Analysis of mouse plasma lipids, lipoproteins and apolipoproteins

Introduction

Lipids and lipoproteins occupy a central role in the development of atherosclerosis. Lipids are more than passive participants that accumulate in fatty streak lesions and atherosclerotic plaques. They exhibit bioactivity that alters the metabolism of several cell types critical to the development of atherosclerosis, such as vascular endothelial cells and macrophages, through a variety of mechanisms as discussed above. The quantitation and compositional analysis of mouse plasma lipoproteins are important, not only to identify those changes in lipoproteins that are known to be associated with increased atherosclerosis, but also to utilize the mouse model to further our understanding of lipoprotein metabolism and the role of lipoproteins in the etiology of atherosclerosis.

A variety of methods have been developed for the isolation and analysis of human plasma lipids, lipoproteins, and apolipoproteins. Many of these classical methods have been successfully adapted for the analysis of lipids in mouse plasma. However, an understanding of important differences between mouse and human lipoproteins and their metabolism is necessary in order to understand some limitations of these methods when applied to the mouse.

This section will first deal with the housing and handling of the mice, factors of importance to accurate plasma lipid determinations, but an aspect that is too often ignored. It will then present those accepted analytical methods that are most amenable to analysis of large numbers of small volume mouse plasma samples. Important differences in plasma lipid/lipoprotein profiles between humans and mice will be discussed, as well as special analytical challenges that will be encountered in analyzing plasma lipids from various genetically altered strains of mice. Finally, the application of more involved procedures to analyze mouse plasma lipoproteins will be briefly discussed, including ultracentrifugation, gel filtration, and electrophoresis.

Housing and handling of the mice

Obtaining accurate plasma lipid data from mice involves proper treatment, handling, and housing of the animals prior to obtaining the blood sample. Stress hormones such as corticosteroids and norepinephrine can dramatically alter plasma lipids, particularly plasma triglycerides and free fatty acids (1). While stress related alterations in plasma lipids are significant for a variety of mammals including man, the mouse is particularly susceptible to stress induced changes in lipid metabolism. This is due in large part to the hyper-excitability of mice and the fact that they are easily stressed (2). No doubt this has occurred as a defense mechanism during evolution, since mice appear on the dinner menu of many predators.

Several studies have demonstrated that stress hormones and associated responses that affect plasma lipids are dramatically altered in rodents by a variety of stressors commonly encountered in their housing and handling. These environmental stress factors include light, noise, temperature, number of mice per cage, location of the cages on the rack, and movement or shipping of the mice (3-17). These factors should be controlled as much as possible, particularly within the critical period prior to obtaining blood samples for analysis. If mice are obtained from an outside source, they should be allowed at least 3 weeks to acclimate to the new facilities before drawing blood for lipid determinations (1,16). The mice and their cages should be handled gently at all times but particularly just prior to and during bleeding. If possible, the animals should not be transported to different rooms to be bled. It should also be pointed out that even maintaining all mice under the same conditions does not guarantee equal amounts of stress, since strain specific differences in response to various stressors have been reported (15,18,19). Despite the best precautions stress effects will occur, however, they can, and should be kept to a minimum.

Fasting

For analysis of plasma lipids in humans, subjects are generally fasted 12-14 hours, usually overnight. This is to insure that no absorption of dietary fats and cholesterol from the gut is occurring at the time of sampling, since plasma lipids would be acutely affected by the size and composition of the last meal consumed rather than representing a “steady state” plasma lipid profile.

Because of similar complications, mice are generally fasted prior to obtaining blood samples for lipid analysis. However, based on a review of the literature there appears to be some controversy as to what the proper length of time to fast a mouse should be. Various studies have fasted anywhere from 5 to 48 hours. While an overnight fast of 14-16 hours has been most commonly used, several studies have chosen fasts of shorter duration, usually 5-7 hours. The argument for the shorter fasting periods in mice is that because of their high metabolic rate an overnight fast results in a “starved”, rather than a fasted mouse. What is not fully appreciated is that in spite of a high metabolic rate, the digestive system of a mouse is comparatively slow.

