|Home | About | Journals | Submit | Contact Us | Français|
Prostanoids play a critical role in clinical areas such as inflammation, thrombosis, immune response and cancer. Although some studies suggest that there are genes that determine variability of some prostanoid-related phenotypes, the genetic influence on these traits has not been evaluated.
The relative contributions of genetic and environmental influences to the prostanoid biosynthetic pathway-related phenotypes, cyclooxygenase isoenzymes, microsomal-PGE-synthase-1 and TxA-synthase expression, and thromboxane-A2 and prostaglandin-E2 production by stimulated whole blood, were assessed in a sample of 308 individuals in 15 extended families. The effects of measured covariates (such as sex, age and smoking) genes, and environmental variables shared by members of a household were quantified. Heritabilities ranging from 0.406 to 0.634 for enzyme expression and from 0.283 to 0. 751 for prostanoid production were found.
These results demonstrate clearly the importance of genetic factors in determining variation in phenotypes that are components of the prostanoid biosynthetic pathways. The presence of such strong genetic effects suggest that it will be possible to localize previously unknown genes that influence quantitative variation in these phenotypes, some of which affect multiple aspects of cell biology, with important clinical implications.
Prostanoids include prostaglandins (PG) and thromboxanes (Tx) and belong to a remarkable group of compounds involved in a variety of clinically important areas such as inflammation, thrombosis, allergic and immune responses and cancer. Particularly, prostanoids are important mediators involved in vascular physiology and pathophysiology .
In the biosynthetic pathways of prostanoid, the first step is conversion of arachidonic acid (AAc) into PGH2, which is catalyzed by cyclooxygenase (COX). Two isoforms of COX have been characterized in humans. COX-1 can be viewed as a constitutive enzyme and prostanoids formed through the action of COX-1 mediate the so-called “housekeeping” functions, such as the regulation of renal function and maintenance of the gastric mucosa integrity and haemostasis. COX-2 is expressed in some tissues and cell types such as endothelium or renal macula densa, but is induced in response to hormones, growth factors, pro-inflammatory cytokines, bacterial endotoxin and tumor promoters. Also, COX-2 is typically over-expressed at inflammatory sites such as atherosclerotic lesions (reviewed in [1,2]).
PGI2, PGE2, PGD2, PGF2α and TxA2 are formed from PGH2 in reactions catalyzed by specific synthases acting on PGH2 . Figure 1 show a simplified scheme of prostanoid biosynthesis. TxA2 is one of the most potent pro-thrombotic agents released by activated platelets and macrophages [3,4] that exhibits two major biologic activities: the stimulation of platelet function and smooth muscle contraction, inducing platelet aggregation, vasoconstriction and broncoconstriction [5,6]. It is also an important pro-inflammatory and pro-atherogenic agent [7,8]. The conversion of PGH2 into TxA2 is catalyzed by the TxA-synthase (TxAS)  a membrane-bound hemoprotein that, according to their spectral characteristics, is a cytochrome P450 protein [9,10].
PGE2 is a major prostanoid produced by many tissues and cell types including leukocytes and vascular smooth muscle cells (VSMC) that mediates some of the cardinal features of inflammation, including pain, edema and fever . It is important to note that PGE2 induces expression of matrix metalloproteinases , enzymes that are crucial in the degradation of extracellular matrix and plaque stability . Also, PGE2 inhibits the production of macromolecules of the extracellular matrix, further favoring plaque fragility . PGE-synthase (PGES) catalyzes conversion of PGH2 to PGE2. The first isoenzyme identified and characterized was the microsomal PGE-Synthase-1 (mPGES-1) that is inducible by pro-inflammatory cytokines. There is convincing evidence of a pathophysiological role of mPGES-1 in cardiovascular pathology [15–19].
Although there are a number of biochemical, functional and pharmacological studies concerning PGE2 and TxA2 little is known about underlying factors involved in the determination of TxA2 and PGE2 biosynthetic machinery. To assess genetic and environmental correlations of proteins involved in prostanoids biosynthesis, we analyzed the phenotypic expression of the proteins related to TxA2 and PGE2 biosynthesis, using data from a set of extended Spanish families. Thus, the present work deals with the genetic and environmental factors that influence the biosynthetic pathways of these pro-inflammatory, pro-thrombotic and pro-atherogenic compounds.
