TLR4 counteracts BVRA signaling in human leukocytes via differential regulation of AMPK, mTORC1 and mTORC2

TLR4 is implicated in diseases associated with chronic low-grade inflammation, yet homeostatic signaling mechanisms that prevent and/or are affected by chronic TLR4 activation are largely uncharacterized. We recently reported that LPS/TLR4 activates in human leukocytes signaling intermediates (SI), abbreviated TLR4-SI, which include mTORC1-specific effectors and targets, and that leukocytes of patients with T2D or after cardiopulmonary bypass (CPB) expressed similar SI. Extending these previous findings, here we show that TLR4-SI expression post-CPB was associated with low serum bilirubin and reduced preoperative expression of biliverdin reductase A (BVRA), the enzyme that converts biliverdin to bilirubin, in patient’s leukocytes. Biliverdin inhibited TLR4 signaling in leukocytes and triggered phosphorylation of mTORC2-specific targets, including Akt, PKCζ, AMPKα-LKB1-TSC1/2, and their association with BVRA. Torin, PP242, and a PKCζ inhibitory peptide, but not rapamycin, prevented these biliverdin-induced responses and TLR4 inhibition. In contrast, LPS/TLR4 triggered decreases in BVRA, AMPKα and PKCζ expression, and an increase in haptoglobin, a heme binding protein, in leukocytes in vivo and in vitro, indicating that activated TLR4 may suppress biliverdin/BVRA signaling. Significantly, compared to non-diabetics, BVRA and PKCζ expression was low and haptoglobin was high in T2D patients leukocytes. Sustained TLR4 activation may deregulate homeostatic anti-inflammatory BVRA/mTORC2 signaling and thereby contribute to chronic inflammatory diseases.

In vitro studies. Blood drawn into heparin-containing tubes was separated into aliquots and treated with LPS (10 ng/ml) or biliverdin (50 μM), unless otherwise indicated. Leukocytes were then isolated as described 19 .
In some experiments, neutrophils and mononucleated cells (which include monocytes and lymphocytes) were isolated, treated or not, and lysed as described 10 . Raw 264.7 cells were obtained from ATCC and cultured for up to 5 passages. Cell lysates were normalized for protein content and analyzed by western blotting. In brief, samples were subjected to SDS-PAGE separation followed by blotting onto polyvinylidene difluoride membrane. Immunoreactive bands were detected using Super Signal Chemiluminescence (Thermo Scientific Pierce) and visualized by autoradiography. All figures shown are an accurate representation of the data and no image was manipulated.
Immunoprecipitation. Pierce crosslink IP kit (Prod #26147, Thermo Scientific) was used to crosslink BVRA antibody to agarose. Leukocyte pellets were washed once with PBS and then lysed for 5 min on ice with 500 μl of RIPA buffer. Cellular debris was removed by centrifugation and supernatants were normalized for protein concentration. Lysates (500 μg protein) were pre-cleared for 2 h at 4 °C with 40 μl of protein A/G agarose www.nature.com/scientificreports www.nature.com/scientificreports/ beads and were then incubated overnight at 4 °C with 2 μg of BVRA-agarose with gentle rotation. Samples were then washed three times with washing buffer (0.025 M Tris, 0.15 M NaCl, 0.001 M EDTA, 1% NP-40, 5% glycerol; pH 7.4). Bound proteins were eluted with 5X NaDodSO 4 sample buffer and analyzed by western blotting as described earlier.

Statistical analyses.
Clinical and laboratory data were analyzed with respect to the detection of TLR4-SI on day 1 post-CPB. t-test was used to analyze continuous variables and Mann-Whitney test for comparison of non-parametric data. Categorical variables were compared using Fisher exact test. Normality was determined using D' Agostino and Pearson test. Demographic and clinical variables recognized in unadjusted comparisons as statistically significant at p-value of less than 0.1 were selected for inclusion in a logistic regression model. Data are presented as mean and 95% confidence interval (95%CI) of mean if data were normally distributed, or median and interquartile (Q25, Q75%) range if data were not normally distributed. Within group values were compared using ANOVA with repeated measure and Tukey's multiple comparisons test, or Friedman test and Dunn's multiple comparisons test if data were not normally distribution. Data and statistical analyses were performed using Prism 7.0 for Mac OSX (Graph Pad software, Inc) and STATISTICA 12.0 for Windows (StatSoft Inc. Oklahoma, USA). p-values less than 0.05 were considered statistically significant.

