Aging is linked to increased matrix metalloproteinase-9 (MMP-9) expression and extracellular matrix turnover, as well as a decline in function of the left ventricle (LV). Previously, we demonstrated that C57BL/6J wild-type (WT) mice > 18 mo of age show impaired diastolic function, which was attenuated by MMP-9 deletion. To evaluate mechanisms that initiate the development of cardiac dysfunction, we compared the LVs of 6–9- and 15–18-mo-old WT and MMP-9 null (Null) mice. All groups showed similar LV function by echocardiography, indicating that dysfunction had not yet developed in the older group. Myocyte nuclei numbers and cross-sectional areas increased in both WT and Null 15–18-mo mice compared with young controls, indicating myocyte hypertrophy. Myocyte hypertrophy leads to an increased oxygen demand, and both WT and Null 15–18-mo mice showed an increase in angiogenic signaling. Plasma proteomic profiling and LV analysis revealed a threefold increase in von Willebrand factor and fivefold increase in vascular endothelial growth factor in WT 15–18-mo mice, which were further elevated in Null mice. In contrast to the upregulation of angiogenic stimulating factors, actual LV vessel numbers increased only in the 15–18-mo Null LV. The 15–18-mo WT showed amplified expression of inflammatory genes related to angiogenesis, including C-C chemokine receptor (CCR)7, CCR10, interleukin (IL)-1f8, IL-13, and IL-20 (all, P < 0.05), and these increases were blunted by MMP-9 deletion (all, P < 0.05). To measure vascular permeability as an index of endothelial function, we injected mice with FITC-labeled dextran. The 15–18-mo WT LV showed increased vascular permeability compared with young WT controls and 15–18-mo Null mice. Combined, our findings revealed that MMP-9 deletion improves angiogenesis, attenuates inflammation, and prevents vascular leakiness in the setting of cardiac aging.
- extracellular matrix
aging is a major risk factor for cardiac mortality and morbidity (11, 30). Elderly patients with acute cardiovascular disease have poor clinical outcomes. Aging is associated with alterations in homeostatic mechanisms that make the vasculature more susceptible to damaging effects of pathophysiological conditions (35). Aging impairs diastolic function in humans, independent of existing comorbidities such as hypertension (18). In mice, aging associates with a subtle but significant decline in function of the left ventricle (LV) (21). Altered LV function is characterized by the development of diastolic dysfunction, as systolic function remains relatively unchanged (7). The cardiac extracellular matrix (ECM) regulates LV diastolic properties, provides structural support for the myocardium, and incorporates homeostatic elements such as growth factors (16). Matrix metalloproteinases (MMPs) are a family of zinc-dependent endopeptidases that regulate ECM turnover (19). MMP-9, in particular, degrades a wide spectrum of ECM substrates, cytokines, and growth factors. MMP-9 plays a key role in cardiac remodeling and contributes to LV dilation and excessive collagen accumulation in both aging hearts and postmyocardial infarction (10, 14).
We have shown that MMP-9 expression increases with age, concomitant with the development of diastolic dysfunction in wild-type (WT) mice after 18 mo of age (7). This decline in diastolic function was attenuated by MMP-9 deletion in old and senescent mice. MMP-9 deletion attenuated age-associated increase in transforming growth factor-β-induced gene and protein levels and attenuated increases in periostin and connective tissue growth factor expression, resulting in reduced cardiac fibrosis and improved diastolic function in MMP-9 null (Null) old and senescent mice (7). Since we have shown that functional differences in cardiac performance between WT and Null groups are not present in mice < 15 mo of age, but are apparent by 19 mo of age, the 15–18-mo-old time point may be a crucial transition stage in the aging process. Studying the extracellular remodeling process within this age group may elucidate mechanisms by which age related cardiac dysfunction develops.
The goal of this study, accordingly, was to examine mice before the development of diastolic dysfunction (i.e., in mice 15–18 mo of age) to identify early mechanisms that lead to the cardiac aging phenotype. Because of the critical role of MMP-9 in LV remodeling and inflammatory responses to aging and myocardial infarction, an improved understanding of MMP-9-dependent changes occurring in the LV during middle age should help identify potential therapeutic targets to preserve LV function.
