Evidence That Anemia Accelerates AS Progression Via Shear-Induced TGF-β1 Activation

Visual Abstract


SUMMARY
The severity of aortic stenosis (AS) is associated with acquired von Willebrand syndrome (AVWS) and gastrointestinal bleeding, leading to anemia (Heyde's syndrome).We investigated how anemia is linked with AS and AVWS using the LA100 mouse model and patients with AS.Induction of anemia in LA100 mice increased transforming growth factor (TGF)-b1 activation, AVWS, and AS progression.Patients age >75 years with severe AS had higher plasma TGF-b1 levels and more severe anemia than AS patients age <75 years, and there was a correlation between TGF-b1 and anemia.These data are compatible with the hypothesis that the blood loss anemia of Heyde's syndrome contributes to AS progression via WSS-induced activation of platelet TGF-b1 and additional gastrointestinal bleeding via WSS-induced AVWS.A ortic stenosis (AS) is a progressive disorder, and so the severity increases with age, with 4% to 13% of individuals over the age 75 years having severe AS. 1,2 As the U.S. population ages, the prevalence of AS is expected to increase dramatically.
AS is a degenerative chronic progressive disorder characterized by thickening of the valve leaflets caused by fibrosis and calcification, leading to narrowing of the aortic orifice with obstruction of left ventricular outflow. 3Narrowing of the valve opening results in increased wall shear stress across the valve during systole, increased cardiac preload, ventricular hypertrophy, and myocardial fibrosis, which ultimately result in heart failure, syncope, and sudden death. 4,5[8] AVWS has been described in 67% to 92% of patients with severe AS, with the severity of the abnormality in vWf being directly related to the severity of AS.
Notably, AVWS in AS patients is reversed after valve replacement. 9Because a diagnosis of AVWS requires specialized testing, Heyde's syndrome is likely underdiagnosed in AS patients. 7,10Anemia itself has been reported to be associated with more severe AS as well as increased all-cause mortality in medically treated AS patients. 11Mild anemia was observed in AS patients older than 75 years of age, and more severe anemia was observed in AS patients older than 80 years, 11 the age range in which Heyde's syndrome is more prevalent. 7,11The negative impact of anemia on AS prognosis has been ascribed to the direct and indirect effects of anemia on oxygen delivery, left ventricular hypertrophy, and cardiac fibrosis. 12,13wever, the mechanistic link between anemia and AS progression is not clear.
Transforming growth factor (TGF)-b1 is a multifunctional cytokine that affects many biological and pathological processes, including cell proliferation and differentiation, the immune response, and pathological tissue fibrosis. 14Polymorphisms in the TGF-b1 gene have been associated with AS, 15 and higher plasma levels of TGF-b1 have been reported in severe AS patients before undergoing valve replacement surgery. 16In addition to profound profibrotic effects, TGF-b1 has also been shown to induce calcification in explanted cardiac valve cells in vitro. 17Platelets contain 40 to 100 times more TGF-b1 than other cell types and rapidly release TGF-b1 upon activation. 18,19atelets release TGF-b1 as a latent complex in which mature TGF-b1 (25 kD) is noncovalently bound to latency-associated peptide (LAP), and LAP forms a disulfide bond via its cysteine residue 33 (C33) with latent TGF-b binding protein 1 (LTBP-1). 20Shear stress activates latent TGF-b1 released from platelets in vitro, and the binding of LAP to LTBP-1 is critical for shear stress-induced TGF-b1 activation. 18iol-disulfide exchange partially contributes to shear stress-dependent latent TGF-b1 activation, and protein disulfide isomerase (PDI) likely facilitates thiol-disulfide exchange during shear stress-induced TGF-b1 activation. 21ce with targeted inactivation of TGF-b1 in their megakaryocytes and platelets are partially protected from developing cardiac hypertrophy, fibrosis, and systolic dysfunction in a pressure overload transverse aortic constriction model, 22 and are partially protected from AS progression in a murine hypercholesterolemic model (low-density lipoprotein receptor [LDLR] À/À apolipoprotein [Apo]B 100 ; LA100) in which mice spontaneously develop AS. 22,23 Because platelets can also be activated by wall shear stress, 24 AS may activate platelets to release latent TGF-b1, and the latter can subsequently be activated by wall shear stress, creating a vicious cycle that contributes to AS progression.
We hypothesized that anemia may bring Heyde's syndrome full circle, potentiating the progression of AS by further increasing wall shear stress, leading to release and activation of platelet-derived TGF-b1.EPARATION OF HUMAN PLASMA SAMPLES.Blood samples were collected in vacuum tubes containing sodium citrate (Vacutainer, BD BioSciences) and immediately centrifuged at 12,000 g for 5 minutes at 4 C to prevent in vitro release of TGF-b1 from platelets as previously described. 26As an additional quality control measure to eliminate samples that had undergone in vitro release of a-granule contents, plasma samples were analyzed by immunoblot for the platelet-specific proteins PF4 and TSP-1, which are contained in a-granules.Plasma samples from healthy control subjects prepared immediately after blood drawing had low TGF-b1 levels (1-3 ng/mL) and nearly undetectable levels of PF4 and TSP-1 antigen. 26Based on this data, we excluded samples with elevated levels of TSP-1 and PF4 from analysis.
and blood counts were assessed using an automated dual-angle laser detection system (HEMAVET HV950; Drew Scientific, Inc).The citrated blood was centrifuged at 12,000 g for 5 minutes at room temperature immediately after blood drawing, and the plasma was frozen at À80 C until assayed for TGF-b1.

