News Article | April 12, 2016
A team of scientists led by Oxford University have made a discovery that could improve our chances of developing an effective vaccine against Tuberculosis. The researchers have identified new biomarkers for Tuberculosis (TB) which have shown for the first time why immunity from the widely used Bacillus Calmette-Guérin (BCG) vaccine is so variable. The biomarkers will also provide valuable clues to assess whether potential new vaccines could be effective. TB remains one of the world's major killer diseases, causing TB disease in 9.6 million people and 1.5 million deaths in 2014. The only available vaccine, Bacillus Calmette-Guérin (BCG), works well (estimated 50 percent effective) to prevent severe disease in children but is very variable (0 percent to 80 percent effective) in how well it protects against lung disease, particularly in countries where TB is most common. While BCG is one of the safest and most widely used vaccines worldwide, there is one key issue: It is currently very difficult to determine whether it will work or not. This also makes it really hard to determine if any new vaccines might work. For many vaccines, medics and scientists can use what are called immune correlates or biomarkers, typically in the blood, which can be measured to determine whether a vaccine has successfully induced immunity. Not only are these correlates useful in measuring the success of existing vaccination programmes, they are also invaluable in assessing whether potential new vaccines could be effective. With a pressing need for a TB vaccine that is more effective than BCG, a research team drawn from a number of groups at Oxford University, working with colleagues from the South African Tuberculosis Vaccine Initiative at the University of Cape Town and the London School of Hygiene & Tropical Medicine, set out to identify immune correlates that could facilitate TB vaccine development. The team, funded by the Wellcome Trust and Aeras, and led by Professor Helen McShane and Dr. Helen Fletcher, studied immune responses in infants in South Africa who were taking part in a TB vaccine trial. Professor McShane said: "We looked at a number of factors that could be used as immune correlates, to try and find biomarkers that will help us develop a better vaccine." The team carried out tests for twenty-two possible factors. One - levels of activated HLA-DR+CD4+ T-cells - was linked to higher TB disease risk. Meanwhile, BCG-specific Interferon-gamma secreting T-cells indicated lower TB risk, with higher levels of these cells directly linked to greater reduction of the risk of TB. Antibodies to a TB protein, Ag85A, were also identified as a possible correlate. Higher levels of Ag85A antibody were associated with lower TB risk. However, the team cautions that other environmental and disease factors could also cause Ag85A antibody levels to rise and so there may not be a direct link between the antibody and TB risk. Professor McShane said: "These are useful results which ideally would now be confirmed in further trials. They show that antigen-specific T cells are important in protection against TB, but that activated T cells increase the risk". Dr. Helen Fletcher from the London School of Hygiene & Tropical Medicine, said: "For the first time we have some evidence of how BCG might work, and also what could block it from working. Although there is still much work to do, these findings may bring us a step closer to developing a more effective vaccine for TB." Dr. Tom Scriba from the South African Tuberculosis Vaccine Initiative said: "TB is still a major international killer, and rates of TB disease in some areas of South Africa are among the highest in the world. These findings provide important clues about the type of immunity TB vaccines should elicit, and bring us closer to our vision, a world without TB." The team is continuing its work to develop a TB vaccine, aiming to protect more people from the disease. The paper was published in the jounral Nature Communications.
