Regulation of Surfactant-Associated Phospholipid Synthesis and Secretion

Introductory Remarks on Mammalian Surfactant Phospholipid Analysis and Functions

Application of the Young-Laplace Equation (1805) by van Neergard in 1929 substantially promoted the understanding of lung physiology, showing that pulmonary retraction is primarily based on surface tension. It took three decades to emphasize its clinical relevance and start the exploration of pulmonary surface active agent (“surfactant”). ^, 1,2-Dipalmitoyl-glycero-3-phosphocholine, also named 1,2-dipalmitoyl-phosphatidylcholine , DPPC, or PC16:0/16:0 (the term used in this chapter), was found critical to lower surface tension. PC16:0/16:0 is a zwitterionic glycerophospholipid with two straight saturated fatty acids, and a phase transition temperature of 41.5°C, optimal to achieve near-zero minimal surface tension upon lateral compression of air-liquid interfaces. Since then, our knowledge about surfactant function, structure, molecular composition, and metabolism has greatly increased. Advancements in phospholipid analysis significantly contributed to this progress.

Initially, phospholipid classes of lung tissue and lung lavage fluid (LLF) extracts were separated by column or thin layer chromatography, followed by gas-chromatographic fatty acid analysis. Techniques of lung tissue explants and culture of type II pneumocytes (also known as alveolar epithelial type 2 cells [AECII] ) were developed. An approximative technique for quantification of PC16:0/16:0, named disaturated PC (DSPC) , was introduced, based on the removal of unsaturated phospholipids with osmium tetroxide (OsO ₄ ). However, Holm and colleagues showed in 1996 that this OsO ₄ technique does not reliably remove monounsaturated PC from samples. Moreover, monounsaturated palmitoyl-palmitoleoyl-PC (PC16:0/16:1) was found to be specific for mammalian surfactant. ^, Irrespective of such lack of validity, DSPC quantification is still in use. ^, However, techniques to quantify intact PC molecular species were developed, using high-performance liquid chromatography. ^, Finally, introduction of tandem mass spectrometry and stable isotope labeled precursors allowed for detailed insight into surfactant phospholipid metabolism in patients ( Fig. 77.1 ).

Surfactant Composition In Relation to Function

Glycerophospholipids account for approximately 80% of surfactant. Cholesterol comprises approximately 10% in mammals, but its concentration can rapidly change in hibernating mammals to adapt surfactant function to body temperature. Its origin from lipoproteins suggests that cholesterol may be essential to the developing lung. Glycerophospholipids comprise a charged polar head group and a hydrophobic “tail” of two fatty acyl residues, ideal to form monolayers or multilayers at air-liquid interfaces, thereby reducing surface tension.

The need of low surface tension in terminal pulmonary air spaces is given by the Young-Laplace equation

for any geometry, where p is pressure, γ is surface tension, and r ₁ and r ₂ are the radii of curvature. When r ₁ = r ₂ (alveolus), p = 2γ/ r . Consequently, when the alveolar radius decreases at end-expiration, high surface tension would increase the pressure forces required to resist effects of surface tension and cause alveolar collapse. In a tubule (bronchiole), r ₂ approximates ∞, resulting in p = γ/ r . Low surface tension is therefore essential to stabilize air-liquid interfaces of alveoli and bronchioles and of tubules and saccules of immature lungs of preterm infants as well ( Fig. 77.2 ). Notably, alveolarization and AECII differentiation are not synchronized: mammals and birds are born with a functioning surfactant system, whereas alveolarization proceeds from fetal age to childhood in humans, is completed in utero in Guinea pigs, starts 3 to 4 days after birth in small rodents, and does not occur in birds possessing air capillaries.

Due to their rapid spreading on surfaces, surfactant phospholipids cover the air-liquid interface throughout the lung. Airways surfactant, from alveolar overspill, accounts for maximally 7% of surfactant turnover, which is ample with respect to small airway compared with alveolar surface. ^{, ,}

Phospholipids and Phosphatidylcholine Species in Surfactant

PC comprises 80% of surfactant phospholipid, with a molecular composition different from that of other phospholipid-containing secretions. ^, In humans, it is mostly diacyl-PC, whereas alkyl-acyl-PCs are abundant in shrews, bats, and marsupials. Anionic phospholipids, mostly phosphatidylglycerol (PG) and some phosphatidylinositol (PI), represent 10% of surfactant, with reciprocal relationship of their fractions. PG is a marker of human lung maturity, whereas in some mammals the switch from PI to PG occurs after term. However, contrary to lung tissue, surfactant comprises little sphingomyelin—a sphingolipid containing a phosphocholine head group like PC—and phosphatidylethanolamine (PE), two major plasma membrane phospholipids (approximately 10% and ∼25%, respectively).

