5th edn. New York: Longman Scienti®c & Tech- nical, pp 142±76. Moreland A (1993) Experimental atherosclerosis of swine. In: Comparative Atherosclerosis.
The Go¨ttingen minipig as a model for postprandial hyperlipidaemia in man: experimental observations A. K. Olsen1,2,3 , E. M. Bladbjerg1 , P. Marckmann2 , L. F. Larsen2 & A. K. Hansen3 1
Department for Thrombosis Research, University of Southern Denmark and Department of Clinical Biochemistry, Ribe County Hospital, Esbjerg, 2 Research Department of Human Nutrition and 3 Department of Pharmacology and Pathobiology, Royal Veterinary and Agricultural University, Frederiksberg, Denmark
Summary Postprandial hyperlipidaemia is believed to be at herogenic. T his study aim ed to establ ish a minipig model to investigate determinants of postprandial lipid metabolism. In a randomized cross-over design seven minipigs were subjected to six different feeding regimens: intragastric fat loads of 1, 2, and 4 g fat (Intralipid1 , 20% ) kg¡1 in two fractions 1.5 h apart (1 =3 ®rst, 2 =3 second), 2 g fat (Intralipid1 ) kg¡1 in one fraction, and 2 g olive oil kg¡1 in two fractions, all after pre-feeding with standard diet, and ®nally 2 g fat (Intralipid1 ) kg¡1 in two fract ions without pre-feeding. Blood was sampled before and hourly for 7 h after gavaging, and plasma triglyc erides were measured. Triglycerides increased signi®cantly in all the feeding regimens (P < 0.001 ), except when olive oil was used as the fat source. A borderline signi®cant dose-response effect of the Intralipid1 dose on the triglyc eride response was observed. We found no signi®cant differences in triglyceride response whether 2 g fat (Intralipid1 ) kg¡1 was given in one or two fract ions, with or without pre-feeding. We conclude that postprandial hyperlipidaem ia in minipigs can be induced by gavaging an emulgated lipid solution (1±4 g fat =kg, Intralipid1 ), while olive oil is not applicabl e. T here is no need to administer the fat fractionated or to withhold food prior to administration. Keywords
Animals; dietary fat ; Intralipid1 ; olive oil; plasm a triglycerides
Postprandial hyperlipidaemia is increasingly recognized as atherogenic (Zilversmit 1979 ), and therefore there is a need to investigate determinants of postprandial lipid metabolism. Laboratory animals may be of great value in the investigation of such determinants, and in this study we sought to establ ish an animal model for the study of postprandial hyperlipidaem ia.
C o rre spond e nc e to : Aa ge Kristian O lsen DVM, Departm e nt o f Clinic a l Bio ch e m istry, Rib e Co unty Ho spita l, é ste rgad e 80, DK-6700 Esb je rg, De nm a rk E-m a il: a a k o @rib ea m t.dk Accepted 21 February 2002
Pig and man are both omnivorous with similarit ies in gastrointestinal anatom y and physiology (Swindle & Smith 1998), and they also have similarit ies in their blood coagulation systems (Olsen e t a l. 1999 ). In humans, high-density lipoprotein (HDL) protects against atherosclerosis, while low-density lipoprotein (LDL) and very low-density lipoprotein (VLDL) are at herogenic (Fekete 1993 ). Lik e humans, pigs are classi®ed as LDL mammals, because LDL is the dominant cholesterol carrier, and pigs may develop spontaneous at herosclerosis similar to humans (Moreland 1993 ).
# Laboratory Animals Ltd. Laboratory Animals (2002) 36, 438–444
Hyperlipidaemia in minipigs
T herefore, pigs may be particularly useful as models for the study of human lipoprotein metabolism and cardiovascular disease (Overturf & Loose-Mitchell 1992, Mortensen 1995). However, hyperlipidaem ia can be obtained in many other species, e.g. mice, rats, rabbits, and primates, but they all have some limitations: mice and rats are classi®ed as HDL mammals, and they are highly resistant to atherosclerosis (Gross 1994 ). Rabbits are VLDL mammals and are not optim al models of atherosclerosis (Overturf & Loose-Mit chell 1992 ). Among primates not all are LDL mammals, and they are not easily available. Methods for obtaining permanently elevated levels of plasma lipids in pigs exist, and these methods are widely used in atherosclerosis research (Schoutan e t a l. 1985, Kobari e t a l. 1991, Prescott e t a l. 1991, Jacobsson 1998 ). However, an evaluated method for obtaining postprandial hyperlipidaemia in pigs has not been established. In the present study we describe the effect of lipid dose, lipid source, lipid administration, time of lipid loading, and the im pact of pre-feeding on postprandial lipidaemia in minipigs.
