A novel approach to oral apoA-I mimetic therapy

Abstract

Transgenic tomato plants were constructed with an empty vector (EV) or a vector expressing an apoA-I mimetic peptide, 6F. EV or 6F tomatoes were harvested, lyophilized, ground into powder, added to Western diet (WD) at 2.2% by weight, and fed to LDL receptor-null (LDLR(-/-)) mice at 45 mg/kg/day 6F. After 13 weeks, the percent of the aorta with lesions was 4.1 ± 4%, 3.3 ± 2.4%, and 1.9 ± 1.4% for WD, WD + EV, and WD + 6F, respectively (WD + 6F vs. WD, P = 0.0134; WD + 6F vs. WD + EV, P = 0.0386; WD + EV vs. WD, not significant). While body weight did not differ, plasma serum amyloid A (SAA), total cholesterol, triglycerides, and lysophosphatidic acid (LPA) levels were less in WD + 6F mice; P < 0.0295. HDL cholesterol and paroxonase-1 activity (PON) were higher in WD + 6F mice (P = 0.0055 and P = 0.0254, respectively), but not in WD + EV mice. Plasma SAA, total cholesterol, triglycerides, LPA, and 15-hydroxyeicosatetraenoic acid (HETE) levels positively correlated with lesions (P < 0.0001); HDL cholesterol and PON were inversely correlated (P < 0.0001). After feeding WD + 6F: i) intact 6F was detected in small intestine (but not in plasma); ii) small intestine LPA was decreased compared with WD + EV (P < 0.0469); and iii) small intestine LPA 18:2 positively correlated with the percent of the aorta with lesions (P < 0.0179). These data suggest that 6F acts in the small intestine and provides a novel approach to oral apoA-I mimetic therapy.

Fig. 1.

A schematic diagram of the full length pBI121-6F cassette. The top panel depicts the pBI121 vector, which is referred to in this manuscript as the EV. The bottom panel shows the vector for expressing the 6F peptide in which the GUS gene has been replaced at the XbaI/SacI site by the plant-derived signal peptide (SP) and the gene encoding 6F under the CaMV35S promoter and nopaline synthase terminator (NOS term) as described in Materials and Methods. The NPT II gene conferring kanamycin resistance is under the control of nopaline synthase promoter (NOS pro).

Fig. 2.

The 6F peptide synthesized from all L-amino acids without end blocking groups (L-6F) when fed to apoE^−/− mice significantly reduced plasma SAA levels and decreased the percent of the aorta with atherosclerotic lesions. A: Groups of female apoE^−/− mice (n = 20) 16–18 months of age were maintained on rodent chow that did not contain peptide (Chow) or contained 1.2 mg of the 6F peptide without end blocking groups per 4 g of chow (Chow + 6F) providing a dose of ∼60 mg/kg/day peptide. The mice in both groups consumed approximately 4 g of the chow per mouse per day. The peptide constituted ∼0.03% of the diet by weight. After 10 days the mice were bled and SAA levels were determined by ELISA as described in Materials and Methods. B: Groups of female apoE^−/− mice 4–6 months of age were fed WD that did not contain or which contained L-6F without end blocking groups (WD + 6F) at a dose of 60 mg/kg/day of peptide. After 6 weeks the mice were bled and plasma SAA levels were determined by ELISA as described in Materials and Methods. C: Groups of female apoE^−/− mice 6–8 months of age (n = 30 per group) were fed WD that did not contain peptide (WD No Peptide) or that contained L-6F without end blocking groups (WD + 6F) at a dose of 60 mg/kg/day of peptide. After seven weeks the percent of the aorta with atherosclerotic lesions was determined by en face analysis as described in Materials and Methods.

Fig. 3.

SDS-PAGE gels from most (but not all) tomato lines that were positive for the 6F gene contained a band that migrated with authentic chemically synthesized 6F without end blocking groups. Proteins were extracted from the tomatoes derived from plant lines shown in the figure. All tomato lines shown except WT were PCR positive for the 6F gene. Protein from each tomato line (100 μg) was added to each lane and run on SDS-PAGE gels using the protocol described in Materials and Methods.

Fig. 4.

Regions of SDS-PAGE gels containing a band migrating with authentic 6F demonstrated the LC-ESI-MS signature of 6F (A) while the same region on gels without a band did not (B). Following HPLC and SDS-PAGE, the region on each lane corresponding to 6F in the inset was excised, in-gel trypsin digested, and subjected to LC-ESI-MS analysis using an LCQAdvantage Max ion trap mass spectrometer (ThermoElectron, Inc.) equipped with electrospray ionization source as described in Materials and Methods. Figure 4A demonstrates that the bands migrating similarly to authentic 6F (arrow in the inset) exhibit the ESI-MS signature for 6F while the same region from those lanes without bands in this region of the gel (arrow in the inset) did not (Fig. 4B). In the insets: EV, empty vector tomato line; 1A, tomato line transgenic for 6F; M, molecular marker.