Mice consume approximately 4 grams of food per day, the majority consumed during the dark cycle, however, appreciable amounts are also consumed at various times during the light cycle (20-23). Although there are definite periods during both the light and dark cycles when most of the food is consumed, individual mice may exhibit considerable daily variation. Also, various strains of mice have been demonstrated to prefer diets with a higher fat content (24-27). Since lipid metabolic studies frequently compare the effects of diets high in fat and cholesterol as well as standard low fat chow diets, a shorter fasting period that may be adequate for a standard chow diet may not be adequate for other diets that the mice prefer and may be consuming in larger quantities and at increased frequencies during the feeding cycle.

We have examined the stomach contents of mice that have been fasted for 5 hours, starting at the beginning of the light cycle. At least 90% of the animals still had food in their stomachs, and the amount of food varied considerably among the animals (data not shown). Our observations are consistent with studies that have examined the rates of gastric emptying in the mouse. In mice that had been fasted for 18-20 hour, then allowed ad libitum access to food for 1 hour, 43, 20, and 10% of the ingested meal remained in the stomach after 2,4, and 7 hours, respectively (28). Based on the customary meaning of “fasting”, an animal that still has food in its stomach cannot be considered as a “fasted” animal. Also, since fats are primarily absorbed in the small intestine and not in the stomach, it is obvious that considerably more than 7 hours is needed before absorbtion of dietary fats is complete. For this reason, an overnight fast would be expected to give more consistent and reproducible plasma lipid values, relatively unaffected by the amount or composition of the last meal ingested, particularly in animals being fed high fat/high cholesterol diets.

Once the duration of fasting is decided upon, the animals should be fasted and bled at the same time each day. As discussed above, since the animals consume most of their food at specific times throughout the dark and light cycles, considerably more food may be consumed in one 8 hour period compared to another. Also, there are diurnal variations in plasma hormones that alter lipid metabolism, as well as diurnal variations in the activity of key enzymes such HMG-CoA reductase, the rate limiting enzyme in cholesterol synthesis.

Lipid Analyses

General Considerations

The most frequently used assays for determination of plasma lipid and cholesterol concentrations are based on chemical reactions that produce colored end products that can be detected spectrophotometrically, with the plasma lipid concentration directly correlating with the optical density determined at a specific wavelength. These spectrophotmetric assays can be conveniently minaturized and the results read in 96 well plates, an important consideration since the amount of plasma that can be obtained from an individual mouse is relatively limited. As with any quantitative assay, standards and blanks must be included with each batch of samples, and for quality control, control samples with known analyte concentrations should be analyzed along with the experimental samples. Ideally, samples should be run in triplicate determinations, but definitely, at least in duplicate determinations. The analyte concentration in the unknown samples should fall on the linear portion of the standard curve, ideally between the lowest and highest standards used to generate the curve. The control sample, which contains a known concentration of the analyte being assayed, is used to confirm that the assay has been done correctly. Control samples with known concentrations of various analytes can be purchased from several companies (Sigma, Beckman, Hoffman-LaRoche).

For analysis of mouse lipids, lipoproteins, and apolipoproteins by the methods described below, we have not observed significant differences between mouse plasma or serum. Also, we have not been able to detect any differences based on the use of EDTA or heparin as the anticoagulant. If EDTA will not interfere with other analyses that the plasma sample may be used for, it is preferred over heparin as an anticoagulant since it also has antioxidant properties. On this note, it is important to verify the effect of any chemicals/drugs, etc. that may be in the plasma sample, in order to ensure that they do not affect the assays. This can be accomplished by analyzing several control samples with and without the “additive” in question, prior to analyzing the unknown experimental samples.

