Fgf23

To study the effects of varying Pi concentrations on phosphatase expression, it was essential to control Pi concentration in the basal mineralizing medium. This ruled out the use of β-glycerophosphate (βGP) as the availability of Pi from βGP requires the action of TNAP (Huesa et al. 2015 (link)) which can itself be modulated by CKD-associated endocrine factors such as Pi, PTH, and FGF23 (Shalhoub et al. 2011 (link), Rendenbach et al. 2014 (link), Houston et al. 2016 (link)). Therefore, upon confluence (day 0), mineralization was induced by supplementing the growth medium (basal concentration: 1.8 mM Ca; 1 mM Pi) with 50 μg/mL l-ascorbic acid (AA) and 1.5 mM CaCl₂ to provide a final medium containing 3.3 mM Ca (Houston et al. 2016 (link)). Cultures were also supplemented with a range of Pi (1–5 mM), PTH (0–50 nM), and FGF23 (0–200 ng/mL) with or without klotho (50 ng/mL) (R&D Systems). Cells were maintained in a 5% CO₂ atmosphere at 37°C and mineralization media was changed every second/third day for 28 days.

Primary Kupffer cells and hepatocytes were isolated from perfusions of mouse livers with Liberase (Roche, Indianapolis, IN), and cultured as previously described [24 (link),26 (link)]. Hepatic stellate cells were isolated from mice by sequential perfusion with pronase (Roche) and collagenase (Crescent Chemicals, Islandia, NY). The liver was manually digested, diluted in DMEM-F12 containing 10 μg/mL DNase, and then mixed with an equal volume of DMEM-F12 containing pronase (0.2 mg/mL) and collagenase (0.5 mg/mL). The mixture was incubated in a shaking water bath at 250 rpm, 37°C for 10 min, strained through sterile gauze, washed with Gey’s Balanced Salt Solution (GBSS) containing 10 μg/mL DNase and pelleted. The pellet was resuspended in GBSS with 10 μg/mL DNase and layered atop a 9–10% Accudenz (Accurate Chemical Co., Tempe, AZ) gradient. After centrifugation at 1,400 g for 17 minutes, and hepatic stellate cells recovered at the GBSS-Accudenz interface. Cultured cells were treated with 10 ng/ml LPS, 25 ng/ml TNF, 25 ng/ml IL-1β, 100 ng/ml IFNγ (R&D Systems, Minneapolis, MN), 100 ng/ml Pam4CSK2, or 100 ng/ml FGF23 (R&D Systems) with 10 ug/ml heparin (MilliporeSigma).

Three different age-matched sources of human aortic endothelial cells (HAECs) (Cat No. 6100; ScienCell Research Laboratory, Carlsbad, CA) and two different age-matched sources of human brain microvascular endothelial cells (HBMECs) (Cat No. ACBRI376; Cell Systems, Kirkland, WA) were obtained. The number (n) corresponds to the total repeat experiments performed collectively by using all 2 or 3 cell sources. Cells were treated with EC medium (ECM, Cat No. 1001; ScienCell Research Laboratory, Carlsbad, CA) containing 5% FBS and 0.5 mM β-glycerolphosphate for HAECs, and CSC medium (Cat No. 4Z3-500; Cell Systems, Kirkland, WA) containing 5% FBS for HBMECs. FGF-23 (Cat No. 2604-FG; R&D Systems, Minneapolis, MN) was used for cell treatments. Passages of cells used in this study were less than or equal to eight. HBMECs were assessed regularly to confirm their central nervous system properties. The cells were stained with antibodies for Von Willebrand factor to confirm EC origin. These cultures were analyzed routinely for astrocyte contamination by staining with anti-glial fibrillary acidic protein. Cell lysates of human pulmonary microvascular endothelial cells (HPMECs, Cat No.3006) and human cardiac microvascular endothelial cells (HCMECs, Cat No. 6006) for Western blot were commercially obtained (ScienCell Research Laboratory, Carlsbad, CA).