Jayachandran, Muthuvel

Objective Blood thrombogenicity and vascular inflammation may influence development of white matter hyperintensities (WMH) in the brain. This study aimed to determine if menopausal hormone treatments (MHT) modified that association in recently menopausal women who were enrolled in the Kronos Early Estrogen Prevention Study (KEEPS) at Mayo Clinic. Methods Women (n=95) underwent MRI prior to and at 18, 36 and 48 months after randomization to either transdermal 17β‐estradiol (tE2, n=30), conjugated equine estrogen (oCEE, n=29) both with progesterone for 12 days per month or placebo pills and patch (PBO, n=36). Fasting blood was analyzed for a set of 14 activation markers of platelets and endothelial cells, inflammation, and thrombogenicity. Principal components (PC) analysis was used to reduce dimensionality of these markers, and the first 5 PC identified were then assessed for association with treatment and changes in WMH. Results WMH increased in all groups over the 48 months of treatment with greater increases in the oCEE compared to tE2 and PBO (P=0.015). The 5 PC explained 62% of the variation in cellular activation markers. A global test of these 5 PC revealed a borderline association with treatment group (P=0.059), with the PC reflecting total platelet count, platelet microaggregates and ATP secretion showing higher scores in the oCEE compared to the other groups (P=0.006). The change in WMH over the 48 months of MHT showed an overall association with the 5 PC (P=0.026), the most prominent of which was the first PC reflecting vascular inflammation by number of blood microveicles (MV) from activated monocytes and endothelial cells and MV positive for intercellular adhesion molecule 1 (P=0.001). There was no evidence that this association differed by MHT. Conclusion Markers of blood thrombogenicity and inflammation may influence development of WMH in the brain but this relationship may be influenced by factors other than the type of MHT used in the KEEPS study. Support or Funding Information NIH P50 AG44170 This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .

Activated platelets release mitogenic factors, vascular remodeling enzymes, inflammatory and cellular adhesive molecules into the vascular microenvironment to alter chemotactic and adhesive properties of vascular endothelium. Platelet‐derived growth factor‐BB (PDGF‐BB) and matrix metalloproteinase‐2 (MMP‐2) increased in the platelets following ovariectomy but decreased with estrogen treatment in pigs. Experiments were designed to identify changes in these factors using custom antibody arrays in lysate of platelets collected from women enrolled in the Kronos Early Estrogen Prevention Study (KEEPS) at Mayo Clinic. Washed platelets were prepared from anticoagulated blood before (baseline) and 48 months after randomization into either placebo or oral or transdermal MHT. Preliminary analysis shows after 48 months compared to baseline, platelet content of C‐reactive protein, intercellular adhesion molecule‐1 (ICAM‐1), vascular cell adhesion molecule‐1 (VCAM‐1), PDGF‐BB, MMP‐2 and transforming growth factor‐beta (TGF‐beta) were lower (20–60%) in two groups but higher (20–74%) in one group. Based on our studies in pigs, it is expected that the two groups showing decreases in platelet proteins would be from women randomized to MHT. However, KEEPS will not be unblinded until February 2012. (Supported by grants from Aurora Foundation, NIH HL90639, 1UL1 RR024150).

Estrogen modulates the prostanoids prostacyclin (PGI 2 ) and thromboxane A 2 (TXA 2 ) which contribute to vascular hemostasis in endothelium and platelets. Estrogen decreases in women at menopause, and menopausal hormone therapy (MHT) may increase risk of thrombosis. This study aimed to determine effects of MHT on serum prostanoid levels and on prostanoid‐related platelet functions in healthy, recently‐menopausal women. Women (age 52.3±2.3 yrs; n=115) enrolled in the Kronos Early Estrogen Prevention Study (KEEPS) were randomized to either: 1) oral conjugated equine estrogen, 2) transdermal 17β estradiol each with intermittent oral progesterone or 3) placebo, for 48 months. Serum PGI 2 and TXA 2 levels were assessed by quantification (ng/ml) of their stable metabolites 6‐k‐PGF 1α and TXB 2 , respectively. Initial TXB 2 levels (46.9±7.5) were unchanged (P>; 0.05) after 48 months in each treatment group. However, 6‐k‐PGF 1α decreased from 2.1±0.2 to 1.4±0.1 at 48 months in combined groups, but was statistically significant (1.5±0.3) only in the transdermal group. Inhibition of platelet ATP secretion by PGE 1 did not correlate with serum 6‐k‐PGF 1α levels. These results suggest that serum PGI 2 levels decrease with age in menopausal women and are unaffected by MHT. Factors other than changes in serum prostanoids may impose thrombotic risk associated with MHT. (Supported by the Aurora Fnd, NIH HL88988 and Mayo Clinic).

