L1 Penalized Estimation in the Cox Proportional Hazards Model

This article presents a novel algorithm that efficiently computes L₁ penalized (lasso) estimates of parameters in high‐dimensional models. The lasso has the property that it simultaneously performs variable selection and shrinkage, which makes it very useful for finding interpretable prediction rules in high‐dimensional data. The new algorithm is based on a combination of gradient ascent optimization with the Newton–Raphson algorithm. It is described for a general likelihood function and can be applied in generalized linear models and other models with an L₁ penalty. The algorithm is demonstrated in the Cox proportional hazards model, predicting survival of breast cancer patients using gene expression data, and its performance is compared with competing approaches. An R package, penalized , that implements the method, is available on CRAN.     

The phenotype and genotype of rheumatoid arthritis in the Democratic Republic of Congo

Introduction

Little is known about rheumatoid arthritis in the black, particularly in Congolese, populations. Our objective was to describe the phenotype and genotype of rheumatoid arthritis (RA) in Congolese.

Methods

All consecutive rheumatoid arthritis (RA) patients attending Kinshasa University Hospital in a three-year time period were included. Demographics, clinical features and tobacco consumption were noted. Disease Activity Score (DAS)-28 based on the erythrocyte sedimentation rate (ESR), Health Assessment Questionnaire (HAQ), anti-citrullinated peptide antibodies (CCP) antibodies and rheumatoid factor (RF) were determined. Radiographs were scored according to Sharp-van der Heijde. On a subset of patients and controls HLA-DRB1 typing was performed.

Results

A total of 114 females and 14 males aged 51.2 ± 14.9 were included. Mean duration of symptoms was four years. Moderate tobacco consumption was reported in a minority of patients. DAS-28 at first visit was >5.1 and HAQ ≥0.5 in all patients. X-rays showed joint erosions and/or joint space narrowing, mostly of a moderate grade in 55.8% of patients. Anti-CCP and/or RF were present in 48.6% of patients with available data (n = 72) and in 3.0% of controls (n = 67). Radiographic changes and nodules were more frequent in RF or anti-CCP positive patients. One copy of the shared epitope was found in 13 patients (35.1%) and 3 controls (12.5%). Two copies were found in one patient (2.7%) and in one control (4.2%).

Conclusion

Congolese patients with RA consult long after disease onset. Despite this delay, the majority presents without major damage and is RF, anti-CCP and SE negative. We put forward the hypothesis that besides different environmental factors there is probably also a particular genetic risk profile in Congolese patients, different from the HLA-DRB1 shared epitope.     

Genes Involved in the Osteoarthritis Process Identified through Genome Wide Expression Analysis in Articular Cartilage; the RAAK Study

Objective

Identify gene expression profiles associated with OA processes in articular cartilage and determine pathways changing during the disease process.

Methods

Genome wide gene expression was determined in paired samples of OA affected and preserved cartilage of the same joint using microarray analysis for 33 patients of the RAAK study. Results were replicated in independent samples by RT-qPCR and immunohistochemistry. Profiles were analyzed with the online analysis tools DAVID and STRING to identify enrichment for specific pathways and protein-protein interactions.

Results

Among the 1717 genes that were significantly differently expressed between OA affected and preserved cartilage we found significant enrichment for genes involved in skeletal development (e.g. TNFRSF11B and FRZB). Also several inflammatory genes such as CD55, PTGES and TNFAIP6, previously identified in within-joint analyses as well as in analyses comparing preserved cartilage from OA affected joints versus healthy cartilage were among the top genes. Of note was the high up-regulation of NGF in OA cartilage. RT-qPCR confirmed differential expression for 18 out of 19 genes with expression changes of 2-fold or higher, and immunohistochemistry of selected genes showed a concordant change in protein expression. Most of these changes associated with OA severity (Mankin score) but were independent of joint-site or sex.

