Establishment of ApoE-knockout mouse model of preeclampsia and relevant mechanisms.

In the present study, we established an ApoE-knockout mouse model of preeclampsia to examine the role of vascular endothelial injury associated with abnormal lipid metabolism in the pathogenesis of preeclampsia. To establish the ApoE-knockout homozygous (ApoE-/-) and heterozygous (ApoE+/-) mouse model, mice were mated with the same genotype and orbital blood on day 19 of conception was collected. The progeny mice were assigned into 3 groups: ApoE-/, ApoE+/- and wild-type (WT) groups. Total cholesterol, triglyceride, low-density and high-density lipoprotein were measured in the serum at the end of conception. During conception, the systolic blood pressure of caudal artery was measured every 4 days. Using bicinchoninic acid protein assay, urinary protein and creatinine ratio was measured with a creatinine kit. We observed the pathological changes of glomerular filtration membrane and macroscopic/microscopic morphological changes of placenta by hematoxylin and eosin (H&E) staining and transmission electron microscope. Take fetal mouse through cesarean section on 19th day, measure the birth weight and placental weight of fetal mouse. Using ELISA we measured the expression levels of toll-like receptor 4 (TLR4) and soluble fms-like tyrosine kinase-1 (sFlt-1). Our results showed that the differences in serum lipid levels were not statistically significant (P>0.05). The mean systolic blood pressure, urinary protein and creatinine in ApoE-/- group were significantly higher than ApoE+/- group and WT group (P<0.05). Thickening and edema of glomerular filtration membrane, capillary thrombosis, significant edema and necrosis of placental villous stroma were observed in ApoE-/- group. No significant change was detected in the ApoE+/- or WT group. The TLR4 and sFlt-1 expression levels in ApoE-/- group were significantly higher than ApoE+/- and WT group (P<0.05). We concluded that ApoE-knockout mouse could simulate the pathologic process of preeclampsia, while the changes in serum lipids were not noteworthy, thus the pathogenesis of preeclampsia may be mediated by TLF4 and sFlt-1.

Introduction

Preeclampsia is a severe complication of gestational hypertension and an important source of perinatal mortality for mothers as well as infants. Impaired cardio-pulmonary function due to proteinuria and severe hypoalbuminemia during treatment are important indications for the termination of pregnancy (1). Vascular endothelial injury hypothesis is a widely accepted hypothesis explaining the pathogenesis mechanisms of preeclampsia relating to endothelial injury, oxidative stress, inflammation, proliferation, apoptosis and immune disorders. It was considered the center piece for the pathogenesis of preeclampsia (2). Previous studies revealed a correlation between abnormal lipid metabolism and abnormal endothelial function (3). The characteristics of preeclampsia have been described using the uteroplacental ischemia model, chronic nitric oxide synthase inhibition model, adriamycin nephropathy model and chronic high insulin model (3,4). In that study, the role of abnormal lipid metabolism and vascular endothelial cell damage in the pathogenesis of preeclampsia was analyzed by establishing an ApoE gene knockout mouse model.

However, to the best of our knowledge, few studies are available in which the preeclampsia model was established based on abnormal lipid metabolism.

Materials and methods

Experimental animals

ApoE−/− mice and wild-type (WT) mice (B6.129) were purchased from the Animal Center of Nanjing University and bred in SPF room. These mice were weaned 20 days after birth and fed with regular feed (4% fat and 0.07% cholesterol). The mice were kept in cages in a quiescent environment with a 12-h light/dark cycle, temperature at (30+0.5), and relative humidity of (55+0.5)%, and had free access to water. ApoE+/− mice were produced by mating ApoE−/− mice and ApoE+/− mice of the same genotype, the progeny mice were assigned into 3 groups (6 mice in each group), i.e., ApoE−/− group, ApoE+/− group and WT group, based on the genotype.