To investigate the susceptibility of different normal human mammary cell types to transformation under the influence of known oncogenes, we isolated CD49f+EpCAMlow basal cells (BCs), CD49f+EpCAM+ luminal progenitors (LPs), CD49f−EpCAM+ non-clonogenic luminal cells (LCs), and non-epithelial CD49f−EpCAM− stromal cells (SCs) at high purity (>97%) by fluorescence-activated cell sorting (FACS) from 37 normal human reduction mammoplasty samples depleted of endothelial and haematopoietic cells4, 7, 8 (Fig. 1a and Extended Data Tables 1 and 2). We then exposed one or more of these subsets to one or more oncogene-encoding lentiviral preparations (encoding complementary DNAs (cDNAs) for TP53R273C and green fluorescent protein (TP53R273C–GFP), PIK3CAH1047R and yellow fluorescent protein (PIK3CAH1047R–YFP)), and KRASG12D and mCherry (KRASG12D–mCherry) and, in some experiments, to a library of biologically neutral, barcoded lentiviral GFP vectors to allow subsequent clonal tracking of their progeny using a DNA sequencing approach5, 6 (Extended Data Fig. 1a). The cells were then embedded in a collagen gel (0.3 × 105 to 16 × 105 cells per gel) and the gels transplanted into highly immunodeficient NOD-SCID Il2rg−/− (NSG) or NOD Rag1−/−Il2rg−/− (NRG) female mice. In initial experiments, 2 × 105 irradiated C3H-10T1/2 fibroblasts were co-embedded in the gels which were then transplanted subrenally, followed by subcutaneous implantation of the recipients with a slow-release capsule containing 17-β-oestradiol and progestin (EP pellets)4, recognizing this would limit follow-up to 8 weeks because of incurred bladder toxicity. BCs and/or LPs isolated from 17 of 27 normal donors and exposed to all three oncogenic vectors produced tumours within 8 weeks (Fig. 1b) at similar overall frequencies (46% of BC isolates and 61% of LP isolates, respectively, Extended Data Fig. 1b). Identical treatment of LCs and SCs isolated from three of these samples did not produce any tumours in the same 8-week period. Both the BC- and LP-derived tumours resembled invasive ductal carcinomas (Fig. 2a and Extended Data Table 3) and were histologically very different from the organized bilayered structures generated in analogous xenografts of unmanipulated or simply barcoded normal human mammary cells4, 6. Secondary female immunodeficient mice transplanted subcutaneously with a small piece (~25–33% of the initial tumour mass)5, 9 from four of nine of these primary tumours (one BC- and eight LP-derived) developed palpable tumours within another 8 weeks (Extended Data Table 4). FACS analysis showed most of the transduced cells in all primary tumours examined co-expressed all three fluorescent reporters, consistent with the high transduction efficiencies measured in separate cell aliquots maintained in vitro for 72 h after virus exposure, and a similar expression profile was maintained in the single secondary tumour similarly analysed (Extended Data Fig. 1c, d). Notably, when the three oncogene-encoding viruses were tested on their own, or in pairs, tumours were obtained with similar efficiency only when the KRASG12D vector was included, and even on its own (64/102 for all transductions that included KRASG12D compared with 1/12 when KRASG12D was not present; for more details, see Extended Data Fig. 1b and Extended Data Tables 1 and 2). PCR and Sanger sequencing confirmed the tumour cells contained the expected oncogene sequences including doubly and triply fluorescent cells isolated separately from tumours arising from cells initially exposed to three oncogenic vectors (Extended Data Fig. 2a–c). Subsequent experiments demonstrated that invasive ductal carcinomas were also obtained at a similar frequency from both BCs and LPs (but not LCs or SCs from the same mammoplasty samples) when the transduced cells were transplanted subcutaneously without irradiated fibroblasts or EP pellets, even when the cells were exposed only to the KRASG12D vector (Extended Data Figs 1e and 2d and Extended Data Tables 1 and 2). These tumours could frequently also be serially passaged (Extended Data Fig. 2e and Extended Data Table 4) and their growth more accurately monitored by luciferase bioluminescence (Fig. 1c and Supplementary Table 1). FACS analysis of 15 tumours showed that 48 ± 5% of the cells were human EpCAM+ and/or HLA+, with similar results for BC- and LP-derived tumours (Extended Data Fig. 1c, f). Immunohistochemical (IHC) analyses of tumour sections (Fig. 2a, b) showed 88% and 55% of primary BC- and LP-derived tumours contained >5% ERα+ cells (median = 58% and 8% ERα+ cells, respectively), but none contained >2% PR+ cells. HER2+ cells were present at similar frequencies (in 88% and 52% of BC- and LP-derived tumours, respectively). Frequencies of Ki67+ cells ranged from 2% to 30%, with only one secondary tumour containing as many as 70% Ki67+ cells. In contrast, cells expressing EGFR, MUC1 and K8/18 were prevalent in almost all tumours examined. High K5 expression, normally exclusive to BCs, was prevalent (median = 90% K5+ cells) in most LP-derived tumours, and less (median = 5% K5+ cells) in BC-derived tumours. Expression of CD44, a marker associated with undifferentiated epithelial cells, was also less prevalent in BC- compared with LP-derived tumour cells (median = 2% and 50%, respectively). K14, another marker of normal human BCs, was also variably detected in both BC- and LP-derived tumours. Gene expression analyses (Extended Data Fig. 2f) showed that transcripts for vimentin (VIM) and N-cadherin (CDH2), normally found exclusively in BCs, were present at high levels in both LP- and BC-derived tumours, with similar results for E-cadherin (CDH1) and ELF5, genes