Genetic Deletion of TLR4 on Platelets Attenuates Experimental Pulmonary Hypertension

Abstract

Rationale

Recent studies demonstrate a role for TLR4 in the pathogenesis of pulmonary hypertension (PH), however, the cell types involved in mediating the effects of TLR4 remain unknown.

Objectives

The objective of this study was to determine the contribution of TLR4 expressed on nonparenchymal cells to the pathogenesis of PH.

Methods and Results

TLR4 bone marrow chimeric mice revealed an equal contribution of TLR4 on nonparenchymal and parenchymal cells in the pathogenesis of PH as determined by measuring right ventricular systolic pressure (RVSP) and right ventricular hypertrophy (RVH). However, deletion of TLR4 from myeloid lineage cells had no effect on the development of PH since we found no difference in RVSP or RVH in WT vs. LysM-TLR4−/− mice. To explore the potential role of platelet TLR4 in the pathogenesis of PH, platelet specific TLR4−/− mice were generated (PF4-TLR4−/− mice). TLR4 −/− platelets from either global TLR4−/− or PF4-TLR4−/− mice were functional but failed to respond to lipopolysaccharide (LPS) demonstrating a lack of TLR4. PF4-TLR4−/− mice demonstrated significant protection from hypoxia-induced PH, including attenuated increases in RVSP and RVH, decreased platelet activation, and less pulmonary vascular remodeling. Deletion of TLR4 from platelets attenuated serotonin release after CH and LPS stimulated platelets released serotonin and promoted pulmonary artery smooth muscle cell proliferation in a serotonin-dependent manner.

Conclusions

Our data demonstrate that TLR4 on platelets contributes to the pathogenesis of PH and further highlights the role of platelets in PH.

INTRODUCTION

Pulmonary hypertension (PH) is a progressive and fatal disease with no cure. Historically, PH research has relied on concepts derived from hypoxic pulmonary vasoconstriction for the development of new therapies. Increasingly, however, thrombotic and inflammatory mechanisms are recognized as complicating and possibly initiating events in PH¹.

The toll-like receptors (TLRs) are one family of receptors that the innate immune system uses to sense pathogens or tissue damage by recognizing molecular patterns in microbial products or endogenous molecules released by damaged tissues². TLR4, perhaps more than any other TLR, interfaces microbial and sterile inflammation by responding to bacterial lipopolysaccharides (LPS) and endogenous ligands, including hyaluronic acid³, heparin sulfate⁴, high-mobility group box 1⁵, and heat shock proteins⁶. Recent studies demonstrate that activation of TLR4 contributes to the pathogenesis of PH^{7, 8}.

The lung is equipped to sense and respond to endogenous TLR4 ligands. The lung consists of parenchymal cells, including fibroblasts, endothelial cells, and smooth muscle cells and nonparenchymal cells (NPC), including alveolar macrophages, dendritic cells, granulocytes, and transiting leukocytes and platelets. Both parenchymal and NPCs express TLR4 and have intact TLR4 signaling pathways. The aim of this study was to determine the role of TLR4 on NPCs in the pathogenesis of PH.

METHODS

Animal procedures were performed in accordance with University of Pittsburgh Institutional Animal Care and Use Committee guidelines. TLR4loxP/loxP mice were generated as described⁹ and interbred with stud males (TLR4loxP/−; Ella-cre, TLR4loxP/−; LysM-cre, or TLR4loxP/−; PF4-cre) to generate the desired genotype. WT mice were TLR4loxP/loxP mice without the introduction of Cre recombinase. Bone marrow was isolated from WT or TLR4−/− mice and transplanted into irradiated WT or TLR4−/− mice as described.¹⁰ Pulmonary hypertension was induced by exposure to chronic hypoxia (CH) for 21 days or the combination of SU5416 (a VEGF receptor inhibitor) plus CH as described.^{7, 11} Right ventricular systolic pressure (RVSP) and right ventricular hypertrophy (RVH) were measured by established methods⁷. Pulmonary vascular muscularization and % wall thickness were assessed as described⁷. Bleeding time was measured as described.¹² Platelet rich plasma was prepared from whole blood anti-coagulated with acid citrate dextrose.¹² To isolate platelets, platelet rich plasma was centrifuged at 800xg, and platelets were resuspended in tyrodes buffer. Serotonin and IL6 were measured by ELISA. Platelet activation was determined using flow cytometry.¹² The effect of platelets on human pulmonary artery smooth muscle cell (HPAMSC) proliferation was assessed by co-culturing LPS activated platelets with HPASMC. Cell proliferation was assessed by [3H]-thymidine incorporation. Statistical analyses were performed using Graphpad Prism software. Data were analyzed by one-way ANOVA and Tukey’s post hoc test. Detailed methods are provided in the Online Supplemental Methods available at http://circres.ahajournals.org.