MATERIALS AND METHODS
All animal procedures were performed based on the Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committees at the University of Texas Health Science Center at San Antonio and the University of Mississippi Medical Center. Adult C57BL/6J WT (n = 20, 10 male and 10 female) and Null (n = 21, 10 male and 11 female) mice of 15–18 mo of age were compared. We also compared the 15–18-mo-old mice to young 6–9-mo-old WT (n = 12, 6 male and 6 female) and Null (n = 12, 7 male and 5 female) mice that were a randomly selected subgroup from our previous study (7). The previous acquired echocardiography and Doppler data were combined with new gene array, histology, and protein expression results obtained on the tissue-banked samples collected for these mice as well as new samples collected for this study. An additional six WT and three Null mice of 6–9 mo of age and three WT and three Null mice of 15–18 mo of age were used for the vascular permeability analysis.
For echocardiography assessment, mice were anesthetized with 1 to 2% isoflurane in an oxygen mix. The heart rate, respiratory rate, and body temperature were continuously monitored to ensure the mice were not too deeply anesthetized to alter physiological variables. To measure the transmitral inflow (the blood flow from the left atrium into the LV via the mitral valve) and aortic outflow (the blood flow from LV via aortal valve), Doppler echocardiography was performed using the Doppler Signal Processing Workstation (Indus Instruments). Transthoracic echocardiography was performed using the Vevo 2100 system (VisualSonics). Measurements were taken from the LV parasternal long and short axes, B- and M-mode views. For each parameter, images from three cardiac cycles were measured and averaged (23).
Ten minutes after acquiring baseline echocardiographic variables, cardiac reserve was evaluated in stress echocardiograms by administering the β-adrenergic receptor agonist dobutamine (4 μg/g body wt ip). Echocardiograms were recorded 10 min after injection.
Mice were euthanized under isoflurane anesthesia. At death, heparin (4 μg/g body wt) was injected intraperitoneally, and 5 min after injection, blood was collected from the common carotid artery and centrifuged for plasma collection. The heart was flushed with cardioplegic solution and removed from the chest cavity (22). The right ventricle was cut away from the LV, each ventricle was weighed separately, and the LV was sectioned into three sections: base, mid, and apex. The base was snap frozen and used for RNA extraction, the middle section was fixed in 10% zinc formalin and paraffin-embedded for histological examination, and the apex was snap frozen and used for protein extraction.
Myocyte cross-sectional areas, myocyte numbers, and intermyocyte space were quantified.
LV sections were stained with hematoxylin-eosin. Five random regions from each slide were scanned at ×40 magnification, and 10 myocytes were measured from each section using Image-Pro Plus (version 7, MediaCybernetics). Only myocytes with central nuclei were measured, as described (21). Myocyte number and intermyocyte space were quantified from hematoxylin-eosin-stained sections using a custom MatLab-based program. Briefly, a color threshold was used to determine cell and nuclei positive pixels. Cell number was determined as the number of connected nuclei regions (as a percentage of total tissue area), whereas intermyocyte white space was calculated as the percentage of pixels that were neither cell nor nucleus.
Collagen content deposition was quantified.
LV sections were stained with picrosirius red as described (7). Six random regions from each slide were scanned at ×60 magnification, and the percentage of collagen area was measured using Image-Pro Plus version 7.0 software.
RNA extraction was performed from the LV using TRIzol reagent (no. 15596, Invitrogen) following the manufacturer's instructions. RNA concentration was quantified using a NanoDrop ND-1000 spectrophotometer (Thermo Scientific). Reverse transcriptase of equal RNA content (0.5 μg) was performed using the RT2 First Strand Kit (no. 330401, Qiagen). We quantified mRNA expression levels using commercially available real-time RT-PCR gene arrays for 84 extracellular matrix (PAMM-011E, list of genes included in the array can be found at http://www.sabiosciences.com/rt_pcr_product/HTML/PAMM-013A.html; Qiagen) or 84 inflammatory cytokines and receptors (PAMM-013E, list of genes included in the array can be found at http://www.sabiosciences.com/rt_pcr_product/HTML/PAMM-011A.html; Qiagen). The relative expression of individual mRNAs was calculated by normalization of the cycle time (Ct) values of the target mRNA to the average Ct of the Hprt1 housekeeping gene.
Plasma Proteomic Profiling
Plasma samples (100 μl) were analyzed using the rodent multianalyte profiling (MAP) version 2.0 (Myriad Rules-Based Medicine, Austin, TX). Concentrations of 60 analytes were measured by clinical laboratory improvement amendments-certified biomarker testing laboratory using multiplexed immunoassays.
