Effect of Sickle Cell Trait and APOL1 Genotype on the Association of Soluble uPAR with Kidney Function Measures in Black Americans

Soluble urokinase plasminogen activator receptor (suPAR) is the circulating form of urokinase plasminogen activator receptor, a glycosyl-phosphatidylinositol–anchored membrane protein expressed in various cell types including kidney podocytes and endothelial cells (¹). suPAR has been associated with a decline in eGFR and the risk of incident CKD or proteinuria in a variety of clinical settings (¹). In the United States, the higher risk of CKD in Black people compared with White people may be at least partially attributable to two genetic susceptibility factors, APOL1 and sickle cell trait, which occur respectively in approximately 13% and 8% of Black people. Hayek et al. (²) recently showed that suPAR levels modify the association between APOL1 genotype and eGFR decline in the African American Study of Kidney Disease and Hypertension and a cohort registry of Black people who underwent cardiac catherization.

The relationship of suPAR and kidney function in a community-based sample of Black adults is unknown, and the potential effect modification of suPAR by sickle cell trait has not been assessed. We measured baseline suPAR levels, along with APOL1 and sickle cell trait genotypes, in 3343 self-identified Black people recruited from urban and rural areas of Jackson, Mississippi (Jackson Heart Study) who have been followed prospectively for a mean 8.0±0.84 years. Kidney function was assessed at baseline (2000–2004) and exam 3 (2009–2013). At baseline, participants’ ages ranged from 20 to 93 (mean 55) years, and 37% were male, 13% were current smokers, 54% were obese, 57% were hypertensive, and 25% were diabetic. Baseline characteristics did not differ by sickle cell or APOL1 genotypes. suPAR levels ranged between 742 and 10,595 (median 2097) pg/ml. At baseline, the mean eGFR was 94±22 ml/min per 1.73 m², and 7% had eGFR<60 ml/min per 1.73 m². In a subset of 1605 participants with spot urine albumin-creatinine ratio (UACR) measurements, baseline mean was 49.5±277 mg/g, and 13% had a UACR>30 mg/g. Of the individuals with both baseline and year 3 exam data, mean eGFR change was −0.004±1.96 ml/min per 1.73 m² per year, 202 of 2240 (7%) developed incident CKD, and 78 of 1054 (11%) developed incident albuminuria.

Higher baseline log-transformed suPAR was associated with lower baseline eGFR and UACR, greater eGFR decline, incident CKD, and incident albuminuria (Table 1). APOL1 was significantly associated with lower baseline eGFR (−1.85±0.93 ml/min per 1.73 m²; P=0.04) and higher UACR (0.21±0.09 mg/g, P=0.01) and sickle cell trait were also associated with lower baseline eGFR (−4.40±1.16 ml/min per 1.73 m²; P<0.001) and higher UACR (0.38±0.11 mg/g; P<0.001).

Table 1. - Association of soluble urokinase plasminogen activator receptor with kidney phenotypes in Jackson Heart Study, overall and stratified by APOL1 risk genotype or sickle cell trait

Phenotype	Overall				APOL1 Low Risk		APOL1 High Risk			P- Interaction	SCT Negative			SCT Positive			P- interaction
	N or Cases/N	Beta or OR (SE or CI) a	P Value	N or Cases/N	Beta or OR (SE or CI) a	P Value	N or Cases/N	Beta or OR (SE or CI) a	P Value		N or Cases/N	Beta or OR (SE or CI) a	P Value	N or Cases/N	Beta or OR (SE or CI) a	P value
Baseline eGFR (ml/min per 1.73 m2)	3343	−7.14 (0.33)	<0.001	2864	−6.65 (0.35)	<0.001	479	−9.54 (0.87)	<0.001	0.002	3060	−6.90 (0.34)	<0.001	283	−8.53 (1.11)	<0.001	0.02
logUACR (mg/g)	1605	0.34 (0.03)	<0.001	1379	0.30 (0.04)	<0.001	226	0.62 (0.09)	<0.001	0.02	1474	0.32 (0.03)	<0.001	131	0.57 (0.15)	<0.001	0.12
eGFR change (ml/min per 1.73 m2 per yr)	2417	−0.27 (0.05)	<0.001	2078	−0.24 (0.05)	<0.001	339	−0.46 (0.13)	0.001	0.04	2214	−0.25 (0.05)	<0.001	203	−0.45 (0.16)	0.005	0.02
Incident CKD	202/2240	1.47 (1.19 to 1.81)	<0.001	172/1940	1.42 (1.13 to 1.77)	0.002	30/300	1.90 (1.06 to 3.39)	0.03	0.29	176/2064	1.44 (1.15 to 1.81)	0.002	26/176	1.58 (0.90 to 2.77)	0.11	0.61
Incident albuminuria	78/1054	1.39 (1.05 to 1.84)	0.02	68/906	1.36 (1.01 to 1.82)	0.04	10/148	2.43 (0.80 to 7.34)	0.11	0.48	73/977	1.49 (1.12 to 1.98)	0.006	5/77	0.54 (0.08 to 3.61)	0.53	0.19

SCT, sickle cell trait; OR, odds ratio; UACR, urinary albumin-creatinine ratio.