Triglyceride assays

Most colorimetric assays for triglycerides are based on chemical reactions involving the glycerol moiety of the molecule, obtained upon complete hydrolysis of the fatty acids, and yielding free glycerol in equimolar concentrations to the original triglyceride concentration (32). Methods that employ chemical hydrolysis of the fatty acids are non-specific, also hydrolyzing fatty acids from other glycerolipids such as phospholipids. These methods require that, prior to hydrolysis, the triglycerides first be isolated by another method, such as thin layer chromatography. Such isolation steps are laborious and time consuming making the analysis of large numbers of plasma samples impractical for most laboratories. For this reason methods that are based on the enzymatic hydrolysis of triglyceride fatty acids are preferable. These methods utilize lipases that specifically hydrolyze triglyceride fatty acids and, therefore, do not require that the triglyceride fraction be isolated prior to hydrolysis. While the necessary reagents, chemicals and enzymes can be purchased separately from a variety of sources, a kit such as that available from Sigma, which contains all of the necessary components is probably the best choice for most laboratories, since optimization and quality control of the individual components are taken care of by the manufacturer. One important difference between mouse and human plasma samples is in the amount of free glycerol present in the plasma. Generally, the glycerol concentration in human plasma is low (~3-4 mg/dl) and is frequently ignored when analyzing plasma triglycerides. However, while mice generally have significantly lower plasma triglyceride levels than humans, their fasting plasma glycerol concentrations are usually several fold higher. For this reason the accurate determination of triglycerides in mouse plasma requires two assays on each sample. One assay to determine the plasma glycerol concentration prior to hydrolysis of the triglyceride fatty acids and a second determination after the hydrolysis. The actual concentration of “triglyceride” glycerol in the plasma can then be calculated.

Total plasma cholesterol

Cholesterol also may be quantitatively determined by either enzymatic or chemical methods (29). As for the triglyceride analyses described above, direct analysis of plasma cholesterol utilizing enzymatic reagent systems is amenable for rapidly and accurately processing large numbers of plasma samples. Several enzymatic procedures are available for the analysis of cholesterol. The initial reaction steps are common to all (29). First, cholesterol esters are hydrolyzed by cholesterol esterase to produce free cholesterol and fatty acids. The cholesterol is then oxidized using oxygen to produce hydrogen peroxide. The most popular methods then involve quantitation of the hydrogen peroxide by formation of a colored oxidation product or a reduced pyridine nucleotide. Sigma makes a convenient kit for analysis of cholesterol that involves a peroxidase catalyzed reaction of the hydrogen peroxide to produce a quinoneimine dye that absorbs at a wavelength of 510 nm. The analysis of total cholesterol in mouse plasma is as straightforward as it is for human plasma. Proper use of standards and controls, as discussed above, should also be employed for the cholesterol determinations.

HDL cholesterol

Several precipitation methods have been used for the determination of HDL cholesterol (30). These methods involve the selective precipitation of apoB-containing lipoproteins using polyanion solutions which precipitate all of the major lipoprotein fractions except for HDL. After precipitation, an aliquot of the supernatant is taken for cholesterol determination, as described above. The main differences among the methods is in the precipitating agent used. Heparin-manganese chloride, dextran sulfate, phosphotungstate, and polyethylene glycol have all been used as precipitating agents in HDL cholesterol assays. Polyethylene glycol is frequently used but gives the poorest accuracy and precision. Phosphotungstate is the most commonly used precipitating agent, however it is know to underestimate HDL concentrations and is temperature sensitive. Dextran sulfate is a commonly used agent and when higher molecular weight dextrans are used this method produces satisfactory results. Heparin-manganese chloride is the precipitating agent of choice, used in the CDC reference method (30). However, the high concentration of manganese in the sample renders the specimen useless for cholesterol analysis by the use of several commonly used enzymatic procedures, a factor that must be considered.

In using precipitation methods to analyze human HDL cholesterol, there is sometimes difficulty in precipitating lipoproteins in samples with triglyceride concentrations in excess of 400 mg/dl. This problem can frequently be circumvented by diluting the plasma sample prior to precipitation. However, care must taken since plasma samples that are diluted too much will also fail to precipitate.

