Age-dependent effects of Igf2bp2 on gene regulation, function, and aging of hematopoietic stem cells in mice.

Increasing evidence links metabolism, protein synthesis, and growth signaling to impairments in the function of hematopoietic stem and progenitor cells (HSPCs) during aging. The Lin28b/Hmga2 pathway controls tissue development, and the postnatal downregulation of this pathway limits the self-renewal of adult vs fetal hematopoietic stem cells (HSCs). Igf2bp2 is an RNA binding protein downstream of Lin28b/Hmga2, which regulates messenger RNA stability and translation. The role of Igf2bp2 in HSC aging is unknown. In this study, an analysis of wild-type and Igf2bp2 knockout mice showed that Igf2bp2 regulates oxidative metabolism in HSPCs and the expression of metabolism, protein synthesis, and stemness-related genes in HSCs of young mice. Interestingly, Igf2bp2 expression and function strongly declined in aging HSCs. In young mice, Igf2bp2 deletion mimicked aging-related changes in HSCs, including changes in Igf2bp2 target gene expression and impairment of colony formation and repopulation capacity. In aged mice, Igf2bp2 gene status had no effect on these parameters in HSCs. Unexpectedly, Igf2bp2-deficient mice exhibited an amelioration of the aging-associated increase in HSCs and myeloid-skewed differentiation. The results suggest that Igf2bp2 controls mitochondrial metabolism, protein synthesis, growth, and stemness of young HSCs, which is necessary for full HSC function during young adult age. However, Igf2bp2 gene function is lost during aging, and it appears to contribute to HSC aging in 2 ways: the aging-related loss of Igf2bp2 gene function impairs the growth and repopulation capacity of aging HSCs, and the activity of Igf2bp2 at a young age contributes to aging-associated HSC expansion and myeloid skewing.

Abundance and size of hyaluronan in naked mole-rat tissues and plasma.

Large amounts of ultra-high molecular weight hyaluronan (HA) have been described as the main cause of cancer resistance in naked mole-rats (Heterocephalus glaber, NMR). Our work examined HA metabolism in these rodents more closely. HA was localized and quantified using HA binding proteins. Its molecular weight was determined using size exclusion chromatography and gel electrophoresis, HA family gene expression using RNAseq analysis, and hyaluronidase activity using zymography. Guinea pigs (Cavia porcellus) and mice (Mus musculus) were used as controls for some experiments. We found that HA localization was similar in NMR, guinea pig, and mouse tissues but NMR had larger amounts and higher molecular weight (maximum, around 2.5 MDa) of HA in serum and almost all tissues tested. We could not find ultra-high molecular weight HA (≥ 4 MDa) in NMR samples, in contrast to previous descriptions. Hyaluronidase-1 had lower expression and activity in NMR than mouse lymph nodes. RNAseq results showed that, among HA family genes, Tnfaip6 and hyaluronidase-3 (Hyal3) were systematically overexpressed in NMR tissues. In conclusion, NMR samples, contrary to expectations, do not harbor ultra-high molecular weight HA, although its amount and average molecular weight are higher in NMR than in guinea pig tissues and serum. Although hyaluronidase expression and activity are lower in NMR than mouse lymph nodes, this not sufficient to explain the presence of high molecular weight HA. A different activity of the NMR HA synthases remains possible. These characteristics, together with extremely high Hyal3 and Tnfaip6 expression, may provide the NMR with a bespoke, and perhaps protective, HA metabolism.

Low sulfide levels and a high degree of cystathionine ß-synthase (CBS) activation by S-adenosylmethionine (SAM) in the long-lived naked mole-rat.

Hydrogen sulfide (H2S) is a gaseous signalling molecule involved in many physiological and pathological processes. There is increasing evidence that H2S is implicated in aging and lifespan control in the diet-induced longevity models. However, blood sulfide concentration of naturally long-lived species is not known. Here we measured blood sulfide in the long-lived naked mole-rat and five other mammalian species considerably differing in lifespan and found a negative correlation between blood sulfide and maximum longevity residual. In addition, we show that the naked mole-rat cystathionine ß-synthase (CBS), an enzyme whose activity in the liver significantly contributes to systemic sulfide levels, has lower activity in the liver and is activated to a higher degree by S-adenosylmethionine compared to other species. These results add complexity to the understanding of the role of H2S in aging and call for detailed research on naked mole-rat transsulfuration.