Little is known about mechanisms by which bacterial infection contributes to thrombotic risk. Experiments were designed to identify the relationship between toll‐like receptor 4 (TLR4) activation by bacterial lipopolysaccharide (LPS) and circulating procoagulant (phosphatidylserine (PS) positive) microvesicles (MV). Blood from wild type (WT), and TLR4‐ gene deleted (dTLR4) female mice was collected prior to and following exposure to either saline or LPS 20ng/g body weight administered as a single intravenous injection (acute exposure) or after administration by subcutaneous osmotic pump (chronic exposure). MV were isolated from blood by differential centrifugation and identified by flow cytometry. MV thrombin generating capacity was measured by solid phase fluorescent assay. Procoagulant MV were significantly lower in dTLR4 compared to WT mice (P<0.007). Platelet derived MV were the most prominent MV in saline treated mice. Following LPS‐treatment, counts of procoagulant MV increased (P<0.006) significantly only in WT mice. The increase was greater after chronic compared to acute exposure. Thrombin generating capacity positively correlated (r 2 =0.9) with PS expression. These results suggest that increased procoagulant MV through TLR4 activation and thus explain in part increased thrombotic risk with bacterial infection. This work is supported by AHA‐SDG 30503Z and Mayo Foundation.

Risk of CAD increases at menopause and platelets contribute to the etiology of CAD. We hypothesized that individual variation in platelet procoagulant activity at menopause correlates with the extent of CAD as reflected by coronary artery calcification ( CAC ). We developed several novel and global assays of platelet procoagulant activity: dense (ATP release) and alpha (P‐selectin expression) granule secretion and GPα IIb β 3 fibrinogen‐binding (PAC1 binding) to ADP stimulation, and prostaglandin E 1 (PEG 1 ) sensitivity as a surrogate for prostacyclin sensitivity. Platelets were prepared from venous blood collected from women screened for the Kronos Early Estrogen Prevention Study (n=125; mean±SD age=52.5±2.4 years) within 19.4±9.2 months of last menses. Baseline CAC score was determined. Of conventional CAD risk factors, only systolic blood pressure correlated with positive CAC score (n=23). ATP secretion tended to be less in women with positive CAC scores and was negatively correlated with blood glucose (P<0.001). P‐selectin expression and PAC1 binding increased with total cholesterol (P<0.05). PEG 1 sensitivity increased with increasing age and waist circumference (P<0.05). These results indicate that platelet functions change with age and in association with conventional risk factors for the disease in women at menopause. Supported by the Kronos Longevity Research Institute and NIH HL90639.

Cell membrane‐derived microvesicles (MV) are shed into the circulation and reflect cell‐cell interaction, intravascular and intercellular signaling process. Infection increases risk for thrombosis and LPS signaling through toll‐like receptor 4 (TLR 4 ) may activate cells to generate procoagulant (phosphatidylserine surface) MV. This study was designed to investigate MV production from leukocytes and platelets exposed to LPS in vitro . Blood from wild type (WT) and TLR 4 gene deleted (dTLR 4 ) mice was incubated with LPS (500ng/ml or 5ug/ml) for 6 hours. Flow cytometer was used to quantify MV using fluorophore conjugated cell specific antibodies plus annexin‐V. Leukocyte‐ and platelet‐derived MV showed a concentration dependent increase in the blood from WT mice; whereas in dTLR 4 mice increased counts of leukocyte‐ and platelet‐derived MV were observed only at the higher concentration of LPS. At the higher concentration of LPS, numbers of leukocyte‐derived MV increased significantly and the mean number of platelet‐derived MV was two fold greater than under control conditions. These results indicate that low concentrations of LPS act through TLR 4 but higher concentrations of LPS act through other receptors in addition to TLR 4 to increase production of leukocyte‐ and platelet‐derived MV which could account in part for increased thrombotic risk with infection. Supported by AHA‐SDG 30503Z and Mayo Foundation.