^aBeta coefficients (or ORs) correspond to estimates of mean difference between (or risk associated with) one SD of soluble urokinase plasminogen activator receptor. Incident CKD was defined as development of eGFR<60 ml/min per 1.73 m² during follow-up with baseline eGFR≥60 ml/min per 1.73 m². Incident albuminuria was defined as development of UACR≥30 mg/g during follow-up at exam 3 with baseline UACR<30 mg/g. All models were adjusted for age, sex, diabetes, and genetic ancestry. Models estimating longitudinal change in eGFR or incident CKD were additionally adjusted for baseline eGFR. Models estimating incident albuminuria were additionally adjusted for baseline UACR. P value for interaction estimated by including a multiplicative interaction term in the regression model.

The association of suPAR with baseline eGFR, UACR, and eGFR decline was significantly stronger among those carrying the APOL1 high-risk genotype compared with the low-risk genotype (Table 1). In the African American Study of Kidney Disease and Hypertension, higher suPAR was associated with worsening proteinuria only in participants with two APOL1 risk alleles (³). These observations are further supported by data from animal models in which APOL1 G1 and G2 risk variants synergize with suPAR to bind and activate αvβ3 integrin on podocytes, leading to foot process effacement and proteinuria (²).

Compared with sickle cell trait noncarriers, the associations of higher suPAR with baseline eGFR and eGFR decline were significantly stronger among sickle cell trait carriers (Table 1). For example, the eGFR decline associated with one SD unit higher suPAR was −0.45±0.16 and −0.25±0.05 ml/min per 1.73 m² per year, respectively, among sickle cell trait carriers compared with noncarriers (P for interaction = 0.02). Interestingly, among the 23 individuals (approximately 1% of the overall Jackson Heart Study cohort) carrying both sickle cell trait and APOL1 high risk genotypes, there was a much stronger association of suPAR with eGFR decline (−1.36±0.59 ml/min per 1.73 m² per year), compared with the remainder of the cohort (−0.26±0.05 ml/min per 1.73 m² per year; P for interaction = 0.003).

The effect modification of sickle cell trait on the association of higher suPAR with lower baseline eGFR, higher UACR, and eGFR decline has not been previously reported. Sickle cell trait confers approximately 1.5 to two-fold higher risk of CKD in Black populations (⁴). It has been hypothesized that subclinical sickling in the hypoxic medulla in the kidney may contribute to vascular disruption, leading to ischemic-reperfusion injury of kidney tubules and glomerulosclerosis (⁴,⁵), which may induce a local or systemic inflammatory response. suPAR release is similarly associated with inflammation and podocyte injury (¹). Therefore, the stronger associations of higher suPAR on CKD-related measures among individuals with sickle cell trait further support a possible role of inflammatory pathways contributing to both suPAR and sickle hemoglobin-induced kidney injury (¹,⁵).

In summary, our results from a community-based Black cohort with a low prevalence of CKD at baseline support the importance of suPAR as a CKD risk factor in the general population and confirm that APOL1 risk genotypes modify the association of suPAR with CKD-related kidney function. Further, our results demonstrate that associations of suPAR with eGFR and UACR are modified by sickle cell trait carrier status.

Disclosures

A. Correa reports receiving honoraria for serving as a member of the United Biosource Corporation Mycophenolate Pregnancy Registry Safety Advisory Committee. L.M. Raffield reports consultancy agreements with University of Mississippi Medical Center. All remaining authors have nothing to disclose.

Funding

This work was supported by National Heart, Lung, and Blood Institute (NHLBI) grant R01HL132947 (to A.P. Reiner). N. Franceschini reports receiving grants from the National Institutes of Health, during the conduct of the study.  L.M. Raffield is supported by NHLBI T32 grant HL129982 and Center for Scientific Review grant KL2TR00249. N. Olson was supported by NHLBI grant R00HL129045.