Analysis of Alpha-2 Macroglobulin from the Long-Lived and Cancer-Resistant Naked Mole-Rat and Human Plasma.

BACKGROUND: The naked mole-rat (NMR) is a long-lived and cancer resistant species. Identification of potential anti-cancer and age related mechanisms is of great interest and makes this species eminent to investigate anti-cancer strategies and understand aging mechanisms. Since it is known that the NMR expresses higher liver mRNA-levels of alpha 2-macroglobulin than mice, nothing is known about its structure, functionality or expression level in the NMR compared to the human A2M. RESULTS: Here we show a comprehensive analysis of NMR- and human plasma-A2M, showing a different prediction in glycosylation of NMR-A2M, which results in a higher molecular weight compared to human A2M. Additionally, we found a higher concentration of A2M (8.3±0.44 mg/mL vs. and 4.4±0.20 mg/mL) and a lower total plasma protein content (38.7±1.79 mg/mL vs. 61.7±3.20 mg/mL) in NMR compared to human. NMR-A2M can be transformed by methylamine and trypsin resulting in a conformational change similar to human A2M. NMR-A2M is detectable by a polyclonal antibody against human A2M. Determination of tryptic and anti-tryptic activity of NMR and human plasma revealed a higher anti-tryptic activity of the NMR plasma. On the other hand, less proteolytic activity was found in NMR plasma compared to human plasma. CONCLUSION: We found transformed NMR-A2M binding to its specific receptor LRP1. We could demonstrate lower protein expression of LRP1 in the NMR liver tissue compared to human but higher expression of A2M. This was accompanied by a higher EpCAM protein expression as central adhesion molecule in cancer progression. NMR-plasma was capable to increase the adhesion in human fibroblast in vitro most probably by increasing CD29 protein expression. This is the first report, demonstrating similarities as well as distinct differences between A2M in NMR and human plasma. This might be directly linked to the intriguing phenotype of the NMR and suggests that A2M might probably play an important role in anti-cancer and the anti-aging mechanisms in the NMR.

Unusual ratio between free thyroxine and free triiodothyronine in a long-lived mole-rat species with bimodal ageing.

Ansell's mole-rats (Fukomys anselli) are subterranean, long-lived rodents, which live in eusocial families, where the maximum lifespan of breeders is twice as long as that of non-breeders. Their metabolic rate is significantly lower than expected based on allometry, and their retinae show a high density of S-cone opsins. Both features may indicate naturally low thyroid hormone levels. In the present study, we sequenced several major components of the thyroid hormone pathways and analyzed free and total thyroxine and triiodothyronine in serum samples of breeding and non-breeding F. anselli to examine whether a) their thyroid hormone system shows any peculiarities on the genetic level, b) these animals have lower hormone levels compared to euthyroid rodents (rats and guinea pigs), and c) reproductive status, lifespan and free hormone levels are correlated. Genetic analyses confirmed that Ansell's mole-rats have a conserved thyroid hormone system as known from other mammalian species. Interspecific comparisons revealed that free thyroxine levels of F. anselli were about ten times lower than of guinea pigs and rats, whereas the free triiodothyronine levels, the main biologically active form, did not differ significantly amongst species. The resulting fT4:fT3 ratio is unusual for a mammal and potentially represents a case of natural hypothyroxinemia. Comparisons with total thyroxine levels suggest that mole-rats seem to possess two distinct mechanisms that work hand in hand to downregulate fT4 levels reliably. We could not find any correlation between free hormone levels and reproductive status, gender or weight. Free thyroxine may slightly increase with age, based on sub-significant evidence. Hence, thyroid hormones do not seem to explain the different ageing rates of breeders and non-breeders. Further research is required to investigate the regulatory mechanisms responsible for the unusual proportion of free thyroxine and free triiodothyronine.

Novel CIC point mutations and an exon-spanning, homozygous deletion identified in oligodendroglial tumors by a comprehensive genomic approach including transcriptome sequencing.