ASSESSMENT OF AS IN MOUSE MODELS BY
ECHOCARDIOGRAPHY.Ultrasound echocardiography imaging was obtained for clear images of aortic valve cusps; an aortic arch view was obtained using a modified method (Supplemental Figure 3A, bottom).
We also used an improved method to measure peak blood flow velocity across the valves (jet) with a combination of color and pulse-wave Doppler that allows alignment along the maximum flow direction (Supplemental Figure 3A, top), as described in detail in our previous study. 23Our echocardiography method directly measures both the area and velocity and is relatively easy to perform even in mice with much faster heart rates than humans.and the RBC-specific protein TER119 and were imaged using a Ziess 710 confocal microscope to obtain z-stacks and tiled pictures for colocalization analysis.
Valve tissues were processed for scanning electron microscopy imaging using a FEI Quanta 600 electron microscope, as previously described. 23ASUREMENT OF TGF-b1.Total TGF-b1 in human and mouse plasma was measured with an antibody enzyme-linked immunosorbent assay (ELISA) specific for the activated form of TGF-b1 (R&D Systems) after converting latent TGF-b1 to active TGF-b1 by acidification (20 minutes incubation at RT with 0.5 volume of 1 N HCl, followed by neutralization with the same volume of 1.2 N NaOH in 0.5 mol/L HEPES), as described previously. 18Active TGF-b1 was measured by directly applying plasma without acid activation and dilution using the same ELISA kit.For mouse samples, the statistical analysis was performed using SAS version 9.4 (SAS Institute, Inc).
We first calculated log 10 transformation for total TGF-b1 and active TGF-b1 values.Then, for the transformed total TGF-b1 and active TGF-b1 variables as well as Hb, cusp separation, fractional valve opening, and WSS, we first used linear mixed effects models to examine their association with bleeding time or scan time, age, Hb, and the corresponding interaction terms.We used the multiple linear regression model for total TGF-b1, active TGF-b1, and Hb.Collinearity among predictors was examined.In addition, we also   Subramani et al  1C).No difference was observed in TGF-b1 levels between age groups in patient control subjects.
We also found that older AS patients had lower Hb levels than younger AS patients (Figure 1D) No difference in Hb levels was observed between the older and younger patient control subjects (P ¼ 0.57).

