Maternal loading with very low-density lipoproteins stimulates fetal surfactant synthesis

Materials

Intermediate-size VLDL with a sedimentation flotation rate of 60–100 (from the density <1.006 g/ml fraction of plasma), heparin, and bovine serum albumin (BSA) were from Sigma (St. Louis, MO). Silica LK5D (0.25 mm × 20 cm × 20 cm) thin-layer chromatography (TLC) plates were purchased from Whatman International (Maidstone, UK). The MRE-040 Micro-Rethane tubing was from Braintree Scientific (Braintree, MA). All radiochemicals were purchased from DuPont New England Nuclear Chemicals (Boston, MA). The radiolabeled tripalmitin [palmitic-1-¹⁴C] (specific activity, 55 mCi/mmol) was obtained from American Radiolabeled Chemicals (St. Louis, MO). Immunoblotting membranes were obtained from Millipore (Bedford, MA). The enhanced chemiluminescence (ECL) Western blotting detection system was from Amersham Pharmacia Biotech (Piscataway, NJ). Anti-CCT rabbit polyclonal antibody to synthetic peptide corresponding to residues 164–176 (22) was generated by Covance Research Products (Richmond, CA). Fatty acid standards were obtained from Nu Chek Prep (Elysian, MN) and Supelco (Bellefonte, PA). The free fatty acids, half-micro test was from Roche Diagnostics (Mannheim, Germany). Polyclonal antibody against LPL was generated in the chicken-against-bovine LPL antigen. The Harvard infusion pump was purchased from Harvard Apparatus (Holliston, MA). The C57BL/6J mice, LDLR knockout mice, and apoprotein E (ApoE) knockout mice were obtained from Jackson Labs (Bar Harbor, ME).

Animals and cell preparation

Timed-pregnant Sprague-Dawley rats at days 17–20 of gestation (day = 0, designated by presence of vaginal sperm plug) were obtained from Harlan Sprague Dawley (Indianapolis, IN). Animals were deeply anesthetized with ketamine (67 mg/kg ip) and xylazine (6.7 mg/kg ip). The anterior cervical area was shaved and prepped with betadine ointment, and under blunt dissection, the left external jugular vein was identified. The distal end of the vein was ligated with 4-0 silk suture, and the proximal visible end was cannulated with 22-gauge MRE-040 tubing. The suture was removed, and we confirmed catheter patency by flushing the catheter with 0.25 ml of heparinized saline. The surgical site was closed with 4-0 silk suture. The pregnant rats were monitored and allowed to recover overnight and fed ad libitum. The next morning, animals were administered BSA in normal saline (1 mg), heparin (2–10 units), VLDL (0.5–3.0 mg/ml), or heparin (2–10 units) in combination with VLDL (1 mg) intravenously over 3 h via a Harvard infusion pump. After infusions were terminated, animals were euthanized with phenobarbital (50 mg/kg ip), and fetal rats were delivered by Cesarean section from their dams. The fetal lungs were resected, pooled, and placed in calcium- and magnesium-free Hanks’ balanced salt solution (HBSS) and subsequently used to isolate fetal pre-type II alveolar epithelial cells as previously described (33). The purity of isolates was >85% as determined by Maclura pomifera lectin binding. With regard to the mice, animals were fed ad libitum and euthanized as above. The tracheas were cannulated with 0.040-diameter plastic tubing and lavaged, and crude alveolar surfactant pellets were isolated as described (42). Lung supernatants were obtained by sequential centrifugation (27). These animal procedures were approved by the University of Iowa Animal Care and Use Committee.

Radiotripalmitination of lipoproteins

VLDL from normal human volunteers (protein concentration, 1.456 mg/ml) were radiolabeled with [¹⁴C] tripalmitin as described (34). For triglyceride labeling, 75 µCi of [¹⁴C]tripalmitin were dried under nitrogen gas, resuspended in medium containing 4 mg/ml BSA, and sonicated in a water bath for 30 min at 25°C. VLDL particles (2 mg protein) were added to the [¹⁴C]tripalmitin solution, incubated at 37°C for 20 min, and then returned to room temperature. The final concentration of radiotripalmitinated VLDL was 75.3 × 10⁶ disintegrations·min⁻¹·ng protein ⁻¹.