Surfactant PC composition changes during development. Surfactant assembly is characterized by sorting processes according to fatty acyl chain length, with specific enrichment of PC16:0/16:0, palmitoyl-myristoyl-PC (PC16:0/14:0), myristoyl-palmitoyl-PC (PC14:0/16:0), and palmitoyl-palmitoleoyl-PC (PC16:0/16:1). Together they comprise approximately 80% of the surfactant PC, and PC16:0/16:0 is inversely related to the fraction of these others. PC16:0/16:0 maximally comprises 50% of surfactant PC (e.g., one third of surfactant). Only in birds, with no lung surface area changes during respiration, PC16:0/16:0 is higher (approximately 75%). In several mammals, particularly those with a high resting respiratory rate, PC16:0/16:0, including its alkyl-acyl-analogue, comprises only 20% of PC, providing good surface tension function ( Fig. 77.3 ). This corrupts the paradigm that PC16:0/16:0 must predominate in mammalian surfactant. ^{, , ,}

During alveolarization, PC16:0/14:0 increases at the expense of PC16:0/16:0 and is a better predictor of human fetal lung maturity than PC16:0/16:0. Aside from the contribution of these components to surface tension function under dynamic conditions, PC16:0/14:0 modulates macrophage functions, possibly relevant to alveolar protection. In essence, mature human surfactant comprises 33% PC16:0/16:0, 15% of the other specific PC species, and 10% to 15% anionic phospholipids. The rest are other fluidic lipids and surfactant proteins (SPs; Fig. 77.4 ).

PC16:0/16:0 and PG are present in other organs such as the brain. On the other hand, AECII and lung explants switch from PC16:0/16:0 synthesis to that of PC16:0/14:0 and PC16:0/16:1. In spite of using hormone and second messenger analogues in AECII cultures, lipid profiles remain different from in vivo conditions. Hence only in vivo conditions reflect physiologic surfactant PC metabolism, in a defined species and at a defined developmental stage.

Pathways of Pulmonary Phospholipid Synthesis

Fig. 77.5 illustrates the pathways of PC, PG, and PI biosynthesis. Synthesis starts with the formation of phosphatidic acid (PA) from dihydroxyacetone phosphate (DHAP), which is derived from glycolysis. Two sequential acylations, using activated fatty acids (acyl-coenzyme A [CoA]), are required. DHAP can be 1-acylated followed by reduction or 1-acylated after reduction to glycerol 3-phosphate. ^, Acylation in position 2 then ends up in PA, the precursor of all other glycerophospholipids (and triglycerides).

Numerous precursors are used for PA synthesis, allowing for their stable isotope labeling to investigate surfactant phospholipid metabolism in patients (see Fig. 77.1 ): glucose, from blood or pulmonary glycogen stores, is used for the glycerol backbone and for fatty acid synthesis. Glycerol is incorporated after phosphorylation to glycerol-3-phosphate. Fatty acids originate from extrapulmonary and intrapulmonary lipids, fatty acid synthesis, and their desaturation or elongation. There is a wide range of experimental options, from deuterated water to fatty acids to address fatty acid synthesis, elongation, and desaturation in AECII. ^,

For PC biosynthesis, PA is dephosphorylated to 1,2-diacylglycerol (DAG; see Fig. 77.5 ). To transfer choline phosphate to DAG, choline is phosphorylated and then activated to cytidine-diphosphocholine (CDP-choline). For this, cytidine monophosphate is transferred from cytidine triphosphate (CTP) to phosphocholine by choline-phosphate cytidylyltransferase (CCT). The highly energetic diphosphoanhydride bond of CDP-choline drives the transfer of phosphocholine to DAG towards PC synthesis.

