Induction of body weight loss through RNAi-knockdown of APOBEC1 gene expression in transgenic rabbits

Abstract

In the search of new strategies to fight against obesity, we targeted a gene pathway involved in energy uptake. We have thus investigated the APOB mRNA editing protein (APOBEC1) gene pathway that is involved in fat absorption in the intestine. The APOB gene encodes two proteins, APOB100 and APOB48, via the editing of a single nucleotide in the APOB mRNA by the APOBEC1 enzyme. The APOB48 protein is mandatory for the synthesis of chylomicrons by intestinal cells to transport dietary lipids and cholesterol. We produced transgenic rabbits expressing permanently and ubiquitously a small hairpin RNA targeting the rabbit APOBEC1 mRNA. These rabbits exhibited a moderately but significantly reduced level of APOBEC1 gene expression in the intestine, a reduced level of editing of the APOB mRNA, a reduced level of synthesis of chylomicrons after a food challenge, a reduced total mass of body lipids and finally presented a sustained lean phenotype without any obvious physiological disorder. Interestingly, no compensatory mechanism opposed to the phenotype. These lean transgenic rabbits were crossed with transgenic rabbits expressing in the intestine the human APOBEC1 gene. Double transgenic animals did not present any lean phenotype, thus proving that the intestinal expression of the human APOBEC1 transgene was able to counterbalance the reduction of the rabbit APOBEC1 gene expression. Thus, a moderate reduction of the APOBEC1 dependent editing induces a lean phenotype at least in the rabbit species. This suggests that the APOBEC1 gene might be a novel target for obesity treatment.

Figure 1. Structure of the rbapobec1-shRNA construct.

The rbapobec1-shRNA construct encompassed the H1-rbapobec1-shRNA gene that expressed the shRNA under the activity of the H1 promoter. A gene expression insulator element (two copies of the chicken ß-GLOBIN gene fragment 5′HS4) and a transcription unit composed of the hEF1alpha – promoter, the rabbit ß-GLOBIN second exon and intron, and the human GH gene polyadenylation signal were expected to protect the shRNA expression from transcriptional extinction that occurs frequently in transgenesis. Transgenic animals were detected by PCR using sets of primers 1, 4, and 5 (Table S1). Moreover, we checked that after PCR amplification using the a/b set, a 864 bp long fragment with the expected sequence was amplified.

Figure 2. rbapobec1-shRNA transgene expression in rabbit intestine.

The amount of shRNA targeting the rabbit APOBEC1 mRNA was measured in RNAs prepared from duodenum cells as described in “materials and methods” section in 3 rbapobec1-shRNA lines (shL21, shL23 and shL27). Values are given in females (F) and males (M) after normalization to the level of Let7c miRNA determined simultaneously as reference gene in each sample. The number of animals in each group is indicated in brackets. Values are given with the standard error of the mean (sem). All shRNA expressing lines harbored one copy of the rbApobec1-shRNA transgene. Note that in shL21 line, the shRNA transgene expression was the hig hest compared with lines shL23 and shL27.

Figure 3. Expression of the rabbit APOBEC1 gene in wild type and rbapobec1-shRNA transgenic rabbits.

The amount of rabbit APOBEC1 mRNA was measured in RNAs prepared from duodenum cells as described in “materials and methods” section in wild type animals (WT) and in two rbapobec1-shRNA lines (shL21, shL27). Values are given in females (F) and males (M) after normalization to the level of expression of three reference genes (RPLT9, YHWAZ, HPRT) determined simultaneously in each sample. The number of animals in each group is indicated in brackets. Values are given with the standard error of the mean (sem). Comparisons were made with control animals of the same sex (*** = p<0.001; * = p<0.05).

Figure 4. Indirect estimation of the level of “CAA” to “UAA” editing.