Oligodendroglial tumors form a distinct subgroup of gliomas, characterized by a better response to treatment and prolonged overall survival. Most oligodendrogliomas and also some oligoastrocytomas are characterized by a unique and typical unbalanced translocation, der(1,19), resulting in a 1p/19q co-deletion. Candidate tumor suppressor genes targeted by these losses, CIC on 19q13.2 and FUBP1 on 1p31.1, were only recently discovered. We analyzed 17 oligodendrogliomas and oligoastrocytomas by applying a comprehensive approach consisting of RNA expression analysis, DNA sequencing of CIC, FUBP1, IDH1/2, and array CGH. We confirmed three different genetic subtypes in our samples: i) the "oligodendroglial" subtype with 1p/19q co-deletion in twelve out of 17 tumors; ii) the "astrocytic" subtype in three tumors; iii) the "other" subtype in two tumors. All twelve tumors with the 1p/19q co-deletion carried the most common IDH1 R132H mutation. In seven of these tumors, we found protein-disrupting point mutations in the remaining allele of CIC, four of which are novel. One of these tumors also had a deleterious mutation in FUBP1. Only by integrating RNA expression and array CGH data, were we able to discover an exon-spanning homozygous microdeletion within the remaining allele of CIC in an additional tumor with 1p/19q co-deletion. Therefore we propose that the mutation rate might be underestimated when looking at sequence variants alone. In conclusion, the high frequency and the spectrum of CIC mutations in our 1p/19q-codeleted tumor cohort support the hypothesis that CIC acts as a tumor suppressor in these tumors, whereas FUBP1 might play only a minor role.

RNA sequencing reveals differential expression of mitochondrial and oxidation reduction genes in the long-lived naked mole-rat when compared to mice.

The naked mole-rat (Heterocephalus glaber) is a long-lived, cancer resistant rodent and there is a great interest in identifying the adaptations responsible for these and other of its unique traits. We employed RNA sequencing to compare liver gene expression profiles between naked mole-rats and wild-derived mice. Our results indicate that genes associated with oxidoreduction and mitochondria were expressed at higher relative levels in naked mole-rats. The largest effect is nearly 300-fold higher expression of epithelial cell adhesion molecule (Epcam), a tumour-associated protein. Also of interest are the protease inhibitor, alpha2-macroglobulin (A2m), and the mitochondrial complex II subunit Sdhc, both ageing-related genes found strongly over-expressed in the naked mole-rat. These results hint at possible candidates for specifying species differences in ageing and cancer, and in particular suggest complex alterations in mitochondrial and oxidation reduction pathways in the naked mole-rat. Our differential gene expression analysis obviated the need for a reference naked mole-rat genome by employing a combination of Illumina/Solexa and 454 platforms for transcriptome sequencing and assembling transcriptome contigs of the non-sequenced species. Overall, our work provides new research foci and methods for studying the naked mole-rat's fascinating characteristics.

Pipeline for large-scale microdroplet bisulfite PCR-based sequencing allows the tracking of hepitype evolution in tumors.

Cytosine methylation provides an epigenetic level of cellular plasticity that is important for development, differentiation and cancerogenesis. We adopted microdroplet PCR to bisulfite treated target DNA in combination with second generation sequencing to simultaneously assess DNA sequence and methylation. We show measurement of methylation status in a wide range of target sequences (total 34 kb) with an average coverage of 95% (median 100%) and good correlation to the opposite strand (rho = 0.96) and to pyrosequencing (rho = 0.87). Data from lymphoma and colorectal cancer samples for SNRPN (imprinted gene), FGF6 (demethylated in the cancer samples) and HS3ST2 (methylated in the cancer samples) serve as a proof of principle showing the integration of SNP data and phased DNA-methylation information into "hepitypes" and thus the analysis of DNA methylation phylogeny in the somatic evolution of cancer.

Statistical inference of allelic imbalance from transcriptome data.

Next-generation sequencing and the availability of high-density genotyping arrays have facilitated an analysis of somatic and meiotic mutations at unprecedented level, but drawing sensible conclusions about the functional relevance of the detected variants still remains a formidable challenge. In this context, the study of allelic imbalance in intermediate RNA phenotypes may prove a useful means to elucidate the likely effects of DNA variants of unknown significance. We developed a statistical framework for the assessment of allelic imbalance in next-generation transcriptome sequencing (RNA-seq) data that requires neither an expression reference nor the underlying nuclear genotype(s), and that allows for allele miscalls. Using extensive simulation as well as publicly available whole-transcriptome data from European-descent individuals in HapMap, we explored the power of our approach in terms of both genotype inference and allelic imbalance assessment under a wide range of practically relevant scenarios. In so doing, we verified a superior performance of our methodology, particularly at low sequencing coverage, compared to the more simplistic approach of completely ignoring allele miscalls. Because the proposed framework can be used to assess somatic mutations and allelic imbalance in one and the same set of RNA-seq data, it will be particularly useful for the analysis of somatic genetic variation in cancer studies.