Materials and methods Anim a ls Animal experiments were performed in accordance with the Danish Animal Experimentation Act on a license granted by the Ministry of Legal Affairs. All housing and procedures were performed according to the Convention ET S 123 of the Council of Europe (Council of Europe 1986 ). We used seven adult GoÈttingen minipigs (hogs, approxim ately 1 12 years old, and 22.0±38.5 kg) from Ellegaard GoÈttingen Minipigs (Dalmose, Denmark ). T he median weight increase was 2.5 kg (1.3±3.8 kg) during the 5-week study period. T he pigs had been health-monitored according to the Federation of European Laboratory Animal Science Associations (FELASA) guidelines (Rehbinder e t a l. 1998 ). T he minipigs were maintained in two groups (3 ‡ 4) in solid concrete ¯oor pens covered with straw bedding, and twice
439
daily the pigs were fed 220 g of a pelleted diet (Altromin No. 9023 Extra, mini-pigs, Brogaarden, Gentofte, Denmark ). On the experimental days the pigs were fed separately under supervision, and the non-pre-fed minipigs were housed separately when the others were fed. Altromin 9023 Extra contains 25 g crude fat kg¡1, 135 g crude protein kg¡1, and 10 500 kJ metabolizable energy kg¡1. Water was available a d lib itum . T he ventilation provided a total air exchange rate 8±10 tim es per hour, and the temperature was maintained at approxim ately 21 C. T here were no speci®c light cycles, and relative humidity was 55±80% . T he pigs were weighed prior to the experiments, and faeces was visually inspected twice daily on the experimental days and the following 2 days after.
Expe rim e nta l d e sign T he study was performed once a week for 6 weeks in a cross-over design. Each of the seven pigs participated in six different feeding regim ens in randomized order as follows: (1 ) At 06:15 h in the morning each minipig was fed 220 g minipig diet. T hereafter, Intralipid1 , 20% (200 g soy oil ‡ 12 g egg yolk ‡ 21.3 g glycerol per litre; Fresenius Kabi AB, Uppsala, Sweden) was adm inistered through a gastric tube in two fracti ons: one-third of the dose, i.e. 0.67 g fat =kg body weight (bw), was given between 07:45 and 08:00 h in the morning, while the remaining two-thirds of the dose, i.e. 1.33 g fat =kg bw was given 1.5 hours lat er. (2 ) As (1 ), but the doses of Intralipid1 were doubled (1.33 and 2.67 g fat =kg bw). (3 ) As (1 ), but the doses of Intralipid1 were halved (0.33 and 0.67 g fat =kg bw). (4 ) As (1 ), but the dose of Intralipid1 was not fracti onat ed. T he entire dose (2.00 g fat =kg bw) was administered between 07:45 and 08:00 h in the morning. (5 ) As (1 ), but olive oil (0.67 and 1.33 g fat =kg bw) (Santagata Lurgi, Genova, Italia) was used instead of Intralipid1 . Prior to administration, Laboratory Animals (2002) 36
440
olive oil was suspended in water (1 ‡ 4) to obtai n the same volume as the Intralipid1 . (6 ) As (1), but the pigs were not fed after 15.00 h the day before. T hese pigs were placed in single boxes without straw bedding, in which they stayed until the last blood sam ple was taken.