Protein Extraction and Immunoblotting
Total LV protein was extracted in protein extraction reagent type 4 (7 M urea, 2 M thiourea, 40 mM Trizma base and the detergent 1% C7BzO; Sigma) and 1× protease inhibitor cocktail (Roche). Protein concentrations were determined by the Quick Start Bradford Protein Assay (Bio-Rad) using a 1:40 dilution of protein extract to dilute the urea to a compatible level. Total protein levels were measured using total protein stain (no, 24580, Pierce reversible protein stain kits for membranes, Thermo Scientific). Protein expression levels were quantified using a rabbit anti-VEGF primary antibody (sc-152-G, 1:1,000; Santa Cruz) or biotinylated anti-griffonia (bandeiraea) simplicifolia lectin I (GSL-I; B-1105, 1:100; Vector).
Total protein (10 μg) of each sample was separated on 4–12% Criterion XT Bis-Tris gels (Bio-Rad), transferred to nitrocellulose membrane (Bio-Rad), and stained with MemCode Reversible Protein Stain Kit (Thermo Scientific) to verify protein concentration and loading accuracy. After blocking with 5% nonfat milk (Bio-Rad), the membrane was incubated overnight at 4°C with primary antibody, followed by incubation with goat anti-rabbit (sc-2004, 1:5,000; Santa Cruz) secondary antibody. Signal detection was performed using Pierce ECL Western Blotting Substrate or SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientifics) and imaged on the GE Image Quant LAS4000 luminescent image analyzer.
Protein expression in each blot was normalized to the total protein levels in the respective lane. The signal intensity of the molecular weight marker was used to normalize results between blots.
Endothelial cells were detected by staining sections with biotinylated GSL-I (100 μg/ml). After slides were incubated with GSL-I, the avidin-biotinylated enzyme complex was added, followed by 3,3′-diaminobenzidine chromogen. The LV sections were scanned at 40× magnification, and five random fields from each section were scanned and quantified. GSL-I staining was quantified by using Image-Pro software (Media Cybernetics) to calculate the total GSL-I area-stained positive (brown 3,3′-diaminobenzidine staining) per total section area. The data are presented as the area of GSL-I staining per total area.
Mice were anesthetized with 1 to 2% isoflurane in an oxygen mix and placed on a heated surgical board (Indus Instruments). FITC-labeled 70 kDa dextran (no. 46945, 50 mg; Sigma-Aldrich) was diluted in 1 ml of sterile saline (50 μg/μl final concentration), and 100 μl (5 mg) was injected into the tail vein (24). Mice were euthanized 5 h after injection, and the LV was divided into two sections. The apex half was snap frozen in liquid nitrogen and later homogenized for immunofluorescent biochemical analysis, and the base half was placed in optimum cutting temperature solution and snap frozen in liquid nitrogen for cryosectioning and histological evaluation.
Snap frozen LV tissue was homogenized in 1× phosphate-buffered saline (16 μl of PBS per 1 mg tissue). Protein concentrations were determined by the Quick Start Bradford Protein Assay (Bio-Rad). Immunofluorescence was measured from 100 μg of protein using SpectraMax M3 plate reader (485 nm excitation, 525 nm emission).
For immunofluorescence, frozen optimum cutting temperature-embedded tissues were sectioned at 10-μm thickness, kept at −20°C, and protected from light. Nuclei were stained with 4′,6-diamidino-2-phenylindole (H-1200; Vector). FITC signal was acquired at 15-ms exposure, and 4′,6-diamidino-2-phenylindole signal was acquired at 797-μs exposure for all sections. All sections were scanned at ×20 magnification, five random fields were scanned per slide, and three slides per tissue were examined.
All analyses were performed by evaluators blinded to experimental groups. Data are expressed as means ± SE. All samples were tested for normality. Multiple group analysis was performed using ANOVA with Student-Newman-Keuls posttest or the non-parametric Kruskal-Wallis ANOVA test with Dunn's posttest. A value of P < 0.05 was considered statistically significant. GraphPad InStat software was used for the statistical analyses.
LV MMP-9 Increased with Age
Consistent with our previous results, we observed a twofold increase in MMP-9 gene expression in 15–18-mo WT mice compared with 6–9-mo WT mice (P < 0.05; Fig. 1). MMP-9 protein levels also increase with age in the LV because of increased macrophage infiltration (6, 7).
Myocyte Hypertrophy, but not LV Dysfunction or Fibrosis, Had Already Developed in the 18-mo-old Mice
We measured multiple echocardiographic variables to comprehensively evaluate LV function in 15–18-mo WT and Null mice compared with 6–9-mo mice. Echocardiographic analyses showed no significant differences in LV dimensions or systolic and diastolic variables between 15–18-mo WT and Null mice compared with each other or their 6–9-mo young controls (Table 1). Our data indicated that functional changes in the LV occur after 15–18 mo of age.