Results
Expression of TLR4 signaling intermediates in leukocytes of patients after CPB is inversely correlated with preoperative serum bilirubin levels. We conducted a prospective blinded study of a cohort of patients scheduled for elective CPB. Patient's demographics are presented in Table 1. Patient's leukocytes were analyzed for expression of previously described TLR4 signaling intermediates (SI), referred to as 'TLR4-SI' , which include MMP9, AMPKα, HIF-1α and TLR4 9 . MMP9, HIF-1α and TLR4 expression was low and AMPKα was intact in leukocytes of patients prior to or immediately after CPB (Fig. 1a, lanes 1 and 2). However, leukocytes of 31 of the 44 patients (70%) exhibited increased MMP9, HIF-1α and TLR4 expression and proteolytically cleaved AMPKα by day 1 postoperatively (Fig. 1a). A similar expression pattern was seen for up to day 4 postoperatively (Fig. 1a).
Next, patient's demographics, clinical parameters, and blood chemistries were all analyzed with respect to the expression of TLR4-SI postoperatively ( Table 2). Patients expressing TLR4-SI had significantly higher (p = 0.028) body surface area (BSA) (Fig. 1b) Biliverdin reductase A (BVRA), the enzyme that reduces biliverdin to bilirubin, is ubiquitously expressed. BVRA expression in CPB patients leukocytes obtained on the morning of the surgery varied among patients (Fig. 1d). Strikingly, BVRA expression in leukocytes of patients who did not express TLR4-SI postoperatively was in general higher than in leukocytes of patients who expressed TLR4-SI postoperatively (see Supplementary  Table 1 for patient's demographics). In addition, whereas TLR4-SI expression postoperatively increased, BVRA expression transiently decreased (Fig. 1e). Preoperative BVRA expression levels and/or presence of TLR4-SI postoperatively did not correlate with patient's intensive care unit length of stay and/or time to discharge. ALT (IU/L)* 21 (13,26) AST (IU/L)* 21 (16,29) Bilirubin (µmol/L)* 0. 5  Neutrophils constitute >60% and lymphocytes ~30% of all leukocytes in humans blood. Naïve T and B cells are less responsive to LPS than neutrophils 20,21 . Analyses of leukocytes and neutrophils from three randomly selected CPB patients showed that the transient decline in BVRA expression seen in leukocytes was reproduced in neutrophils (Fig. 1f,g). Together, the data established that postoperative TLR4-SI expression in CPB patient's leukocytes/neutrophils was inversely associated with both preoperative serum bilirubin and BVRA expression, suggesting that biliverdin/BVRA may counter-regulate TLR4 signaling.
Like biliverdin, metformin, the most widely used drug for treating type 2 diabetics, activates LKB1 and AMPK 31 . Therefore, we assumed that there could be additional targets that both biliverdin and metformin engage. As shown in cardiomyocytes 32 , metformin induced Akt phosphorylation at Ser473 in leukocytes (Fig. 2c). Both metformin 33 and biliverdin induced tuberous sclerosis complex proteins (TSC1/2 complex) phosphorylation at Ser1387, but not at Thr1462 (Fig. 2b,c). The distinction between Ser1387 and Thr1462 is important; Ser1387 is phosphorylated by AMPK and involved in mTORC1 inhibition 34 whereas TSC1/2 phosphorylation at Thr1462 is Akt dependent and contributes to mTORC1 activation 13,35,36 . In metformin treated human and bovine umbilical vein endothelial cells, PKCζ contributed to LKB1 Ser428 (Ser431 in humans) and AMPKα Thr172 phosphorylation 37 . Other showed that mTORC2-dependent phosphorylation at Thr410 activates PKCζ 38 . Here, both biliverdin and metformin induced PKCζ phosphorylation at Thr410 (Fig. 2b,c). LPS, metformin, and biliverdin all induced mTOR Ser2481 phosphorylation, which is associated with both mTORC1 and mTORC2 39 (Fig. 2c). The data established that multiple SI implicated in mTORC2 and AMPK activation, and on the other hand, mTORC1 inhibition are associated with metformin and biliverdin signaling in leukocytes.
