Metabolic products of the intestinal microbiome and extremes of atherosclerosis

BACKGROUND AND AIMS

There is increasing awareness that the intestinal microbiome plays an important role in human health. We investigated its role in the burden of carotid atherosclerosis, measured by ultrasound as total plaque area.

METHODS

Multiple regression with traditional risk factors was used to identify three phenotypes among 316/3056 patients attending vascular prevention clinics. Residual score (RES; i.e. the distance off the regression line, similar to standard deviation) was used to identify the 5% of patients with much less plaque than predicted by their risk factors (Protected, RES <-2), the 90% with about as much plaque as predicted (Explained, RES -2 to 2), and the 5% with much more plaque than predicted (Unexplained RES >2). Metabolic products of the intestinal microbiome that accumulate in renal failure - gut-derived uremic toxins (GDUT) - were assayed in plasma by ultra-performance liquid chromatography coupled to quadrupole time-of-flight mass spectrometry.

RESULTS

Plasma levels of trimethylamine n-oxide (TMAO), p-cresyl sulfate, p-cresyl glucuronide, and phenylacetylglutamine were significantly lower among patients with the Protected phenotype, and higher in those with the Unexplained phenotype, despite no significant differences in renal function or in dietary intake of nutrient precursors of GDUT. In linear multiple regression with a broad panel of risk factors, TMAO (p = 0.011) and p-cresyl sulfate (p = 0.011) were significant independent predictors of carotid plaque burden.

CONCLUSIONS

The intestinal microbiome appears to play an important role in atherosclerosis. These findings raise the possibility of novel approaches to treatment of atherosclerosis such as fecal transplantation and probiotics.

Expanding the UniFrac Toolbox

The UniFrac distance metric is often used to separate groups in microbiome analysis, but requires a constant sequencing depth to work properly. Here we demonstrate that unweighted UniFrac is highly sensitive to rarefaction instance and to sequencing depth in uniform data sets with no clear structure or separation between groups. We show that this arises because of subcompositional effects. We introduce information UniFrac and ratio UniFrac, two new weightings that are not as sensitive to rarefaction and allow greater separation of outliers than classic unweighted and weighted UniFrac. With this expansion of the UniFrac toolbox, we hope to empower researchers to extract more varied information from their data.

Ras Signaling Is a Key Determinant for Metastatic Dissemination and Poor Survival of Luminal Breast Cancer Patients

Breast cancer is associated with alterations in a number of growth factor and hormone-regulated signaling pathways. Mouse models of metastatic breast cancer typically feature mutated oncoproteins that activate PI3K, Stat3, and Ras signaling, but the individual and combined roles of these pathways in breast cancer progression are poorly understood. In this study, we examined the relationship between oncogenic pathway activation and breast cancer subtype by analyzing mouse mammary tumor formation in which each pathway was activated singly or pairwise. All three oncogenes showed cooperation during primary tumor formation, but efficient dissemination was only dependent on Ras. In addition, transcriptional profiling demonstrated that Ras induced adenocarcinomas with molecular characteristics related to human basal-like and HER2(+) tumors. In contrast, Ras combined with PIK3CA(H1047R), an oncogenic mutant linked to ERα(+)/luminal breast cancer in humans, induced metastatic luminal B-like tumors. Consistent with these data, elevated Ras signaling was associated with basal-like and HER2(+) subtype tumors in humans and showed a statistically significant negative association with estrogen receptor (ER) signaling across all breast cancer. Despite this, there are luminal tumors with elevated Ras signaling. Importantly, when considered as a continuous variable, Ras pathway activation was strongly linked to reduced survival of patients with ERα(+) disease independent of PI3K or Stat3 activation. Therefore, our studies suggest that Ras activation is a key determinant for dissemination and poor prognosis of ERα(+)/luminal breast cancer in humans, and hormone therapy supplemented with Ras-targeting agents may be beneficial for treating this aggressive subtype.