Conclusion

We provide further insights into the ongoing OA pathophysiological processes in cartilage, in particular into differences in macroscopically intact cartilage compared to OA affected cartilage, which seem relatively consistent and independent of sex or joint. We advocate that development of treatment could benefit by focusing on these similarities in gene expression changes and/or pathways.     

Clinical Associations of Human T-Lymphotropic Virus Type 1 Infection in an Indigenous Australian Population

Introduction

In resource-poor areas, infectious diseases may be important causes of morbidity among individuals infected with the Human T-Lymphotropic Virus type 1 (HTLV-1). We report the clinical associations of HTLV-1 infection among socially disadvantaged Indigenous adults in central Australia.

Methodology and Principal Findings

HTLV-1 serological results for Indigenous adults admitted 1^st January 2000 to 31^st December 2010 were obtained from the Alice Springs Hospital pathology database. Infections, comorbid conditions and HTLV-1 related diseases were identified using ICD-10 AM discharge morbidity codes. Relevant pathology and imaging results were reviewed. Disease associations, admission rates and risk factors for death were compared according to HTLV-1 serostatus. HTLV-1 western blots were positive for 531 (33.3%) of 1595 Indigenous adults tested. Clinical associations of HTLV-1 infection included bronchiectasis (adjusted Risk Ratio, 1.35; 95% CI, 1.14–1.60), blood stream infections (BSI) with enteric organisms (aRR, 1.36; 95% CI, 1.05–1.77) and admission with strongyloidiasis (aRR 1.38; 95% CI, 1.16–1.64). After adjusting for covariates, HTLV-1 infection remained associated with increased numbers of BSI episodes (adjusted negative binomial regression, coefficient, 0.21; 95% CI, 0.02–0.41) and increased admission numbers with strongyloidiasis (coefficient, 0.563; 95% CI, 0.17–0.95) and respiratory conditions including asthma (coefficient, 0.99; 95% CI, 0.27–1.7), lower respiratory tract infections (coefficient, 0.19; 95% CI, 0.04–0.34) and bronchiectasis (coefficient, 0.60; 95% CI, 0.02–1.18). Two patients were admitted with adult T-cell Leukemia/Lymphoma, four with probable HTLV-1 associated myelopathy and another with infective dermatitis. Independent predictors of mortality included BSI with enteric organisms (aRR 1.78; 95% CI, 1.15–2.74) and bronchiectasis (aRR 2.07; 95% CI, 1.45–2.98).

Conclusion

HTLV-1 infection contributes to morbidity among socially disadvantaged Indigenous adults in central Australia. This is largely due to an increased risk of other infections and respiratory disease. The spectrum of HTLV-1 related diseases may vary according to the social circumstances of the affected population.     

Transcriptional profiling of human familial longevity indicates a role for ASF1A and IL7R

The Leiden Longevity Study consists of families that express extended survival across generations, decreased morbidity in middle-age, and beneficial metabolic profiles. To identify which pathways drive this complex phenotype of familial longevity and healthy aging, we performed a genome-wide gene expression study within this cohort to screen for mRNAs whose expression changes with age and associates with longevity. We first compared gene expression profiles from whole blood samples between 50 nonagenarians and 50 middle-aged controls, resulting in identification of 2,953 probes that associated with age. Next, we determined which of these probes associated with longevity by comparing the offspring of the nonagenarians (50 subjects) and the middle-aged controls. The expression of 360 probes was found to change differentially with age in members of the long-lived families. In a RT-qPCR replication experiment utilizing 312 controls, 332 offspring and 79 nonagenarians, we confirmed a nonagenarian specific expression profile for 21 genes out of 25 tested. Since only some of the offspring will have inherited the beneficial longevity profile from their long-lived parents, the contrast between offspring and controls is expected to be weak. Despite this dilution of the longevity effects, reduced expression levels of two genes, ASF1A and IL7R, involved in maintenance of chromatin structure and the immune system, associated with familial longevity already in middle-age. The size of this association increased when controls were compared to a subfraction of the offspring that had the highest probability to age healthily and become long-lived according to beneficial metabolic parameters. In conclusion, an "aging-signature" formed of 21 genes was identified, of which reduced expression of ASF1A and IL7R marked familial longevity already in middle-age. This indicates that expression changes of genes involved in metabolism, epigenetic control and immune function occur as a function of age, and some of these, like ASF1A and IL7R, represent early features of familial longevity and healthy ageing.     