Leukocytes are not amenable to genetic manipulations. Therefore, to further examine the role of mTORC2 relative to biliverdin signaling in leukocytes, blood samples were co-treated with biliverdin and one of three well-characterized pharmacologic response modifiers: rapamycin, an acute mTORC1 inhibitor, torin, an mTORC1 and mTORC2 inhibitor 40 , and GSK2334470 (GSK), an inhibitor of 3-phosphoinositde-dependent protein kinase (PDK1) 41 . In LPS treated leukocytes ( Supplementary Fig. 3), rapamycin and torin inhibited HIF-1α expression and S6K1 phosphorylation at Ser389; GSK also inhibited Akt phosphorylation at Thr308. In marked contrast, while rapamycin and GSK had not effect, torin inhibited all biliverdin-induced responses (Fig. 2a,b). Taken together, the data established that mTORC2 is central to biliverdin signaling in leukocytes.
Biliverdin inhibits TLR4 signaling in leukocytes via mTORC2. Next, we examined the role of biliverdin relative to TLR4 signaling in leukocytes. As demonstrated in Raw264.7 cells 28 , biliverdin inhibited TLR4 signaling in leukocytes in a dose-dependent manner (Fig. 3a). In LPS-treated leukocytes, the increase in MMP9 and TLR4 expression, AMPKα cleavage and Raptor dephosphorylation at Ser792 are all upstream to mTORC1 10 . Therefore, by blocking BVRA-mTORC2 signaling, PP242 and torin, chemically distinct ATP-competitive www.nature.com/scientificreports www.nature.com/scientificreports/ inhibitors of mTORC1 and mTORC2 40,42 , reversed the effect of biliverdin and restored the expression pattern of SI associated with TLR4 signaling (( Fig. 3b (i)). In contrast, mTORC1 phosphorylates S6K1 at Ser389 and triggers, though indirectly, Akt phosphorylation at Thr308 10 . Thus, since torin and PP242 blocked both BVRA-mTORC2 and TLR4-mTORC1 signaling, S6K1 and Akt phosphorylation remained suppressed in leukocytes co-treated with biliverdin and LPS (( Fig. 3b (ii)). Torin and PP242 inhibited AMPKα Thr172, LKB1 Ser431, and Akt Ser473 phosphorylation ( (Fig. 2c, (iii)). The data suggested that mTORC2 plays a central role relative to biliverdin-inhibition of TLR4. mTORC2 is a key regulator of biliverdin/BVRA intermolecular interactions. In leukocytes and neutrophils treated with LPS, SI appeared within 10-90 min 9 . In contrast, time-course studies conducted in biliverdin-treated leukocytes (Fig. 4a) revealed co-appearance of all phosphorylated proteins studied by 60 minutes. Numerous BVRA interacting proteins were identified [43][44][45] . This suggested the hypothesis that it may take approximately 60 min for BVRA to establish stable intermolecular interactions. Then, we sought to determine which of the SI activated in biliverdin treated leukocytes interact with endogenous BVRA. To this end, untreated and biliverdin-treated leukocytes were lysed and subjected to BVRA immunoprecipitation followed by western blotting to identify interacting proteins (Fig. 4b). Phosphorylated LKB1, AMPKα, PKCζ, Akt and mTOR were all detected in biliverdin-treated leukocyte's co-immunoprecipitates (Fig. 4b). Torin 37 , but not GSK2334470, the PDK1 inhibitor, prevented all BVRA's intermolecular interactions (Fig. 4b). A well-characterized PKCζ inhibitory peptide previously used to determine the role of PKCζ in neutrophils 46 also prevented complexes formation. The data established that BVRA interacts either directly or indirectly with multiple SI in biliverdin treated leukocytes. mTORC2 and PKCζ may facilitate formation of and/or stabilize BVRA complexes once formed.
TLR4 is a negative regulator of biliverdin/BVRA signaling. Having found that biliverdin/BVRA inhibits TLR4 signaling, the inverse correlation between TLR4-SI and BVRA expression in CPB patient's leukocytes described earlier (Fig. 1) required explanation. We hypothesized that TLR4 counter-regulates biliverdin/ BVRA signaling. In line with this, we found that LPS triggered a decline in BVRA expression in leukocytes in vitro (Fig. 5a). Inverse TLR4 and BVRA expression trends were also detected in leukocytes of subjects challenged with a bolus dose (1 ng/kg) of LPS in vivo 8 (Fig. 5b). As reported 47 , total BVRA expression remained unchanged in LPS-treated Raw 264.7 cells (Supplementary Fig. 4a). LPS also did not induce an increase in MMP9 expression, AMPKα cleavage or Raptor dephosphorylation at Ser792 in Raw264.7 cells (Supplementary Fig. 4b), establishing www.nature.com/scientificreports www.nature.com/scientificreports/ that LPS and biliverdin signaling in human leukocytes differ from those previously documented in Raw264.7 cells.