ApoE genotyping

ApoE−/− mice were produced by replacing a part of exon 3 and intron 3 in ApoE gene by neo gene, thus, ApoE gene was inactivated. WT specific DNA fragment (155 bp) was amplified using sense primer P2 and antisense primer P3. DNA fragment of homozygous allele (255 bp) of ApoE−/− was amplified with sense primer P1 and antisense primer P3. Primers used in this part were: P1, 5-GCCTAGCCGAGGGAGAGCCG-3′; P2, 5′-TGTGACTTGGGAGCTCTGCAGC-3′; P3, 5′-GCCGCCCCGACTGCATCT-3′.

PCR reaction system used: ddH2O 11.55 µl, buffer 1.5 µl, dNTPs (10 µM) 0.5 µl, P1 (180) (10 µM) 0.25 µl, P2 (671) (10 µM) 0.25 µl, P3 (672) (10 µM) 0.25 µl; Taq enzyme (5 U/µl): 0.2 µl, gDNA (1 µg/µl) 0.5 µl (4). PCR conditions were: 94°C, for 3 min; 94°C, for 20 sec, 68°C, for 40 sec, 72°C, for 2 min, 35 cycles; 72°C, for 10 min. PCR product was characterized using 2% agarose gel electrophoresis, with loading volume of 8 µl. The electrophoresis set-up was 90 V for 30 min.

Specimen preparation

Twelve-week mice of the same genotype (male:female =1:1) were housed in a cage. Vaginal plug indicated successful mating. Ten days after vaginal plug, non-pregnant mice were re-housed in the cage. The day after successful mating was recorded as day 1 of gestation. Blood pressure was measured every 4 days starting from day 0, mouse urine was collected on day 4, 8, 12 and 16 in metabolic cages. On day 19, mice were anaesthetized using intraperitoneal injection of pentobarbital sodium (3 g/l, 30 mg/kg) after 12-h fasting, and orbital blood was collected. The placenta was removed after cesarean section, placed in an ice bath and stored at −80°C until use. The number and weight of fetal mice were measured and recorded.

Observational measurements and analysis methods

Using ELISA and an automatic biochemical analyzer, total cholesterol (TC), triglyceride (TG), low-density lipoprotein (LDL) and high-density lipoprotein (HDL) levels measured at the end of gestation. The systolic blood pressure in caudal artery was measured using the non-invasive tail-cuff method with a CODA non-invasive mouse-tail blood pressure gauge (Kent Scientific Corporation, Torrington, CT, USA). The mouse was placed in the CODA fixation device and tail was exposed and immersed in 40°C water for 30 min, after which the tail was soft enough and dilated adequately, and fixed at the tail root. The caudal artery was closely contacted to the pulse sensor of CODA non-invasive mouse-tail blood pressure gauge. The resting blood pressure was measured 3 times and the mean blood pressure was calculated. Urinary protein was measured with using bicinchoninic acid (BCA) protein kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA) and urinary creatinine was measred with a creatinine kit (ab65340; Abcam, Cambridge, MA, USA). The pathological changes of glomerular filtration membrane and macroscopic/microscopic morphological changes of placenta were evaluated using hematoxylin and eosin (H&E) staining and TEM. Cardiac perfusion of PBS was performed for 5 min, and then 2% paraformaldehyde for 2 min. One kidney was removed and dissected longitudinally and fixed in 2% paraformaldehyde for 4 h, dehydrated in 30% sucrose/PBS at 4°C for 24 h, embedded in cool embedding medium (OCT compound), cut into sections and stained using H&E staining. The morphological changes in glomerular slices were observed under a microscope. For TEM observation, the sections were fixed in 2% glutaraldehyde for 2 h, washed 3 times with 0.1 mol/l phosphate buffer, fixed in 1% osmium tetroxide, dehydrated in ethanol and acetone subsequently, replaced with oxypropylene, embedded in epoxy resin (Epon 812), cut into ultrathin sections with an LKB-III microtome, dual-stained by uranyl acetate and lead citrate, and observed with JME-1200-EX transmission electron microscope (Olympus, Tokyo, Japan).