In this study we have included 308 individuals belonging to 15 extended families. The depth and complexity of these pedigrees is illustrated in the supplemented Table I. For a detailed description of the methods, see the data supplement (available online at htto://atvb.ahjournals.org).
Table 1 shows the characteristics of the sample and the values of the phenotypes analyzed. The ages ranged from 5 to 93 years and the number of male and female subjects was similar with a similar mean age and range.
Since TxA2 is produced mainly by platelets, in agreement with a statistical correlation found in this study between TxAS levels and platelet number (ρ = 0.12, p=0.015) the production of TxA2 after A23187 challenge was normalized by the number of platelets and by the total platelet volume. Similarly, leukocytes, and particularly monocytes were probably the main contributors to PGE2 formation. This is consistent with the fact that COX-2 expression correlated with the total number of leukocytes (ρ = 0.16, p=0.003). Therefore, data concerning PGE2 formation in response to A23187 and LPS was normalized by the total number of leukocytes and number of monocytes.
Significant lower levels of PGE2 production were observed in females, irrespective of the absolute or relative expression of this parameter. No statistically significant differences were observed between males and females in any of the other phenotypes analyzed. Age effects were significant only for parameters concerning PGE2 production by whole blood in response to LPS stimulus. A significant negative dependence on age was found for absolute production of PGE2 (p=0.0001), relative to the number of leukocytes (p=0.0349), and relative to the number of monocytes (p=0.0003). In addition, the absolute production of PGE2 in response to LPS was significantly higher in smoking subjects (p=0.003), although it was not statistical significant when the parameter was expressed in relative terms, either as leukocyte or monocyte number. Also, smokers tended to express more TxAS (p=0.037) than non-smokers.
The components of variance are shown in Table 2 based upon the most parsimonious model (i.e., the model that best fits the observed data and exhibits the minimum complexity) for each phenotype, including only significant sources of variation. The remaining variance not accounted for in Table 2 is attributable to individual-specific random environmental influences and random error. The levels of expression of COX-1, COX-2, mPGES-1 and TxAS, showed highly significant heritabilities, ranging from 0.406 to 0.634 after correcting for covariate effects. The proportion of the residual phenotypic variability accounted for shared household effects tended to be considerably smaller than that accounted for genetic effects. This indicates that genes are important in determining the expression of these traits.
Both parameters of platelet function (PFA-collagen/epinephrine and PFA-collagen/ADP) showed significant heritabilities of 0.511 and 0.197, respectively.
Also, TxA2 production, in terms of TxB2, by whole blood challenged with calcium ionophore exhibit highly significant heritability, either in terms of absolute production or relative to the number and volume of platelets (Table 3). When the results of TxA2 production were expressed in terms of relative production, the value of heritability was very high and similar in platelet number and platelet volume.
Heritability of PGE2 production in response to A23187 achieved statistical signification only when expressed relative to monocyte number (Table 3). The household effects of A23187-stimulated PGE2 production were significant for the absolute and relative expression, ranging from 0.197 to 0.251, and quantitatively similar to those of heritability. In contrast, heritability of PGE2 production in response to LPS, which involve induction of COX-2 and mPGES-1 expression, showed highly significant heritabilities after correcting for covariate effects, ranging from 0.355 to 0.493, depending on the absolute or relative expression of the phenotype (Table 2). There were no appreciable household influences on these traits.
The expression value of all of the enzymes (COX-1, COX-2, mPGES-1 and TxAS) showed significant positive phenotypical correlation among all of them (Table 3). Nevertheless, when phenotypical correlations were partitioned using a bivariant variance component model in terms of genetic and environmental correlations, the scenario was different. Indeed, no genetic correlation was observed between the expression of COX-1 and the expression of any other enzyme analyzed. In contrast, it is important to note that a significant genetic correlation was observed between COX-2 expression and expression of mPGES-1 (p= 0.02). Surprisingly, a strong genetic correlation was found between COX-2 expression and TxAS (p= 9.4×10−4). Both, COX-1 and COX 2 expression exhibited positive significant environmental correlation only with TxAS expression. TxAS and mPGES-1 also exhibited a significant genetic positive correlation (p= 3.2×10−7).