In an earlier study 7 , leukocytes of subjects challenged with LPS in vivo were subjected to genome-wide transcriptome analyses. Data from that study, available through GEO Dataset GSE3284, showed that LPS-induced an increase in TLR4 mRNA (Fig. 5c), and decreases in BVRA (Fig. 5d) and PKCζ (Fig. 5e) mRNA expression. Consistent with the latter, LPS triggered a decrease in PKCζ protein expression in leukocytes in vitro (Fig. 5a) and in vivo (Fig. 5b). Li and Gao reported 38 that following phosphorylation by mTORC2, PKCζ was more resistant to proteosomal degradation. Therefore, it is possible that in addition to the decline in transcriptional expression, mTORC2 inhibition in LPS treated leukocytes contributed to unphosphorylated PKCζ degradation.
Heme oxygenase(s) (HO) convert heme to biliverdin, the substrate of BVRA. Two HO isoforms were described in leukocytes: HO-1 and HO-2. The level of HO-1 is resting leukocytes was below detection (Fig. 5), suggesting that biliverdin production at steady state is primarily HO-2 dependent. Consistent with this, others demonstrated www.nature.com/scientificreports www.nature.com/scientificreports/ that bilirubin production in mice neuronal cells required HO-2 but not HO-1 48 . As in macrophages 49,50 , LPS triggered an increase in HO-1 expression, but not HO-2, in leukocytes in vitro and in vivo (Fig. 5a,b). Despite evidence that HO-1 may regulate interferon beta production in LPS treated bone-marrow derived macrophages 51 , the role of HO-1 in leukocytes remains unclear. Furthermore, contrary to the prevailing notion that HO-1 has an anti-inflammatory function, knockout of HO-1 in mice liver and myeloid cells contributed to a decrease in mice susceptibility to diet-induced insulin resistance and inflammation 52 . As proposed 52 , it is possible that HO-1 drives rather than inhibits inflammation in the context of metabolic diseases.
Haptoglobin is a high-affinity heme binding protein that regulates heme availability. A recent study found that monocytes of subjects with Gilbert's syndrome, a syndrome associated with higher than normal bilirubin levels, had significantly lower intracellular haptoglobin levels as compared to matched controls 53 . Neutrophils synthesize and store haptoglobin in their granules and release it following activation 54 . Leukocytes treated with LPS in vitro (Fig. 5a) and in vivo (Fig. 5b) exhibited elevated intracellular haptoglobin protein expression. Data from GEO Dataset GSE3284 showed that LPS also triggered an increase in haptoglobin mRNA in leukocytes in vivo (Fig. 5f). Therefore, TLR4 may deregulate biliverdin/BVRA signaling in leukocytes by suppressing expression of pathway SI and heme availability. Our work model, summarizing both BVRA and TLR4 signaling in leukocytes, is presented in Fig. 6.
Can a decline in BVRA expression alter leukocyte's sensitivity to biliverdin? To address this question, leukocytes were first treated for 4 hours with LPS to induce a decline in BVRA expression (Fig. 5g lanes 4-6). Then, leukocytes were treated for 1 hour with biliverdin ( Fig. 5g lane 5) or metformin (Fig. 5g lane 6). Absence of Akt Ser473, AMPKα Thr172, and LKB1 Ser431 phosphorylation all indicated that LPS blunted leukocyte's responses to biliverdin. In contrast, leukocytes pretreated with LPS remained responsive to metformin. These findings suggested that leukocytes with reduced BVRA expression are less responsive to biliverdin. www.nature.com/scientificreports www.nature.com/scientificreports/ Expression of BVRA signaling intermediates in T2D patient's leukocytes. Then, we sought to explore whether our findings had clinical correlates. We reported that leukocytes of patients with T2D expressed TLR4-SI 9 . Here, using TLR4 as a surrogate for TLR4-SI expression, leukocytes of a previously described cohort of patients without or with T2D 10 were analyzed for BVRA, PKCζ and haptoglobin expression (Fig. 7). Reproducing the pattern seen in LPS-treated leukocytes (Fig. 5), leukocytes of patients with T2D exhibited reduced BVRA and PKCζ and increased haptoglobin expression. Consistent with increased HO-1 expression in liver and visceral fat of obese, insulin resistant subjects 52 , leukocytes of patients with T2D exhibited elevated HO-1 expression. Collectively, the data suggested that biliverdin/BVRA signaling is deregulated in T2D patient's leukocytes.