The observational measurements included the morphology of capillary endothelial cells, thickening or thinning of basal membrane, precipitation of electron dense material, the location, number and shape of the deposits, the changes of podocytes and foot process, the proliferation of mesangial cells and mesangial matrix in mesangial region. Slices of placenta were prepared and observed in the same way.

The expression levels of toll-like receptor 4 (TLR4) and soluble fms-like tyrosine kinase-1 (sFlt-1) at the end of gestation were measured with an ELISA kit (MVR100; R&D Systems, Inc., Minneapolis, MN, USA).

Statistical analysis

SPSS 20.0 software (Chicago, IL, USA) was used for statistical analyses. Quantitative data were presented as mean ± SD. One-way ANOVA was used for the comparison among multiple groups. Qualitative data are presented as number and percentage. The Chi-square test was used for inter-group comparison. P<0.05 was considered as statistically significant.

Results

Serum lipids

As shown in Table I, serum lipid levels in the ApoE−/−, ApoE+/− and WT groups were not significantly different (P>0.05).

Table I.

Serum lipids (mmol/l).

Groups	TC	TG	LDL	HDL
ApoE−/−	4.6±1.2	1.2±0.6	2.7±0.8	0.6±0.2
ApoE+/−	4.5±1.3	1.3±0.7	2.6±0.9	0.7±0.3
WT	4.4±1.2	1.2±0.5	2.5±0.7	0.6±0.3
F-value	0.526	0.427	0.326	0.528
P-value	0.411	0.325	0.241	0.427

TC, total cholesterol; TG, triglyceride; LDL, low-density lipoprotein; HDL, high density lipoprotein.

Blood pressure

As shown in Fig. 1, mean systolic blood pressures in ApoE−/− group on day 12 and 16 were significantly higher than the ApoE+/− and WT groups (P<0.05).

Figure 1.

Changes in trend of SBP.

Urinary proteins and creatinine

As shown in Table II, urinary proteins and creatinine levels in ApoE−/− group on day 12 and 16 were significantly higher than the ApoE+/− and WT groups (P<0.05).

Table II.

Urinary proteins and creatinine.

Groups	Urinary proteins (mg/l)					Urninary creatinine (µmol/l)
	0 day	4 days	8 days	12 days	16 days	0 day	4 days	8 days	12 days	16 days
ApoE−/−	50.2±13.6	65.6±17.2	73.2±16.6	104.7±23.6	232.5±46.5	45.2±13.2	56.5±14.6	66.8±18.2	154.7±36.4	246.7±52.7
ApoE+/−	53.3±14.5	64.7±13.6	75.3±12.3	82.0±15.8	94.5±26.5	44.3±12.6	53.2±15.4	65.9±13.7	83.2±23.1	96.3±26.5
WT	51.7±15.2	63.2±14.8	72.4±14.9	81.6±11.3	96.5±24.7	43.7±13.6	52.7±16.5	64.2±18.9	80.5±24.7	94.5±26.5
F-value	0.526	0.432	0.857	4.847	9.655	0.541	0.329	0.502	5.302	8.657
P-value	0.214	0.303	0.768	0.034	0.000	0.236	0.214	0.548	0.031	0.000

WT, wild-type.

Structural changes of glomerular filtration membrane and placenta

As shown in Fig. 2, thickening and edema in the glomerular filtration membrane as well as capillary thrombosis was evident in the ApoE−/− group, while no significant changes were detected in the ApoE+/− or WT group.

Figure 2.

Structural changes of glomerular.

Significant edema and necrosis of placental villous stroma, irregular nuclear morphology, degeneration of cytoplasmic membrane structures, fat deposition in placenta and mitochondrial swelling and deformation were observed in in the ApoE−/− group (Fig. 3), while no significant changes were detected in the ApoE+/− or WT group.

Figure 3.

Structural change of placenta. (B) is the magnified version of (A), red arrow shows the swelling and deformation of the trophoblast mitochondria.