Blo o d sa m pling Before the experiments all the minipigs were accustomed to the blood sam pling procedure, and therefore no or only minimal signs of stress were observed during the sam pling. Blood samples were taken 15 min before (t ˆ ¡15 min; baseline) and at t ˆ 1 h, 2 h, 3 h, 4 h, 5 h, 6 h, and 7 h after the ®rst fat load dose. Venepunctures were performed as a blind procedure with the pigs in dorsal recumbency. A 21 gauge, 38 mm long needle was inserted into the cranial vena cava, and blood was taken using a Vacutainer1 (Terumo, Leuven, Belgium ). T hree millilitres of blood were collected in tubes containing 0.06 ml 0.235 mol =l tripotassium-ethylenediaminetetraac etic acid (K3-EDT A). Samples were stored at room temperature. Within 2 h after blood sampling the tubes were centrifuged for 15 min at 3000 6 g. Plasma was pipetted into plastic vials, rapidly frozen and stored at ¡18 C until analysis within less than 2 months.
Blo o d a na lyse s All samples were thawed at room temperature and analysed in one run. Total plasma triglyc erides (mmol =l) were determined by a commercial enzymatic method (GPO-PAP, Boehringer Mannheim Gm bH, Germany) on a Cobas Mira S (Roche Diagnostic Systems, Basel, Switzerland). T he intraserial coef®cient of variation (CV) was 1.9% based on duplicate measurements of 17 porcine plasma sam ples. Sta tistic s We wanted suf®cient power to detect a difference in peak postprandial triglyc eride concentration of 0.3 mmol =l between the six feeding regimens at a signi®cance level of 5% and a power of 85% . According to a previous Laboratory Animals (2002) 36
Olsen et al.
study (Ellegaard e t a l. 1995) the population standard deviati on of baseline triglyc eride concentrat ion in minipigs is 0.12 mmol =l. We decided on a group size of seven minipigs, which was based on an ANOVA sample size calculation. T he triglyceride responses for regimens 1±6 were compared by means of summary measures (Mat thews e t a l. 1990 ): area under the curve (AUC ), postprandial peak triglyceride concentration increment (peak ), and time from ®rst gavaging to peak postprandial triglyceride concentration (time to peak ). Differences between dietary regimens (baseline values and summary measures) and triglyceride changes over tim e within each dietary regimen were compared by Friedm an’s ANOVA. If these were signi®cant (P < 0.05 ), pairs of regimens or time points were compared with Wilcoxon’s test.
Results Baseline and postprandial triglyc eride concentrations are shown in Fig 1, and calculated values of AUC, peak, and time to peak are presented in Table 1. T he overall median baseline triglyceride concentration was 0.8 mmol =l (25±75 percentiles: 0.7±0.9 mmol =l), and there were no stati stically signi®cant differences in baseline between the feeding regimens. Triglycerides increased signi®cantly after all the feeding regimens (P < 0.001 ), except when olive oil was used as the fat source (Fig 1). Triglyceride AUC (ANOVA P < 0.029 ) and peak (ANOVA P < 0.007 ), but not the tim e to peak, were signi®cantly different between treat ments (Table 1). Triglyc eride AUC and peak did not differ signi®cantly between the ®ve Intralipid1 regim ens, but the highest response was observed with the highest dose of Intralipid1 . T his suggests a dose±response effect, which was supported by ®nding of a borderline signi®cant linear relationship between the Intralipid1 dose and the triglyceride response expressed as AUC (P ˆ 0.16 ) or peak value (P ˆ 0.051 ) (Fig 2). T here were no signs of steatorrhoea during the experiment and on the 2 days after.
Hyperlipidaemia in minipigs
441
Fig 1 Concentration of triglycerides in porcine plasma before and after administration of fat. L: lipid gavaging. Values are medians (n ˆ 7) (25–75 percentiles). 1 g, 2 g, and 4 g: 1, 2, and 4 g fat (Intralipid1 ) kg¡1 in two fractions; 2 g; unfractionated: 2 g fat (Intralipid1 ) kg¡1 in one fraction; 2 g; olive oil: 2 g fat (olive oil) kg¡1 in two fractions; 2 g; no pre-feeding: fasting before 2 g fat (Intralipid1 ) kg¡1 in two fractions; TG: triglycerides, *:different from baseline values (P < 0.05)
Laboratory Animals (2002) 36
442
Olsen et al.
