorphisms in APOC3 are also associated with triglyceride levels in humans.1520
The inverse effect of APOA5 and APOC3 on plasma triglycerides in vivo raises the question whether they mediate their effect through either a shared or an independent mechanism. One hypothesis is that alteration in APOA5 or APOC3 levels in transgenic or knockout mice simply disrupts the plasma protein level of the other apolipoprotein and accounts for the abnormal triglyceride phenotype. For instance, in APOA5 transgenic mice, apociii levels could be reduced by displacement caused by the overexpression of APOA5, and this reduction in apociii is the mechanism behind the lower triglycerides (an apociii-dependent pathway). In contrast, in APOC3 transgenic mice, apoav levels could be reduced because of the overexpression of APOC3 and this reduction in apoav is in fact the mechanism behind the lower triglycerides (an apoav-dependent pathway). Finally, apoav and apociii could function independently to modulate triglyceride concentrations in an opposing fashion.
To begin to explore this issue, we pursued a genetic approach and generated 2 independent lines of mice with simultaneous alterations in both ApoAV and ApoCIII levels. The first line of mice was engineered to completely lack both genes ("double knockouts"), whereas the second line was created to overexpress both genes ("double transgenics"). We hypothesized that if ApoAV and ApoCIII affect a common pathway in an interdependent fashion, then the resulting phenotype in the "double knockouts" should be similar to that in either of the single knockouts alone (+400% or 30% triglyceride concentrations). In contrast, if apoav and apociii function independently, then the loss of both genes should result in intermediate triglyceride concentrations (somewhere between that found in either of the single knockouts alone). A similar logic was applied to the usefulness of analyzing "double APOA5/APOC3 transgenic" mice.
| Methods |
|---|
|
|
|---|
Gene Expression Analysis
Animals were euthanized and tissues harvested for total RNA isolation using the Trizol RNA reagent (Gibco-BRL, Gaithersburg, Md). Approximately 24 µg of total RNA was separated in 1.0% agarose by gel electrophoresis and the RNA was transferred to a charged nylon membrane (Ambion, Austin, Tex). The RNA blots were hybridized with [
-32P]dCTP random-primed apoa5 or apoc3 probes in ULTRAhyb buffer (Ambion). The primers used to generate probe templates can be found online (http://atvbaha.org). Filters were washed in 2x saline sodium citrate at room temperature for 20 minutes and in 0.1x saline sodium citrate (SSC) at 42°C for 20 minutes, followed by autoradiography visualization.
Plasma Apolipoprotein and Lipid Analysis
All animals were fed standard chow diet and were analyzed between 8 and 12 weeks of age. Blood samples were collected after a 5-hour fast by retro-orbital bleeding using heparinized microhematocrit tubes. Total cholesterol and triglyceride concentrations were measured using enzymatic methods on a Gilford System 3500 analyzer (Gilford Instruments, Oberlin, Ohio).29 Plasma levels of human apo C-III were measured by kinetic immunonephelemetric system (Immage; Beckman Coulter) using polyclonal antibodies produced with total synthetic apoCIII in goats. Plasma levels of human apoAV were quantitated by sandwich enzyme-linked immunosorbent assay (ELISA) using polyclonal antibodies generated against a human apoAV synthetic peptide. The variation coefficients were <6% for both methods, including within-day and between-day variation. Plasma lipoproteins from pooled mouse plasma were separated by gel filtration chromatography using a Superose 6HR 10/30 column; Pharmacia LKB Biotechnology). The gel was equilibrated with phosphate-buffered saline (10 mmol/L) containing 0.1 g/L sodium azide and plasma were eluted with the buffer at room temperature at a flow rate of 0.2 mL/min. Elution profiles were monitored at 280 nm and recorded with an analog-recorder chart tracing system (Pharmacia LKB biotechnology). The elution fraction numbers (0.24 mL for each) of the plasma lipoproteins separated by FPLC were very-low-density lipoprotein, 10 to 18; intermediate-density lipoprotein/low-density lipoprotein, 20 to 30; and high-density lipoprotein, 30 to 40. Lipids and apolipoproteins in the recovered fractions were assayed as described.