TGF-b1 IS INVERSELY CORRELATED WITH HB IN OLDER AS PATIENTS AND IS ASSOCIATED WITH AS
PROGRESSION.We also observed an inverse correlation between plasma TGF-b1 levels and Hb in older AS patients (Figure 1E) (r ¼ À0.67; P ¼ 0.008), but not younger patients.The inverse correlation between plasma TGF-b1 levels and Hb in older AS patients remained significant even after adjusting for age and sex (r ¼ À0.57; P < 0.01) (Figure 1F).
Using a logistic regression model in older AS patients (age >75 years), factors found independently associated with AS were increasing age (OR: 1.32; P ¼ 0.020) and total TGF-b1 levels (OR: 4.08; In terms of TGF-b1 levels, this OR means that going from 1 to 1.5 ng/mL (roughly the first and third quartiles of the observed values for TGF-b1), the risk increases 2.88-fold.Age also was a risk factor in patients age #75 years; age was still associated with AS (OR: 1.13; P < 0.001) but not TGF-b1.A unified risk COMPARED WITH CONTROL MICE.To assess whether plasma TGF-b1 was also elevated in murine models of AS, blood was collected from hypercholesterolemic LA100 mice which spontaneously develop AS with aging, most commonly after approximately one year.
We found that average plasma TGF-b1 levels in LA100  2D).(Figure 3B).LA100 mice that underwent repeated large-volume (300-500 mL) phlebotomies developed progressive anemia with time, with Hb levels that were negatively correlated with the number of phlebotomies in both old and young mice (P < 0.0001) (Figure 3C).Plasma TGF-b1 levels (both total and active) increased progressively with phlebotomy number, as did anemia in both old and young groups (P < 0.0001) (Figure 3C).Anemia was associated with significant AS progression (ie, decreased fractional valve opening, increased aortic valve thickness and increased peak velocity, and wall shear stress) in both young and old LA100 mice after the sixth and eighth phlebotomies, respectively, compared with baseline (P < 0.001) (Figure 3D, Supplemental Figure 3).To evaluate the platelet contribution to the increased plasma TGF-b1 levels induced by phlebotomy, we induced anemia using low-iron diet and large-volume (300-500 mL) phlebotomies (Figure 4A) in a cohort of WT-C57Bl/6 mice (n ¼ 7) and mice deficient in platelet TGF-b1 (PF4CreTgfb1 flox/flox ; n ¼ 8). 22Hb levels decreased equally in both groups with phlebotomy (Figure 4B).In agreement with the data from our previous study, 22 TGF-b1 levels were 35% to 50% lower in the PF4CreTgfb1 flox/flox mice at baseline and after anemia induction (Figure 4C).
These data suggest that platelets are a major source of the increased TGF-b1 induced by phlebotomy.

LOSS OF PLATELET TGF-b1 REDUCES AS PROGRESSION IN THE PRESENCE OF MILD TO MODERATE ANEMIA.
Whereas age-related anemia in both LA100 without phlebotomy and AS patients is in the mild to moderate range, we also tested the effect of loss of platelet TGF-b1 in the presence of anemia.We induced mild to moderate anemia by feeding a low-iron diet for  8 weeks without phlebotomy to both LA100 mice and LA100 mice with deficiency in platelet TGF-b1 (LA100;PF4CreTgfb1 flox/flox ) (Figure 4D).Average Hb levels decreased in both groups (from average 12.5 at baseline to 8.3 g/dL in 5 LA100 mice and 12.3 to 9.8 g/dL in 4 LA100;PF4CreTgfb1 flox/flox mice after 8 weeks of low-iron diet) (Figure 4E).In association with the development of anemia, TGF-b1 levels increased significantly in the LA100 mice after 8 weeks of low-iron diet (Figure 4F) (P ¼ 0.0006).In contrast, TGF-b1 levels were 50% lower in the PF4CreTgfb1 flox/flox mice at baseline and after anemia induction (Figure 4F) (P ¼ ns).The LA100 mice showed significant progression of both AS indicators during the 8 weeks of a low-iron-diet, whereas LA100;PF4CreTgfb1 flox/flox mice showed less AS progression despite these mice being 5 weeks older than control LA100 mice (LA100 mice had lower fractional valve opening [P ¼ 0.005] and higher wall shear stress [P ¼ 0.004] values at baseline than the LA100; PF4CreTgfb1 flox/flox mice) (Figure 4G).

PRODUCING ANEMIA IN LA100 MICE INDUCES SEVERE
ACQUIRED vWf SYNDROMES.Since the acquired form of von Willebrand syndromes (AVWS) is characterized by a loss of the HMW vWf multimers, we quantified vWf multimer size in older LA100 mice before and after phlebotomy.We found a dramatic reduction of vWf multimer size as the mice became anemic, with almost complete loss of HMW vWf multimers after the eighth phlebotomy (Figures 5A and 5B and Supplemental Figure 4).We found negative correlations between HMW vWf multimers and both total and active TGF-b1 (Figures 5C and 5D).We also found a positive correlation between HMW vWf multimers and Hb levels (r ¼ 0.561; P ¼ 0.020).