Table 1 Measures of triglyceride responses (AUC, peak, time to peak) for six feeding regimens in minipigs (n ˆ 7) Feeding regimens
AUC (mmol=l.h)
Peak (mmol=l)
Time to peak (h)
1 g fat=kg bw 2 g fat=kg bw 4 g fat=kg bw 2 g fat=kg bw, unfractionated 2 g fat=kg bw, olive oil 2 g fat=kg bw, fractionated, no pre-feeding
12.96 12.43 13.82 11.16
1.24 1.34 2.30 1.34
5 5 6 5
(8.41–13.67) (10.49–13.63) (10.70–18.85) (9.56–16.50)
7.87* (7.18–10.83) 11.30 (8.30–16.46)
(0.90–1.50) (1.10–2.26) (1.36–2.86) (1.29–2.29)
(4–6) (4–5) (5–6) (4–5)
0.66* (0.49–0.79)
6 (3–6)
1.51 (0.75–1.80)
4 (4–5)
AUC: area under the curve; peak: postprandial peak triglyceride concentration substracted the baseline value; time to peak: time from rst gavaging to postprandial peak triglyceride concentration; bw: body weight * : different from the other feeding regimens (P < 0.05). Median values (25–75 percentiles)
Fig 2 Individual triglyceride dose-response curves presented as AUC and peak triglyceride concentration after administration of three different doses of Intralipid1 (1, 2, and 4 g fat=kg body weight). AUC: area under the curve; peak: postprandial peak triglyceride concentration substracted the baseline value
Discussion We found a dose-dependent postprandial increase in plasm a triglyc eride concentration in minipigs after gavaging Intralipid1 (1, 2, and 4 g=kg). In contrast, olive oil instillati on was not associated with a postprandial triglyc eride increase (Fig 1). We found no signi®cant differences in triglyceride response whether 2 g fat (Intralipid1 ) kg¡1 was given in one or two fracti ons, with or without prefeeding. T he observed minipig baseline triglyc eride concentrations, and triglyc eride responses aft er Intralipid1 loading were of Laboratory Animals (2002) 36
similar magnitude as in human beings eating high-fat diets (42% of energy from fat ) (Larsen e t a l. 1997, 1999 ). We observed a borderline signi®cant linear relationship between the Intralipid1 dose and the triglyc eride response expressed as AUC or peak triglyceride concentration (Fig 2). T he dose of Intralipid1 (1.0±2.0 g fat =kg) needed to reach a certain response was at the same level or only a little higher than in similar human studies (high-fat diets containing 1.0 g fat =kg) (Larsen e t a l. 1997, 1999). T he small difference may be explained by a higher clearance of triglyc erides in minipigs, since
Hyperlipidaemia in minipigs
metabolism are supposed to be fast er in minipigs than in humans due to the smaller size of the minipigs (Hau & Poulsen 1988). We cannot exclude that faeces contained small quant a of fat. However, there were no signs of steatorrhea during the experiment, and therefore fat absorption seems to have been suf®cient. Furthermore, it is unlikely that fat adm inistration by gavaging differs signi®cantly from oral feeding in relation to fat absorption: In humans as much as 30% of triglyc erides can be digest ed by lingual lipase (Ganong 1991 ), but in most animal species digest ion in the mouth is mainly mechanical (McDonald e t a l. 1995 ), and the saliva does not play an important role for fat digesti on (Lewis & Hill 1983). As the sam ple size was small, it cannot be excluded that a true difference between the dietary regimens has been overlooked. However, the im portant thing is that postprandial hyperlipidaem ia can be obtai ned with all dietary regimens involving Intralipid1 , the highest response apparently being observed with the highest dose of Intralipid1 . It is therefore not necessary to deprive the pigs of their normal food in the morning, and there is no need to administer the Intralipid1 in more than one fract ion. T he postprandial hyperlipidaem ia was signi®cantly higher after administration of Intralipid1 than after administrat ion of olive oil. T his can be