This problem is much more frequent in analyzing genetically altered strains of mice, particularly on high fat diets, where it is not rare to encounter animals in which the plasma samples fail to precipitate. In such instances, an overestimation of plasma HDL cholesterol concentrations will occur. We have observed that plasmas from apoE knockout mice and LDL receptor knockout mice that have both been maintained on high fat diets exhibit problems with precipitating lipoproteins in some samples. Furthemore, we frequently encounter problems precipitating lipoproteins in two hypertriglyceridemic strains, the apoAII transgenic mice and HcB/Dem19 mice, even when they are maintained on a low fat chow diet. This problem in the mice appears to involve changes in the lipoprotein composition rather than just the hypertriglyceridemia itself, since precipitation sometimes fails to occur even when total plasma triglyceride concentrations are ~200 mg/dl.

For these reasons it would be advisable to check lipoprotein precipitations when analyzing HDL in genetically altered strains of mice for the first time, particularly if they are maintained on high fat diets. This can be accomplished by subjecting the supernatant obtained after precipitation to PAGE followed by western blot analysis of apoB. The presence of apoB in the supernatant would indicate incomplete precipitation. If precipitation is a problem it can be circumvented by using a preparative ultracentrifugation step to remove the triglyceride rich lipoprotein fractions.

Analysis of tissue lipids

In some instances it may be desirable to analyze lipid accumulation in various tissues such as liver, adrenals, etc. There are a variety of well documented techniques for extraction of total lipid from tissues, such as those described by Folch et al. or Bligh and Dyer (31-33). One problem presents itself in utilizing the enzymatic methods described above to analyze lipids extracted from tissue. Lipids are soluble in plasma as part of the lipoprotein particle or, in the case of free fatty acids, bound by albumin. The extracted tissue lipids are soluble in a variety of organic solvents such as chloroform, however, they are not soluble to any appreciable extent in aqueous media. In adding the total lipid extract to our enzymatic reaction mixtures we have observed that while some organic solvents have less of an effect than others, all, even in relatively low concentration, interfere to some extent with the enzymatic steps, as might be expected. To circumvent this problem we dry aliquots of the total lipid extract in chloroform in micro test tubes. We then add the enzymatic reaction mixtures to the test tube and proceed with the assays in the same manner as for the plasma samples. After testing the cholesterol, triglyceride, and fatty acid analyses with serial dilutions of total lipid extract, we have confirmed that we are able to completely hydrolyze and accurately detect all of the lipid fractions that were dried in the test tubes. While this method works well in our hands, it would be advisable to test the method under your conditions, since it is likely that “overloading” the tubes with a large amount of lipid may prevent all of the lipid from reacting during the normal assay incubation periods.

Ultracentrifugation

Mouse lipoproteins can be separated by any of the several standard ultracentrifugation methods described for the analysis of human lipoproteins. However, because of the rotors and larger tube sizes used for analysis of human plasma samples, it will usually be necessary to increase the volume of the mouse plasma sample by the addition of a 1.0063 g/ml density solution prior to beginning the isolation. Alternatively, plasma from several animals can be pooled. While the same density fractions for human lipoproteins (VLDL, <1.0063; IDL,1.063-1.019; LDL,1.019-1.063; HDL, 1.063-1.21 g/ml) are generally used for mice and are usually satisfactory, in some genetically engineered strains, or in mice maintained on high fat/high cholesterol diets, unusual lipoprotein profiles develop that do not allow separation of distinct lipoprotein classes as defined above, and the usual density fractions above may contain lipoproteins of another class.