Haplotyping and copy number estimation of the highly polymorphic human beta-defensin locus on 8p23 by 454 amplicon sequencing.

BACKGROUND: The beta-defensin gene cluster (DEFB) at chromosome 8p23.1 is one of the most copy number (CN) variable regions of the human genome. Whereas individual DEFB CNs have been suggested as independent genetic risk factors for several diseases (e.g. psoriasis and Crohn's disease), the role of multisite sequence variations (MSV) is less well understood and to date has only been reported for prostate cancer. Simultaneous assessment of MSVs and CNs can be achieved by PCR, cloning and Sanger sequencing, however, these methods are labour and cost intensive as well as prone to methodological bias introduced by bacterial cloning. Here, we demonstrate that amplicon sequencing of pooled individual PCR products by the 454 technology allows in-depth determination of MSV haplotypes and estimation of DEFB CNs in parallel. RESULTS: Six PCR products spread over approximately 87 kb of DEFB and harbouring 24 known MSVs were amplified from 11 DNA samples, pooled and sequenced on a Roche 454 GS FLX sequencer. From approximately 142,000 reads, approximately 120,000 haplotype calls (HC) were inferred that identified 22 haplotypes ranging from 2 to 7 per amplicon. In addition to the 24 known MSVs, two additional sequence variations were detected. Minimal CNs were estimated from the ratio of HCs and compared to absolute CNs determined by alternative methods. Concordance in CNs was found for 7 samples, the CNs differed by one in 2 samples and the estimated minimal CN was half of the absolute in one sample. For 7 samples and 2 amplicons, the 454 haplotyping results were compared to those by cloning/Sanger sequencing. Intrinsic problems related to chimera formation during PCR and differences between haplotyping by 454 and cloning/Sanger sequencing are discussed. CONCLUSION: Deep amplicon sequencing using the 454 technology yield thousands of HCs per amplicon for an affordable price and may represent an effective method for parallel haplotyping and CN estimation in small to medium-sized cohorts. The obtained haplotypes represent a valuable resource to facilitate further studies of the biomedical impact of highly CN variable loci such as the beta-defensin locus.

A subtle alternative splicing event of the Na(V)1.8 voltage-gated sodium channel is conserved in human, rat, and mouse.

The voltage-gated sodium channel subtype Na(V)1.8 (SCN10A) is exclusively expressed in dorsal root ganglia (DRG) and plays a critical role in pain perception. We isolated mRNA from human, rat, and mouse DRGs and screened for alternatively spliced isoforms of the SCN10A mRNA using 454 sequencing. In all three species, we found an event of subtle alternative splicing at a NAGNAG tandem acceptor that results in isoforms including or lacking glutamine 1030 (Na(V)1.8+Q and Na(V)1.8-Q, respectively) within the cytoplasmic loop between domains II and III. The relative amount of Na(V)1.8-Q mRNA in adult DRG was measured with 14.1 +/- 0.1% in humans and 11.2 +/- 0.2% in rats. This is in contrast to an abundance of 64.3 +/- 0.3% in mouse DRG. Thus, the NAGNAG tandem acceptor in SCN10A is conserved among rodents and humans but its alternative usage apparently occurs with species-specific abundance. Analysis of human Na(V)1.8+Q and -Q isoforms in whole-cell patch-clamp experiments after heterologous expression in the neuroblastoma cell line Neuro-2A revealed no obvious impact of the splicing event on channel function.

Genetic variants of the copy number polymorphic beta-defensin locus are associated with sporadic prostate cancer.