Cooperation between Pik3ca and p53 mutations in mouse mammary tumor formation

PIK3CA, which codes for the p110α catalytic subunit of phosphatidylinositol 3-kinase, is one of the most frequently mutated genes in human breast cancer. Here, we describe a mouse model for PIK3CA-induced breast cancer by using the ROSA26 (R26) knock-in system, in which targeted Pik3ca alleles can be activated through transgenic expression of Cre recombinase. We mated Pik3ca(H1047R) and Pik3ca(wt) knock-in lines with MMTV-Cre transgenics, which express Cre in mammary epithelium. Starting at approximately 5 months of age, female R26-Pik3ca(H1047R);MMTV-Cre mice, but not control R26-Pik3ca(wt);MMTV-Cre mice, developed mammary tumors, as well as lymphoid and skin malignancies. R26-Pik3ca(H1047R);MMTV-Cre mammary tumors were typically either adenosquamous carcinoma or adenomyoepithelioma. As p53 is the most commonly mutated gene in breast cancer, we tested for genetic interaction between Pik3ca(H1047R) and p53 loss-of-function mutations in R26-Pik3ca(H1047R);p53(loxP/+);MMTV-Cre mice. This led to decreased survival of double-mutant animals, which developed lymphoma and mammary tumors with rapid kinetics. Mammary tumors that formed in p53(loxP/+);MMTV-Cre conditional mutants were either poorly differentiated adenocarcinoma or spindle cell/EMT, whereas R26-Pik3ca(H1047R);p53(loxP/+);MMTV-Cre mammary tumors were mostly adenosquamous carcinoma or spindle cell/EMT indicating that double-mutant mice develop a distinct spectrum of mammary tumors. Thus, an oncogenic variant of PIK3CA implicated in multiple human breast cancer subtypes can induce a very diverse spectrum of mammary tumors in mice. Furthermore, Pik3ca(H1047R) shows cooperation with p53, which altered the specific tumors that formed. Thus, the two most frequently mutated genes in human breast cancer show cooperation in mammary tumor formation.

Nerve growth-promoting factor produced in culture media conditioned by specific CNS tissues of the snail Helisoma

Medium conditioned by tissue from the CNS of the snail, Helisoma, is capable of promoting neurite outgrowth in isolated neurons from adult central ganglia. The conditioning factor(s) (CF), contained in conditioned medium (CM), is produced only by central ganglionic rings and buccal ganglia and not by other tissues, including hemolymph. CF requires a minimum of 24 h to be produced or released into the medium. At 12 h growth-promoting activity was not detectable. CF binds tightly to the polylysine substratum and its activity is not mimicked by addition of various sera, NGF or fibronectin. CF activity is abolished by chymotrypsin, trypsin or heating to 100 degrees C, but is stable to DNase and RNase treatment. The percentage of cells exhibiting neurite outgrowth is approximately linear with the amount of neural tissue used to condition the medium up to 2 ganglionic rings/ml. Addition of more ganglia fails to stimulate a greater response. This apparent plateau of CM activity appears to be a function of production and/or release of CF, rather than a saturation effect on plated cells, since dose-response curves for dilutions of CM are approximately linear regardless of the number of ganglia used for conditioning. In addition, anisomycin inhibits 35% of CF appearance under conditions of over 90% protein synthesis inhibition in the ganglia used to produce the CM. Under these conditions anisomycin has no apparent effect on the maintenance of electrical excitability. The inhibitor data suggest that 65% of CF is derived from a pre-existing storage pool and that the remainder is synthesized during the 72 h conditioning period.

Formation of novel central and peripheral connections between molluscan central neurons in organ cultured ganglia

An in vitro organ culture system for buccal ganglia of the adult snail, Helisoma, is described. The system supports: (1) maintenance of characteristic electrophysiological parameters of identified neurons over seven days of culture; (2) choline metabolism including uptake and synthesis over the same duration; (3) sprouting and growth of neurons in response to axotomy; (4) the formation of novel central electrotonic connections between identified neurons as a result of sprouting and growth. These observations on neuronal growth and the formation of connections are similar to those made with in vivo culture. The use of in vitro culture allows precise manipulations not previously possible. When buccal ganglia are cultured in vitro with the cut distal ends of peripheral nerve trunks held closely apposed, axons of neurons 5R and 5L in the nerve trunks are capable of forming electrotonic connections similar to central connections. The capability of these neurons to form electrotonic connections via their peripheral axons implies that special structures (i.e., central neurites) are not required for the formation of connections; and neither are special environments (i.e., the central neuropile) required for these connections.

Separate factors produced by the CNS of the snail Helisoma stimulate neurite outgrowth and choline metabolism in cultured neurons

Neurons from the snail Helisoma require a brain-derived factor(s) for neurite outgrowth in both organ and isolated cell culture. This factor is released from the CNS of Helisoma when brains are incubated in defined medium, producing a conditioned medium (CM). In addition to its growth-promoting activity, CM also enhances total uptake of 3H-choline and the incorporation of 3H-choline into specific metabolites: acetylcholine, phosphorylcholine and lipid. This choline metabolism-enhancing factor(s) is distinct and separable from neurite growth-promoting factor: 1. Over 95% of neurite growth-promoting activity can be removed from CM by adsorption to a polylysine surface while there is no loss of choline metabolism-enhancing activity. 2. When central ganglia were treated with anisomycin, a potent inhibitor of molluscan protein synthesis, the choline metabolism-enhancing activity was completely absent from the resulting CM, while the growth promoting activity was reduced by only 35%. These results suggest that the Helisoma CNS produces a variety of trophic factors that are involved in regulating the interaction between neuronal growth and metabolism.