Human HDL can be separated into several discrete sub-fractions that differ in size and density. As described above, HDL is normally considered to consist of those lipoprotein particles which fall within the density range 1.063-1.21 g/ml. Within this density region in humans are found a variety of subspecies of HDL, each of characteristic particle size and apolipoprotein and lipid composition rather than a continuous spectrum of HDL. Human HDL contains two major subfractions, HDL2 and HDL3, of densities 1.063 to 1.125 g/ml and 1.125 to 1.21 g/ml, respectively. Lipoprotein particles with the characteristics of HDL but with densities in the region 1.055 to 1.063 g/ml are also present in the plasma of some subjects (referred to as HDL1 and in certain animal species that have been subjected to a diet high in cholesterol (designated HDLc). In contrast to humans, mice exhibit a monodisperse distribution of HDL, except for certain genetically modified “humanized” strains, which show distinct HDL subfractions. Nevertheless, genetic and dietary factors have been demonstrated to alter the mean HDL particle size, as well as HDL composition in several common inbred strains.

Unlike the lipid determinations described above which can be performed on fresh or frozen samples, it is strongly recommended to use fresh and non-frozen samples of plasma to avoid the occurrence of changes in the composition and properties of the HDL subfractions; in some cases, however, (e.g. when subsequent analysis is to be made by gradient gel electrophoresis only), satisfactory results can be obtained with frozen plasma. Since strong salt solutions and high gravitational forces that are inherent in the methods result in substantial dissociation of some apolipoproteins (apoAIV, apo E, and apo Cs) from the native HDL particles, it is vital that standard conditions, especially centrigugation times, are strictly adhered to in order that a uniform product may be obtained from different preparations. Precipitation procedures are no less destructive to the lipoprotein particles than is ultracentrifugation with respect to the dissociation of apolipoproteins.

Ultracentrifugal isolation of mouse lipoproteins also has been described using a Beckman airfuge (34). The advantage of the airfuge is that is uses very small tubes, appropriate for ultracentrifugation of 100 ul of plasma, and because of the very high G forces that it generates, the time required to float the various lipoprotein fractions is dramatically reduced compared to classical methods. However, there are several disadvantages in using the airfuge for isolation of plasma lipoproteins. First, due to the very high G forces generated, the loss of various components of the lipoproteins, such as the apoC’s, would be expected to differ considerably compared to losses observed using lower G forces associated with the more classical methodologies. Secondly, density fractions isolated by the airfuge have been demonstrated to vary significantly compared to lipoproteins isolated using traditional centrifuges and rotors (35).

Electrophoresis

Much of the early work in classifying the human hyperlipoproteinemias involved the use of paper electrophoresis and later agarose gel electrophoresis. These methods allow identification of HDL, LDL, and VLDL bands, with chylomicrons remaining at the origin. While these methods are barely semiquantitative, they are occasionally of value in the investigation of unusual hyperlipidemias and can be used in the analysis of mouse lipoproteins. Suitable agarose gels are commercially available from Helena Laboratories (Beaumont, TX).

Polyacrylamide gradient gel electrophoresis (PAGE) separates lipoproteins according to their size and is widely used in the identification of lipoproteins and apolipoproteins. Human HDL2 and HDL3, described above, have each been demonstrated to contain more than ten fractions by the use of a combination of isoelectric focusing and immunological techniques that are not appropriate for routine assay. The resolution of human HDL into five subpopulations by gradient gel electrophoresis has provided a useful system for the definition of some of these components.

For analyses in the mouse, PAGE is primarily used to determine apolipoprotein concentrations. While several automated antibody based methods are available for the analysis of human apolipoproteins, at present no automated commercially supported methods exist for the mouse. However, anti-sera against several mouse apolipoproteins are commercially available and are suitable for western blot analysis (Biodesign International, Research Diagnostics, Cascade Bioscience). A wide variety of polyacrylamide gels that are suitable for the separation of apolipoproteins are available from Invitrogen/Novex and Biorad. The convenience, quality, and reproducibility of these commercially available gels makes them the best choice for most laboratories.