ACTIVATED PLATELETS ARE PHYSICALLY ASSOCIATED
WITH VALVULAR ENDOTHELIAL CELLS.We imaged whole-mount aortic valves of AS patients and LA100 mice and found that platelet aggregates and RBCs attached to valvular cells (Figures 5E and 5F).
Scanning electron microscopy imaging of the same We previously demonstrated that latent TGF-b1 can be activated by shear, but that a mutant latent TGF-b1 protein that cannot form the disulfide bond with the LTBP-1 via its Cys33 residue because of a C33S mutation in the TGF-b1 precursor did not undergo shearinduced activation to the same extent as WT latent TGF-b1. 18To better assess the role of shear-induced TGF-b1 activation in AS progression, we generated LA100 mice expressing a heterozygous mutant Figure 1A).Serum-containing latent TGF-b1 from LA100;C33S þ/À mice older than 60 weeks of age produced w17% lower TGF-b1 activation by shear (17.16% in LA100 vs 14.25% in LA100;C33S þ/À ) than littermate control subjects (Figure 6A, Supplemental Figure 1B).They also showed significantly lower AS progression (Figure 6B) as measured by cusp separation (fractional valve opening) and wall shear stress (Figure 6C) compared with littermate LA100 mice.
Aortic valves of these mice had lower collagen expression and p-Smad2 levels compared with We previously found that shear-induced activation of TGF-b1 involves thiol-disulfide exchange and PDI contributes to the thiol disulfide exchange and latent TGF-b1 activation in vitro. 21To assess the role of platelet-derived PDI in AS progression, we genetically depleted PDI expression in megakaryocytes/platelets by crossing LA100 mice with PF4CrePDI flox/flox mice to generate LA100;PF4CrePDI flox/flox mice as shown in Supplemental Figure 2. 23 Immunoblotting of platelet lysates demonstrated that PDI levels were more than 90% lower (Figure 6D) in LA100;PF4CrePDI flox/flox mice than littermate LA100 control mice.The latter mice also had less AS progression (Figure 6E) as measured by fractional valve opening and wall shear stress (Figure 6F).we induced anemia in LA100 mice to better define the potential impact of anemia on TGF-b1 levels and AS.