Background: The identification of blood-borne biomarkers for the diagnosis and prognosis of Alzheimer’s disease and related dementias is more feasible at the population level than obtaining cerebrospinal fluid or neuroimaging markers. Objective: This study determined the association of blood microvesicles, derived from cells of the neurovascular unit, with brain amyloid-β deposition in menopausal women. Methods: A subset of women from the Kronos Early Estrogen Prevention Study underwent brain amyloid-β positron emission tomography three years following cessation of study treatment with placebo (PL, n = 29), transdermal 17β-estradiol (tE2; n = 21), or oral conjugated equine estrogen (oCEE; n = 17). Isolated peripheral venous blood microvesicles were analyzed by digital flow cytometry using fluorophore conjugated antibodies directed toward total tau, amyloid-β 1–42 (Aβ1–42), neuron specific class III β-tubulin (Tuj1), microglia ionized calcium -binding adaptor molecule 1(Iba1), glial fibrillary acid protein (GFAP), and low density lipoprotein receptor-related protein1 (LRP1). Principal components analysis reduced the dimensionality of these selected six markers to two principal components (PCs). Proportional odds ordinal logistic regression analysis was used with amyloid-β deposition regressed on these PCs. Results: Only the number of microvesicles positive for Aβ1–42 differed statistically among prior treatment groups (median [IQR]: 6.06 [2.11, 12.55] in PL; 2.49 [0.73, 3.59] in tE2; and 4.96 [0.83, 10.31] in oCEE; p = 0.032). The joint association between the 2 PCs and brain amyloid-β deposition was significant (p = 0.045). Conclusion: Six selected markers expressing peripheral blood microvesicles derived from cells of the neurovascular unit, when summarized into two principal components, were associated with brain amyloid-β deposition.

Abstract Backgrounds: Previous studies have demonstrated that excretion of urinary extracellular vesicles (EVs) from different nephron segments differs between kidney stone formers and non-stone formers (NSFs), and could reflect pathogenic mechanisms of urinary stone disease. In this study we quantified selected populations of specific urinary EVs carrying protein markers of immune cells and calcium/phosphorus physiology in calcium stone formers (CSFs) compared to non-stone formers (NSFs). Methods Incident CSFs (n = 24) and age- and sex- matched NSFs (n = 21) were studied. Clinical data were abstracted and biobanked cell-free urine samples were used to quantify specific urinary EV populations. EVs carrying proteins related to renal calcium/phosphorus physiology (phosphorus transporters (PiT1 and PiT2), Klotho, and fibroblast growth factor 23 (FGF23)); markers associated with EV generation (anoctamin-4 (ANO4) and Huntington interacting protein 1 (HIP1)), and markers shed from activated immune cells were quantified by standardized and published method of digital flow cytometry. Results The urine pH of CSFs was lower than NSFs (P < 0.05), whereas urine excretion of calcium, phosphorus, and calcium oxalate and uric acid supersaturation (SS) were significantly higher in CSFs compared to NSFs (P < 0.05). Urinary excretion of EVs with markers of total leukocytes (CD45), neutrophils (CD15), macrophages (CD68), Klotho, FGF23, PiT1, PiT2, and ANO4 were each markedly lower in CSFs than NSFs (P < 0.05) whereas excretion of those with markers of monocytes (CD14), T-Lymphocytes (CD3), B-Lymphocytes (CD19), plasma cells (CD138 plus CD319 positive ) were not different between the groups. Urinary excretion of EVs expressing PiT1 and PiT2 negatively (P < 0.05) correlated with urinary phosphorus excretion whereas excretion of EVs expressing FGF23 correlated negatively (P < 0.05) with both urinary calcium and phosphorus excretion. Conclusions Urinary excretion of EVs derived from specific types of activated immune cells and EVs with proteins related to calcium/phosphorus regulation were different between CSFs and NSFs. Thus, further validation of these and other populations of urinary EVs could identify biomarkers that elucidate novel pathogenic mechanisms of calcium stone formation in specific subsets of patients.