Gel filtration

Separation of lipoproteins by size using gel filtration as described by Jiao et al, is a widely used method that provides the complete lipoprotein profile of the major lipoprotein classes in mouse plasma (36). As detailed in the original method, lipoproteins are collected in 0.5 ml fractions as they are eluted through two Pharmacia Superose 6 columns connected in series, during a total run time of approximately two hours. A variety of analyses can be performed on the isolated fractions including cholesterol, triglyceride, and apolipoprotein analyses. The usual modifications to the original method that have been reported in the literature include the use of one column rather than two and the collection of larger fractions, usually one ml. Both of these modifications decrease the resolving capabilities of the method, a consequence that could be important in some circumstances. Subtle, but significant and reproducible changes in the mean particle diameter that can be demonstrated with two columns and 0.5 ml fractions, may be obscured with the decreased resolution that results from using a single column and collecting larger fractions. In general, gel filtration is “gentler” than ultracentrifugation with respect to loss of certain apolipoproteins as discussed above, however, remodeling of the lipoproteins still occurs to some extent during the isolation.

Literature cited

Olubadewo JO, Wingard M, Washington B, Tsai CS, Robinson TJ, Ochillo RF. The effects of stress on lipoproteins and catecholamines in rats. Cell Mol Biol. 1994;40(8):1201-6.
Whitney G. Timidity and fearfulness of laboratory mice: an illustration of problems in animal temperament. Behav Genet. 1970;1(1):77-85.
Irwin J, Ahluwalia P, Zacharko RM, Anisman H. Central norepinephrine and plasma corticosterone following acute and chronic stressors: influence of social isolation and handling. Pharmacol Biochem Behav. 1986;24(4):1151-4.
Misslin R, Herzog F, Koch B, Ropartz P. Effects of isolation, handling and novelty on the pituitary--adrenal response in the mouse. Psychoneuroendocrinology. 1982;7(2-3):217-21.
Nichols DJ, Chevins PF. Effects of housing on corticosterone rhythm and stress responses in female mice. Physiol Behav. 1981;27(1):1-5.
Goldsmith JF, Brain PF, Benton D. Effects of the duration of individual or group housing on behavioural and adrenocortical reactivity in male mice. Physiol Behav. 1978;21(5):757-60.
Pfaff J. Noise as an environmental problem in the animal house. Lab Anim. 1974;8(3):347-54. Review.
Del Pup J, Palmes ED. Effect of housing conditions on corticosterone levels in mice. Arch Environ Health. 1971;22(4):493-5.
Treiman DM, Fulker DW, Levine S. Interaction of genotype and environment as determinants of corticosteroid response to stress. Dev Psychobiol. 1970;3(2):131-40.
Peng X, Lang CM, Drozdowicz CK, Ohlsson-Wilhelm BM. Effect of cage population density on plasma corticosterone and peripheral lymphocyte populations of laboratory mice. Lab Anim. 1989;23(4):302-6.
Mazelis AG, Albert D, Crisa C, Fiore H, Parasaram D, Franklin B, Ginsberg-Fellner F, McEvoy RC. Relationship of stressful housing conditions to the onset of diabetes mellitus induced by multiple, sub-diabetogenic doses of streptozotocin in mice. Diabetes Res. 1987;6(4):195-200.
Naidu S, Winget CM, Jenner JW, Mele G, Holley DC. Effects of housing density on mouse physiology and behavior in the NASA Animal Enclosure Module simulators. J Gravit Physiol. 1995;2(1):P140.
Haemisch A, Gartner K. Effects of cage enrichment on territorial aggression and stress physiology in male laboratory mice. Acta Physiol Scand Suppl. 1997;640:73-6.
Ader DN, Johnson SB, Huang SW, Riley WJ. Group size, cage shelf level, and emotionality in non-obese diabetic mice: impact on onset and incidence of IDDM. Psychosom Med. 1991;53(3):313-21.
