Platelet TGF-β1 contributions to plasma TGF-β1, cardiac fibrosis, and systolic dysfunction in a mouse model of pressure overload

Abstract

Circulating platelets contain high concentrations of TGF-β1 in their α-granules and release it on platelet adhesion/activation. We hypothesized that uncontrolled in vitro release of platelet TGF-β1 may confound measurement of plasma TGF-β1 in mice and that in vivo release and activation may contribute to cardiac pathology in response to constriction of the transverse aorta, which produces both high shear and cardiac pressure overload. Plasma TGF-β1 levels in blood collected from C57Bl/6 mice by the standard retro-bulbar technique were much higher than those obtained when prostaglandin E₁ was added to inhibit release or when blood was collected percutaneously from the left ventricle under ultrasound guidance. Even with optimal blood drawing, plasma TGF-β1 was lower in mice rendered profoundly thrombocytopenic or mice with selectively low levels of platelet TGF-β1 because of megakaryocyte-specific disruption of their TGF-β1 gene (Tgfb1(flox)). Tgfb1(flox) mice were also partially protected from developing cardiac hypertrophy, fibrosis, and systolic dysfunction in response to transverse aortic constriction. These studies demonstrate that plasma TGF-β1 levels can be assessed accurately, but it requires special precautions; that platelet TGF-β1 contributes to plasma levels of TGF-β1; and that platelet TGF-β1 contributes to the pathologic cardiac changes that occur in response to aortic constriction.

Figure 1

The ultrasound-guided LV puncture method of blood drawing minimizes in vitro release of TGF-β1, PF4, and TSP-1 from platelet α-granules into plasma. (A) Blood samples were drawn into sodium citrate anticoagulant by either the RB capillary technique in the absence (RB; n = 25) or presence of 1μM PGE₁ (RB + PGE₁; n = 10), or by the ultrasound-guided LV method (n = 15). Plasma samples were prepared by centrifuging blood at 3000g for 15 minutes at RT. Total TGF-β1 was measured by ELISA. (B) TSP-1 and (C) PF4 antigen levels were estimated by measuring immunoreactive band intensities on immunoblots. Each dot represents an individual mouse sample, and each color represents data obtained from the same mouse. P values reflect RB values compared with LV values. (D) Representative immunoblots with antibodies to TSP-1 and PF4 of plasma prepared by the RB method from 6 mice and by the LV methods from 5 different mice. Bottom panel: A nonspecific band, indicating that similar amounts of protein were loaded in each lane. (E) Total TGF-β1 was measured in plasma prepared from blood collected first by the LV method and subsequently by the RB method (left panel) or from blood collected in the opposite sequence (right panel).

Figure 2

Induction of profound thrombocytopenia by mAb 1B5 injection reduces plasma TGF-β1 levels in mice. WT C57BL/6 mice were injected intraperitoneally with the indicated amount of hamster mAb 1B5 against mouse αIIbβ3. (A) Platelet counts at the indicated time points after injection of 1B5. (B) Total TGF-β1 levels in plasma samples at the indicated time points as measured by ELISA. Each dot represents an individual mouse, and the different colors indicate samples from different mice. (C) Bleeding over the thorax and abdomen 24 hours after the 1B5 injection.

Figure 3

Megakaryocyte-specific targeted deletion of TGF-β1 reduces platelet, platelet releasate, serum, and plasma levels of total TGF-β1 but does not affect platelet function in vitro or in vivo. (A-B) Agonist-induced platelet aggregation was similar in control and Tgfb1^flox mice. (A) Platelet-rich plasma from control and Tgfb1^flox mice was stimulated with 1μM ADP, and light transmission was measured over time (1 of 3 similar experiments is shown). (B) Washed platelets from control and Tgfb1^flox mice were stimulated with 0.125 U/mL thrombin, and the aggregation response was measured over time (1 of 3 similar experiments). (C) FeCl₃-induced carotid artery thrombosis in vivo was similar in control and Tgfb1^flox mice. Thrombosis was initiated by applying 20% FeCl₃ to the carotid arteries of control and Tgfb1^flox mice, and the time to complete cessation of blood flow was recorded as the occlusion time (n = 4). (D) Total TGF-β1 levels in platelet lysates (PLT-Lys) and platelet releasates (PLT-Rel) from control mice (n = 4) or from Tgfb1^flox (n = 4) were measured by ELISA. (D) Serum total TGF-β1 levels in control (n = 19) and Tgfb1^flox mice (n = 10) were measured by ELISA. (E) Platelet lysates from control and Tgfb1^flox mice were immunoblotted with antibodies to TGF-β1 (top panel), αIIb (middle panel), and actin (bottom panel). (F) Serum samples from 4 control and 4 Tgfb1^flox mice were immunoblotted with antibody to TGF-β1. (G) Plasma total TGF-β1 levels in control (n = 45) and Tgfb1^flox (n = 25) mice were measured by ELISA.

Figure 4

Mice with megakaryocyte-specific gene targeting of TGF-β1 are partially protected from developing cardiac hypertrophy and fibrosis 4 weeks after TAC surgery. (A) Heart weight/body weight ratios as calculated by dividing the weight of the perfused heart by total body weight at the time of harvest (4 weeks after the surgery). (B-C) Mice were euthanized and hearts were perfused and fixed in formalin. Heart sections were stained with trichrome and scored for fibrosis by microscopy on a scale from 0 to 4. (B) Perivascular fibrosis scores. (C) Interstitial fibrosis scores. (D) Pretreatment and postsurgery values of plasma TGF-β1 in experimental animals measured by ELISA. (E) The correlations between plasma total TGF-β1 and heart weight/body weight ratio in individual WT mice (r = 0.66; P < .001). (F) The correlation between plasma total TGF-β1 and the combination of interstitial and perivascular fibrosis grade in individual TAC mice of both genotypes.

Figure 5

Mice with megakaryocyte-specific gene targeting of TGF-β1 have better preservation of systolic function than WT mice. (A) EF, (B) FS, (C) Vcf_c, and (D) SV were calculated from echocardiographic measurements in WT and Tgfb1^flox mice before (basal) and 1 and 4 weeks after TAC surgery. Total TGF-β1 levels at 1 week after surgery were inversely correlated with (E) FS and (F) Vcf_c.