RESULTS

Equal contribution of NPC and parenchymal cell TLR4 to hypoxia-induced PH

To deduce a role for TLR4 in NPCs vs. parenchymal cells we performed bone marrow (BM) transplantation into irradiated mice to create WT and TLR4−/− BM chimeric mice. Following transplantation and recovery, mice were exposed to CH to induce PH. WT mice that received WT BM developed PH as determined by measuring RVSP and RVH. PH was attenuated in TLR4−/− mice that received TLR4−/− bone marrow (Fig 1A–B). Interestingly, WT mice that received TLR4−/− BM, or TLR4−/− that received WT BM showed the same attenuation of PH as whole body TLR4 −/− chimeric mice.

Figure 1. TLR4 on BM derived cells, but not those of myeloid lineage, contributes to the pathogenesis PH.

PH was assessed in (A–B) WT and TLR4−/− BM chimeric mice or (C–D) WT and LysM-TLR4−/− mice by measuring RVSP and RV hypertrophy. n=6–8 mice per group. (*P<0.05 vs. WT/WT chimeric mice. n.s. = not significant. 917-863-7755.

TLR4 on myeloid-derived cells does not contribute to hypoxia-induced PH

TLR4 is highly expressed on cells of myeloid lineage and these cells are often associated with response to TLR4 agonists. To investigate a role for TLR4 on myeloid-derived cells we exposed WT or LysM-TLR4−/− mice to 3- weeks CH. The loss of TLR4 on myeloid-derived cells failed to attenuate the development of PH as assessed by RVSP (Fig 1C) and RVH (Fig 1D).

Functional characterization of platelet TLR4 knockout mice

To investigate a role for platelet TLR4 in PH, we utilized platelet-specific TLR4−/− (PF4-TLR4−/−) mice. Platelets were isolated from WT, TLR4−/−, and PF4-TLR4−/− mice and stimulated with LPS. WT platelets upregulated surface P-selectin after LPS stimulation, whereas platelets from TLR4−/− or PF4-TLR4−/− mice did not (Fig. 2A–C). Collagen induced P-selectin expression on platelets isolated from all three genotypes demonstrating that TLR4−/− and PF4-TLR4−/− platelets were functional. Additionally, serotonin was increased in the supernatant of WT platelets, but not TLR4−/− or PF4-TLR4−/− platelets, after stimulation by LPS (Fig. 2E).

Figure 2. TLR4 −/− platelets are functional but do not respond to LPS.

(A–C) Representative flow cytometry histograms of platelets from (A) WT, (B) TLR4 −/− or (C) PF4-TLR4−/− mice stimulated with LPS. (D) Quantitation of platelet surface P-selectin on LPS or collagen treated platelets. (E) Serotonin measured in supernatant of LPS or Collagen treated platelets. (F) Fluorescent images of isolated ECs stained for CD31, SMA and nuclei. (G) Quantitation of IL-6 in the media of WT, TLR4−/−, PF4-TLR4−/− ECs treated with LPS. Data represent the mean ± SEM of 3 independent experiments. *P<0.05 vs. WT unstimulated.

To demonstrate the specificity of the platelet knockout we tested the effect of LPS on endothelial cells (ECs) isolated from the lungs of WT, TLR4−/− or PF4-TLR4−/− mice. The ECs expressed CD31 but not smooth muscle alpha actin (SMA) demonstrating purity (Fig 2E). LPS induced interleukin-6 release from ECs derived from WT or PF4-TLR4−/− mice, but not from ECs derived from TLR4−/− mice (Fig 2F).

Genetic deletion of platelet TLR4 attenuates hypoxia-induced PH

We next exposed WT, TLR4−/−, and PF4-TLR4−/− mice to 3-weeks of CH to induce PH. The deletion of TLR4 globally or on platelets significantly attenuated PH as assessed by measuring RVSP and RVH (Fig. 3A–B).

Figure 3. Genetic deletion of platelet TLR4 attenuates PH.