DISCUSSION
We found that anemia in these mice was associated with higher TGF-b1 levels and with more rapid progression of AS.Moreover, mice with targeted deletion of platelet TGF-b1 were protected from AS progression and elevation of plasma TGF-b1 levels associated with induced anemia, either severe or moderate.We also studied the potential platelet contribution to AS progression by analyzing the impact of impairing the ability of latent platelet TGF-b1 to be activated by shear force because it is not bound to an LTBP-1. 18ese mice also showed less AS progression compared with control subjects, as did mice with a targeted deletion of PDI from platelets.These data build on our previous studies showing the following: 1) shear force can activate latent TGF-b1 released from platelets in vitro; 18 2) plasma TGF-b1 levels increase over time in parallel with wall shear stress in the LA100 mouse model of AS; 23 3) megakaryocyte and platelet-derived TGF-b1 accounts for w45% of plasma TGF-b1 as judged by mice with targeted deletion of platelet TGF-b1; 22 and 4) thiol-disulfide exchange mediated by PDI contributes to activation of TGF-b1. 18,21 trying to link the anemia to the development of increased TGF-b1 levels and AS progression, it is notable that anemia also led to altered wall shear stress, 31 which may contribute to the release and activation of latent TGF-b1.One mechanism potentially linking anemia to increased wall shear stress is a compensatory increase in cardiac output (Supplemental Figure 5) and decreased systemic vascular resistance, 32 because blood velocity is a key component of wall shear stress.Support for the physiologic significance of the increased shear stress was its correlation with loss of HMW vWf multimers, because shear stress alters the conformation of vWf, making it more susceptible to cleavage by ADAMTS-13. 33Together, these human and mouse data are in agreement with the report by Kearney et al. that anemia is an independent risk factor for progression to severe AS. 34 Our findings thus convert Heyde's syndrome from a linear model consisting of AS leading to increased shear, leading to AVWS via unfolding of vWf and cleavage by ADAMTS-13, 33 leading to gastrointestinal hemorrhage and anemia, to a vicious circle in which the anemia itself further increases shear, leading to both release of TGF-b1 from platelets and activation of released TGF-b1, resulting in further AS progression (Figure 7).
Both inherited and acquired forms of von Willebrand disease have been associated with vascular abnormalities, especially angiodysplasia of the gastrointestinal tract. 35It has been proposed that vWf interactions with endothelial cell aVb3 inhibits signaling by VEGF through VEGFR2, and so loss of vWf results in abnormal proliferative angiogenesis.
Inherited von Willebrand disease may, in addition, result in angiodysplasia via a mechanism related to loss of vWf in Weibel-Palade bodies. 35In fact, 2 of We recognize that in addition to gastrointestinal hemorrhage, intravascular hemolysis of RBCs as they traverse a stenotic valve at a very high shear stress may also contribute to anemia. 36This intravascular hemolysis may also result in the release of iron, which has been found to be deposited on aortic valves in association with calcification. 36If correcting anemia diminishes wall shear stress, it may have the added advantage of decreasing any ongoing hemolysis.Serum lactate dehydrogenase and plasma haptoglobin levels are valuable biomarkers of AS-associated hemolytic anemia. 37atomic support for a role for platelets comes from our finding of platelet aggregates in both human AS patient and LA100 mouse valve leaflets by whole mount immunofluorescent and scanning electron microscopy imaging.Our data are also consistent with previous observations that activated platelets are physically associated with valvular cells. 23,38In fact, Ozawa et al 39 recently showed that LDLR À/À (without ApoB100) mice lacking ADAMTS13 have increased aortic endothelial vWF, enhanced platelet adhesion, and evidence for TGF-b1 pathway activation in association with the development of hemodynamically significant AS.In patients with critical aortic stenosis, increased wall shear stress reduces large vWf multimers and leads to the development of acquired von Willebrand syndrome, 9 which is also common in patients with left ventricular assist device implantation.
It is postulated that both AS and left ventricular assist device generate high shear force, resulting in vWf unfolding and cleavage by ADAMTS13 or perhaps other mechanisms leading to impaired hemostasis. 40 fact, our previous study found an inverse correlation with platelet-released TGF-b1 and loss of vWf multimer size in heart failure patients implanted with LVAD. 26UDY LIMITATIONS.First, the experimental anemia produced by both an iron-deficient diet and phlebotomy was more severe than the anemia experienced by patients with AS.To address this, we also tested animals subjected to more moderate anemia and found similar findings with regard to the protective effect of targeted loss of platelet TGF-b1.
We did not study the potential contribution of other cells, including monocytes/macrophages, which can also secrete TGF-b1.Because both AS and anemia increase with age, other age-related metabolic and inflammatory conditions may also contribute to AS progression, and future studies are needed to address these issues.Our study is also limited by the relatively small AS patient sample size and the retrospective nature of the data collection.Similarly, the under-representation of women (n ¼ 11, 20%) in this study limits its generalizability.We attempted to partially address these limitations by prospectively studying several mouse models in which we included more than 40% female mice in the analysis.

CONCLUSIONS
We demonstrate that anemia, higher TGF-b1 levels, and AVWS are observed in older AS patients and LA100 mice, and the strong negative correlation between Hb and TGF-b1 and vWf multimers suggests that the anemia-induced increase in wall shear stress might be responsible for elevated activated TGF-b1 levels that then contribute to AS progression.This (J Am Coll Cardiol Basic Trans Science 2024;9:185-199) © 2024 The Authors.Published by Elsevier on behalf of the American College of Cardiology Foundation.This is an open access article under the CC BY license (http://creativecommons.org/ licenses/by/4.0/).
vWf = von Willebrand factorThe authors attest they are in compliance with human studies committees and animal welfare regulations of the authors' institutions and Food and Drug Administration guidelines, including patient consent where appropriate.For more information, visit the Author Center.Manuscript received May 3, 2023; revised manuscript received September 7, 2023, accepted September 7, 2023.

