Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Oral iron acutely elevates bacterial growth in human serum


Iron deficiency is the most common nutrient deficiency worldwide and routine supplementation is standard policy for pregnant mothers and children in most low-income countries. However, iron lies at the center of host-pathogen competition for nutritional resources and recent trials of iron administration in African and Asian children have resulted in significant excesses of serious adverse events including hospitalizations and deaths. Increased rates of malaria, respiratory infections, severe diarrhea and febrile illnesses of unknown origin have all been reported, but the mechanisms are unclear. We here investigated the ex vivo growth characteristics of exemplar sentinel bacteria in adult sera collected before and 4 h after oral supplementation with 2 mg/kg iron as ferrous sulfate. Escherichia coli, Yersinia enterocolitica and Salmonella enterica serovar Typhimurium (all gram-negative bacteria) and Staphylococcus epidermidis (gram-positive) showed markedly elevated growth in serum collected after iron supplementation. Growth rates were very strongly correlated with transferrin saturation (p < 0.0001 in all cases). Growth of Staphylococcus aureus, which preferentially scavenges heme iron, was unaffected. These data suggest that even modest oral supplements with highly soluble (non-physiological) iron, as typically used in low-income settings, could promote bacteremia by accelerating early phase bacterial growth prior to the induction of immune defenses.


Iron deficiency (ID) remains the most pervasive nutritional deficiency worldwide. The prevalence of ID in mothers and young children frequently exceeds 50% in low-income countries. Insufficient iron impairs growth and cognitive development in childhood1,2.

Low cost iron supplements are effective for the treatment of ID and in countries with ID rates of >40%, the World Health Organization recommends universal iron supplementation of pregnant women and young children3,4,5. To overcome perceived limitations in the ability to absorb iron, supplements usually employ highly soluble forms of iron (ferrous sulfate or fumarate) given in rather large non-physiological bolus doses. The wisdom of these policies has long been questioned6,7 and has come under serious scrutiny starting in 2006 with the premature termination of a large trial in Pemba, Tanzania after significant increases in serious adverse outcomes (hospitalizations and deaths) in young children receiving iron-folate supplements were seen8. The emphasis was originally focused on malaria as the causative agent for the increases in morbidity and mortality during iron supplementation9 but subsequent trials have described excesses of other infections in groups randomized to iron or multiple micronutrients containing iron10,11,12,13,14,15. These findings have paralyzed iron supplementation policies.

The underlying mechanisms and the types of organisms responsible for these clinical and epidemiological observations remain unclear. Recent field studies suggest that oral iron supplementation in children increases susceptibility to bacterial infections, particularly diarrhea13,16, alters the gut microbiota16,17 and increases the virulence of many common bacterial enteropathogens18,19,20.

Numerous animal studies over many decades have shown that administration of iron in diverse forms accelerates the growth of peritoneally-injected pathogens, causing a septicemia with rapidly fatal outcomes21,22. We here examine the possibility in humans that a simple oral dose of supplemental iron could promote bacterial growth in serum. We used a series of ex vivo bacterial growth assays with sentinel organisms that were selected on the basis of their varying modes of pathogenesis and abilities to scavenge iron from the host.


Oral iron supplementation increases iron parameters in serum

To determine the effects of oral iron supplementation on bacterial growth in human serum, we enrolled 48 normal healthy non-anemic male subjects [mean ± SD: Hemoglobin (Hgb) = 14.5 ± 1.13 g/dL; Mean Corpuscular Volume (MCV) = 83.8 ± 5.5 fL; Ferritin = 62.8 ± 53.2 ng/mL]. Volunteers donated serum immediately before and then four hours after, oral ingestion of 400 mg ferrous sulfate (containing the equivalent of 130 mg of elemental iron). Transferrin saturation (TSAT) increased from 42.1% (±12.5%, SD) to 75.7% (±18.1%, SD) and total serum iron increased from 30.3 μmol/L (± 10.2μmol/L, SD) to 53.0μmol/L (± 15.8μmol/L, SD) four hours after iron supplementation.