We further investigated whether aged 15–18-mo mice displayed signs of myocyte hypertrophy by analysis of hematoxylin-eosin-stained sections. WT and Null 15–18-mo mice showed an increase in nuclei count, and both WT and Null 15–18-mo mice showed a decrease in intermyocyte space compared with young controls (Table 2). Myocyte cross-sectional areas were increased in both WT and Null 15–18 mo mice compared with young controls (Fig. 2). The increase in individual myocyte cross-sectional area was not accompanied by an increase in LV mass-to-body weight ratio (Table 2). The increases in nuclei number and individual myocyte size in the WT and Null 15–18-mo mice suggest myocyte hypertrophy that has not yet progressed to a level reflected by a global increase in LV mass.
Collagen deposition was not altered among the groups (P = 0.68), indicating that cardiac fibrosis develops after 18 mo of age. We have reported that collagen deposition was increased in mice over 26 mo of age, and this increase was attenuated by MMP-9 deletion (7).
MMP-9 Deletion Improved Angiogenesis in Aged Mice
An increase in myocyte size would suggest a need for additional blood vessels to supply the increase in oxygen demand. We investigated whether angiogenic stimuli were increased in WT and Null 15–18-mo mice compared with young controls.
Out of 84 ECM genes measured, we observed an increase in the expression of 17 genes and a decrease in 13 genes (Fig. 3A). Aging in WT 15–18-mo mice involved upregulation of cadherin-1 (Cdh1) and downregulation of integrin αV (Itgav), both of which were MMP-9 dependent as the Null 15–18-mo mice did not show similar changes. Increased expression of Cdh1 in the WT 15–18-mo mice suggests an increasing attempt to preserve vascular permeability in the aged LV. Integrins are essential components for angiogenesis, and the age-dependent decrease in Itgav expression suggests an inhibitory effect of MMP-9 on angiogenesis. We also observed an age-dependent increase in the expression of tissue inhibitor of metalloproteinase (TIMP)-3 only in Null 15–18-mo mice (Fig. 3). TIMP-3 inhibits angiogenesis, suggesting a need to attenuate angiogenesis in the aged Null mice.
Out of 60 plasma analytes measured, null mice showed age-dependent increases in the levels of stem cell factor (SCF) and vascular cell adhesion molecule 1 (VCAM1). Both WT and Null 15–18-mo mice showed higher von Willebrand factor (vWF) levels, which were enhanced in Null mice (Fig. 4). The increase in vWF is indicative of an increased stimulation for angiogenesis in the WT and Null 15–18-mo mice, with a more pronounced effect in the Null, whereas SCF and VCAM1 showed increased angiogenesis stimulation only in the Null 15–18-mo mice. These results indicate an increase in angiogenic stimuli as a function of age, and this increase is actually enhanced by MMP-9 deletion.
Because there was a net increase in angiogenic signaling in both the WT and Null, we quantified LV blood vessel numbers. With aging, WT mice did not show an increase in the number of vessels as evidenced by GSL-1 immunostaining or immunoblotting. In contrast, Null mice showed significant increases in vessel numbers in the aged LV (Fig. 5, A–C). Both WT and Null LV showed age-associated increases in VEGF protein expression, which was significantly higher in the Null aged mice compared with corresponding WT and 6–9-mo controls (Fig. 5D). These data suggest a disconnect between angiogenic potential and realization, as the age-associated increase in angiogenic potential was not matched by an concomitant increase in vessels numbers in the WT mice.
MMP-9 Deletion Attenuated Age-Associated Increase in Vascular Permeability
To further investigate the effects of aging on vascular function, we injected mice with FITC-labeled dextran to measure vessel permeability. We hypothesized that WT mice would have higher vascular permeability with age and that MMP-9 deletion would prevent this loss of vessel integrity. Aged WT mice showed significantly increased vascular permeability compared with that in 6–9-mo WT mice, evidenced by increased immunofluorescent histochemical and spectrophotometric signals (Fig. 6). There was a 37% increase in vascular permeability in the aged WT mice that was significantly attenuated by MMP-9 deletion. These results indicate the increased vascular leakiness in the aged WT mice occurred through an MMP-9-mediated mechanism.
