A novel promoter of endothelial dysfunction in African Americans: Relevance to sickle cell anaemia

Vascular endothelial cell dysfunction (ECD) is a prominent feature of sickle cell anaemia (SCA), and it likely is promoted by multiple acquired aberrancies of the blood typically found in this disease.1 In addition, however, it is relevant that African Americans (AA) in general have an increased likelihood of developing ECD, compared to Caucasian Americans (CA).2-6 In the literature, this is identified not only via functional studies such as flow-mediated dilation but also by examination of cultured endothelial cells. Notably, ECD can be identified even in AA adolescents and is independent of their body mass index.4 Correspondingly, individuals of African ancestry exhibit an increased incidence of the varied cardiovascular sequelae of ECD.7 In our laboratory, using known individuals as blood donors, we can survey endothelial cell gene expression by studying their blood outgrowth endothelial cells (BOEC), our term for the ex vivo cultured progeny of a circulating, transplantable endothelial progenitor.8 (BOEC has been re-labelled “late outgrowth endothelial cells” in the more recent literature.9 BOEC are unrelated to the unfortunately mislabelled “EPC”, intended for endothelial progenitor cells, that actually are not of endothelial lineage.9) Importantly, our method is to examine BOEC for gene expression at a degree of in vitro cell number expansion at which acquired gene expression effects have washed out, yet gene expression instability has not yet started.10, 11 So, observed BOEC gene expression presumably identifies inherent traits of the donor endothelium. Thus, a study of BOEC can potentially identify a linkage between heritable endothelial gene expression and endothelial function and clinical phenotype. Applying this approach in a new study of CA subjects undergoing coronary catheterization, we discovered and recently reported that there is a significant association between early development of coronary artery ECD and BOEC displaying elevated HMGB1 (high-mobility group box 1) expression linked inversely to lowered LAMC1 (laminin gamma chain 1) expression; correlation coefficient r = −0.844.12 Verification was obtained by re-establishing the (frozen) BOEC in culture and performing functional assays of LAMC1 deposition into a collagen matrix and lipopolysaccharide-induced endothelial HMGB1 release. Knowing this, we looked back at our data sets from previous studies, and we realized that this high-HMGB1/low-LAMC1 inverse expression linkage was also evident for the BOEC from the three other, independent subject groups we previously had studied: a group of random 23–69 year olds of mixed gender and race (r = −0.584)11, 12; a group of children and adolescents with SCA (r = −0.383)11, 12; and a group of healthy 20–29 years olds who self-identified as being either CA or AA (r = −0.569).12, 13 Because this newly-discovered inverse HMGB1/LAMC1 expression relationship is putatively heritable, and the range of subject expression of each gene is very broad, we have now looked in greater detail at the data previously obtained from the latter group of healthy CA and AA subjects.12 We now realize that, compared to CA subjects, the AA subjects exhibited an overlapping but significantly higher range of BOEC HMGB1 expression and an overlapping but significantly lower range of LAMC1 expression (Figure 1A). This also revealed that the inverse linkage of high-HMGB1 with low-LAMC1 expression for the overall group of subjects is also evident for both the AA and CA groups separately (Figure 1B). Thus, it is not surprising that the BOEC gene expression for the previously studied group of children and adolescents with SCA11 also segregates into two groups: one tending to exhibit higher-HMGB1/lower-LAMC1 expression and the other tending to exhibit lower-HMGB1/higher-LAMC1 expression (Figure 1C).13 Data sources: These Figure 1A data are newly extracted from Wei et al.13 and were originally deposited in Gene Expression Omnibus GSE22688. Figure 1B here is modified from Figure 1B in Hebbel et al.,12 and data were originally deposited as Gene Expression Omnibus GSE22688. Figure 1C data were extracted from Milbauer et al.11 and previously published as Figure 2D in Hebbel et al.12 Data were originally deposited in Gene Expression Omnibus GSE9877. Hebbel et al.12 is open access and permission for 1A and 1B and 1C reproduction is granted. These observations are presumably clinically significant because multiple, known biological effects of high-HMGB1 and low-LAMC1 expression can be predicted to synergistically compromise endothelial cell sensing of and/or responsiveness to shear stress created by flowing blood. We have cited such specific harmful effects in some detail elsewhere.12 Here we present a cartoon and tabular summary of some known effects of these gene expression shifts that are directly relevant to ECD (see Figure S1 and Supplemental Information). Notably, the considerable arteriopathic impacts of HMGB1 have been discussed in detail by Cai et al.14 And specific relevance to SCA is directly supported by the fact that Xu et al.15 detected elevated plasma HMGB1 concentrations in the blood of both sickle transgenic mice and humans with SCA. Thus, we conclude that having higher HMGB1 expression linked to lower LAMC1 expression in endothelium is likely a heritable risk factor for the development of ECD, a risk embedded in the SCA population simply due to their African ancestry. Robert P. Hebbel planned original experiments and coordinated the study; Liming Milbauer and Peng Wei executed experiments'; Robert P. Hebbel, Liming Milbauer and Peng Wei contributed to the writing of the manuscript. The original studies underlying this report were supported by the National Institutes of Health: HL116720 to PW; HL68970, HL76540 and HL55552 to RPH. HL116720 to PW; HL68970, HL76540, HL55552 to RPH. The authors have no conflicts of interest. The original source studies were approved by our institutional committee on the use of human subjects. The source studies included approval for blood draw and the publication of de-identified data. Figure 1C is reproduced from an open-access publication (J Amer Heart Assoc) that does not require approvals for re-use. Data plotted here were obtained previously and deposited in Gene Expression Omnibus (access codes in Figure 1 legend). Data S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.