| Results |
|---|
|
|
|---|
35 kb of each other on mouse chromosome 9 and are thus "genetically linked," it was not feasible to directly intercross each of the previously generated single knockout mice to create double knockout animals (Figure 1). Rather, we chose to sequentially target each apolipoprotein gene in embryonic stem cells through a double-selectable marker scheme. We first disrupted the apoa5 gene with a hygromycin-based targeting vector and subsequently deleted the apoc3 gene with a neomycin-based targeting vector. We used properly targeted embryonic stem cell clones to generate chimeric mice and subsequently intercrossed heterozygous double-targeted animals to generate double knockout mice. We obtained viable offspring at the expected Mendelian genotype ratio of 1:2:1 (15:23:15), and no obvious differences were noted in their outward physical appearance. To confirm proper targeting of the 2 genes, we determined the RNA levels for apoa5 and apoc3 from liver tissue (Figure 2A). Northern analysis revealed that double knockout mice completely lacked detectable apoa5 and apoc3 transcripts in comparisons to wild-type controls, supporting the proper targeting of the 2 genes.
|
|
APOA5/APOC3 Double Knockouts Display Normal Plasma Triglyceride Concentrations
We next examined plasma lipid levels in these genetically engineered mice. For both triglycerides and cholesterol, we found no significant differences between double knockouts and wild-type littermates. Plasma triglyceride concentrations were 53±20 mg/dL and 44±14 mg/dL for wild-type controls and double knockouts, respectively (Figure 2B) (Student t test P=0.4). Furthermore, no significant difference was found in cholesterol concentrations (Figure 2B). Finally, we characterized plasma lipoprotein particles by fast protein liquid chromatography (FPLC) to determine whether there was dramatic repartitioning of lipoprotein particles despite the fact that there were no significant differences in plasma lipid concentrations. Overall, we found the FPLC profiles quite similar between the controls and "double knockouts," with no obvious differences in the various plasma particle sizes (Figure 2C).
APOA5/APOC3 Double Transgenics Display Normal Plasma Triglyceride Concentrations
To further examine the relationship between apoAV and ApoCIII, we overexpressed human genes in transgenic mice and examined their resulting plasma triglyceride concentrations. Our first model was obtained by intercrossing previously established human APOA5 and APOC3 transgenic mice.2,13 We found that these double transgenic mice had normal plasma triglyceride and cholesterol concentrations compared with controls (data not shown).
One potential complicating factor in our analysis of the described double transgenic mice was the variable copy number and genome integration sites for each of the independently generated APOA5 and APOC3 transgenic lines. To reduce the possibility of artifact caused by this scenario, we generated a second type of double transgenic mice using a different experimental design. Specifically, we selected a human bacterial artificial chromosome (BAC) containing both the APOA5 and APOC3 genes, as well as 2 interspersed apolipoprotein genes (APOA1 and APOA4) (Figure 1). Our rationale for this approach was to control for identical APOA5 and APOC3 copy number and integration events that would not be possible through the generation of independent APOA5 and APOC3 transgenes. In addition, whereas all the genes in this cluster have been well-studied in vivo, only the APOA5 and APOC3 genes affect triglyceride concentrations, thus overexpression of APOA1 or APOA4 was unlikely to confound our interpretation of triglyceride concentrations in these mice. Finally, this strategy was expected to better-mimic the physiological expression of APOA5 and APOC3 because several regulatory elements are embedded within the larger gene cluster (Figure 1).
We generated 2 independent founder lines of human BAC transgenic mice and determined that all 4 genes (APOA5, APOC3, APOA4, and APOA1) expressed in a pattern consistent with their endogenous levels in liver and intestines as determined by Northern blot analysis (Figure 3A). Predominant expression was found in liver tissue for APOA5, APOC3, and APOA1, and in the intestines for APOA4. Examination of plasma lipid concentrations revealed no differences in triglycerides between the 2 transgenic lines and controls (Figure 3B). For 1 of 2 double transgenic lines, a slight increase in cholesterol concentrations was noted (Student t test P=0.016) (Figure 3B). This effect might be explained by the APOA1 or APOA4 transgenes in this model because it has been well-established that overexpression of either gene increases cholesterol concentrations and protects against atherosclerosis.5,21 FPLC analysis indicated the lack of repartitioning of lipoprotein particles between transgenics and controls, with subtle differences found in the lipid concentrations from the peak elution volume (Figure 3C).