BACKGROUND/AIMS: Prostate cancer represents the cancer with the highest worldwide prevalence in men. Chromosome 8p23 has shown suggestive genetic linkage to early-onset familial prostate cancer and is frequently deleted in cancer cells of the urogenital tract. Within this locus some beta-defensin genes (among them DEFB4, DEFB103, DEFB104) are localized, which are arranged in a gene cluster shown to exhibit an extensive copy number variation in the population. This structural variation considerably hampers genetic studies. In a new approach considering both sequence as well as copy number variations we aimed to compare the defensin locus at 8p23 in prostate cancer patients and controls. METHODS: We apply PCR/cloning-based haplotyping and high-throughput copy number determination methods which allow assessment of both individual haplotypes and gene copy numbers not accessible to conventional SNP-based genotyping. RESULTS: We demonstrate association of four common DEFB104 haplotypes with the risk of prostate cancer in two independent patient cohorts. Moreover, we show that high copy numbers (>9) of the defensin gene cluster are significantly underrepresented in both patient samples. CONCLUSIONS: Our findings imply a role of the antibacterial defensins in prostate cancerogenesis qualifying distinct gene variants and copy numbers as potential tumor markers.

High-resolution mapping of the 8p23.1 beta-defensin cluster reveals strictly concordant copy number variation of all genes.

One unexpected feature of the human genome is the high structural variability across individuals. Frequently, large regions of the genome show structural polymorphisms and many vary in their abundance. However, accurate methods for the characterization and typing of such copy number variations (CNV) are needed. The defensin cluster at the human region 8p23.1 is one of the best studied CNV regions due to its potential clinical relevance for innate immunity, inflammation, and cancer. The region can be divided into two subclusters, which harbor predominantly either alpha- or beta-defensin genes. Previous studies assessing individual copy numbers gave different results regarding whether the complete beta-defensin cluster varies or only particular genes therein. We applied multiplex ligation-dependent probe amplification (MLPA) to measure defensin locus copy numbers in 42 samples. The data show strict copy number concordance of all 10 loci typed within the beta-defensin cluster in each individual, while seven loci within the alpha-defensin cluster are consistently found as single copies per chromosome. The exception is DEFA3, which is located within the alpha-defensin cluster and was found to also differ in copy number interindividually. Absolute copy numbers ranged from two to nine for the beta-defensin cluster and zero to four for DEFA3. The CNV-typed individuals, including HapMap samples, are publicly available and may serve as a universal reference for absolute copy number determination. On this basis, MLPA represents a reliable technique for medium- to high-throughput typing of 8p23.1 defensin CNV in association studies for diverse clinical phenotypes.

Polymorphic segmental duplications at 8p23.1 challenge the determination of individual defensin gene repertoires and the assembly of a contiguous human reference sequence.

BACKGROUND: Defensins are important components of innate immunity to combat bacterial and viral infections, and can even elicit antitumor responses. Clusters of defensin (DEF) genes are located in a 2 Mb range of the human chromosome 8p23.1. This DEF locus, however, represents one of the regions in the euchromatic part of the final human genome sequence which contains segmental duplications, and recalcitrant gaps indicating high structural dynamics. RESULTS: We find that inter- and intraindividual genetic variations within this locus prevent a correct automatic assembly of the human reference genome (NCBI Build 34) which currently even contains misassemblies. Manual clone-by-clone alignment and gene annotation as well as repeat and SNP/haplotype analyses result in an alternative alignment significantly improving the DEF locus representation. Our assembly better reflects the experimentally verified variability of DEF gene and DEF cluster copy numbers. It contains an additional DEF cluster which we propose to reside between two already known clusters. Furthermore, manual annotation revealed a novel DEF gene and several pseudogenes expanding the hitherto known DEF repertoire. Analyses of BAC and working draft sequences of the chimpanzee indicates that its DEF region is also complex as in humans and DEF genes and a cluster are multiplied. Comparative analysis of human and chimpanzee DEF genes identified differences affecting the protein structure. Whether this might contribute to differences in disease susceptibility between man and ape remains to be solved. For the determination of individual DEF gene repertoires we provide a molecular approach based on DEF haplotypes. CONCLUSIONS: Complexity and variability seem to be essential genomic features of the human DEF locus at 8p23.1 and provides an ongoing challenge for the best possible representation in the human reference sequence. Dissection of paralogous sequence variations, duplicon SNPs ans multisite variations as well as haplotypes by sequencing based methods is the way for future studies of interindividual DEF locus variability and its disease association.