Fig. 5.

Feeding transgenic 6F tomatoes to LDLR^−/− mice for two weeks improved a number of plasma biomarkers. Female LDLR^−/− mice 10 weeks of age were housed four in each cage and each cage was given each night compacted WD containing no lyophilized tomatoes (n = 20), compacted WD containing 2.2% by weight ground lyophilized WT tomatoes (n = 8), or compacted WD containing 2.2% by weight ground lyophilized transgenic 6F (Tg6F) tomatoes. The mice in all cages ate all of the diet each night. The mice receiving the Tg6F tomatoes received 800 μg of 6F per mouse per day (40 mg/kg/day). After two weeks the mice were bled and the following plasma biomarkers were measured: SAA (A); PON (B); LPA 16:0 (C); LPA 18:0 (D); LPA 18:1 (E); LPA 20:4 (F); free 5-HETE levels (G); free 15-HETE levels (H); free PGD2 levels (I); free PGE2 levels (J); and HDL cholesterol (K). Measurements were made as described in Materials and Methods (LPA levels were determined by LC-ESI-MS/MS).

Fig. 6.

Lycopene content of transgenic 6F tomatoes was slightly but significantly less than that of WT or empty vector tomatoes. The lycopene content of ripened tomatoes that were WT, transgenic for 6F (1A; 17A), or empty vector (108; 110) was determined as described in Materials and Methods.

Fig. 7.

Feeding transgenic 6F tomatoes to LDLR^−/− mice for 13 weeks improved a number of plasma biomarkers. Female LDLR^−/− mice 7–9 months of age were housed four in each cage and each cage was given each night compacted WD containing no lyophilized tomatoes (n = 28), compacted WD containing 2.2% by weight ground lyophilized EV tomatoes (from line 110) (n = 20), or compacted WD containing 2.2% by weight ground lyophilized transgenic 6F (Tg6F) tomatoes (from line 17A) to provide 900 μg of 6F per mouse per day (45 mg/kg/day). The mice in all cages ate all of the diet each night. After 13 weeks the mice were bled and the following measurements were made as described in Materials and Methods: SAA (A); total cholesterol (B); triglycerides (C); PON (D); HDL cholesterol (E); LPA 18:1 (F); LPA 18:2 (G); LPA 20:4 (H); and body weight (I).

Fig. 8.

Feeding EV and transgenic 6F tomatoes decreased some biomarkers and increased others. The plasma from the mice described in Fig. 7 was analyzed as described in Materials and Methods for: free arachidonic acid (A); free 5-HETE (B); free 15-HETE (C); free DHA (D); and free EPA (E).

Fig. 9.

Feeding transgenic 6F tomatoes but not EV tomatoes significantly decreased the percent of the aorta with atherosclerotic lesions. The aortas from the mice described in Fig. 7 were harvested and analyzed to determine the percent of the aorta with atherosclerotic lesions as described in Materials and Methods. The aorta from one of the mice fed transgenic 6F was severely damaged during the harvest and was not processed. The aortas from all other mice were successfully harvested, processed, and analyzed.

Fig. 10.

The percent of the aorta with atherosclerotic lesions in mice receiving WD and transgenic 6F tomatoes was positively and significantly correlated with plasma total cholesterol and triglycerides, and was significantly and inversely correlated with PON activity and HDL cholesterol; there was no correlation with body weight. Linear regression of data from individual mice described in Figs. 7–9 that received WD and transgenic 6F tomatoes is shown for the percent of the aorta with atherosclerotic lesions and plasma total cholesterol (A); plasma triglycerides (B); plasma PON activity (C); plasma HDL-cholesterol levels (D); and body weight (E).

Fig. 11.

Addition of transgenic 6F tomatoes (Tg6F) to the WD significantly reduced the levels of LPA in the small intestine, while addition of the EV tomatoes did not. The levels of LPA 18:2 and LPA 20:4 were determined by LC-ESI-MS/MS in a random subset of the mice described in Figs. 7–9 as described in Materials and Methods. A: LPA 18:2 in the duodenum. B: LPA 20:4 in the duodenum. C: LPA 18:2 in the jejunum. D: LPA 20:4 in the jejunum. E: LPA 18:2 in the ileum. F: LPA 20:4 in the ileum.

Fig. 12.

The levels of LPA in the small intestine significantly correlated with the percent of the aorta with atherosclerotic lesions. The levels of LPA in the small intestine of the mice described in Fig. 11 were plotted against the percent of the aorta with lesions for each mouse, and linear regression was performed as described in Materials and Methods. A: LPA 18:2 duodenum. B: LPA 20:4 duodenum. C: LPA 18:2 jejunum. D: LPA 20:4 jejunum. E: LPA 18:2 ileum. F: LPA 20:4 ileum.