Phospholipid and DSPtdCho analysis

Lipids were extracted by the method of Bligh and Dyer (7). Lipids dried under nitrogen gas were spotted on silica LK5D plates and resolved in chloroform/methanol/petroleum ether/acetic acid/ boric acid [40:20:30:10:1.8, vol/vol (13)] for resolution of phospholipids or chloroform/methanol/7 M ammonium hydroxide [65:35:5, vol/vol/vol (31)] for DSPtdCho analysis. Samples that comigrated with phospholipid standards as detected by iodine exposure were scraped from the silica gel. In the case of PtdCho, lipids were reacted with osmium tetroxide before a run in the second dimension to isolate DSPtdCho (29). For individual phospholipid determination, cellular PtdCho was not consistently resolved from phosphatidylinositol (PI), and thus these lipids were quantitated and reported together. Phospholipid content was quantitated by a phosphorus assay (9). In selected studies, spots corresponding to PtdCho were scraped from the gel and quantitated directly by scintillation counting.

Enzyme assays

The activities of choline kinase and choline-phosphotransferase were assayed exactly as described (28). We determined the activity of CCT by measuring the rate of incorporation of [methyl-¹⁴C]phosphocholine into CDP-cho-line by a charcoal extraction method (29). Assays were performed without the inclusion of a lipid activator, PtdChooleic acid, in the reaction mixture. All activities were assayed in total cellular lysates.

Immunoblot analysis

For immunoblot analysis, equal amounts of protein from pre-type II cell lysates were used. Each sample was adjusted to give a final solution of 60 mM Tris·HCl, pH 6.8, 2% SDS, 10% glycerol, 0.1% bromphenol blue, and 5% mercaptoethanol, heated at 100°C for 5 min, electrophoresed through a 10% SDS-polyacrylamide gel, and transferred to polyvinylidene difluoride membrane. We probed for CCT or LPL mass with the ECL Western blotting detection system. The dilution factor for anti-CCT and anti-LPL antibodies was 1:1,000.

Fatty acid analysis

Lipids were extracted from cell lysates (1 mg protein) according to the method of Bligh and Dyer (7). Fatty acid methyl ester (FAME) was prepared by transmethylation in the presence of 10% boron trifluoride. The FAME derivatives were separated by gas liquid chromatography and detected by flame ionization. The gas chromatography column packing was 10% SP-2330 on 100/120 Chromosorb W AW (Supelco). The initial column temperature was held at 165°C for 8 min and then increased at a rate of 3°C/min to a final temperature of 210°C, which was maintained for 12 min. We identified individual fatty acids by comparing the retention times with those of known standards. Free fatty acid content was determined by a free fatty acids, half-micro test per the manufacturer’s instructions with palmitic acid as a standard. Levels of fatty acids in extracts were within the linear range of testing.

Statistical analysis

The data are expressed as means ± SE. Unless otherwise stated, statistical analysis was performed by one-way ANOVA with the Bonferroni adjustment for multiple comparisons (41).

Distribution of radiolabeled VLDL

We first determined whether parenterally administered radiolabeled VLDL particles in the mother transfer across the placenta and become distributed systemically in fetal tissues. We were specifically interested in determining whether these lipoproteins accumulate in fetal lung. We also examined whether lipid components contained within lipoproteins serve directly as substrates used for fetal surfactant synthesis. To address this, we gave VLDL particles containing radiolabeled [¹⁴C]tripalmitin to the pregnant dams and measured radioactivity in various maternal and fetal organs and isolated alveolar type II epithelial cells (Fig. 1). In addition, we assayed radioactivity contained within the PtdCho fraction isolated from pre-type II cells, whole lung, and maternal lavage (Fig. 1, insets). As shown in Fig. 1, giving dams either radiolabeled VLDL or radiolabeled VLDL in combination with heparin resulted in variable levels of activity that was distributed systemically in both maternal and fetal organs. In the mother, the bulk of radioactivity was recovered in the lung and type II cells (Fig. 1A). This was an anticipated finding because radiolabeled VLDL was given via the external jugular vein. VLDL with heparin significantly increased activity in the lung compared with other maternal whole organs. In the fetus, [¹⁴C]tripalmitin VLDL particles were incorporated in several organs, including the fetal lung, although the highest activity tended to be observed in the fetal kidney with coad-ministration of heparin (Fig. 1B). Perhaps more importantly, radiolabeled [¹⁴C]tripalmitin particles were incorporated into the PtdCho fraction isolated from maternal and fetal type II cells and whole lung (Fig. 1 and Fig 2 and insets). Moreover, [¹⁴C]tripalmitin activity was readily detected within the PtdCho fraction isolated from adult surfactant in alveolar lavage (Fig. 1A, inset). Collectively, these results provide evidence supporting transplacental distribution of native VLDL or its remnant or lipid components, with distribution to the fetal lung, where they are incorporated into surfactant lipid within fetal type II cells. Some of these effects also appear to be accentuated by heparin.