Discussion
Leukocytes form the first line of defense against invading microorganisms. In humans, minute LPS concentrations and multiple endogenous ligands that are produced by damaged tissues activate leukocyte's TLR4. Since activated leukocytes may cause sever damage to hosts, we hypothesized existence of physiological mechanisms that prevent unwarranted TLR4 activation, and furthermore, that failure of these mechanisms may contribute to chronic, low-grade inflammatory diseases. We recently reported that insulin is a TLR4-signaling regulator/inhibitor in leukocytes 10 . In this study, analyses of CPB patient leukocytes and leukocytes exposed to LPS in vitro and in vivo uncovered an additional TLR4 regulatory mechanism that is controlled by biliverdin and BVRA. We found that biliverdin/BVRA inhibits TLR4 signaling using multiple SI, including PKCζ, LKB1, AMPK, TSC1/2 and Akt. We conclude that mTORC2, but not mTORC1, is central to biliverdin/BVRA signaling in human leukocytes. Two observations support this conclusion: (i) biliverdin induced Akt phosphorylation at Thr473, the mTORC2 phosphorylation site 22,55 , but failed to induce S6K1 phosphorylation at Ser389, the mTORC1 phosphorylation Our data suggest that biliverdin facilitates formation of BVRA signaling complex(s) that include mTORC2. mTORC2 induces phosphorylation of two targets: PKCζ at Thr410 and Akt at Ser473. PKCζ is involved in LKB1 Ser431 and AMPKα Thr172 phosphorylation. AMPK phosphorylates TSC1/2 at Ser1387 and Raptor at Ser792. Therefore, BVRA activation may contribute to mTORC1 inhibition via two SI: phosphorylated Raptor, which inhibits mTORC1 complex formation, and TSC1/2 mediated inhibition of mTORC1 activity. Because mTORC1 is central to TLR4 signaling in leukocytes 10 , BVRA inhibition of mTORC1 suppresses TLR4. Pharmacological inhibitors are shown in blue. Torin and PP242, but not rapamycin, inhibited BVRA-mediated protein-protein interactions. TLR4 induced response are shown in red. Once activated, TLR4 down-regulates BVRA and PKCζ expression (dashed lines). TLR4 upregulates expression of MMP9 9 , which then triggers AMPKα cleavage (dashed line), leading to Raptor Ser792 dephosphorylation 9 . This enables mTORC1 activation, which then triggers an increase in HIF-1α expression and S6K1 phosphorylation at Thr389. TLR4 also triggers an increase in haptoglobin expression, thus limiting heme availability. Using these parallel mechanisms TLR4 ensures biliverdin/BVRA and mTORC2 signaling inhibition, and on the other hand, mTORC1 activation. www.nature.com/scientificreports www.nature.com/scientificreports/ site 16 , and (ii) torin and PP242, which act on mTORC1 and mTORC2, suppressed responses to biliverdin, whereas rapamycin, which under acute conditions inhibits only mTORC1, did not. These findings are significant since they demonstrate that bilirubin/BVRA inhibits TLR4/mTORC1 signaling in leukocytes via mTORC2. A similar mechanism may prevent TLR4 activation at steady state.
It is interesting to compare the mechanisms by which insulin and biliverdin/BVRA inhibit TLR4 signaling in leukocytes. Contrasting biliverdin/BVRA signaling dependence on mTORC2, insulin signaling in leukocytes, and in general, involves both mTORC1 and mTORC2 activation 10 . Consequently, whereas PKCζ-LKB1-AMPK-TSC1/2 inhibited mTORC1 in the context of biliverdin/BVRA signaling in leukocytes, insulin suppressed TLR4 signaling via mTORC2-Akt Ser473 and Foxo1/3 without LKB1-AMPK-TSC1/2 involvement. These data demonstrate that mTORC2 can engage distinct SI subsets to inhibit TLR4. Whereas the insulin and biliverdin/BVRA signaling pathways that converge on mTORC1 differ, we noted striking mechanistic similarities between biliverdin/BVRA and metformin signaling in leukocytes. As one of the oldest and most commonly used drugs for treating type 2 diabetics, metformin's mechanism of action has been studied extensively. In general, metformin activates LKB1-AMPK-TSC1/2 signaling 33 . Our data now demonstrate that both metformin and biliverdin use mTORC2 to inhibit mTORC1. How biliverdin/BVRA activates mTORC2 remains unanswered. Maines and colleagues suggested that BVRA acts as a scaffold that supports oligomers formation 45,56 . Our data support this possibility and suggest that in leukocytes, BVRA facilitates formation of signaling complexes that include mTORC2 and it's signaling targets.