normally expressed exclusively in LPs and LCs. However, transcript levels of SLUG (SNAI2), another BC marker, were strongly decreased in BC-derived tumours whereas transcript levels of both GATA3 and NOTCH3, two markers of LPs, showed little change. Cyclin-dependent kinase 1 (CDK1) was also highly expressed in all tumours, but other proliferation-associated genes, such as cyclin B1 (CCNB1) and PCNA, were highly expressed only in the LP-derived tumours. TERT transcripts remained at a similar level to that seen in the parental normal cell populations whereas those for VEGFA, HIF1A and MAPK3 were more variable. RNA sequencing analysis was conducted on FACS-purified human cells isolated from six primary tumours generated from triply transduced cells (three from BCs and three from LPs) and the matched starting cells. Unsupervised clustering showed a closer relation of coding gene transcript levels in all six tumour populations to each other than to the normal cells from which the tumours had arisen (Fig. 2c). This prominent sharing of transcriptome changes in tumours derived both from BCs and from LPs suggests a key role of their mode of creation on their resultant molecular features. Specific differences in the gene expression changes that distinguished BC- and LP-derived tumours and their respective starting cell populations showed shared increased and decreased expression of 146 and 22 genes, respectively in both, indicative of a common gene signature in the transformants (Fig. 2d, top). Further analyses using either PAM50 (ref. 10) or AIMS classifier methodologies11 indicated the transcriptional profiles of the de novo tumours most closely resembled those of spontaneous human breast cancers classified as ‘normal-like’ (Fig. 2e). However, the unsupervised clustering also indicated that the three BC- and three LP-derived tumours formed separate groups, suggesting some retained influence of their different origins. This was further supported by the finding that >20% (72) of the differentially expressed genes in the BC- and LP-derived tumours were similarly differentially expressed in the cells from which the tumours had arisen (Fig. 2d, bottom). Nevertheless, genes whose expression was upregulated in BC-derived tumours included several that are normally highly expressed in LPs and LCs but not BCs (for example, AR, ESR1, FOXA1, TOX3, EPCAM, EHF and ELF5). Conversely, the genes whose expression was upregulated in LP-derived tumours included several recognized BC-specific genes (for example, VIM, TP63, ACTA2, THY1 and CDH2, Supplementary Table 2). Clonal analyses were performed on primary tumours obtained from 45 isolates of BCs and LPs both from DNA extracted directly from tumour tissue and from FACS-purified human cells (Extended Data Fig. 3). The results showed a high variability in the clone content of different tumours (up to 1,700 using a threshold of 70 cells per clone), regardless of the protocol used for their generation (Extended Data Fig. 4a, b). Calculated (minimal) frequencies of tumorigenic clone-forming cells (T-CFCs) using the total number of initial cells transplanted as the denominator, ranged from 1/23,000 to 1/150. Paired comparisons for tumours produced from BCs and LPs from the same donor also did not reveal any effect on T-CFC frequency (Fig. 3a). To estimate clone sizes, we first derived ‘relative’ clone size values by normalizing each tumour to the sum of its absolute clone sizes. We then pooled the data for all tumours in each group being assessed. The overall distribution of relative clone sizes, like the clone frequencies, was very broad and showed no evidence of any effect of the cell of origin, oncogene(s) used or the transplantation site (Fig. 3b). Analysis of 15 secondary tumours showed their clonal content was often high but again very variable, regardless of their origin (Fig. 3c and Extended Data Fig. 5a, b). Calculated frequencies of secondary clones (with respect to the number of cells initially transplanted into primary mice) also yielded highly variable secondary T-CFC values but with no consistent difference from the calculated primary T-CFC frequencies. However, >75% of the clones detected in each secondary tumour were ‘new’; that is, not detected in the matching primary tumour (Fig. 3d). Moreover, most of the clones present in multiple sibling secondary tumours produced from a common primary tumour were also different from one another (two primary tumours analysed, Extended Data Fig. 5c). Overall the total measured T-CFC frequencies (calculated from the total number of different clones in the primary or secondary tumours combined) ranged from ~1/5,700 to ~1/120 (Extended Data Fig. 5d). The relative sizes of the clones present in secondary tumours were also highly variable (Fig. 3c). Interestingly, in secondary tumours, the median size of the ‘continuing’ clones (evident in both primary and secondary tumours) was significantly larger than the clones that first became detectable upon tumour passaging (P = 4 × 10−12, Mann–Whitney U-test, Fig. 3e right panels). We then analysed the clonal composition of the cells produced from oncogene-transduced BCs and LPs after just 2 weeks in subrenal transplants, before tumours become grossly evident. The results showed the sizes as well as the numbers of clones detected at this time to be similar to those detected 6 weeks later in tumours derived from the same input cells (Fig. 4a and Extended Data Fig. 6a, b). The distributions of the relative clone sizes measured in the 2-week transplants both of BCs and of LPs were also similar (Fig. 4b). However, after 2 weeks, the absolute sizes of the clones derived from the KRASG12D-transduced LPs