Phospholipid analysis

The maternal administration of lipoproteins significantly elevated PtdCho and DSPtdCho mass in fetal pre-type II cells. In head-to-head studies with animals receiving saline-BSA, heparin, VLDL, or VLDL plus heparin, VLDL with heparin increased DSPtdCho content by 45% (P < 0.05 vs. control). Infusion of VLDL increased DSPtdCho levels by 30%, although these effects did not reach statistical significance (Fig. 2). In comparison, maternal treatment with heparin did not alter DSPtdCho content in the fetus compared with control (Fig. 2). In addition to testing for these effects of VLDL in day 20 fetuses, we infused lipoproteins in mothers earlier in gestation. Compared with results in gestationally matched controls, VLDL given maternally at days 17, 18, and 19 of gestation increased DSPtdCho levels by 11, 15, and 78%, respectively. Thus these results suggest that effects of lipoprotein loading on DSPtdCho are most pronounced during the canalicular phase (days 19–20) of lung development. VLDL elevated total PtdCho-PI content by ~54%, and it did not change the composition of other major phospholipids (Fig. 2, inset), nor did it alter surfactant apoprotein (surfactant protein A) levels in preliminary studies (data not shown). Together, these results indicate that prenatal treatment with VLDL selectively increases DSPtdCho content, the major surfactant lipid. These effects of VLDL are enhanced by concurrent infusion with heparin and are most apparent during the later stages of gestation.

Enzyme assays

We next pursued studies to determine whether maternal VLDL loading increases DSPtdCho content in fetal lung by altering the activity of enzymes of the CDP-choline pathway, the principle pathway for de novo synthesis of DSPtdCho. Maternal VLDL treatment increased CCT activity, the rate-limiting enzyme for DSPtdCho synthesis (Fig. 3). The administration of heparin or VLDL with heparin also significantly increased CCT activity. However, maternal VLDL administration increased enzyme activity by nearly three- to fourfold compared with control, and effects were significantly greater than effects of heparin alone or VLDL with heparin (P < 0.05). Effects of VLDL were dose dependent, reaching a plateau for maximal enzyme activation at about threefold between 1 and 3 mg of administered lipoprotein (Fig. 4A). Concurrent administration of heparin significantly increased VLDL activation of CCT above control at all doses tested (Fig. 4B). As with DSPtdCho, we tested VLDL effects on CCT activity at other stages of fetal development. VLDL tended to produce stimulatory effects at day 19 of gestation and significantly increased CCT activity in fetal pre-type II cells when administered to day 20 dams. VLDL did not alter enzyme activity when given to mothers at days 17 or 18 of gestation (Fig. 4C). Together, these results indicate that VLDL increases surfactant content by stimulating a key regulatory enzyme required for surfactant synthesis.

Finally, in contrast to effects on CCT, in three separate experiments VLDL particles had no significant effect on activities of choline kinase, the first committed enzyme of the CDP-choline pathway [49 ± 17 pmol·min⁻¹ ·mg⁻¹ (control) vs. 45 ± 5 pmol·min⁻¹ · mg⁻¹ (VLDL)], or cholinephosphotransferase, the terminal enzyme of the pathway [503 ± 131 pmol·min⁻¹ · mg⁻¹ (control) vs. 664 ± 204 pmol·min⁻¹ ·mg⁻¹ (VLDL)].