Haemisch A, Gartner K. The cage design affects intermale aggression in small groups of male laboratory mice: strain specific consequences on social organization, and endocrine activations in two inbred strains (DBA/2J and CBA/J). J Exp Anim Sci. 1994;36(4-5):101-16.
Tuli JS, Smith JA, Morton DB. Stress measurements in mice after transportation. Lab Anim. 1995;29(2):132-8.
Hoffman-Goetz L, MacNeil B, Arumugam Y. Effect of differential housing in mice on natural killer cell activity, tumor growth, and plasma corticosterone. Proc Soc Exp Biol Med. 1992;199(3):337-44.
De Souza J, Butler AA, Cone RD. Disproportionate inhibition of feeding in A(y) mice by certain stressors: a cautionary note. Neuroendocrinology. 2000;72(2):126-32.
Griffiths J, Shanks N, Anisman H. Strain-specific alterations in consumption of a palatable diet following repeated stressor exposure. Pharmacol Biochem Behav. 1992;42(2):219-27.
Kurokawa M, Akino K, Kanda K. A new apparatus for studying feeding and drinking in the mouse. Physiol Behav. 2000;70(1-2):105-12.
Poirel C, Larouche B. Murine circadian patterns of eating and drinking behaviors revisited: rhythmometric analyses and neuro-ethological comments. Chronobiologia. 1986;13(4):345-50.
Sigdestad CP, Connor AM, Sharp JB Jr, Ackerman MR. Evaluating feeding cycles of small rodents. Lab Anim Sci. 1974;24(6):919-21.
Neuhausser-Wespy F, Konig B. Living together, feeding apart: how to measure individual food consumption in social house mice. Behav Res Methods Instrum Comput. 2000;32(1):169-72.
Bachmanov AA, Reed DR, Tordoff MG, Price RA, Beauchamp GK. Nutrient preference and diet-induced adiposity in C57BL/6ByJ and 129P3/J mice. Physiol Behav. 2001;72(4):603-13.
Smith BK, Andrews PK, West DB. Macronutrient diet selection in thirteen mouse strains. Am J Physiol Regul Integr Comp Physiol. 2000;278(4):R797-805.
Arafa MS, Salit AM, Maher A, Abd-El-Gawad K. Food preference and consumption by murine rodents. Acta Vet Acad Sci Hung. 1975;25(2-3):275-8.
Imaizumi M, Takeda M, Suzuki A, Sawano S, Fushiki T. Preference for high-fat food in mice: fried potatoes compared with boiled potatoes. Appetite. 2001;36(3):237-8.
Wang L, Martinez V, Rivier JE, Tache Y. Peripheral urocortin inhibits gastric emptying and food intake in mice: differential role of CRF receptor 2. Am J Physiol Regul Integr Comp Physiol. 2001; 281(5):R1401-10.
Warnick, G.R., Enzymatic methods for quantification of lipoprotein lipids, In Methods in Enzymology (Segrest, J.P. and Albers, J.J., eds.) 129, 101-123. Academic Press. (1986).
Puppione, D. L. and Charugundla, S., A microprecipitation technique suitable for measuring alpha-lipoprotein cholesterol. Lipids. 29, 595-597 (1994).
Iverson SJ, Lang SL, Cooper MH. Related Articles. Comparison of the Bligh and Dyer and Folch methods for total lipid determination in a broad range of marine tissue. Lipids. 2001;36(11):1283-7.
Bligh EG and Dyer WJ. A rapid method of total lipid extraction and purification. Can. J. Biochem. 1959; 37:911-917.
Folch J, Lees M, Sloane-Stanley G. A simple method for the isolation and purification of total lipides from animal tissues. Biol. Chem. 1957;226:497-509.
Kupke IR, Worz-Zeugner S. Sequential microultracentrifugation of lipoproteins in 100 microliters of serum. J Lipid Res. 1986;27(9):988-95.
Shireman RB, Williams D. Related Articles. Evaluation of the rapid micromethod for ultracentrifugal separation of labeled plasma lipoproteins. Lipids. 1983;18(9):636-40.
Jiao S, Cole TG, Kitchens RT, Pfleger B, Schonfeld G. Genetic heterogeneity of lipoproteins in inbred strains of mice: analysis by gel-permeation chromatography. Metabolism. 1990;39(2):155-60.