Anoctamins typically function as Ca 2+ activated Cl − channels (CaCC) or as specific, phospholipid scramblases. GWAS of miniature schnauzers (>40% with recurrent calcium oxalate (CaOx) kidney stones) revealed a risk locus including an anoctamin 4 (ANO4) variant. ANO4 in mammals was initially reported as a phosphatidylserine (PS) scramblase. Urinary extracellular vesicles (EV) show decreased ANO4 protein in human stone formers. However, in Drosophila renal structures (Malpighian tubules, MTs), subdued (the ANO4‐homolog), functions as both a CaCC and scramblase. Since subdued also functions to defend against gram‐negative bacteria, it was knocked down to determine whether CaOx crystals were altered by bacterial infection. Subdued was knockdown (KD) in MT principle cells by crossing Uro:Gal4 flies with UAS: subdued ‐RNAi flies. F1‐flies were evaluated using CaOx crystallization in dissected MTs as well as after 4d NaOx feeding protocols. To evaluate infection, uropathic E. coli (UPEC):eGFP were fed to flies or added to NaOx solution before an overnight time‐lapse. Interestingly, the canine ANO4 variant has a 0.01 allele variant frequency at a residue conserved in humans, canines and Drosophila . Rapid crystallization experiments show that MTs with subdued‐KD have fewer CaOx crystals than control MTs. However, prolonged NaOx feeding assays showed that subdued‐KD in MTs produces enlarged CaOx crystals. These feeding assays also showed an unexpected phenotype: crystal clustering and matting structures (e.g., bacterial biofilms). These structures have similar patterning to human kidney stones [Sivaguru et al., PMID: 30213974]. Following a four day feeding of NaOx, when UPEC + NaOx is introduced overnight ( ex vivo ), a three hour bacterial event occurs, causing crystals to more tightly pack. Finally, when UPEC is introduced during a rapid crystallization experiment, the bacteria heavily localize in subdued‐KD MT lumens while control MTs have little eGFP‐fluorescence. Thus, UPEC facilitates CaOx crystal aggregates in subdued‐KD MTs while the nature of this interaction is unknown. In human kidney, ANO4 localizes in thick ascending limb (NKCC2 + ), while in murine kidney localizes more proximal in the Loop of Henle (UMOD + ). As mice are resistant to CaOx stones, these slight location differences may reveal altered TAL function by ANO4. This Drosophila avatar indicates that subdued (ANO4) normally limits CaOx crystal aggregation and limits UPEC growth. As an apparent pathogenic amino acid is conserved in mammals and flies; subdued’ s role in fly bacterial infection suggests an additional ANO4 role in human kidney stone formation. These results indicate that examining antibacterial treatments with UPEC infection may allow an antibiotic approach to more efficiently eliminate kidney stones. Future experiments determine what functional aspect of subdued alters CaOx crystallization. Moreover, this simple kidney stone avatar‐the fly – continues to provide new and mechanistic insights to CaOx stone formation. Support or Funding Information Support: Oxalosis & Hyperoxaluria Foundation, R25‐DK101405, Mayo Foundation Drosophila MT (Uro:Gal4 x UAS‐ subdued ‐RNAi) viewed with (A) DIC for CaOx birefringence, (B) fluorescence to visualize UPEC:eGFP and (C) merged to show coincidence of UPEC with CaOx crystals. The red line indicates the length of the CaOx/UPEC mass. This is not observed in control MTs. Figure 1

Infection increases propensity for thrombotic events. Gram negative pathogens release lipopolysaccharides (LPS), which initiate intracellular signaling cascades through activation of toll‐like‐receptor 4 (TLR4). Platelets contribute to defense from infection and to thrombosis, but little is known about how LPS affects platelet production, life span and reactivity. Platelet life span was determined by in vivo biotinylation in wild type (C57BL/10SnJ; WT) and TLR4 deficient (C57BL/10ScN; dTLR4) mice before and after a single intravenous injection of LPS (1ng – 500ng/25g body weight). Prior to LPS injection, the number of platelets and percentage of reticulated platelets in dTLR4 mice was less than in WT mice but there were no differences in platelet life span, indicating that loss of TLR4 affects platelet production and not life span. A single injection of LPS dose dependently increased percentages of P‐selectin positive platelets and of reticulated platelets but decreased the number of biotin positive platelets within 24hrs in WT but not in dTLR4 mice. Results suggest that LPS–simulated infection reduces life span of circulating platelet and increases production of more reactive (reticulated) platelets from megakaryocytes through TLR4 mechanisms. Thus, a single exposure to LPS represents a “priming” mechanism which could increase risk for thrombosis with infection. (Funded by NIH‐HL78638 and Mayo Fnd.)