Global or platelet deletion of TLR4 attenuated PH as assessed by measuring (A) RVSP and (B) RVH. Global or platelet specific deletion of TLR4 prevented in vivo platelet activation as assessed by measuring (C) bleeding time, (D) or platelet surface P-selectin. n=6–8 mice per group. ^†P<0.05 vs. WT CH. ^‡P<0.05 vs. WT SUH. ^§Not significant compared to TLR4−/− normoxic. ^¶Not significant compared to PF4-TLR4−/− normoxic.

Genetic deletion of platelet TLR4 attenuates platelet activation in hypoxia-induced PH

Exposing mice to CH leads to decreased bleeding time and increased platelet surface P-selectin expression, indicative of platelet activation¹². In both the CH and SUH model, bleeding time decreased in WT mice but not in TLR4−/− or PF4-TLR4−/− mice (Fig 3C). Similarly, platelet surface P-selectin expression was increased in CH and SUH WT but not TLR4−/−, or PF4-TLR4−/− mice (Fig 3D).

Genetic deletion of platelet TLR4 attenuates vascular remodeling in hypoxia-induced PH

TLR4−/− or PF4-TLR4−/− CH or SUH mice exhibited less pulmonary vascular remodeling compared to WT mice (Fig 4A). Morphometric analysis of peripheral arterioles revealed less thickening of the vessel wall in TLR4−/− or PF4-TLR4−/− mice vs. WT mice (Fig 4B). Likewise, there was less muscularization of peripheral arterioles in TLR4−/− or PF4-TLR4−/− mice vs. WT mice (Fig 4C–D). Plasma serotonin was increased in CH WT mice but not TLR4−/− or PF4-TLR4−/− mice. Coculture of HPASMC with LPS-stimulated WT platelets, but not TLR4−/− platelets promoted HPASMC proliferation (Fig 4E). This effect was blocked by GR127935, a 5HT1B receptor antagonist (Fig 4F). The diagram in figure 4G shows the proposed conceptual framework for how platelet TLR4 fits in to TLR4 and serotonin signaling in PH.

Figure 4. Genetic deletion of platelet TLR4 attenuates pulmonary vascular remodeling.

(A) Representative photomicrographs of SMA stained arterioles (scale bar = 25μm) from WT, TLR4−/− and PF4-TLR4−/− mice. (B) Quantification of wall thickness and the number of (C) partially and (D) fully muscularized arterioles (<100 μm) per 20x field (200x magnification). (E) Plasma serotonin levels in WT vs. TLR4−/− or PF4-TLR4−/− mice. Bars represent the mean ± SD. n=6–8 mice per group. *P<0.05 vs. WT normoxic. ^†P<0.05 vs. WT CH. ^‡P<0.05 vs. WT SUH (F) Effect of LPS stimulated WT or TLR4−/− platelets on HPASMC proliferation with or without the 5HT1B receptor antagonist GR127935. Data represent the mean ± SD of three experiments. *P<0.05. (G) Conceptual model of the role of TLR4 in PH and the serotonin hypothesis of PH.

DISCUSSION

In addition to their role in hemostasis and thrombosis, platelets are mediators of inflammation and immune responses¹³. They express several TLRs (TLR1,2,4,6,8,9)^{14, 15}, pattern-recognition receptors involved in the innate immune response by recognizing microbial structures and endogenous molecules released from damaged, stressed, or activated cells¹⁶. LPS, a TLR4 agonist, induces platelet P-selectin expression and IL-1β and ATP secretion and primes platelets to aggregate in response to low-dose thrombin¹⁷. Thus, platelet TLRs bridge innate immunity and coagulation. Previous studies from our laboratory demonstrate a role for TLR4 in the pathogenesis of PH^{7, 8}. The major finding of this study is that genetic deletion of TLR4 on platelets, but not myeloid cells, attenuates the pathogenesis of PH. Furthermore, platelets are activated to express surface P-selectin and release serotonin during hypoxia by TLR4-dependent mechanism(s). Thus, deletion of TLR4 globally or on platelets prevented CH or SUH-induced PH, in vivo platelet activation, and pulmonary vascular remodeling. Importantly, this study is the first to show that genetic deletion of a platelet surface receptor can attenuate PH.

There is an emerging role of platelet-derived mediators, such as serotonin, thromboxane-A2 (TxA2) and growth factors in patients with severe PH. These vasoactive mediators promote vasoconstriction (TxA2, serotonin), thrombosis (TxA2), and proliferation of vascular smooth muscle cells, endothelial cells, and fibroblasts (serotonin, platelet-derived growth factor). Further, platelet aggregation is enhanced by the altered balance of pro-aggregatory molecules (TxA2) and anti-aggregatory molecules (nitric oxide, prostacyclin).