ING2 recruits histone methyltransferase activity with methylation site specificity distinct from histone H3 lysines 4 and 9

p33ING2 belongs to the ING-gene family that is involved in tumor suppression, DNA repair, cell cycle regulation, and cellular senescence. Most functions are dependent on the tumor suppressor p53. p33ING2 was also shown to bind to trimethylated lysine 4 of histone H3. Here, we show that p33ING2 contains a transferable silencing function, which is independent of p53. p33ING2-mediated gene silencing is resistant to the HDAC-inhibitor trichostatin A indicating that p33ING2 uses a non-HDAC class I or II pathway for gene repression in reporter assays. In line with that we show that p33ING2 is associated with histone methyltransferase (HMT) activity in vitro and in vivo, methylating specifically histone H3. Interestingly, the specificity is distinct from the MeCP2-recruited HMT. Mutation or methylation of lysine 9, a mark well known for repression, abrogates histone methylation by MeCP2 but not by the p33ING2 complex. Instead, the ING2-associated HMT shows an increased methylation activity if lysine 9 is methylated. In contrast, mutation or methylation of lysine 4, a methylation preferentially detected at active genes, led to a reduction of the ING2-associated HMT. Notably, also p33ING1 recruits HMT activity suggesting a more general biochemical interaction between members of p33ING family and HMT activity. Deletion analyses revealed that the ING2 C-terminus recruits HMT activity, which correlates with silencing function.     

Autocorrelated logistic ridge regression for prediction based on proteomics spectra

This paper presents autocorrelated logistic ridge regression, an extension of logistic ridge regression for ordered covariates that is based on the assumption that adjacent covariates have similar regression coefficients. The method is applied to the analysis of proteomics mass spectra.     

Low tryptophan diet increases stress-sensitivity, but does not affect habituation in rats

Cerebral dysfunction of 5-HT (serotonin) has been associated with stress response and with affective disorders. Stress alone is insufficient to induce depression, since only a minor proportion of subjects that have experienced stressful life events develop depressive episodes. We investigated whether long-term brain 5-HT depletion induced in rats by a diet with low content of its precursor tryptophan affects stress-responsiveness in rats. Stress-sensitivity was measured through various physiological parameters and by measuring the rats' response to acoustic stimuli. One group of rats was subjected to daily acoustic stimulus sessions for 5 days. Other groups received both immobilization stress and acoustic stimulus sessions daily for either 9 days (chronic experiment) or 1 day (acute experiment). A low tryptophan diet led to decreases in plasma tryptophan levels, low ratio of tryptophan/large neutral amino acid, whole blood 5-HT, and neuronal 5-HT content in the Dorsal and Median Raphe Nuclei, as well as altered c-fos expression in the brain. Without concomitant immobilization, the diet alone did not affect reactivity and habituation to acoustic stimuli, although plasma corticosterone levels, but not the adrenal weights, were increased on day 5. Low tryptophan and chronic immobilization stress together with the acoustic testing procedure increased adrenal weight, plasma corticosterone levels and reactivity to the acoustic stimuli, but not the rate of habituation to acoustic stimuli. These results show that cerebral dysfunction of serotonin achieved through a low tryptophan diet, increases the sensitivity of rats to external and stressful stimuli, but does not impair the capacity to adapt to these stimuli. Accordingly, brain-serotonin modulates reactivity to stress, but not stress coping.