Table 4 presents the significant p values of correlations between the expression of enzymes and LPS-induced PGE2 production. No significant correlation between any enzyme and prostanoid production by whole blood was found, with the exception of COX-2 expression and LPS-induced PGE2 production, which showed a significant phenotypic correlation. It is note worthy that this phenotypic correlation was mainly genetic in nature, since no environmental correlation was found in the partitioned analysis of the variance.
Table 5 shows the regression coefficients and p values of significant correlations between production of prostanoids by whole blood. Production of PGE2 strongly correlated with TxA2, irrespective of the parameters that were expressed in absolute or relative terms. The p values from the bivariant variance component model indicated that environmental correlations were stronger than genetic ones in all cases.
Identifying the genetic and environmental factors that influence susceptibility to complex human traits has been challenging, owing to the complexity of both genetic and environmental factors. In this context, the prostanoid biosynthetic pathway has been implicated in a variety of clinically important areas (complex traits) manly because of its pro-inflammatory and pro-atherogenic role. To our knowledge, this is the first report that evaluates the relative contributions of genetic and environmental influences to the prostanoid biosynthetic pathway-related phenotypes.
One of the most important components of prostanoid-pathways is the TxA2, where its production showed a high heritability in response to calcium ionophore and particularly when the phenotype is expressed in relative terms to platelet number or volume assuming that platelets are the major cells involved in the production of TxA2 in the whole blood test. This was consistent with the high heritabilities that were found for the phenotypes concerning the enzymes involved in its biosynthetic pathways. This was also consistent with data reported by other authors [20,21]. Bray et al.  estimated heritabilities of several parameters related with platelet function in Caucasian and African American families with premature coronary artery disease. They found that aggregation in response to epinephrine and ADP was significant in both populations, whereas in response to collagen platelet aggregation was only highly heritable in the African American population. In our study, based on a Caucasian population, we found that PFA-collagen/epinephrine has also a quite high heritability.
In addition, the same group found high significant heritabilities of TxA2 production, both by whole blood stimulated with collagen  and after low dosis aspirin administration . Interestingly, they found that platelet function phenotypes, not strictly dependent on TxA2 biosynthesis, after aspirin administration were strongly heritable across races . Our study reinforces the notion that the ability to synthesize TxA2 is strongly determined by genes, at least in Caucasian populations.
Intake of COX-2 selective inhibitors celecoxib or MK-966 decreased urinary excretion of PGI2 metabolites in healthy subjects; whereas no significant effect was observed in the urinary excretion of TxA2 metabolites was observed. Non-selective NSAIDs indomethacin and ibuprofen significantly reduced the levels of TxA2 metabolites in urine and only these drugs were able to inhibit TxA2-dependent platelet aggregation [22,23]. These findings suggest that whereas COX-2 is involved in the systemic PGI2 biosynthesis and COX-1 is the main isoenzyme linked to systemic production TxA2. However, recent reports provide evidence that COX-2 could also contribute to TxA2 [24–27]. In our study no correlation between any enzyme and TxA2 production by whole blood was found. It is important to take into account that the production of TxA2 by whole blood is probably due to platelets and a minor contribution of monocytes. Unlike platelets, monocytes are able to express COX-2 and mPGES-1 in addition to COX-1 and TxAS. This could account for the strong genetic correlation between COX-2 and TxAS expression that we found. Interestingly, we observed that TxAS and mPGES-1 also exhibited a significant positive genetic correlation, indicating that both enzymes might be regulated by a common set of genes in certain cell type, probably monocytes. In our study, enzyme expression was determined in terms of mRNA. Hence, unlike TxA2 production, the contribution of monocytes was likely the most relevant regarding enzyme expression. In addition, basal levels of mRNA encoding COX-2, mPGES-1 and TxAS, showed highly significant heritabilities.