Earlier studies have focused on the possibility that bilirubin and biliverdin act primarily as antioxidants and may reduce oxidative stress damage by suppressing intracellular oxides levels 57,58 . Consistent with this possibility, a recent study demonstrated an increase in oxidative stress indicators in BVRA KO mice 59 . Others reported that bilirubin is a ligand of PPARα, a regulator of lipids metabolism in mice livers 60 . On the other hand, BVRA was implicated in insulin, IGF-1 and TNF-α signaling 28,[43][44][45]47 . In IGF-1 treated HEK293 cells, recombinant BVRA co-immunoprecipitated with Akt1, Akt2 and PDK1 and facilitated Akt phosphorylation at Thr308, the PDK1 phosphorylation site 44,45 . BVRA also interacted with PKCδ in this model cell system 45 . However, in TNFα treated HEK293 cells, BVRA regulated PKCζ 45 . Of these, only PKCζ activation was reproduced in biliverdin-treated leukocytes. Consistent with our findings, Wegiel and colleagues reported that biliverdin triggered TLR4 signaling inhibition in Raw 264.7 cells 28 . However, the TLR4 inhibitory mechanism in Raw 264.7 cells involved endothelial nitric oxide synthase (eNOS) activation by CaMKΚβ and increased nitric oxide generation. Although we observed similar biliverdin-induced responses in Raw264.7 cells, i.e., eNOS and CaMKΚβ activation, biliverdin did not induce CaMKΚβ/eNOS signaling in leukocytes, neutrophils or mononucleated cells. The data establish existence of cell-type specific biliverdin/BVRA signaling mechanism(s), and identify a novel biliverdin/BVRA signaling pathway in leukocytes.
Prior clinical and epidemiologic studies found that low serum bilirubin levels were associated with increased risk of pre-diabetes, T2D, metabolic syndrome, as well as stroke [61][62][63][64][65] . Low serum bilirubin were also shown to correlate with higher C reactive protein levels in patients with T2D and impaired glucose tolerance 66 . It is also known that patients with Gilbert's syndrome, a syndrome associated with higher than normal bilirubin, have in general a lower incidence of coronary heart disease and arteriosclerosis 67 . Together, these clinical studies suggested that whereas high bilirubin levels are health beneficial, low bilirubin levels could be detrimental to health. We hypothesize that higher bilirubin levels are associated with decreased risk for human diseases, and vice-a-versa, since these levels reflect, at least in part, how well the biliverdin/ BVRA signaling network operates. When fully functional, biliverdin//BVRA may contribute to higher serum bilirubin levels and at the same time suppress TLR4 signaling in leukocytes.
Notably, our data are first to show that activated TLR4 can counter-regulate BVRA signaling. One of the key biliverdin/BVRA signaling components targeted by TLR4 is BVRA itself. TLR4 may also disrupt biliverdin/ BVRA signaling by reducing AMPKα/AMPK, PKCζ, and heme availability. Significantly, leukocytes of patients with T2D expressed low BVRA and PKCζ levels and elevated haptoglobin, reproducing the expression pattern seen in LPS treated leukocytes. Compared to untreated leukocytes, LPS-treated leukocytes with reduced BVRA expression were less responsive to biliverdin. Higher BVRA expression may enhance leukocyte's resistance to stressors that act on TLR4, or in other words, that biliverdin/BVRA impact leukocyte's activation threshold. Our data suggest that activated TLR4 deregulates this safety mechanism.
In summary, we discovered a novel TLR4 and biliverdin/BVRA counter regulatory mechanism that controls TLR4 activation in human leukocytes. Furthermore, our data suggest that the balance between TLR4 and biliverdin/BVRA signaling in leukocytes of patients with T2D is deregulated and chronically shifted in favor of TLR4. A fully functional biliverdin/ BVRA signaling network may, in general, suppress unwarranted TLR4 signaling, whereas suboptimal network activity may fail to prevent TLR4 activation thus contributing to non-resolving chronic low-grade inflammation.