|
Vast Differences in ApoAV and ApoCIII Plasma Protein Concentrations in Double Transgenic Mice
To determine the relative plasma levels of human apoAV and apoCIII proteins in the double transgenic mice, we performed immunoassays. Because our BAC double transgenics have identical copy number and integration sites, we chose to examine these animals for plasma protein levels. We found that ApoAV and ApoCIII levels were
500-fold different in both founder lines, with double transgenics having
0.027 mg/dL of ApoAV and
13 mg/dL of ApoCIII.
| Discussion |
|---|
|
|
|---|
Our analyses of the various engineered animals revealed they all have normal triglyceride concentrations compared with either single gene transgenic or knockout. This effect is most striking in comparison of wild-type controls to the previously published APOA1/C3/A4 transgenics14 versus our APOA1/C3/A4/A5 transgenics (+500% to 800% versus no change in triglyceride levels). These data indicate that the addition of APOA5 alone to APOA1/C3/A4 transgenic mice can dramatically reduce triglycerides to within the normal physiological range (but not to the extreme level found in single APOA5 transgenics that have 70% triglycerides). This study supports the hypothesis that ApoAV and ApoCIII function on plasma triglycerides in an independent but opposing manner.
Although these data support that ApoAV and ApoCIII function independently, we cannot completely rule out that there is some overlap in their functional relationship. For instance, in our analysis of apoa5/apoc3 double knockout mice, triglycerides levels are within the physiological range; however, these levels are not perfectly intermediate to apoa5 knockouts (+400%) and apoc3 knockouts (30%). These data raise the possibility that part of the phenotype found in apoa5 knockouts could be caused by displacement or knockout of apoCIII function.
An argument against significant overlap between ApoAV and ApoCIII function is provided from single transgenic data. The fact that APOA5 transgenic mice have lower triglycerides (70% reduced) than the previously described apoC3 knockout mice (30%) indicate that changes in apoCIII alone cannot explain the entire effect of apoAV. If this were true, then APOA5 transgenics should only reach the triglyceride levels reached in mice completely lacking ApoCIII. The contrary is also true, if the apoCIII transgenic effect (+200% to 2000% triglycerides) were solely through the displacement and abolition of ApoAV function, the lowest triglyceride levels could reach would be to that found in the apoa5 knockouts (+400%). These observations further support that ApoAV and apoCIII function independently in triglyceride metabolism.
In humans and mice, APOC3 is expressed predominantly in the liver and intestines whereas APOA5 expression is limited to the liver. The ApoCIII and ApoAV proteins are both found in plasma with ApoCIII localized primarily to chylomicrons and VLDL particles and ApoAV mostly to high-density lipoprotein particles. Although it remains unclear how the newly identified ApoAV protein affects plasma triglycerides, the mechanism of action of ApoCIII has been reported. Lipid metabolism studies in APOC3 transgenic and knockout lines of mice indicate that these triglyceride changes are caused by alterations in triglyceride catabolism by lipoprotein lipase and by chylomicron remnant clearance from the liver.12,13,22
Although the mechanism by which ApoAV acts on triglycerides remains to be determined, our study reveals that ApoAV is a potent triglyceride regulator relative to ApoCIII. The finding of a
500-fold difference in ApoAV and ApoCIII plasma levels in double transgenic mice, and yet triglyceride levels are normal compared with either transgenic alone suggests significant activity of small amounts of ApoAV. The fact that both APOA5 and APOC3 are contained in a 1:1 ratio as a component of the single transgenic BAC fragment indicates that these protein level differences are not caused by differences in copy number or position effects. In addition, Northern blot analysis in wild-type as well as transgenic mice (Figures 2A and 3A
) show only small differences in the expression of apoa5 and apoc3 in the liver. Thus, this difference is caused by posttranscriptional events, such as the rate of liver secretion of each protein. It is also worth noting that the level of each plasma protein in these transgenic mice is similar to that found in humans.2325 Previous data indicate that human ApoCIII levels range from 10 to 50 mg/dL, which is similar to the 15 mg/dL found in our transgenic mice.2325 Furthermore, we find human ApoAV levels are
0.005 to 0.05 mg/dL (manuscript in preparation), similar to the
0.025 mg/dL found in our APOA1/C3/A4/A5 BAC transgenic mice. Thus, these animals may serve as a useful model for studying these human plasma apolipoproteins in their physiological concentrations within an in vivo model. Subsequent interbreeding of these transgenics to double apoa5/apoc3 knockouts could also provide a "humanized mouse" with respect to these 2 apolipoprotein genes and would provide a system to study human gene regulation such as in response to pharmacological agents.