Although PC is primarily synthesized de novo, acyl remodeling contributes to approximately 50% of surfactant-specific PC species (see Fig. 77.5 ): deacylation of 1-palmitoyl-2-unsaturated-PC by phospholipase A ₂ results in 1-palmitoyl-2-lyso-PC. Lyso-PC: acyl-CoA acyltransferase 1 (LPCAT-1) then reacylates lyso-PC to PC, mainly PC16:0/16:0 and PC16:0/14:0. ^{, ,} Consequently, absence of the remodeling pathway increases oleated and polyunsaturated PC in surfactant. ^,

Sorting mechanisms with a preference for PC containing two fatty acids of 14 or 16 carbon units assure the enrichment of these components in surfactant ^, but are not absolutely specific. In infants and newborn rats, surfactant comprises approximately 5% to 10% PC containing an arachidonic or docosahexaenoic acid residue. Notably, high contents of polyunsaturated fatty acids in surfactant correlates inversely with development of bronchopulmonary dysplasia (BPD) in preterm infants. ^, The responsible adenosine triphosphate (ATP)-binding cassette transporter A3 (ABCA3), located at the outer membrane of lamellar bodies (LBs), is regulated by glucocorticoids and triggers surfactant-PC and PG enrichment in LB of AECII. Consequently, several, but not all, ABCA3 mutations result in the absence of normal LB and acute neonatal respiratory failure of term infants. ^, In addition, ABCA3 knockout decreases the expression of proteins involved in lipid synthesis and trafficking, including LPCAT1, ABCA1, and steroidogenic acute regulatory domain protein 2 (stard2), highlighting the complexity of factors contributing to pulmonary PC metabolism. ^,

In contrast to PC, for PG and PI synthesis the phosphate group of PA is activated by CTP, resulting in CDP-DAG (see Fig. 77.5 ). Transfer of inositol or glycerol-3-phosphate to this intermediate yields PI, or PG-3-phosphate, which is immediately dephosphorylated to PG. PG and PI are enriched in unsaturated species and do not contain myristic acid, with the exception of rodent surfactant comprising 20% to 30% PG16:0/16:0. Differences in molecular species composition suggest different DAG/PA pools or different selection principles in trafficking. ^,

Normal Development of the Surfactant System

During fetal development, PC and PG increase at the expense of other glycerophospholipids and sphingomyelin. ^, PC16:0/14:0 and PC16:0/16:1 continuously increase, whereas the fraction of PC16:0/16:0 decreases from 34 weeks postmenstrual age onwards (i.e., when lung alveolarization has started). As the fetal lung secretes water, surfactant is carried out from the lung into amniotic fluid. Gastric and upper airway samples of newborns are similarly useful to characterize surfactant composition and function, as predictors of neonatal respiratory distress development. Increases in PG, in the ratio between PC (“lecithin”) and sphingomyelin (L/S ratio), in PC16:0/14:0, PC16:0/16:1, and the PC16:0/14:0 to PC16:0/16:0 ratio are specific predictors of human lung maturity. D9-choline labeling showed that airway samples reflect alveolar PC composition and metabolism, whereas inhibition after meconium aspiration is not necessarily reflected by altered metabolism. ^, However, none of these functional or biochemical analyses was tested in randomized trials against clinical practice, of prophylactic surfactant application in very low-birth-weight preterm infants, or according to clinical status with measurement of oxygen saturation of hemoglobin, pressures required for adequate ventilation, and X-ray diagnostics.

Prenatal increase of surfactant goes hand in hand with increased PC synthesis, due to lung parenchymal growth and AECII proliferation. However, lung PC has a rapid turnover, and a major fraction is secreted into the circulation via basolateral ABCA1 transporters rather than into the airspaces, ^, thereby contributing to lipoprotein homeostasis and the choline/PC shuttle between liver and lung. Consequently, PC synthesis of the lungs and AECII does not simply represent surfactant metabolism but must be regarded in a systemic context. ^{, ,} Moreover, while incorporation of labeled precursors has been used to assess surfactant PC synthesis and secretion in humans, ^{, ,} isotope enrichment of precursors at the moment of synthesis and analysis of individual molecular species with their different turnover is necessary to assess synthesis rates.

