Directed Transport of CRP Across In Vitro Models of the Blood-Saliva Barrier Strengthens the Feasibility of Salivary CRP as Biomarker for Neonatal Sepsis

Abstract

C-reactive protein (CRP) is a commonly used serum biomarker for detecting sepsis in neonates. After the onset of sepsis, serial measurements are necessary to monitor disease progression; therefore, a non-invasive detection method is beneficial for neonatal well-being. While some studies have shown a correlation between serum and salivary CRP levels in septic neonates, the causal link behind this correlation remains unclear. To investigate this relationship, CRP was examined in serum and saliva samples from 18 septic neonates and compared with saliva samples from 22 healthy neonates. While the measured blood and saliva concentrations of the septic neonates varied individually, a correlation of CRP levels between serum and saliva samples was observed over time. To clarify the presence of active transport of CRP across the blood-salivary barrier (BSB), transport studies were performed with CRP using in vitro models of oral mucosa and submandibular salivary gland epithelium. The results showed enhanced transport toward saliva in both models, supporting the clinical relevance for salivary CRP as a biomarker. Furthermore, CRP regulated the expression of the receptor for advanced glycation end products (RAGE) and the addition of soluble RAGE during the transport studies indicated a RAGE-dependent transport process for CRP from blood to saliva.

Keywords: HTB-41 (A-253); TR146; immunology; infectious disease; neonatal sepsis; neonatology; saliva diagnostics.

Figure 1

CRP concentration in serum [µg/mL] and corresponding concentration in saliva [ng/mL] of 18 neonates (gestational age at birth 33.10 ± 5.44 weeks, six females) over 1–4 time points (n = 33). Serum concentration of CRP was measured with 53.12 ± 72.82 µg/mL (mean ± SD, maximum at 320 µg/mL and minimum at 0.3 µg/mL). Salivary CRP was measured with 65.73 ± 137.70 ng/mL (mean ± SD, maximum at 489 ng/mL and minimum at 0.12 ng/mL). Pearson’s correlation was calculated as r = 0.72 (p < 0.001) with y = −6.388x + 1.358, R² = 0.52. Concentrations indicating severe sepsis (>80 µg/mL CRP in serum) were circled in red (A). Time course of measured salivary [ng/mL] and serum [µg/mL] CRP concentration over 4–9 days from three selected neonates (NN1–NN3) with gestational age of 29 + 1, 35 + 0, and 30 + 1 (week + day) (B).

Figure 2

Apparent permeability (P_app) of 10 µg/mL CRP applied on the apical (A/B) or basolateral (B/A) side for 24 h in the oral mucosa model, based on TR146 cells, and salivary gland model, based on clone B2 of HTB-41 cells. Results calculated as x-fold referred to P_app A/B, shown as mean ± SEM after referring to P_app values for A/B of each individual experiment of in total three independent experiments (N = 11–13). Statistical analysis was performed as Student’s t-test with α = 0.05, p < 0.01 **, p < 0.001 ***.

Figure 3

Messenger RNA expression of RAGE of the oral mucosa model (TR146, A) and salivary gland model (B2 clone, B) after applying CRP on the apical (api) or basolateral (baso) side for 24 h, Ct values were referred to the endogenous control (18sRNA for oral mucosa or PPIA for salivary gland) and normalized to untreated control samples. Results shown as mean ± SEM of two–three independent experiments (N = 4–7). Statistical analysis was performed as one-way ANOVA on ranks with α = 0.05, * p < 0.05. Representative western blots of RAGE and ß-actin of the oral mucosa model (TR146, C) and the salivary gland model (B2 clone, D) after treatment with 10 µg/mL CRP for 24 h on the apical and basolateral side, compared to control samples. Densitometric values of western blots from two—three independent experiments (N= 3) for the oral mucosa model (TR146, E) and salivary gland model (B2 clone, F) shown as mean ± SEM. Statistical analysis was performed as Student’s t-test, with α = 0.05, p < 0.05 *.

Figure 4

Apparent permeability (P_app) values of the oral mucosa and salivary gland model upon applying 10 µg/mL CRP or 10 µg/mL CRP + 5 µg/mL sRAGE on the apical (A/B) or basolateral (B/A) side for 24 h. Results shown as mean ± SEM of two–six independent experiments (N = 7–24) upon normalization to P_app CRP values from A/B. Outliers were excluded applying Grubb’s test. Statistical analysis was performed as two-way ANOVA, post-hoc Holm-Sidak test with α = 0.05, p < 0.01 **, p < 0.001 ***.