An important clinical implication of our findings of the independent activity of ApoAV and ApoCIII relates to human genetic association studies between APOA5 and APOC3 polymorphisms and triglyceride concentrations. For both genes, numerous common variants have been identified that are strongly associated with triglyceride concentrations.2,811,1517,19,20,2628 However, the close neighboring relationship of these 2 genes, separated by only
35 kb, have precluded the determination of whether both genes have functional variants or whether polymorphisms in one of the genes are in significant linkage disequilibrium with functional polymorphisms in the other gene. Although either is a possibility, the precise molecular mechanism behind this effect could ultimately be caused by a single apolipoprotein (ApoAV or ApoCIII). Our mouse data support that ApoAV and ApoCIII function in an independent manner and suggest that common human variation in both genes can genetically and mechanistically contribute to triglyceride alterations. Future detailed haplotype analysis and association studies with polymorphisms throughout this cluster in a wide-range of populations should reveal the ultimate functional variants in this interval and how they relate to triglyceride metabolism.
| Acknowledgments |
|---|
We thank James Priest for critical feedback and Michael Collier and Ingrid Plajzer-Frick for mouse engineering support. This work was funded in part by the Fondation Leduc, the NIH-NHLBI Programs for Genomic Application Grant HL66681 (E.M.R.) and HL071954A (L.A.P.) through the US Department of Energy under contract number DE-AC03-76SF00098 and an American Heart Association Postdoctoral Fellowship (N.B.).
Received March 11, 2004; accepted April 19, 2004.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Dichlberger, L. A. Cogburn, J. Nimpf, and W. J. Schneider Avian apolipoprotein A-V binds to LDL receptor gene family members J. Lipid Res., July 1, 2007; 48(7): 1451 - 1456. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Qu, G. Perdomo, D. Su, F. M. D'Souza, N. S. Shachter, and H. H. Dong Effects of apoA-V on HDL and VLDL metabolism in APOC3 transgenic mice J. Lipid Res., July 1, 2007; 48(7): 1476 - 1487. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Moreno-Luna, F. Perez-Jimenez, C. Marin, P. Perez-Martinez, P. Gomez, Y. Jimenez-Gomez, J. Delgado-Lista, J. A. Moreno, T. Tanaka, J. M. Ordovas, et al. Two Independent Apolipoprotein A5 Haplotypes Modulate Postprandial Lipoprotein Metabolism in a Healthy Caucasian Population J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2280 - 2285. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. G. Schaap, M. C. Nierman, J. F. P. Berbee, H. Hattori, P. J. Talmud, S. F. C. Vaessen, P. C. N. Rensen, R. A. F. M. Chamuleau, J. A. Kuivenhoven, and A. K. Groen Evidence for a complex relationship between apoA-V and apoC-III in patients with severe hypertriglyceridemia J. Lipid Res., October 1, 2006; 47(10): 2333 - 2339. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. E. Alborn, M. G. Johnson, M. J. Prince, and R. J. Konrad Definitive N-terminal protein sequence and further characterization of the novel apolipoprotein a5 in human serum. Clin. Chem., March 1, 2006; 52(3): 514 - 517. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Arbones-Mainar, M. A. Navarro, S. Acin, M. A. Guzman, C. Arnal, J. C. Surra, R. Carnicer, H. M. Roche, and J. Osada Trans-10, cis-12- and cis-9, trans-11-Conjugated Linoleic Acid Isomers Selectively Modify HDL-Apolipoprotein Composition in Apolipoprotein E Knockout Mice J. Nutr., February 1, 2006; 136(2): 353 - 359. [Abstract] [Full Text] [PDF] |
||||
![]() |