Effects of oral iron supplementation on ex vivobacterial growth in serum

We next measured the growth of the five species of sentinel bacteria in the baseline and post-dose sera. To account for the between-subject variance in the starting transferrin saturation levels, we used mixed statistical models to allow two nested higher levels of variation: patient and bleed (pre- and post-iron supplementation). This enabled us to independently analyze the effects of TSAT and iron supplementation.

S. aureus, an organism with a strong preference for heme-derived iron23, behaved differently from the other four bacteria (Fig. 1A). Iron supplementation had no impact on the general pattern of the growth curve (p = 0.3). Both the time to reach peak doubling time (p = 0.21) and the doubling time during the exponential growth phase (p = 0.78) were also unchanged by iron supplementation (Table 1). Growth did not correlate with TSAT (p = 0.08) (Table 2).

Table 1 Statistical analysis of ex vivo bacterial growth assays.
Table 2 Statistical testing of the independent effects of iron supplementation and transferrin saturation (TSAT).
Figure 1

Growth of sentinel gram-positive bacteria in human serum before and after oral iron supplementation.

S. aureus (A) and S. epidermidis (B) were grown in serum from subjects before (blue) and after (red) oral iron supplementation with 400 mg ferrous sulfate (containing the equivalent of 130 mg of elemental iron). The thicker central lines represent the fitted curves for an average individual and the thinner lines the 95% confidence intervals for these estimates. The points show individual values of OD620. Curves were derived from mixed effects models fitting degree-four orthogonal polynomials in time to log(OD620), where OD620 is the mean optical density of the three replicates at time (t).

S. epidermidis (Fig. 1B) demonstrated an initial delay in growth in comparison with the other species. Iron supplementation influenced the overall pattern of the growth curve (p < 0.0001). Specifically, iron supplementation reduced the lag phase, the time to reach peak doubling time (p = 0.001) and increased the doubling time during the exponential growth phase (p < 0.001) (Table 1). The very strong effect of iron on the overall increase in bacterial growth (X2 = 55 (approximately), p < 0.0001) was equally explained using pre/post supplementation as a dichotomized variable or by TSAT (Table 2).

S. Typhimurium (Fig. 2A) and E. coli (Fig. 2B) both showed highly significant differences in their growth curves after iron supplementation (p < 0.0001). The doubling times post-iron supplementation were significantly shorter than pre-iron supplementation (p < 0.0001), but time to reach peak doubling time was unaffected by treatment (Table 1). Transferrin saturation had a very strong effect on overall growth rates of both S. Typhimurium (X2 = 348, p < 0.0001) and E. coli (X2 = 300, p < 0.0001), comparable to, indeed a little larger than, that of iron supplementation in both S. Typhimurium (X2 = 213, p < 0.0001) and E. coli (X2 = 221, p < 0.0001). For E. coli, iron supplementation (X2 = 35, p < 0.0001) and TSAT (X2 = 69, p < 0.0001) each had significant effects on bacterial growth after controlling for the other. The same held true for S. Typhimurium growth where both iron supplementation (X2 = 22, p = 0.0004) and TSAT (X2 = 105, p < 0.0001) each had significant effects on bacterial growth after controlling for the other (Table 2). This is likely to be because, in addition to capturing the effect of supplementation, transferrin saturation also explains differences between individuals.

Figure 2

Growth of sentinel gram-negative bacteria in human serum before and after oral iron supplementation.

S. Typhimurium (A), E. coli (B) and Y. entercolitica (C) were grown in serum from subjects before (blue) and after (red) oral iron supplementation with 400 mg ferrous sulfate (containing the equivalent of 130 mg of elemental iron). The thicker central lines represent the fitted curves for an average individual and the thinner lines the 95% confidence intervals for these estimates. The points show individual values of OD620. Curves were derived from mixed effects models fitting degree-four orthogonal polynomials in time to log(OD620), where OD620 is the mean optical density of the three replicates at time (t).

For Y. enterocolitica (Fig. 2C), the maximum growth rate was difficult to locate. In fact, although the growth curves clearly differ significantly (p < 0.0001), the doubling times at one hour did not differ significantly. As was the case for S. Typhimurium and E. coli, TSAT had an impact on growth (X2 = 120, p < 0.0001) as did iron supplementation (X2 = 108, p < 0.0001). Additionally, iron supplementation (X2 = 33, p < 0.0001) and TSAT (X2 = 37, p < 0.0001) each had significant effects on bacterial growth after controlling for the other (Table 2).