The levels of serum TLR4 and sFlt-1

TLR4 and sFlt-1 levels in ApoE−/− group were significantly higher than the ApoE+/− and WT groups (P<0.05) (Table III).

Table III.

Levels of serum TLR4 and sFlt-1 (ng/ml).

Groups	TLR-4	sFlt-1
ApoE−/−	45.7±4.7	32.4±5.6
ApoE+/−	5.3±1.2	4.2±1.3
WT	5.2±1.3	3.6±1.4
F-value	10.524	12.645
P-value	0.000	0.000

TLR4, toll-like receptor 4; sFlt-1, soluble fms-like tyrosine kinase-1; WT, wild-type.

Discussion

The theory of vascular endothelial injury during gestational hypertension suggested that vascular endothelial injury increased the synthesis and release of vasoconstrictor factors, decreased the synthesis and release of endothelium-derived relaxing factor, leading to the disturbance of vasoactive factors and subsequent vasoconstriction, impaired connection of vascular endothelial cells, increased vascular permeability, extravasation of intravascular proteins and fluid, platelet aggregation and activation of coagulation system (5). Previous findings showed that lipid peroxidation and inflammatory response due to abnormal lipid metabolism played a major role in inducing vascular endothelial dysfunction (6). The possible mechanism may be that the activation of neutrophils and adherence of neutrophils to vascular endothelium, led to vascular injury (7). Placenta could secrete various cytokines, leading to vascular endothelial injury, thus it could mediate various pathology and play important roles in the pathogenesis of gestational hypertension (8).

The present study established the ApoE−/− mouse model of preeclampsia and found the following symptoms in these mice: i) thickening and edema in glomerular filtration membrane; ii) capillary thrombosis; iii) significant edema and necrosis of placental villous stroma; iv) irregular nuclear morphology; v) degeneration of cytoplasmic membrane structures; and vi) fat deposition in placenta and mitochondrial swelling and deformation. This model could simulate the pathological process of preeclampsia and indicated that dyslipidemia is important in preeclampsia pathogenesis. In the ApoE−/− group, we observed no changes in serum lipids, which was inconsistent with previous studies (9,10). Those studies showed that preeclampsia occurred in parallel with dyslipidemia. On the other hand, our results were consistent with the results of the study by Belo et al (11). Their results revealed that ApoE gene polymorphism was not a risk factor for preeclampsia (11,12).

The results of this study showed that both TLR4 and sFlt-1 expression levels in the ApoE−/− group were considerably higher than those of the ApoE+/− and WT groups. Results from previous studies reported that TLR4 was involved in ischemic and hypoxic pulmonary hypertension, pulmonary edema and cerebral edema due to barrier dysfunciton of endothelial cells, and proteinuria due to barrier dysfunction of glomerular filtration membrane. TLR4 recognized the pathogen-associated molecular pattern and endogenous ligands, induced intracellular signal transduction, and led to an inflammatory response (13). TLR4 was overexpressed in immune cells, renal podocytes, trophoblast cells, and vascular endothelial cells (14). In addition, overexpression of sFlt-1 was also associated with the pathogenesis of preeclampsia. It was shown that sFlt-1 bound to VEGF, reduced the expression of glomerular slit diaphragm protein, promoted endothelial isolation and hypertrophy, and led to proteinuria (15,16). The mean systolic blood pressures in the ApoE−/− group on day 12 and 16 were significantly higher than those of the ApoE+/− and WT groups, indicating the probability of preeclampsia in the middle and late phase of gestation. This was consistent with natural disease course. A previous study established TLR4-knockout pregnant mouse, in which the proteinuria in preeclampsia mice were reversible, and the hypertention in preeclampsia mice was also relieved, indicating that TLR4 may be involved in the pathological process of preeclampsia due to sFlt-1 (15).

In conclusion, ApoE-knockout mouse simulated the pathologic process of preeclampsia, while the change of serum lipids was not significant, thus the pathogenesis of preeclampsia may be mediated by TLF-4 and sFlt-1.