MMP-9 Deletion Attenuated Age-Associated Inflammation
Out of 84 inflammatory genes measured, we observed age-associated increases in 10 inflammatory genes and decreases in 12 inflammatory genes in WT and Null 15–18 mo mice (all P < 0.05; Fig. 7A). MMP-9 deletion attenuated age associated increase in C-C chemokine receptor (CCR7), CCR10, IL-1f8, IL-13, and IL-20 (Fig. 7B). These data indicate increased inflammation in 15–18 mo WT mice, which was blunted by MMP-9 deletion. Interestingly, several of these inflammatory genes also play roles in angiogenesis, which ties this result in with the above findings.
In this study, we investigated whether MMP-9-dependent mechanisms initiate the cardiac aging phenotype before the development of diastolic dysfunction occurs. The major findings were 1) both WT and MMP-9 Null mice showed increases in myocyte hypertrophy and expression of proangiogenic stimuli; however, an actual increase in vessel numbers to accommodate the increase in myocyte size was observed only in the Null mice; 2) WT mice demonstrated compromised vascular permeability, indicative of endothelial dysfunction, which was preserved in the Null LV; and 3) aging triggered an inflammatory response in WT because of the inadequate angiogenic environment. Combined, our results illustrate the mechanism through which an early myocyte-vessel mismatch leads to increased inflammation, fibrosis, and the development of diastolic dysfunction in old mice (Fig. 8).
As others have reported, we observed an increase in myocyte hypertrophy with age, indicated by an increase in cross-sectional area and a decrease in intermyocyte space (3, 4). While individual myocyte size increased, the LV mass-to-body weight ratio was not different among genotypes and age groups. The increase in nuclei number and myocyte size in the WT and Null 15–18-mo mice suggest that cardiac myocyte hypertrophy has not yet progressed to a level reflected by a global change in LV mass. Another potential explanation could be a dropout of myocyte numbers with age that is compensated by the increase in individual myocyte size (38). We observed an age-associated increase in myocyte nuclei numbers, indicating nuclear multiplication occurred (27). The age-associated myocyte hypertrophy increases oxygen demand that, if not met by an increase in supply, could lead to an imbalance in the myocardium that yields a hypoxic environment.
Because of the increase in myocyte hypertrophy, we measured angiogenic stimuli to see whether signaling was enhanced with aging as a means to increase oxygen supply. Indeed, we observed an increase in several proangiogenic stimuli circulating in the plasma, namely, vWF, SCF, and VCAM-1. Levels of vWF increased in both WT and Null 15–18-mo mice, and this increase was significantly higher in the aging Null mice compared with the aging WT mice. vWF is an essential angiogenesis regulator, stimulating activation of the endothelium (37). Additionally, Null 15–18-mo mice showed increased levels of SCF and VCAM-1 compared with young and age-matched controls, suggesting that the angiogenic stimulus was greater in the absence of MMP-9 (8, 33). These data demonstrate an increased systemic proangiogenic potential, which is enhanced by MMP-9 deletion.
In addition to the circulating angiogenic stimuli observed in plasma, we also showed an age-associated increase in Cdh1 and decrease in Itgav in WT LV. Cadherins help to maintain blood vessel integrity (5). MMP-9 has been reported to downregulate cadherins through proteolytic shedding, which would increase vascular permeability (9, 32). The absence of a similar increase of Cdh1 in the Null mice suggests that the increase in the WT may be a compensatory mechanism to improve impaired vascular permeability. Integrins are an essential component for angiogenesis; in particular, Itgav plays a supporting role in angiogensis. Itgav expression decreased in WT 15–18-mo mice compared with young controls, whereas its expression levels unchanged in null mice, suggesting that MMP-9 in the WT mice may serve to block Itgav synthesis or increase its degradation (1).
TIMP-3 has been shown to be a negative regulator of angiogenesis in diseases with prominent neovascularization (2, 28). Specifically, TIMP-3 inhibits chemotaxis of vascular endothelial cells to vascular endothelial growth factor and basic fibroblast growth factor, inhibits capillary morphogenesis in vitro, and inhibits basic fibroblast growth factor-induced angiogenesis in vivo (2). We observed an increase in TIMP-3 expression in the Null LV, which could be a turnoff signal as a result of the increase in angiogenesis (2).