The biologically useful redox characteristics of the Fe(II) to Fe(III) transition place iron apart from other nutrients. Additionally, it lies at the epicenter of the host-pathogen battle for resource control. Host defense mechanisms to withhold iron from invading pathogens are some of the most evolutionarily conserved innate strategies against infection24, but most bacterial species have evolved counter-acting strategies for pirating host iron including: (1) receptors that bind transferrin, lactoferrin or hemoglobin; and (2) low molecular weight siderophores that acquire iron from host proteins or from low molecular weight iron compounds25.

The potential health threat posed by exogenous iron, repeatedly demonstrated in animal models25, has tended to be overlooked in clinical settings. The recent iron trials with adverse outcomes in children in developing countries8,10,11,12,13,14,15 have prompted new mechanistic studies providing experimental verification that oral iron adversely modifies the gut microbiome17,26 and increases the virulence of pathogenic enteric bacteria18,20. In this study we focused on the issue of systemic, as opposed to enteric, bacterial infections building upon prior knowledge that iron can precipitate septicemias (for instance, based on the disastrous outcomes of intramuscular iron-dextran administration to Polynesian neonates27).

The ex vivo assays we describe here show that customary oral supplementation with highly-soluble iron as ferrous sulfate can profoundly affect the growth dynamics of four of the five sentinel species we studied. This could potentially undermine a key component of innate immunity allowing such organisms to achieve overwhelming numbers by the time adaptive immune defense mechanisms are up-regulated. Note that the very strong correlations between TSAT and growth rates emphasizes the importance of this variable even in the presence of likely inter-individual differences in other iron-related (e.g. lipocalin-2, haptoglobin) and other (e.g. defensins) non-cellular defense mechanisms within the sera.

Recent molecular insights into human iron metabolism have challenged the basic pillars on which public health strategies involving highly soluble iron supplements have been developed. The prior belief that humans are constitutionally inefficient at absorbing iron and hence require large non-physiological doses taken apart from food, is now overturned by the knowledge that hepcidin actively down-regulates iron acquisition especially in the presence of an infectious threat24. The dual regulation of hepcidin by iron and infection (inflammation) underscores the threat posed by exogenous iron. An increase of hepcidin caused by an infection might have the evolutionary function of decreasing further iron uptake from the intestine to reduce circulating iron fuel for microorganisms. This suggests that we should not interfere via high dose iron supplements.

Transferrin saturation is homeostatically controlled with a normal range between 15–50% in males. Our data show that increasing TSAT from a mean of 42% to 76% profoundly stimulated bacterial growth with a continuous association across all levels of TSAT. The role of TSAT in mediating host susceptibility to infection has been known for almost half a century28. However, neither the strength of this association, nor the ability of acute increases in TSAT following iron doses to so rapidly favor bacterial growth, have been previously appreciated. The dose level selected for the adults in this study was based on that most frequently used for young children with iron deficiency (2 mg/kg/day), however our subjects were iron replete. In iron deficient children, hepcidin would be down-regulated to allow maximal iron absorption29 and hence TSAT would be expected to rise even further. Although a higher percentage of increase can be expected, the TSAT baseline will be lower and may still end up lower than in iron replete men. In the event of accidental ingress of pathogens through a cut, abrasion or leaky gut, these high levels of TSAT could precipitate a fulminant bacteremia before other cognate immune defenses have time to respond.

Iron absorbed from a natural food matrix, or even when ferrous sulfate is given with food30, is released much more slowly and causes minor deviations in TSAT and in non-transferrin bound iron (NTBI)30 which, according to the data presented here, would be much less likely to promote bacterial growth. In regions of the world where iron-rich foods are scarce, or too costly, supplementation with slow release nano-molecular formulations that mimic a food matrix31 may provide a safer option and might additionally reduce the adverse effects of unabsorbed iron on the gut microbiota. Field trials of such compounds are warranted. Additional safe approaches to supplement iron include the use of oral bovine lactoferrin32.