were already significantly larger than the sizes of the clones produced by matching transplants of control vector-transduced cells (median = 206 and 93, respectively, P = 3.3 × 10−8, Mann–Whitney U-test), with a slightly smaller effect apparent in the progeny of BCs from the same two donors (median = 112 and 94, respectively, P = 3.6 × 10−7, Mann–Whitney U-test, Fig. 4c). These studies provide new insights into the earliest phases of malignant transformation in vivo of cells isolated directly from normal human mammary tissue. Four findings are particularly noteworthy. The first is the rapidity and efficiency, albeit with high variability, with which this process can be induced in prospectively purified, biologically distinct types of normal human mammary epithelial cells using a single transducing oncogene (KRASG12D). This finding challenges previous assumptions of a requirement for a slow, multi-step selective process to accrue the genetic and/or epigenetic changes needed to obtain a continuously growing tumour. Interestingly, we did not obtain tumours from LCs or SCs subjected to the same protocols, in contrast to a recent report of highly ERα+ tumours generated by transduction of EpCAM+CD49f− LCs with SV40/Ras12. The second important finding was the considerable heterogeneity displayed in the numbers, phenotypes and growth behaviour of clonally tracked human cells with tumorigenic activity in vivo within 2–8 weeks. This result suggests that a similar range and speed of perturbations may accompany the spontaneous development of some breast cancers in patients. A third and unexpected finding was the lack of a strong influence of the human mammary cell type initially transduced with the frequency of clones generated, the histopathology of the tumours produced or their loss of lineage-specific expression profiles. Taken together, this suggests a greater effect of the potent transforming role of the KRASG12D oncogene in these cells. The fourth finding was the frequent delayed activation of clonal growth observed in secondary tumours. This latency could either be biologically determined, reflecting an origin of these late appearing clones from their normal counterparts with similar features5, or simply reflective of a stochastic process, as previously indicated for established human breast cancer cell lines passaged in vivo4. These results set the stage for examining the molecular basis of the biological heterogeneity now revealed that can occur during the earliest stages of breast cancer formation, the role of additional modifiers and how these may influence the acquisition of treatment response and resistance13, 14.
News Article | April 10, 2016
Our modern human ancestors had interbred with Neanderthals hundreds of thousands of years ago, passing on about 99.5 percent of the same DNA. Despite the overlapping genes, the lineage of both modern humans and Neanderthals have been kept apart, and it's all rooted to genetic incompatibilities. A study led by researchers from Stanford University revealed that modern humans are genetically mismatched with our ancient relatives because of a missing chromosome fragment in males: the Neanderthal Y chromosome. The differences may have caused a "dead end" between Neanderthals and modern humans, genetically separating both groups. Neanderthal genes have been incorporated into our own, specifically on our X chromosomes. Neanderthal DNA has been linked to people's susceptibility to allergies, as well as to the increased risks for depression and nicotine addiction. Now, in the new study, Stanford researchers examined the Y chromosome from a 49,000-year-old male Neanderthal discovered in the El Sidrón cave in Spain. They compared it with that of chimps, and archaic and modern humans. Turns out, the Neanderthal Y chromosome has seemingly gone extinct without leaving a trace in human DNA. Although the missing gene could have just drifted out of the modern human gene pool, scientists say there is also another possible explanation for this distinction. If Neanderthals and modern humans often interbred, the Neanderthal Y chromosome could have created conditions that might have often led to miscarriages, experts said. The hybrid offspring who carried the Neanderthal Y chromosome could have also been infertile. Stanford researcher Fernando Mendez and his colleagues found mutations in four genes that could have hindered the process of passing the Y chromosome onto hybrid children. Mendez said these mutations could have played a role in the loss of Neanderthal Y chromosome. "We should pay attention to the potential role of immune incompatibilities in population isolation," said Mendez. Mendez and his team said male fetuses who were fathered by Neanderthal males and carried in the womb by human females would have miscarried. One of the mutations was found in the gene KDM5D, which contributes to cancer suppression. It has been linked to higher risks of miscarriage because it can trigger an immune response in pregnant women. Mendez said a woman's immune system may target a male fetus that carried the Neanderthal Y chromosome, specifically the H-Y genes, which are minor antigens that resemble HLA antigens. The latter is often transplanted by surgeons to ensure that recipients and organ donors have similar immune profiles. Evolutionary biologist Mark Pagel said this could very well be the reason why the Neanderthal Y chromosome is not present in modern humans. It could be the factor that keeps both species separate from each other. The findings of the study are featured in The American Journal of Human Genetics. © 2016 Tech Times, All rights reserved. Do not reproduce without permission.
News Article | April 10, 2016
Although the human genome has ancient fragments of Neanderthal DNA, researchers of a new study have not found evidence of the Neanderthal genes in modern males' Y chromosomes backing up theories that the male offspring of Neanderthals and Homo sapiens may have faced more challenges than their female counterparts. Earlier studies have sequenced mitochondrial DNA, which is passed to children from their mother, from fossils of Neanderthal women. The new study, which was published in The American Journal of Human Genetics on April 7, is the first to examine Neanderthal Y chromosome, which is passed exclusively from father to son. The researchers are not sure if the disappearance of the Y chromosome happened by chance or if it occurred by evolutionary circumstances but they think that genes in the Neanderthal's Y chromosome may not have been compatible with human genes. "The functional nature of the mutations we found suggests to us that Neanderthal Y chromosome sequences may have played a role in barriers to gene flow, but we need to do experiments to demonstrate this and are working to plan these now," said study researcher Carlos Bustamante, Stanford's School of Medicine. Study researcher Fernando Mendez, also from Stanford, explained that a woman's immune system may attack a male fetus that carries Neanderthal H-Y genes, minor histocompatibility antigens that resemble the HLA antigens that transplant surgeons have to check to ensure that the organ donor and recipient have similar immune profiles. The absence of the Neanderthal Y chromosomes could be attributed to women consistently miscarrying male babies that carry it. Although this is an unproven idea, modern women's immune system sometimes reacts to male offspring when there is genetic incompatibility. For instance, a mutation in the KDM5D gene, which plays a role in suppressing cancer, has earlier been linked to greater risk of miscarriage because it can elicit immune response in pregnant women. "We identify protein-coding differences between Neanderthal and modern human Y chromosomes, including potentially damaging changes to PCDH11Y, TMSB4Y, USP9Y, and KDM5D. Three of these changes are missense mutations in genes that produce male-specific minor histocompatibility (H-Y) antigens," the researchers wrote in their study. "It is possible that incompatibilities at one or more of these genes played a role in the reproductive isolation of the two groups." The analysis also allowed researchers to identify when the Neanderthals and Homo sapiens split, which is around 590,000 years ago. © 2016 Tech Times, All rights reserved. Do not reproduce without permission.
In India, the concepts of Ayurvedic medicine are first laid out in sacred texts known as the Vedas. The goal of the medicine is to maintain balance between body, mind and spirit. Treatments are tailored to each person's prakriti, or constitution. Hippocrates establishes the scientific practice of medicine in Greece. He believes everyone has four humours — blood, phlegm, and black and yellow bile — and treats people as individuals of unique age and health status. The 'father of taxonomy', Carl Linnaeus, catalogues diseases based on symptoms into groups such as 'feverish' or 'painful'. This does not always lead to better treatments. “The true sanctuary of medical science is in a laboratory,” writes French physiologist Claude Bernard in his Introduction to the Study of Experimental Medicine. Aided by a new understanding of bacteriology, and emerging technologies such as the blood-pressure cuff and X-rays, scientists start to develop better treatments. Doctors focus on assessing symptoms, taking shorter medical histories, and quickly categorizing patients according to their diagnosis. US physicians use arsenic, in Fowler's solution, to improve white-blood-cell counts in a subset of leukaemia patients. It does not help most patients, and can have side effects such as liver disease, but becomes the standard treatment. In his influential textbook The Principles and Practice of Medicine, Canadian physician and pathologist William Osler (pictured) writes: “If it were not for the great variability among individuals medicine might as well be a science and not an art.” Three European botanists — Hugo DeVries, Carl Correns and Erich von Tschermak — work out the laws of genetic inheritance, by using flowering plants, maize (corn) and peas. They realize that Gregor Mendel reported the same laws in 1865, although his publication had largely been ignored. This time, the scientific community takes notice. London physician Archibald Garrod studies alkaptonuria, which causes dark urine. It arises in siblings, and he links it to Mendel's laws of heredity, the cause being genetic variation in metabolism: “Just as no two individuals of a species are absolutely identical in bodily structure neither are their chemical processes carried out on exactly the same lines” ( Lancet 2, 1616–1620; 1902). Reuben Ottenberg of Mount Sinai Hospital in New York becomes the first to record testing a patient and donor for matching blood type before a transfusion. The practice makes transfusions safer, although it will be decades until this becomes common practice. Arthur Fox of DuPont Laboratories allows phenylthiocarbamide (PTC) powder to drift around his lab. A colleague complains of a bitter taste, which Fox cannot detect ( Proc. Natl Acad. Sci. USA 18, 115–120; 1932). Geneticist Laurence Snyder finds that PTC “taste blindness” is recessively inherited, indicating a genetic origin for responses to chemicals ( Ohio J. Sci. 32, 436–440; 1932). US geneticist Arno Motulsky proposes that inherited traits explain why people react differently to medications. As the use of medicines grows in the 1950s, adverse reactions are more common. Variation in the enzyme glucose-6-phosphate dehydrogenase explains sensitivity to the antimalarial primaquine, for example, and a prolonged response to succinylcholine, used in anaesthesia, results from pseudocholinesterase deficiency. Later, Friedrich Vogel coins the term 'pharmacogenetics'. US researchers initiate a gene-therapy trial in two girls with severe combined immunodeficiency (SCID) by inserting a functional adenosine deaminase gene into their blood cells. Both girls develop stronger immune systems. The US Food and Drug Administration (FDA) approves the first matched drug and diagnostic test: trastuzumab for breast-cancer patients whose tumours overexpress the HER2 protein. This is the first major application of precision medicine to fight a type of cancer. Scientists discover why 1 in 20 patients with HIV (pictured) are sensitive to the reverse transcriptase inhibitor abacavir. Certain HLA types — the protein markers that allow the immune system to distinguish self cells from foreign invaders, and that are used to match organ transplants — predict who will respond badly ( et al. Lancet 359, 727–732; 2002, and et al. Lancet 359, 1121–1122; 2002). A large clinical trial shows that genotyping patients eliminates hypersensitivity reactions ( et al. N. Engl. J. Med. 358, 568–579; 2008). After 13 years of effort, and at a cost of around US$3 billion, scientists collaborating across 6 nations complete the Human Genome Project, producing a sequence of human chromosomes encompassing more than 20,000 genes from the DNA of several volunteers. Roche Diagnostics' AmpliChip CYP450 test — a microarray that classifies patients according to their cytochrome P450 (CYP) enzymes — is approved in both the United States and Europe. These enzymes are known to partly determine how a person metabolizes and reacts to medications. The test can help doctors to select the right drug and dosage for medications such as antipsychotics and cancer treatments. However, the high cost — frequently upwards of $1,000 — means it is not widely used. The FDA approves a genetic test to improve the prescription of warfarin, a blood-thinning drug that 2 million US people start taking every year. Responses to the drug vary widely, and taking the wrong dose risks blood clots or excessive bleeding. The test checks two gene variants that encode CYP2C9 and VKORC1, the latter of which is the enzyme targeted by the drug that affects blood clotting. The FDA updates the labels on warfarin, explaining that genetics partly determines an individual's response to the drug. Massachusetts General Hospital says it will be the first clinic to profile tumour genes for all cancer patients. It will check 13 genes for 110 mutations that predict which drugs will be most effective. The cost is about $2,000, and if insurers won't pay, the hospital or its patients may have to. UK Prime Minister David Cameron launches the 100,000 Genomes Project to read the DNA of people with cancer or rare diseases and their families. Other large-scale sequencing projects also try to advance precision medicine into the clinic. In 2015, US President Barack Obama outlines the Precision Medicine Initiative, a $215-million project to study the genomes and health status of 1 million volunteers and develop the required databases and privacy standards. In early 2016, China reveals plans to sequence genomes and connect them with clinical data. US biotech company Illumina announces a system capable of sequencing a genome for only $1,000. However, the $10-million initial cost of the ten-instrument arrangement effectively limits its use to institutions that do high-volume sequencing.