PATIENTPOPULATION.
Informed consent was obtained from all participants in accordance with a protocol approved by the Rockefeller University, Mount Sinai Medical Center, and Oklahoma Medical Research Foundation Institutional Review Boards.Patients with severe AS (n ¼ 35) were recruited from Mount Sinai Medical Center based on either a mean aortic valve pressure gradient $40 mm Hg and/or a calculated aortic valve area #1 cm 2 .A separate patient control group (n ¼ 47) consisted of patients with cardiovascular disease undergoing cardiac catheterization or an electrophysiological study who did not have severe AS by echocardiography.Healthy control subjects (n ¼ 20) without a history of heart disease were recruited at Rockefeller University and Oklahoma Medical Research Foundation.ASSESSMENT OF AS IN HUMAN PATIENTS BY ECHOCARDIOGRAPHY AND CARDIAC CATHETERIZATION.2-dimensional transthoracic echocardiograms and cardiac catheterization were performed on AS patients and patient control subjects.Transaortic valve gradients were calculated by the modified Bernoulli equation or by direct measurement during cardiac catheterization.Aortic valve orifice areas were calculated either by the continuity equation from transthoracic echocardiogram data or the Gorlin equation from cardiac catheterization data.

TISSUE
PROCESSING AND WHOLE MOUNT STAINING FOR LIGHT, CONFOCAL, AND SCANNING ELECTRON MICROSCOPY IMAGING.Whole aortic valve tissues were fixed in 4% paraformaldehyde and pinned along with surrounding tissues on a dissecting petri dish using a high-power surgical microscope to open and flatten the aortic valve leaflets.Tissues were stained with antibodies to the platelet-specific protein PF4 vWf MULTIMERS ANALYSIS.Plasma vWf multimers were analyzed by electrophoresing plasma samples in discontinuous 1.5% agarose gels, followed by immunoblotting with an anti-vWf monoclonal antibody and detection by imaging and densitometry.High molecular weight (HMW) vWf multimers were defined as the area under the curve of all multimers above the 7th multimer band (starting from the bottom) expressed as a percentage of the area under the curve of all multimer bands as described previously.26STATISTICALANALYSIS. Statistical calculations were performed using SPSS 20 (International Business Machines Corp) and R version 2.1.12.For comparisons between case and control subjects, demographic characteristics were presented using the mean AE SD or median with 25th and 75th percentiles (Q1, Q3) for continuous variables and count (percentage) for dichotomous variables.Continuous variables were compared using Student's t-test for normally distributed data and the Mann-Whitney U test for non-normally distributed data.The Shapiro-Wilk test was used to assess whether the variables initially followed the normal distribution.Categorical variables were compared using the Fisher exact test.Linear mixed-effects models were used to compare changes over time across different groups.Specifically, we used mixed-effect models with time, group, and its interactions as fixed effects and a random intercept for each mouse, under a compound symmetric correlation structure.Comparisons of interest were assessed through contrasts, and the estimated marginal means and SEM are presented in summary plots.Correlation among continuous values was assessed using Pearson's correlation coefficient (r) or linear regression analysis.Because TGF-b1 levels were measured by enzyme-linked immunosorbent assay and the values were log normally distributed, we converted all values to log 10 for analysis.Data are shown as median (Q1, Q3), mean AE SD, or mean AE SEM.Multivariable linear regression models were used to evaluate TGF-b1 association with other variables, such as hemoglobin (Hb) levels, age, and gender as covariates.Variable selection was carried out using backward stepwise regression, and the optimal model was based on the Akaike information criteria (AIC). 29For the risk model, a logistic regression was fitted with the factors of interest through a generalized linear model and a similar variable selection procedure was applied.ORs and 95% CIs were estimated from this model.A 2-sided P value <0.05 was considered significant.Correlation coefficients referred to are Pearson's correlation unless otherwise specified.Results for Spearman's correlation were largely the same, as presented in the Results section.

J
A C C : B A S I C T O T R A N S L A T I O N A L SCIENCE VOL. 9, NO. 2, 2024

FIGURE 1
FIGURE 1 AS Patients Have Higher Plasma Levels of Total TGF-B1 Than Patient Control Subjects and TGF-B1 Levels Correlate With Age in AS Patients

FIGURE 2
FIGURE 2 LA100 Mice, Which Spontaneously Develop AS, Have Higher Plasma Total TGF-b1 and Lower Hb Levels Than Control Mice

J
A C C : B A S I C T O T R A N S L A T I O N A L SCIENCE VOL. 9, NO. 2, 2024 LA100 MICE >60 WEEKS OF AGE DEVELOPED SEVERE AS WHEREAS CONTROL MICE DID NOT.LA100 mice >60 weeks had more severe AS (n ¼ 34) as measured by decreased fractional valve opening (calculated by dividing cusp separation distance by left ventricular outflow tract distance) as previously described23 (Figure2E), and increased wall shear stress (Figure2F) compared with younger LA100 mice (n ¼ 17).No AS was observed in control mice (ApoE À/À or C57Bl/6), even at ages over 100 weeks (n ¼ 17-49) (Figures2E and 2F).