Materials and Methods


Forty-eight male Gambian subjects (averaging 40y; range 21–64y) were recruited. Subjects had no history of fever, illness or anti-microbial use during the preceding seven days; were malaria rapid test negative and were non-anemic (Hemoglobin >12 g/dL). All patients donated blood between 9 and 10 am, on an empty stomach, immediately prior to and four hours after taking 400 mg ferrous sulfate orally. Food was provided two hours after iron supplementation.

Informed consent was obtained from all subjects. This study was approved by the Gambian Government/MRC Joint Ethics Committee (SCC1312v2) and by the University of North Carolina Institutional Review Board (protocol #143044). All experiments were performed in accordance with the approved guidelines.

Biochemical Parameters

Complete blood counts were obtained using a Medonic M series (Boule Diagnostics Int AB, Stockholm, Sweden) hematology analyzer. Serum biochemical parameters including serum iron, transferrin, ferritin and transferrin saturation were obtained using a Cobas Integra 400 plus (Roche, Basel, Switzerland) biochemistry analyzer.

Bacterial Growth assays

Staphylococcus aureus (strain NCTC8325), Staphylococcus epidermidis (FDA strain PCI1200, ATCC12228), Salmonella enterica serovar Typhimurium (strain LT2, ATCC19585) and Escherichia coli (strain Crooks, ATCC8739) were grown overnight for 18 hours at 37 °C in 5 mL iron free minimal growth media, Iscove’s Modified Dulbecco’s Medium (IMDM, Invitrogen). This was conducted in air with continuous shaking (250 rpm). A high-virulence, siderophore producing Yersinia enterocolitica (strain WA-314, ATCC51871) was grown in IMDM containing 10 mM ethylene glycol tetraacetic acid (EGTA, Sigma) (pH7). All growth assays were run in triplicate in IMDM containing 50% heat-inactivated human serum. Bacterial growth was monitored by measuring the optical density at 620 nm (OD620) hourly for 12 hours (Staphylococcus aureus, Salmonella enterica serovar Typhimurium and Escherichia coli) and then at 20, 28, 36 hours (Staphylococcus epidermidis and Yersinia enterocolitica) using a Multiscan FC ELISA plate reader (Thermo Scientific).

Statistical analysis

We analyzed the growth assays with a mixed model using a quartic polynomial in Stata12 (StataCorp, College Station, TX). We used the logarithm of the mean of the three replicate OD readings at each time point as the response variable and modeled growth trajectories by fitting this to orthogonal polynomials (to degree 4) in time using mixed effects models with random intercept and coefficients due to patient and, nested within patient, bleed (pre and post). In effect, therefore, the growth trajectory for the ith blood sample was modeled by an equation of the form:

where t is time since inoculation, the βs are estimated coefficients and the τs and ε are the random effects.

For each species of bacterium, we compared its growth pre- and post-iron supplementation by examining differences in (1) Growth Curve i.e. general pattern of the overall growth trajectories; (2) Time to Reach Peak Doubling Time, i.e. the time at which the rate of increase in log[OD] was at a maximum and (iii) Doubling Time During Exponential Growth Phase, i.e. when the doubling time was at its greatest, or, if the lag phase was not detectable at 1 hour after inoculation. We estimated the timing of the maximum slope by setting the second derivative with respect to time of the deterministic component of the model to zero. We calculated the doubling time at time t as log(2)/(slope of log[OD(t)]). We employed the likelihood ratio test to compare models for which the growth patterns were and were not dependent on the transferrin saturation or iron supplementation. We used the delta method to obtain the 95% confidence intervals for the time at which maximum slope occurred and the doubling time.

Additional Information

How to cite this article: Cross, J. H. et al. Oral iron acutely elevates bacterial growth in human serum. Sci. Rep. 5, 16670; doi: 10.1038/srep16670 (2015).


  1. Black, M. M., Quigg, A. M., Hurley, K. M. & Pepper, M. R. Iron deficiency and iron-deficiency anemia in the first two years of life: strategies to prevent loss of developmental potential. Nutr. Rev. 69 Suppl 1, S64–70 (2011).