A: schematic representation of the rabbit APOB mRNA from the AUG translation initiation codon until the STOP codon. At the 2177^th codon, the “C” residue is edited in a “U” residue. Using reverse transcribed RNA as template, the LApob48F/LApoB48R set of primers amplifies a 455 bp long amplicon encompassing the 2177^th codon. When using the APOBR4 primer as sequencing primer, the chromatogram shows the antisense sequence. B: detail of a characteristic chromatogram showing how the heights of the peaks were measured at the level of the 2177^th codon. Here, the “A” residue was the major one (a1), and the “G” the minor one (g1). Consequently, a large majority of DNA strands in this mixture encompassed the edited TAA (STOP) codon at position 2177. (a2) and (g2) are measured as references. C: standard equations obtained by plotting the a1/a2 and g1/g2 ratios against the amount of “A” or “G” containing DNA 455 bp fragment in the sequenced sample. Amounts are given as percentage of “A” or “G” containing DNA.

Figure 5. Indirect estimation of editing in wild type and rbapobec1-shRNA transgenic rabbits.

APOBEC1 dependent editing was measured in the intestine of wild type and transgenic animals. Values were deduced from sequence chromatograms of a PCR fragment encompassing the edited codon as described in “material and methods” section and in Figure 4. The amount of DNA with a “A” residue was representative of the amount of APOB mRNA with a 2177^th STOP/edited codon; the amount of DNA with a “G” residue was representative of the amount of full length APOB mRNA. The number of studied animals in each group is indicated in brackets. Mean values are given as percentages with the standard error of the mean (sem). Comparisons were made with control animals (** = p<0.001).

Figure 6. Plasma concentration of APOB48 in rabbits challenged by a high fat/high cholesterol regimen.

Plasma concentration of APOB48 was assayed by a specific ELISA kit. Four wild type rabbits and four transgenic rabbits expressing the rbapobec1-shRNA transgene were fed ad libitum with a high fat/high cholesterol regimen for 9 days. Blood samples were collected before the high fat/high cholesterol regimen (D0) and 9 days after the starting of the regimen (D9). Each point indicates the plasma concentration of APOB48 (in ng/ml) in one animal.

Figure 7. Plasma concentrations of triglycerides and cholesterol (total, free and esterified) in rabbits fed with a normal diet.

Triglycerides and cholesterol were assayed in the plasma, and in three lipoproteic compartments separated by ultracentrifugation. Blood samples were collected in rabbits (6 wild-type and 7 transgenic rabbits from line shL21) fed with a normal diet and starved for 20 hours (white bars) and 4 hours after re-feeding with the normal diet (black bars). Values are given in mg/ml, with the standard error of the mean. Comparisons were made between starved and fed animals within each group (** = p<0.001).

Figure 8. Plasma concentration of triglycerides and cholesterol in rabbits fed with a high fat/high cholesterol regimen.

Rabbits (4 wild type, and 3 transgenic rabbits from line shL21) were fed for 8 days with a high fat/high cholesterol diet. Plasma samples were collected before the diet (D0, white bars), after feeding for 8 day with the diet (D8, black bars), after 20 hours starvation (D9 starved, grey bars) and 4 hours after re-feeding with the high fat diet (D9 fed, dotted bars). Triglycerides and cholesterol were assayed as in Figure 7. Values are given in mg/ml, with the standard error of the mean. Comparisons were made between transgenic and wild type animals for each day of the challenge (* = p<0.05).

Figure 9. Total content of body lipids and growth curves of wild type and transgenic rabbits from lines shL21 and shL27.

The total content of body lipids and growth curves were established on the same rabbits. All rabbits (mothers during pregnancy and lactation and their litters after weaning) were fed with the normal diet. Wild type mothers nourished all newborns (transgenic or wild type ones). The total content of body lipids, expressed as the percentage of the body weight, was measured in transgenic (shL21 and shL27, black bars) and wild type (white bars) rabbits at around 12–16 weeks after birth. Three animals at least were considered for each point. Values are means +/− sem. Note that the percentage was always the lowest in shRNA expressing animals, and the highest in wild type animals. Growth curves were established by weighing weekly each rabbit from 3–5 weeks to 12–18 weeks after birth. Males and females are shown in separate graphs. * = p<0.05 comparison of shRNA expressing animals and wild type ones.