Neurite outgrowth in molluscan organ and cell cultures: the role of conditioning factor(s)

Isolated neurons from adult central ganglia of the snail Helisoma were cultured in vitro in modified Liebowitz L-15 medium. Such neurons displayed electrical excitability comparable to that in acutely dissected ganglia. Isolated neurons remained spherical in defined medium throughout culture durations up to 2 weeks. This static morphology was contrasted by the significant neuritic outgrowth which occurred from neurons maintained in medium with co-cultured intact Helisoma brains or in brain conditioned medium. A morphological sequence of growth cone formation and neurite extension occurred only in the presence of a conditioning factor(s) with a mode of action which included tight binding of the conditioning factor to the substratum. Under these conditions, the two primary neuronal phenotypes, electrical excitability and complex neuronal architecture, could be affected independently in adult molluscan neurons cultured in vitro.

Studies on the mechanism of action of calciferol. VIII. The effects of dietary vitamin D and the polyene antibiotic, filipin, in vitro, on the intestinal cellular uptake of calcium

Intestinal transport and cellular uptake of calcium were studied in vitro in ileal segments obtained from calciferol (vitamin D)-deficient (minus D3) chicks or in calciferol-treated chicks (+ D3). When calcium flux (J) was measured in both the mucosal yields serosal (J-ms) and serosal yields mucosal (J-sm) directions it was found that calcium transport in the +D3 system was an active process. Calcium was not actively transported in the minus D3 system or by either the +D3 or minus D3 serosal tissues which underlie the intestinal mucosa tissue. It was also found that the serosal tissue was not the rate-limiting step in calcium translocation. When calcium uptake was measured at the mucosal or brush border surface it was found to be a cation-oriented, saturable process in both the +D3 and minus D3 systems and was enhanced by calciferol supplementation. The rate of uptake was found to exhibit a two component transport phenomenon, one a saturable process at low Ca2+ concentrations and the other a linear of diffusion process at higher Ca2+ concentrations (greater than 5 mM). The polyene antibiotic, filipin, was used to study the transport and uptake of calcium in both the +D and minusD chick ileum in vitro. Filipin (10 mug/ml), when added in vitro to solutions bathing the mucosal surface of the ileum, stimulated the calcium flux, J-ms, of minus D ileal tissue by 150 to 200%, but had little or no stimulatory effect on J-ms of +D3 ileal tissue. However, in contrast it stimulated calcium uptake across the mucosal membrane in both the +D3 and minus D3 ileal tissue (50 to 100%). The effect of filipin is specific for calcium uptake as compared to Rb+, P-i, SO4(2-), glycerol, thiourea, and urea uptake. Isolated brush borders from +D3 chicks were found to bind calcium to a greater extent than the minus D3 brush borders, but both membrane fractions had the same cholesterol content. The sum of this and other evidence suggests that of uptake of calcium is a calciferol-mediated event which is not the rate-limiting step in the total transport of this ion across the intestinal epithelial cell. The filipin effect appears to affect a structural reorganization of the brush border membrane in both the +D3 and minus D3 membranes in a manner specific for calcium translocation. It now appears that filipin treatment somehow makes both transport systems more efficient in the uptake of calcium, but in a manner independent of the mode of action of calciferol. However, these data suggest that other components associated with calcium uptake are calciferol-dependent in an as yet unknown fashion.

Biologic effects of 1,25-dihydroxycholecalciferol (a highly active vitamin D metabolite) in acutely uremic rats

The development of a vitamin D-resistant state in the course of renal failure may be responsible for reduced intestinal absorption of calcium and an impaired response of skeletal tissue. Moreover, the kidney has been shown to carry out the conversion of 25-hydroxycholecalciferol (25-OH-CC) to a highly biologically active metabolite, 1,25-dihydroxycholecalciferol (1,25-diOH-CC). In the present studies, vitamin D-deficient rats, made acutely uremic by either bilateral nephrectomy or urethral ligation, received physiological doses of cholecalciferol (vitamin D(3)) (CC), 25-OH-CC or 1,25-diOH-CC; 24 hr later intestinal calcium transport, in vitro, and bone calcium mobilization, in vivo, were assessed. Whereas CC and 25-OH-CC stimulated calcium transport in sham-operated controls, they were without effect in the uremic animals. In contrast, administration of 1,25-diOH-CC stimulated calcium transport in both groups of uremic animals. Administration of 1,25-diOH-CC also stimulated calcium mobilization from bone in each group of animals. However, CC and 25-OH-CC were only effective in the sham controls and the uremic group produced by urethral ligation and had little or no effect in animals without kidneys. These results indicate that renal conversion of calciferol to a more biologically active form is necessary for the stimulation of intestinal calcium absorption and calcium mobilization from bone, and that 1,25-diOH-CC may bypass a possible defect in vitamin D metabolism in uremia. From these studies it is likely that uremia, per se, may also impair intestinal calcium transport.