Cardiac aging is associated with a decline in the maximal oxygen uptake and, as a consequence, a decline in cardiac performance (26). One possible mechanism to improve oxygen supply to the myocardium is to increase the number of blood vessels. Our study showed that both WT and Null 15–18-mo mice had an increased angiogenic potential; however, only Null mice showed an actual increase in blood vessel numbers in the LV. The fewer blood vessels in the aged WT mice was not necessarily explained by an age-associated increase in MMP-9 proteolytic activity, since blood vessel numbers were similar between the 6–9-mo and 15–18-mo WT groups. MMP-9 degrades ECM, including the basement membrane surrounding blood vessels, and could thus contribute to the loss in blood vessel integrity. We did observe an increase in vessel leakiness in the 15–18-mo WT LV, indicating that the function of those vessels was impaired. A deficit in vessel number and integrity could trigger an inflammatory reaction. This finding indicated that WT mice were desensitized to angiogenic stimuli, whereas MMP-9 deletion resulted in improved angiogenesis.
In our study, we showed that WT 15–18-mo LV had increased levels of CCR7, CCR10, IL-1f8, IL-13, and IL-20 inflammatory gene expression, which were attenuated by MMP-9 deletion. CCR7 and CCR10 interact with chemokine (C-C motif) ligand 27 to promote T cell-mediated inflammation (13). IL-1f8 is associated with an increased production of inflammatory mediators (17, 34). Expression of IL-13 is associated with increased inflammation (39). IL-20 is produced by monocytes as well as nonimmune cells during inflammation (31). These data demonstrate that aging increases the inflammatory environment within the LV, and MMP-9 deletion attenuates this age-associated inflammation. Increased inflammation could explain the endothelial dysfunction observed (36).
There was feedback between the angiogenesis and inflammation pathways, as CCR7 and IL-13 have been shown to directly inhibit angiogenesis (12). This indicates that a lack of angiogenesis could stimulate inflammatory mediators that further block angiogenic signaling. CCR7 has also been reported to upregulate MMP-9 expression, thus exacerbating the proteolytic effects of MMP-9 (13, 29). CCR10 and IL-20 both promote angiogenesis, suggesting there was a continual attempt to stimulate angiogenesis in the aging WT LV (25). IL-20 promotes angiogenesis in vitro and in vivo by inducing migration of endothelial cells and promoting vascular endothelial growth factor (15). IL-20 has also been reported to increase MMP-9 expression (20).
Combined, these data suggest that the MMP-9 dependent desensitization of WT mice to proangiogenic stimuli led to an inability to increase the number of blood vessels and resulted in subsequent increased inflammation, which in turn further impaired endothelial function. An important future direction will be to investigate the role of NF-kB and other transcription factors in aged myocardium, and the effects MMP-9 deletion has on their expression, to further explore the intracellular signaling pathways involved.
The age-associated increase in MMP-9 contributes to increased proangiogenic signaling in an attempt to stimulate new vessel formation in response to increased myocyte hypertrophy. This attempt is not successful, as there is no change in vessel numbers, resulting in increased inflammation and vascular permeability. These changes occur before and thus may contribute to the development of LV dysfunction in old and senescent mice. MMP-9 deletion improves angiogenesis, attenuates inflammation, and prevents vascular leakiness in the setting of cardiac aging.
This study was supported by National Institutes of Health San Antonio Cardiovascular Proteomics Center HHSN-268201000036C (N01-HV-00244), HL-075360, HL-101430, H-L095852, HL-051971, and GM-104357; and Biomedical Laboratory Research and Development Service of the Veterans Affairs Office of Research and Development Award 5I01BX000505.
No conflicts of interest, financial or otherwise, are declared by the author(s).
A.Y., Y.M., Y.A.C., E.F.L., M.L.L., and Y.-F.J. conception and design of research; A.Y., Y.M., Y.A.C., E.F.L., A.P.V., and H.T. performed experiments; A.Y., Y.M., Y.A.C., E.F.L., A.P.V., H.T., M.L.L., and Y.-F.J. analyzed data; A.Y., Y.M., Y.A.C., E.F.L., A.P.V., H.T., M.E.H., H.-C.H., M.L.L., and Y.-F.J. interpreted results of experiments; A.Y. and Y.-F.J. prepared figures; A.Y. drafted manuscript; A.Y., Y.M., Y.A.C., E.F.L., A.P.V., H.T., M.E.H., H.-C.H., M.L.L., and Y.-F.J. edited and revised manuscript; A.Y., Y.M., Y.A.C., E.F.L., A.P.V., H.T., M.E.H., H.-C.H., M.L.L., and Y.-F.J. approved final version of manuscript.
We thank Wesley Lowell and Jianhua Zhang for technical assistance in this study.