Immunoblot analysis

The observation that maternal lipoprotein administration stimulates CCT activity in fetal pre-type II cells led us to investigate whether this is secondary to an increase in the amount of enzyme (Fig. 5A). Overall, the levels of immunoreactive CCT remained unchanged compared with control in fetal pre-type II cells isolated from mothers given heparin or VLDL; VLDL in combination with heparin tended to increase levels of CCT mass, but these effects were not significant (Fig. 5, A and B). In contrast, immunoblot studies revealed that heparin in combination with VLDL increased LPL levels in maternal serum relative to VLDL alone without affecting levels in the fetus (Fig. 5C). These results suggest that VLDL-activation of CCT occurs by posttranslational mechanisms. Heparin, however, appears to regulate LPL activity by increasing its release into serum.

Mouse surfactant analysis

To further test our hypothesis that triglycerides carried within VLDL regulate surfactant synthesis, we assayed alveolar DSPtdCho levels and CCT activity in ApoE and LDLR knockout mice. We compared expression of these parameters with those in wild-type (C57BL/6J) controls. Unlike the wild-type and LDLR knockout mice, the ApoE knockout mice have increased VLDL-triglyceride levels (21, 39). The results show nearly two- and three-fold increases in CCT activity in ApoE knockout mice compared with wild-type controls and LDLR knockout mice, respectively (Fig. 6A). In addition, ApoE knockout mice exhibited a similar, approximately twofold increase in levels of alveolar DSPtdCho content compared with mice with normal serum triglyceride levels (Fig. 6B). Overall, these studies suggest that modification of some genes involved in lipoprotein catabolism results in hypertriglyceridemia; these elevated lipids in turn appear to be associated with altered surfactant synthesis.

Fatty acid analysis

Palmitic acid is esterified to the glycerol backbone of DSPdCho, whereas other unsaturated fatty acids are potent activators of CCT (12, 30). We determined whether maternal lipoprotein loading increases the availability of specific fatty acid species in pre-type II cells that could be used to synthesize surfactant. Treatment with VLDL and heparin increased the proportion of palmitic (16:0) and stearic (18:0) acids (P < 0.05) and tended to increase oleic acid (18:1, P = 0.09) compared with saline-BSA (Table 1). However, lipoprotein administration also decreased the relative amounts of the long-chain polyunsaturates, docosahexanoic (22:6) and arachidonic acids (20:4, P < 0.05 and P = 0.07, respectively). The results suggest that a predominant mechanism by which exogenous VLDL plus heparin increases surfactant lipid is by increasing the relative amounts of saturated species that are incorporated into the DSPdCho molecule. Maternal VLDL loading also produces a modest increase in oleic acid, a potent stimulator of CCT activity. We next extended our fatty acid analysis to wild-type and ApoE knockout mice (Table 2), the latter of which exhibit higher surfactant levels (Fig. 6). We observed that ApoE knockout mice compared with control mice had a significant increase in docosahexanoic (22:6) and arachidonic (20:4) acids vs. control (P < 0.05). These polyunsaturates are potent activators of CCT (31). Thus both exogenous VLDL loading in rats and elevated VLDL-triglycerides in mice harboring genetic defects in LDL catabolism exhibit increases in DSPdCho levels that may be due, in part, to the selective induction of key fatty acids required for CCT activation and surfactant lipid synthesis.

Finally, our present studies show that maternal VLDL treatment stimulates CCT activity to a greater extent than VLDL plus heparin (Fig. 3). VLDL with heparin, however, increases DSPtdCho levels more than VLDL alone in pre-type II cells (Fig. 2). One possible explanation for these results is that VLDL increases more free fatty acids that are available for CCT activation. The addition of heparin to VLDL might trigger hydrolysis of triglycerides, also releasing free fatty acids; however, these will be less abundant, resulting in lower CCT activity but higher phospholipid content as they are rapidly esterfied into the DSPtdCho molecule. To test this hypothesis, we assayed free fatty acids in the lung and observed that levels were 5.03 ± 0.9 nmol/mg protein in dams given VLDL alone vs. 2.42 ± 0.64 nmol/mg protein in dams administered VLDL plus heparin (P < 0.05, n = 6, Student’s t-test). These results suggest that higher levels of CCT activity induced by VLDL may be attributed to greater availability of free fatty acid activators, whereas after heparin treatment these fatty acids originating from VLDL are promptly acylated as phosphoglycerides resulting in higher DSPtdCho levels.