INDUCED
SEVERE IRON DEFICIENCY ANEMIA IN LA100 MICE LEADS TO HIGHER LEVELS OF PLASMA TGF-b1 AND AS PROGRESSION.The observed inverse correlation between plasma total TGF-b1 and Hb levels in AS patients and in LA100 mice led us to consider whether anemia itself results in increased plasma TGF-b1 levels.To test this hypothesis, we induced anemia by administering an iron-deficient diet and performing 7 repeated large-volume (300-500 mL) phlebotomies at the times described in the Methods section and shown in Figure 3A for up to

FIGURE 3
FIGURE 3 Anemia in LA100 Mice Is Associated With Higher Plasma Levels of Total and Active TGF-B1 and AS Progression b1 LEVELS IN THE SEVERE ANEMIA MODEL.

FIGURE 6
FIGURE 6 LA100;C33S þ/À and LA100;PF4CrePDI flox/flox Mice Are Partially Protected From AS Progression 10 anemia mice died and postmortem examinations revealed signs of bleeding in the gut as evident by whole mount immunostaining showing some areas with discontinuous blood vessels, RBC leakage, and fibrin deposition (data not shown).However, it is not clear whether the bleeding is a result of postmortem effects or spontaneous; thus, more work is needed to screen for angiodysplasia in older anemic mice with severe AS.Our data provide a potential mechanism to account for the clinical observation of very rapid progression of AS as it reaches critical stenosis.7,9This construct has potential therapeutic implications, because, if correct, anemic patients with AS may benefit from aggressive attempts to improve their anemia.Moreover, because the severity of the AVWS is a biomarker for increased shear stress, monitoring vWf multimers may provide a guide as to when the maximum benefit has been achieved by correcting the anemia.Our model also identifies plasma TGF-b1 as another potential biomarker of AS progression and inhibiting release of platelet TGF-b1 as a potential adjunct to slow the progression of AS.Future clinical studies will be required to test this hypothesis.Because valve replacement normalizes the shear gradient, it is expected to normalize TGF-b1 levels and the loss of HMW vWf multimers, which we plan to test in future studies.Since the current majority of valve replacement is by transcatheter aortic valve replacement, it would be interesting to compare TGF-b1 levels in patients before and after undergoing transcatheter aortic valve replacement vs surgical aortic valve replacement.

FIGURE 7
FIGURE 7 Heyde's Syndrome hypothesis has the potential to be tested clinically by studying the effect of aggressive correction of anemia in AS patients, perhaps using vWf multimers and/or TGF-b1 levels as biomarkers.PERSPECTIVES COMPETENCY IN MEDICAL KNOWLEDGE:AS leads to increased shear force, which in turn leads to cleavage of high molecular weight vWf multimers, which in turn leads to a bleeding diathesis, gastrointestinal bleeding, and anemia (Heyde's syndrome).Our data indicate that anemia itself may then accelerate AS progress via increasing shear-induced platelet-derived TGF-b1 release and activation, converting Heyde's syndrome into a vicious circle that may account for the rapid progression of AS near the end of its course.TRANSLATIONAL OUTLOOK: This proof-ofconcept study demonstrates that anemia itself may contribute to AS progression via its impact on release and activation on TGF-b1.Thus, correcting anemia and targeting platelet-derived TGF-b1 release and activation may slow AS progression.Plasma TGF-b1 and vWf multimers may be valuable biomarkers of AS progression.Additional studies are required to test these hypotheses.

TABLE 1
Patient Demographics 9alues are mean AE SD or n (%).Statistical analysis performed by either a Student's t-test or b Mann-Whitney U test.c Shear stress calculated according to the formula of Vincentelli et al.9AS ¼ aortic stenosis; TTE ¼ transthoracic echocardiography.