Figure 10. Analysis of rIFABP-hapobec1 transgenic and double transgenic rabbits.

10A: Structure of the recombinant gene to express the human APOBEC1 cDNA in the intestine of transgenic rabbits. The rIFABP-hAPOBEC1 construct encompassed two copies of the chicken ß-GLOBIN gene fragment 5′HS4 (gene expression insulator element, dotted box), the promoter of the rat intestinal fatty acid binding protein gene (rIFABP; grey box), the rabbit (rb) ß-GLOBIN second intron (black boxes and thick line), the human APOBEC1 cDNA (white box) produced by PCR amplification from reverse transcribed RNA of HT29 cells (derived from a human colon tumor that have retained the ability to express the APOBEC1 gene), and the human growth hormone polyadenylation sequences (box with vertical bars). The horizontal black arrow points the position of the transcription start site. ATG =  translation initiation site of hAPOBEC1 cDNA. Numbers and small horizontal arrows represent the sets of primers. All studied transgenic animals were PCR positive for the sets 1–4. 10 B: gene expression. The levels of human APOBEC1 mRNA (left panel) and shRNA (middle panel) were measured in RNAs prepared from duodenum cells in two rIFABP-hapobec1 lines (L01 and L02) and in double transgenic animals (shL21+L01; shL21+L02; shL27+L01). Values are given in females (F) and males (M) after normalization to the level of reference gene expression determined simultaneously in each sample: Let7c miRNA in the case of shRNA, and RPL19, YHWAZ, HPRT in the case of human APOBEC1. In double transgenic animals, males and females were not distinguished, considering the small number of animals in these groups. The number of animals in each group is indicated in brackets. Values are given with the standard error of the mean (sem). The mean level of shRNA in shL21 and shL27 as presented in Figure 2 is indicated with a horizontal line. The level of expression of the human APOBEC1 transgene measured in the liver is given in L02. In L01, this level was not significantly detected. The level of rabbit APOBEC1 mRNA (right panel) was measured in intestinal RNAs as described in Figure 3. The level found in wild type rabbits and in lines shL21 and shL27 is indicated with a horizontal line. 10 C: APOBEC1 dependent editing in intestine and liver and Plasma concentration of APOB48 in rabbits expressing the human APOBEC1 gene. The estimation of editing was made as described in the legend of Figure 5. Plasma concentration of APOB48 was performed as described in legend of Figure 6 in 3 transgenic rabbits from line L02. Wild type animals are the same than those in Figure 6.

Figure 11. Total mass of body lipids and growth curves in double transgenic rabbits.

Double transgenic animals (shL21+L01; shL21+L02; shL27+L01) were produced by breeding rIFABP-hAPOBEC1 (L01 or L02) and rbapobec1-shRNA transgenic lines (shL21 or shL27). In these litters, the total mass of lipids was significantly lower in shL21 or shL27 transgenic animals than in animals from all other groups (* = p<0.05). Numbers in brackets indicate the number of animals in each group. Growth curves were established by weighing weekly each rabbit from 3–5 weeks to 12–18 weeks after birth. Males and females are shown in separate graphs. * = p<0.05 comparison of shRNA expressing animals and wild type ones.

Figure 12. Intestinal and liver regulation of APOB48, APOB100, and chylomicron production.

Four schematic representations are given, to simulate the regulation of APOB mRNA editing and the consequence upon the phenotype in wild type and transgenic rabbits. The models depict the situation after a diet challenge with normal of high fat/high cholesterol diet. Blue characters are used for the transgene expression (shRNA targeting the rabbit APOBEC1 mRNA, and human APOBEC1 gene); red characters indicate the measured parameters with significant modifications; the size of the letters is related to the level of production.