The “serotonin hypothesis” of PH was postulated in the 1960s when it was discovered that women taking an indirect serotonergic agonist developed PH. More recently, it has been recognized that PH patients have markedly elevated plasma serotonin.¹⁸ Data demonstrate that serotonin released from ECs binds to serotonin receptors on PASMC or is taken up by PASMC via the serotonin transporter stimulating PASMC proliferation, migration and contraction, thus contributing to vascular remodeling in PH.¹⁹ Furthermore, mice deficient in bone morphgenetic protein receptor 2 (BMPR2) are more sensitive to serotonin-induced PH, which was associated with inhibition of Smad1/5 phosphorylation.²⁰ These data suggest that, in humans, increased serotonin could provide a “second hit” necessary for the development PH caused by BMPR2 haploinsufficiency. In this study we found that genetic deletion of TLR4 on platelets abrogated platelet activation and prevented the increase in plasma serotonin in two experimental models of PH. Furthermore, coculture of HPASMC with LPS-stimulated WT platelets, but not TLR4−/− platelets, promoted HPASMC proliferation via a mechanism that was dependent on the 5HT1B receptor. Together our data suggest that TLR4 plays a role in platelet activation and serotonin release in PH and that platelets are an important source of serotonin in PH.

It was surprising to find that loss of TLR4 on myeloid cells did not influence the disease course since myeloid cells are important responders to TLR4 ligands. It is possible that TLR4 on myeloid cells works towards counter purposes in PH thus masking the role of TLR4 on individual cell types. It may also be a limitation of the CH mouse model that the role of these cells in sensing endogenous TLR4 ligands is diminished or absent due to the mild inflammatory phenotype. Future studies will be necessary to sort out the role of TLR4 on myeloid cells in the pathogenesis of PH.

In summary, this study demonstrates the importance of platelet TLR4 in PH as its deletion from platelets improved disease outcome. These data suggests that platelet TLR4 is a proximate promoter of platelet activation and serotonin release in PH. We proffer that drugs interrupting TLR4 interaction with its endogenous ligands may limit platelet activation and inflammation, and lead to better therapies for PH.

Novelty and Significance.

What is Known?

• Toll-like receptor 4 contributes to the pathogenesis of PH via largely unknown mechanisms.

• Platelets participate actively in thrombus formation and produce (e.g. Thromboxane A2), store (e.g. Serotonin), and release mediators that may contribute to the initiation or aggravation of PH.

• Platelets express functional TLR4 and TLR4 agonists stimulate release of soluble mediators from platelets.

What New Information Does This Article Contribute?

• TLR4 on both parenchymal and non-parenchymal cells contribute to the pathogenesis of PH.

• Platelet TLR4 is a proximate promoter of platelet activation and serotonin release in PH.

• Serotonin released from platelets may contribute to increased pulmonary vascular smooth muscle cell proliferation and pulmonary vascular remodeling.

Despite evidence for thrombotic and inflammatory mechanisms in PH, no drugs have been developed to target these mechanisms. Recent data from out laboratory demonstrate a role for the innate immune receptor TLR4 in PH. We sought to determine the role of TLR4 on bone marrow derived cells in PH. Bone marrow chimera experiments revealed an equal contribution of TLR4 on non-parenchymal and parenchymal cells. While myeloid lineage cells are most often associated with TLR4 responses, deletion of TLR4 from these cells had no effect on PH. Platelets, in addition to their role in thrombosis, are mediators of inflammation and express functional TLR4. We, therefore, hypothesized that platelet TLR4 contributes to PH. Indeed, deletion of TLR4 on platelets attenuated platelet activation, serotonin release, pulmonary vascular remodeling and PH in two experimental models of the disease. These data demonstrate that platelet TLR4 is a proximate promoter of platelet activation and serotonin release in PH and suggests that drugs interrupting TLR4 interaction with its endogenous ligands may limit platelet activation and inflammation, and lead to better therapies for PH.

Nonstandard Abbreviations and Acronyms

PH

pulmonary hypertension

RV

right ventricle

RVSP

right ventricular systolic pressure

RVH

right ventricular hypertrophy

PF4

platelet factor 4

TLR

toll-like receptor

LPS

lipopolysaccharide

LysM

lysozyme

BM

bone marrow

TxA2

thromboxane-A2

NPC

nonparenchymal cells

ECs

endothelial cells

SMA

smooth muscle alpha actin