explained by the emulsion state of the Intralipid1 . One litre of Intralipid1 20% contains 200 g soy oil, 12 g egg yolk phospholipids, and 21.3 g glyc erol. T he diameter of each lipid particle is approximat ely one millimetre, which is similar to the size of intestinal micelles (Lewis & Hill 1983), and faci litates fat absorption. Probabl y the concentration of bile salts in minipig intestines was too low to emulsify the olive oil suf®ciently. T he concentration of bile salts and lipase in pancreas juice changes in response to the nature of the diet, and the emulsifyi ng capacit y is the rate-lim iting stage in fat absorption (McDonald e t a l. 1995). T he normal minipig diet has a low fat content (25 g crude fat =kg). In another study using 2 12 g fat (Intralipid1 ) kg¡1 in minipigs (seven hogs, median weight 24 kg) we found a signi®cant
443
postprandial triglyc eride increase in the chylomicron fraction (from 0.4 mmol =l at baseline to 1.2 mmol =l 3.5 h postprandial ), but no increase in the non-chylomicron fraction (0.2 mmol =l at baseline) (Larsen, Olsen, Hansen, Bukhave, Marck mann, unpublished observation, 2000 ). Accordingly, the postprandial lipoprotein pro®le in minipigs resembles that seen in man (Pedersen e t a l. 1999 ). T he similarities between human and porcine gastrointestinal physiology, lipoproteins, and atherosclerosis mak e our porcine model of postprandial hyperlipidaemia attrac tive. Com pared to farm pigs GoÈttingen minipigs are easy to handle, are de®ned free of a long range of microorganisms, and are genetically homogenous (Bollen e t a l. 2000). Besides, we investigated whether feeding regimens No. 1 was capable of inducing hyperlipidaemia in farm pigs (six sows, Landrac e =Yorkshire, 30±31 kg), but found no postprandial hyperlipidaemia (1.1 mmol =l at baseline), probably due to exceedingly effective plasm a lipid clearance. Our model (feeding regim ens No. 1) therefore cannot be transferred to farm pigs (Olsen, Bladbjerg, Hansen, Marc kmann, unpublished observation, 2000 ). In a newly published paper a three-point test for human postprandial hyperlipidaem ia was evaluated (Guerci e t a l. 2001 ). T hey found that a simpli®ed model (three blood sam ples) correlates well with a conventional model (nine blood sam ples). We found nearly the same trends in our porcine results whether three (t ˆ 0, 4, and 7 h) or eight (t ˆ 0, 1, 2, 3, 4, 5, 6, and 7 h) time points were used. T herefore, the three-point test also seems useful for the analysis of postprandial hyperlipidaemia in minipigs. We conclude that gavaging an emulgated lipid solution (Intralipid1 ) induces an apparently dose-related postprandial hyperlipidaemia in minipigs similar to that following human consumption of high-fat meals. Ac k no w le d gm e nts We wish to thank Marie Pedersen, Preben Lund, Klaus Switon, Pia Adolfsen, Kit Jùrgensen, and Jens Flesher for technical assistance with the animals. Kirsten Ebbesen is thanked for helping with the triglyceride analyses. Laboratory Animals (2002) 36
444
References Bollen PJA, Hansen AK, Rasmussen HJ (2000) Th e La b o ra to ry Sw ine. Boca Raton, Florida: CRC Press, 135 pp Ellegaard L, Jùrgensen KD, Klastrup S, Hansen AK, Svendsen O (1995) Hematologic and clinical chemical values in 3 and 6 month old GoÈttingen minipigs. Sca nd ina via n Jo urna l o f La b o ra to ry Anim a l Sc ie nce 3, 239±48 Fekete S (1994) Animal models in experimental atherosclerosis: a critical review. Ac ta Ve te rina ry Hungary 41, 3±9 Ganong WF (1991) Digestion and Absorption In: Re vie w o f Me d ic a l Ph ysio lo gy, 15th edn. USA: Prentice-Hall International Inc, pp 437±47 Gross DR (1994) Anim a l Mo d e ls in C a rd io va scula r Re se a rch . Dordrecht: Kluwer Academic Publishers, 494 pp Guerci B, Paul JL, Hadjadj S, Durlach V, VergeÁs B, Attia