In addition to the cardinal features of inflammation, including pain, edema and fever , PGE2 induces expression of matrix metalloproteinases , enzymes considered crucial in the degradation of plaque stability . It is also known that PGE2 inhibits the production of macromolecules in the extracellular matrix, further favoring plaque fragility . The first PGES isoenzyme identified and characterized was a 16 KDa protein with glutathione-dependent PGES activity, now called “mPGES-1”. This enzyme was inducible by pro-inflammatory cytokines. Functional coupling of mPGES-1 with COX-2 was reported earlier , although now this cannot be generalized . Conversely, a cytosolic-PGES (cPGES) seems to act functionally coupled with COX-1 , even though this cannot be generalized also . The latter enzyme is ubiquitously expressed and identical to p23, a protein somewhat related to steroid hormone receptor-mediated signal transduction. Another type of microsomal-PGES called “mPGES-2” was further characterized and was reported to use PGH2 generated by both COX-1 and COX-2 activities .
Our previous results indicate that mPGES-1 is the main isoenzyme involved in PGE2 biosynthesis under inflammatory conditions [18,31–33]. It is widely accepted that during inflammation there is an increased production of PGE2 due to the action of COX-2/mPGES-1. COX-2 is detectable after vascular damage and is highly expressed in atherosclerotic lesions . In addition, expression of COX-2 in VSMC has contributed to abdominal aortic aneurism (AAA) in mice . Suppression of mPGES-1 increases PGI2 biosynthesis in VSMC, when COX activity is not the limiting step . This is consistent with the fact that suppression of mPGES-1 depresses systemic PGE2 biosynthesis . Up-regulation of mPGES-1 has been found in symptomatic atherosclerotic carotid plaques, associated with up-regulation of metallo-proteinase-2 and -9 [15,17]. Increased mPGES-1 levels were found in atherosclerotic plaques in diabetic patients when compared to non-diabetics . Recently it has been reported that mPGES-1 deletion suppresses experimental AAA in mice . Hence the data point to a pathophysiological role of COX-2/mPGES-1 pathway in cardiovascular disease. Our data showed a significant phenotypic correlation between basal expression of COX-2 and LPS-induced PGE2 production. It is clear that this phenotypic correlation was mainly genetic in nature. Unfortunately, we could not determine mRNA levels of any enzyme after LPS stimulation of whole blood, since mRNA was highly degraded during incubation. Nevertheless, our results suggest that initial levels of COX-2 were essential for further LPS-induced PGE2 biosynthesis.
Our results document the importance of genetic factors that influence prostanoid-related phenotypes in this Spanish population. Genes appear to be the largest identifiable determinant of quantitative variation for all of the traits. The use of extended pedigrees and household-sharing information yielded precise information on the correlations among family members. Shared environment had not a substantial effect on phenotypes. These prostanoid-related phenotypes are similar to other cardiovascular risk factors such as hemostasis-related phenotypes; in which shared environmental effects also appear to be of minor importance . In addition, we have limited the estimation of genetic components to that attributable to additive effects. If other non-additive sources of genetic variance exist (e.g., dominance or epistasis), then our heritabilities will have been underestimated. Therefore, our estimates are conservative at best.
We would like to emphasize that the utility of genetic studies of quantitative intermediate risk factors is manifold. Intermediate risk factors are more proximal to gene action and thus provide less attenuated genetic signals than when a discrete clinical end point such as disease or no disease are analyzed. Also, susceptibility to disease is primarily a quantitative process that reflects an unobservable continuous liability. Evidence for the continuous relation between several of the phenotypes considered in this study and the risk of cardiovascular disease has been reported widely. Our data provide evidence that not only factors involved in the hemostasis or lipid profile are regulated genetically, but also expression of some mediators committed to platelet function, stenosis, and plaque fragility that might be determined by genes. In this regard, our findings reinforce the idea that susceptibility to vascular events is partly determined genetically. These data also provide excellent support for our plan to perform a genomic search to identify and assess the importance of genes related to prostanoid biosynthesis in the Spanish population.
Sources of Funding
This study was supported partially by grants No. 2 R01 HL070751-05 from the USA NIH, PI-08/0420, PI-08/0756, SAF2008/01859 and RECAVA-RD06/0014. J.M. Soria was supported by “Programa d’Estabilització d’Investigadors de la Direcció d’Estrategia i Coordinació del Departament de Salut” (Generalitat de Catalunya).
We are indebted to all of the families who participated in the GAIT Project.
Conflicts of interest
There are not conflicts of interest.