U. Hodoglugil, S. Tanyolac, D. W. Williamson, Y. Huang, and R. W. Mahley Apolipoprotein A-V: a potential modulator of plasma triglyceride levels in Turks J. Lipid Res., January 1, 2006; 47(1): 144 - 153. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. A. Ruiz-Narvaez, Y. Yang, Y. Nakanishi, J. Kirchdorfer, and H. Campos APOC3/A5 haplotypes, lipid levels, and risk of myocardial infarction in the Central Valley of Costa Rica J. Lipid Res., December 1, 2005; 46(12): 2605 - 2613. [Abstract] [Full Text] [PDF] |
||||
![]() |
X. Prieur, F. G. Schaap, H. Coste, and J. C. Rodriguez Hepatocyte Nuclear Factor-4{alpha} Regulates the Human Apolipoprotein AV Gene: Identification of a Novel Response Element and Involvement in the Control by Peroxisome Proliferator-Activated Receptor-{gamma} Coactivator-1{alpha}, AMP-Activated Protein Kinase, and Mitogen-Activated Protein Kinase Pathway Mol. Endocrinol., December 1, 2005; 19(12): 3107 - 3125. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. E. Schultze, W. E. Alborn, R. K. Newton, and R. J. Konrad Administration of a PPAR{alpha} agonist increases serum apolipoprotein A-V levels and the apolipoprotein A-V/apolipoprotein C-III ratio J. Lipid Res., August 1, 2005; 46(8): 1591 - 1595. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Merkel, B. Loeffler, M. Kluger, N. Fabig, G. Geppert, L. A. Pennacchio, A. Laatsch, and J. Heeren Apolipoprotein AV Accelerates Plasma Hydrolysis of Triglyceriderich Lipoproteins by Interaction with Proteoglycan-bound Lipoprotein Lipase J. Biol. Chem., June 3, 2005; 280(22): 21553 - 21560. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Genoux, H. Dehondt, A. Helleboid-Chapman, C. Duhem, D. W. Hum, G. Martin, L. A. Pennacchio, B. Staels, J. Fruchart-Najib, and J.-C. Fruchart Transcriptional Regulation of Apolipoprotein A5 Gene Expression by the Nuclear Receptor ROR{alpha} Arterioscler. Thromb. Vasc. Biol., June 1, 2005; 25(6): 1186 - 1192. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Gao, Y. Wei, Y. Huang, D. Liu, G. Liu, M. Wu, L. Wu, Q. Zhang, Z. Zhang, R. Zhang, et al. The Expression of Intact and Mutant Human apoAI/CIII/AIV/AV Gene Cluster in Transgenic Mice J. Biol. Chem., April 1, 2005; 280(13): 12559 - 12566. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Nowak, A. Helleboid-Chapman, H. Jakel, G. Martin, D. Duran-Sandoval, B. Staels, E. M. Rubin, L. A. Pennacchio, M.-R. Taskinen, J. Fruchart-Najib, et al. Insulin-Mediated Down-Regulation of Apolipoprotein A5 Gene Expression through the Phosphatidylinositol 3-Kinase Pathway: Role of Upstream Stimulatory Factor Mol. Cell. Biol., February 15, 2005; 25(4): 1537 - 1548. [Abstract] [Full Text] [PDF] |
||||
![]() |
V. Charlton-Menys and P. N. Durrington Apolipoprotein A5 and Hypertriglyceridemia Clin. Chem., February 1, 2005; 51(2): 295 - 297. [Full Text] [PDF] |
||||
![]() |
C. P. Oliva, L. Pisciotta, G. L. Volti, M. P. Sambataro, A. Cantafora, A. Bellocchio, A. Catapano, P. Tarugi, S. Bertolini, and S. Calandra Inherited Apolipoprotein A-V Deficiency in Severe Hypertriglyceridemia Arterioscler. Thromb. Vasc. Biol., February 1, 2005; 25(2): 411 - 417. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. J. O'Brien, W. E. Alborn, J. H. Sloan, M. Ulmer, A. Boodhoo, M. D. Knierman, A. E. Schultze, and R. J. Konrad The Novel Apolipoprotein A5 Is Present in Human Serum, Is Associated with VLDL, HDL, and Chylomicrons, and Circulates at Very Low Concentrations Compared with Other Apolipoproteins Clin. Chem., February 1, 2005; 51(2): 351 - 359. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.-Q. Lai, S. Demissie, L. A. Cupples, Y. Zhu, X. Adiconis, L. D. Parnell, D. Corella, and J. M. Ordovas Influence of the APOA5 locus on plasma triglyceride, lipoprotein subclasses, and CVD risk in the Framingham Heart Study J. Lipid Res., November 1, 2004; 45(11): 2096 - 2105. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||