Regulation of Fetal Lung Phosphatidylcholine Biosynthesis by Enzyme Activity and Substrate Availability

The critical enzymes for PC synthesis de novo are choline kinase (CK) and CCT (see Fig. 77.5 ). CCT catalyzes the synthesis of CDP-choline. All CCT isoforms (α, β ₁ , and β ₂ ) are expressed in fetal lung. The β isoforms are splice variants of the same gene. CCTα, the only isoform relevant to surfactant development and expressed in adult lung, is encoded by a different gene ( Pcyt1a ) and is found in association with cytosol and endoplasmic reticulum in the lung. Increased CCT expression during development has become paradigmatic to explain for surfactant-PC enrichment. However, this is questionable because mice overexpressing lung CCT four-fold have only modest or no increases in DSPC synthesis. CCT activity increases during developmental lung growth and cell proliferation, whereas choline incorporation into PC remains unchanged relative to tissue weight. Moreover, accumulation of PC16:0/16:0 and PC16:0/14:0 at late gestation is due to decreased turnover and sequestration into LB rather than increased synthesis. ^,

Sphingosine and fibroblast growth factor 7 (FGF-7) enhance gene expression and/or activation of existing enzyme, whereas expression is inhibited by transforming growth factor-β1 (TGF-β1) and Zn deficiency. ^{, , ,} However, CCT activity is disconnected from gene expression because the lipid-free cytosolic form is inactive, whereas translocation to the endoplasmic reticulum or other surrounding lipid (e.g., from fatty acid synthesis) activates the enzyme. Curvature and physical properties of lipids are critical to CCT activation. Notably, CCT stimulation by glucocorticoids, estrogens, progesterone, and thyroid hormones does not result from translocation but from increased fatty acid synthesis and presence. This alone results in maximal activation, so that hormones do not cause additional stimulation, whereas lipid removal abolishes any stimulatory effect.

However, factors beyond CCT may be similarly important. CK ( K _m = 13 μmol/L) for de novo PC synthesis is regulatory in other organs. Lung tissue depends on exogenous choline supply, as choline synthesis by PE- N -methyltransferase (PEMT) is virtually absent from lung tissue and hepatic activity is low in the fetus and preterm infant. The ubiquitous choline transporters have K _m values of greater than 30 μmol/L. Consequently, plasma choline concentration affects PC synthesis via cellular uptake, which may apply to the lungs as well. Notably, fetal plasma choline is approximately 40 μmol/L, which is three- to four-fold over adult values. Plasma PC concentration rapidly and untimely decreases to 20 μmol/L or less after preterm delivery. ^, Moreover, in choline deficiency PC is recruited from the lungs to supply the liver, possibly impacting on pulmonary PC and surfactant homeostasis.

Fatty acid availability is similarly essential to lung tissue and surfactant PC synthesis as that of choline. Three sources of fatty acids are available: synthesis from carbohydrates via fatty acid synthase (FAS) within AECII, supply from pulmonary neutral lipid stores, and uptake from plasma, namely free fatty acids and lipoprotein lipids. FAS expression is induced by glucocorticoids in fetal lungs but is down-regulated in the absence of ABCA3. Notably, although retinoic acid (RA) is essential to lung maturation, it antagonizes FAS expression by glucocorticoids. Increased surfactant PC pools in neonatal lungs after FGF-7 treatment go in line with increased pulmonary adipose triglyceride lipase (ATGL) rather than FAS expression. ATGL is a ubiquitous lipase primarily acting on triglycerides, expressed in lungs together with hormone-sensitive lipase (HSL) and monoacylglycerol lipase (MGL). Incorporation of fatty acids into surfactant-PC, derived from pulmonary lipofibroblasts, plasma lipoproteins, and free fatty acids, as well as from intrapulmonary fatty acid elongation or desaturation, suggests that lung tissue and surfactant PC synthesis depends on both intrapulmonary synthesis and the specific use of other fatty acid sources. ^{, ,} In this context, myristic acid (C14:0) is enriched in the lungs, whereas lauric acid (C12:0) is elongated to C14:0 for PC16:0/14:0 synthesis. ^, These complex, timely adjusted mechanisms may be corrupted, for example, by lipofibroblast to myofibroblast transdifferentiation by hyperoxia. ^{, ,}