N, Girard-Globa A, Drouin P (2001) Analysis of the postprandial lipid metabolism: use of a 3-point test. Dia b e te s Me ta b o lism 27, 449±57 Hau J, Poulsen OM (1988) Doses for laboratory animals based on metabolic rates. Sc a nd ina vian Jo urna l o f La b o ra to ry Anim a l Sc ie nce 2, 81±4 Jacobsson L (1998) Experimental hyperlipidemia and atherosclerosis in mini-pigs; in¯uence of certain drugs. Sc a nd ina vian Jo urna l o f La b o ra to ry Anim a l Sc ie nce Supple m e nt 1, 85±91 Kobari Y, Koto M, Tanigawa M (1991) Regression of diet-induced atherosclerosis in GoÈttingen miniature swine. La b o ra to ry Anim a ls 25, 110±16 Larsen LF, Bladbjerg EM, Jespersen J, Marckmann P (1997) Effects of dietary fat quality and quantity on postprandial activation of blood coagulation factor VII. Arte rio sc le ro sis Th ro m b o sis Va sc ula r Bio lo gy 17, 2904±9 Larsen LF, Jespersen J, Marckmann P (1999) Are olive oil diets antithrombotic? Diets enriched with olive, rapeseed, or sun¯ower oil affect postprandial factor VII differently. Am e ric an Jo urna l o f C linic a l Nutritio n 70, 976±82 Lewis D, Hill KJ (1983) The provision of nutrients. In: Nutritio na l Ph ysio lo gy o f Fa rm Anim a ls (Rook JAF, T homas PC, eds). Essex: Longman Group Ltd, pp 3±40 Matthews JNS, Altman DG, Campbell MJ, Royston P (1990) Analysis of serial measurements in medical research. British Me d ic a l Jo urna l 300, 230±5 McDonald P, Edwards RA, Greenhalgh JFD, Morgan CA (1995) Digestion. In: Anim a l Nutritio n, 5th edn. New York: Longman Scienti®c & Technical, pp 142±76
Laboratory Animals (2002) 36
Olsen et al.
Moreland A (1993) Experim ental atherosclerosis of swine. In: C o m para tive Ath e ro scle ro sis. Th e Mo rph o lo gy o f Sponta ne o us a nd Ind uc ed Ath ero sle ro tic Le sio ns in Anim a ls a nd its Re la tio n to Hum a n Dise a se (Robert J, Straus M, eds). Hoeber Medical Division, pp 21±4 Mortensen A (1995) Transgenic laboratory animal models in atherosclerosis research. Sca nd ina via n Jo urna l o f La b o ra to ry Anim a l Scie nc e 2, 133±43 Olsen AK, Hansen AK, Jespersen J, Marckmann P, Bladbjerg EM (1999) T he pig as a model in blood coagulation and ®brinolysis research. Sca nd ina via n Jo urna l o f La b o ra to ry Anim a l Scie nc e 4, 214±24 Overturf ML, Loose-Mitc hell DS (1992) In vivo model system s: the choice of the experimental animal model for analysis of lipoproteins and atherosclerosis. C urre nt O pinio n in Lipid o lo gy 3, 179±85 Pedersen A, Marckmann P, SandstroÈm B (1999) Postprandial lipoprotein, glucose and insulin responses after two consecutive meals containing rapeseed oil, sun¯ower oil or palm oil with or without glucose at the ®rst meal. British Jo urna l o f Nutritio n 82, 97±104 Prescott MF, McBride CH, Hasler-Rapacz J, von Linden J, Rapacz J (1991) Development of complex atherosclerotic lesions in pigs with inherited hyper-LDL cholesterolemia bearing mutant alleles for apolipoprotein B. Am e rica n Jo urna l o f Pa th o lo gy 139, 139±47 Rehbinder C, Baneux P, Forbes D, van Herck H, Nicklas W, Rugaya Z, Winkler G (1998) FELASA recommendations for the health monitoring of breeding colonies and experimental units of cats, dogs and pigs. Report of the Federation of European Laboratory Animal Science Associations (FELASA) Working Group on Animal Health. La b o ra to ry Anim a ls 32, 1±17 Schoutan JA, Beynen AC, Hoitsma HFW, Mulder C, Havekes LM, van Gent CM (1985) Serum lipoproteins in cholesterol-fed pigs after partial ileal bypass. La b o ra to ry Anim a ls 19, 98±105 Swindle MM, Smith AC (1998) Comparative anatomy and physiology of the pig. Sc a nd ina via n Jo urna l o f La b o ra to ry Anim a l Sc ie nce Supple m ent 1, 11±21 Zilversmit DB (1979) Atherogenesis: a postprandial phenomenon. C irc ula tio n 60, 473±85