    Article  Google Scholar 

  2. Miller, J. L. Iron deficiency anemia: a common and curable disease. Cold Spring Harb. Perspect. Med. 3, (2013).

  3. Peña-Rosas, J. P., De-Regil, L. M., Dowswell, T. & Viteri, F. E. Daily oral iron supplementation during pregnancy. Cochrane Database Syst. Rev. 12, (2012) CD004736.

    PubMed  PubMed Central  Google Scholar 

  4. Okebe, J. U., Yahav, D., Shbita, R. & Paul, M. Oral iron supplements for children in malaria-endemic areas. Cochrane Database Syst. Rev. Online CD006589 (2011), 10.1002/14651858.CD006589.pub3.

  5. Nutritional anaemias. Report of a WHO scientific group. World Health Organ. Tech. Rep. Ser. 405, 5–37 (1968).

  6. Murray, M. J., Murray, A. B., Murray, M. B. & Murray, C. J. The adverse effect of iron repletion on the course of certain infections. Br. Med. J. 2, 1113–1115 (1978).

    CAS  Article  Google Scholar 

  7. Oppenheimer, S. J. Iron and its relation to immunity and infectious disease. J. Nutr. 131, 616S–633S (2001).

    CAS  Article  Google Scholar 

  8. Sazawal, S. et al. Effects of routine prophylactic supplementation with iron and folic acid on admission to hospital and mortality in preschool children in a high malaria transmission setting: community-based, randomised, placebo-controlled trial. Lancet 367, 133–143 (2006).

    CAS  Article  Google Scholar 

  9. Spottiswoode, N., Duffy, P. E. & Drakesmith, H. Iron, anemia and hepcidin in malaria. Front. Pharmacol. 5, 125 (2014).

    Article  Google Scholar 

  10. Esan, M. O. et al. Iron supplementation in HIV-infected Malawian children with anemia: a double-blind, randomized controlled trial. Clin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am. (2013), 10.1093/cid/cit528.

  11. Jonker, F. A. M. et al. Iron status predicts malaria risk in Malawian preschool children. PloS One 7, e42670 (2012).

    CAS  ADS  Article  Google Scholar 

  12. Manno, D. et al. Effect of multiple micronutrient-fortified food on mild morbidity and clinical symptoms in Zambian infants: results from a randomised controlled trial. Eur. J. Clin. Nutr. 65, 1163–1166 (2011).

    CAS  Article  Google Scholar 

  13. Soofi, S. et al. Effect of provision of daily zinc and iron with several micronutrients on growth and morbidity among young children in Pakistan: a cluster-randomised trial. Lancet 382, 29–40 (2013).

    CAS  Article  Google Scholar 

  14. Veenemans, J. et al. Effect of supplementation with zinc and other micronutrients on malaria in Tanzanian children: a randomised trial. PLoS Med. 8, (2011) e1001125.

    CAS  Article  Google Scholar 

  15. Zlotkin S. et al. Effect of iron fortification on malaria incidence in infants and young children in ghana: A randomized trial. JAMA 310, 938–947 (2013).

    CAS  Article  Google Scholar 

  16. Jaeggi, T. et al. Iron fortification adversely affects the gut microbiome, increases pathogen abundance and induces intestinal inflammation in Kenyan infants. Gut (2014), 10.1136/gutjnl-2014-307720.

  17. Zimmermann, M. B. et al. The effects of iron fortification on the gut microbiota in African children: a randomized controlled trial in Cote d’Ivoire. Am. J. Clin. Nutr. 92, 1406–1415 (2010).

    CAS  Article  Google Scholar 

  18. Kortman, G. A. M., Boleij, A., Swinkels, D. W. & Tjalsma, H. Iron availability increases the pathogenic potential of Salmonella typhimurium and other enteric pathogens at the intestinal epithelial interface. PloS One 7, (2012) e29968.

    CAS  ADS  Article  Google Scholar 

  19. Kortman, G. A. et al. Low dietary iron intake restrains the intestinal inflammatory response and pathology of enteric infection by food-borne bacterial pathogens. Eur. J. Immunol. (2015), 10.1002/eji.201545642.

  20. Dostal, A. et al. Salmonella adhesion, invasion and cellular immune responses are differentially affected by iron concentrations in a combined in vitro gut fermentation-cell model. PloS One 9, e93549 (2014).

    ADS  Article  Google Scholar 

  21. Bullen, J. J., Leigh, L. C. & Rogers, H. J. The effect of iron compounds on the virulence of Escherichia coli for guinea-pigs. Immunology 15, 581–588 (1968).

    CAS  PubMed  PubMed Central  Google Scholar 

  22. Bullen, J. J., Rogers, H. J. & Griffiths, E. Role of iron in bacterial infection. Curr. Top. Microbiol. Immunol. 80, 1–35 (1978).

    CAS  PubMed  Google Scholar 

  23. Skaar, E. P., Humayun, M., Bae, T., DeBord, K. L. & Schneewind, O. Iron-source preference of Staphylococcus aureus infections. Science 305, 1626–1628 (2004).

    CAS  ADS  Article  Google Scholar 

  24. Drakesmith, H. & Prentice, A. M. Hepcidin and the iron-infection axis. Science 338, 768–772 (2012).

    CAS  ADS  Article  Google Scholar 

  25. Weinberg, E. D. Iron availability and infection. Biochim. Biophys. Acta 1790, 600–605 (2009).

    CAS  Article  Google Scholar 

  26. Dostal, A. et al. Effects of iron supplementation on dominant bacterial groups in the gut, faecal SCFA and gut inflammation: a randomised, placebo-controlled intervention trial in South African children. Br. J. Nutr. 112, 547–556 (2014).

    CAS  Article  Google Scholar 

  27. Barry, D. M. & Reeve, A. W. Increased incidence of gram-negative neonatal sepsis with intramuscula iron administration. Pediatrics 60, 908–912 (1977).

    CAS  PubMed  Google Scholar 

  28. Kochan, I. Mecahnism of tuberculostasis in mammalian serum. I. Role of transferrin in human serum tuberculostasis. J. Infect. Dis. 119, 11–18 (1969).

    CAS  Article  Google Scholar 

  29. Prentice, A. M. et al. Hepcidin is the major predictor of erythrocyte iron incorporation in anemic African children. Blood 119, 1922–1928 (2012).

    CAS  Article  Google Scholar 

  30. Schümann, K. et al. Impact of oral iron challenges on circulating non-transferrin-bound iron in healthy Guatemalan males. Ann. Nutr. Metab. 60, 98–107 (2012).

    Article  Google Scholar 

  31. Pereira, D. I. A. et al. Nanoparticulate iron(III) oxo-hydroxide delivers safe iron that is well absorbed and utilised in humans. Nanomedicine Nanotechnol. Biol. Med. 10, 1877–1886 (2014).

    CAS  Article  Google Scholar 

  32. Nappi, C. et al. Efficacy and tolerability of oral bovine lactoferrin compared to ferrous sulfate in pregnant women with iron deficiency anemia: a prospective controlled randomized study. Acta Obstet. Gynecol. Scand. 88, 1031–1035 (2009).

    CAS  Article  Google Scholar 

Download references


The authors would like to acknowledge Bakary Darboe, Pierre Coulin, Karamo Camara, Ebrima Sise, Morgan Goheen, Daniel Gardner, Abdoulie Faal and all of the lab staff at the MRC-ING Keneba lab for their endless help, support and enthusiasm during this project. In addition, we would like to thank all of the blood donors who participated in the study. Funding: This work was supported by MCA760-5QX00 to the MRC International Nutrition Group by the UK Medical Research Council (MRC) and the UK Department for International Development (DFID) under the MRC/DFID Concordat agreement and by the National Institutes of Health and the National Institute of Child Health and Human Development (U01 HD061234 05S2 to CC).

Author information




J.H.C., R.S.B., A.P. and C.C. designed the experiments. J.H.C. and A.T.J. performed experiments. J.H.C., R.S.B. and R.W. recruited patients; A.F., A.P. and C.C. analyzed data; A.P. and C.C. wrote the manuscript. All authors approved the final manuscript.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Cross, J., Bradbury, R., Fulford, A. et al. Oral iron acutely elevates bacterial growth in human serum. Sci Rep 5, 16670 (2015).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing