Thirty mosquitoes in a tube, and thirty-seven washed forearms

Stefano Fantini 30 Ott 2026



A Beijing group measured how many mosquitoes flew towards the forearms of thirty-seven volunteers, and found more Bacillus on the skin of those who drew fewer. It then smeared those bacteria on the backs of three macaques and tested four molecules on the shaved skin of mice. The work appeared in Nature Communications on 19 October 2026, received on 26 January 2026.

 

Rise in abundance of the four volatile compounds identified in the study, measured by gas chromatography and mass spectrometry on the skin of animals treated with Bacillus licheniformis (blue) and Bacillus paralicheniformis (orange), against the control. The scale is logarithmic and a value of one means no increase. The four compounds were then tested on their own on the shaved skin of female mice, with laboratory-reared Aedes aegypti, and not on human skin. Chart produced by ScienceOnline from data in Gu Z. et al., Nature Communications 2026, 17:10869, doi 10.1038/s41467-026-78356-5. The article carries a CC BY-NC-ND 4.0 licence, which does not allow derivative works, and for that reason no figure from the paper has been reproduced. Original newsroom chart.


Thirty female Aedes aegypti, one hour and a single-channel olfactometer, a tube with two exits and one odour in play. On one side the forearm of a volunteer, washed with unscented neutral soap and rinsed with deionised water, on the other nothing. The volunteers were thirty-seven, twenty-three women and fourteen men, aged twenty-two to thirty-two, with no history of skin disease and no antibiotics or repellents in the previous month. Twenty-two drew mosquitoes to a statistically significant degree, fifteen did not.
Zhenyu Gu, Heting Gao, Hongxia Guo, Zhihua Fan, Weixian Zhang and their colleagues work at the State Key Laboratory of Pathogen and Biosecurity of the Academy of Military Medical Sciences in Beijing, at Guizhou Medical University in Guiyang and at Anhui Medical University in Hefei. The study sits in Nature Communications, volume 17, article 10869, received on 26 January 2026, accepted on 22 September 2026 and posted online on 19 October 2026, open access under a CC BY-NC-ND 4.0 licence.


Right after the mosquito assay the researchers wiped a sterile swab over forearms and palms for two minutes and sequenced the DNA collected. In the group that drew more mosquitoes Proteobacteria were worth 65.49 per cent and Firmicutes 10.93 per cent, while in the other group the proportions turn over, 27.65 and 65.57 per cent. The genus Bacillus goes from 0.57 per cent on the skin that attracts to 62.72 per cent on the skin that attracts less. The authors write that 60 per cent and more is not the real share of Bacillus on human skin. The DNA collected was too scarce and they amplified it before sequencing, and the amplification inflated Bacillus in every sample.
The next step was to smear those bacteria on living skin. Three male cynomolgus macaques, Macaca fascicularis, aged seven and eight, with a back shaved over roughly seven centimetres by twenty, and ten millilitres of suspension holding a hundred million cells per millilitre, applied on three days running. The controls received culture broth alone. The mosquitoes were fifty in an acrylic chamber, the counts were made by two researchers blind to which animal was treated, five replicates each time. The index used compares the counts in the treated group and the control before and after application, and below one means fewer mosquitoes. It stayed below one on days two, five and eight, without fading.


The metabolites on the macaques' skin changed as well. Bacillus licheniformis raised sixty-seven of them and lowered twelve, Bacillus paralicheniformis twenty-two and four. The pathways most affected involve phenylalanine and fatty acid biosynthesis. From there the authors picked four volatile compounds, p-cresol, decanoic acid, N,N-dibutylformamide and 2-chloroethyl benzoate. In the group treated with B. licheniformis 2-chloroethyl benzoate came out 29.6 fold the control and the other three between 2.15 and 2.25 fold. With B. paralicheniformis decanoic acid rises to 4.08 fold.
The four molecules were tested on their own, and not on people. Female Kunming mice of six to eight weeks, a shaved square of back two centimetres by two, two hundred microlitres of solution in ethanol at one, ten and a hundred milligrams per millilitre. After five minutes the animal was exposed to mosquitoes and the time to the first landing or biting attempt was measured, with an observation limit of six hours. The control with solvent alone held for less than two minutes. Four replicates per concentration, and protection time grew with the dose. The standard asked of a repellent is four hours of protection, and decanoic acid, N,N-dibutylformamide and 2-chloroethyl benzoate reach it, p-cresol does not.


The limits are listed by the authors. The mosquito tested is one species, a laboratory-reared Aedes aegypti of the Xishuangbanna strain, and no check was run on Anopheles or Culex. No human cohort received the bacterial treatment. The two strains smeared on were bought from commercial collections rather than isolated from the volunteers' skin, and metabolic differences can run between a commercial strain and a native one. How long those bacteria last on human skin remains to be determined. The comparison with DEET, the reference repellent, is left to future work.
One point bears closely on the skin. The introduction presents Bacillus as probiotics able to act without causing skin health issues, and the discussion reports that after application Streptococcus and other taxa potentially associated with skin infection and scarlet fever rose as well. A dermatological assessment of the treated animals does not appear in the text. Bacillus licheniformis cut back Staphylococcus epidermidis, a steady resident of healthy skin that other work links to greater attractiveness to mosquitoes.
The protocol was approved by the ethics committee of Guizhou Medical University and the volunteers signed informed consent. The authors declare no relevant funding and no competing interests, and no patent declaration appears in the text. In the acknowledgements they state that they used artificial intelligence tools to refine the language of the manuscript.


On the skin of a person the Bacillus has not been smeared yet.

 

References
Gu Zhenyu, Gao Heting, Guo Hongxia, Fan Zhihua, Zhang Weixian, Xing Dan, Zhou Xinyu, Yu Haotian, Zhao Teng, Wu Jiahong, Li Chunxiao, «Mosquito repellency mediated by skin commensal Bacillus species via volatile metabolite reprogramming», Nature Communications, volume 17, article 10869, pages 1–13, received 26 January 2026, accepted 22 September 2026, online 19 October 2026, open access under CC BY-NC-ND 4.0. doi: 10.1038/s41467-026-78356-5
De Obaldia M.E. et al., «Differential mosquito attraction to humans is associated with skin-derived carboxylic acid levels», Cell, 185, 4099–4116, 2022, cited in the paper as the basis for the role of skin volatiles in host choice.
Liu F. et al., «Engineered skin microbiome reduces mosquito attraction to mice», PNAS Nexus, 3, 267, 2024, cited in the paper among earlier work on skin microbiota and mosquitoes in mice.
Kim D. et al., «Quorum sensing gene regulation in Staphylococcus epidermidis reduces the attraction of Aedes aegypti», Frontiers in Microbiology, 14, 1208241, 2023, cited in the paper as the source for the effect of Staphylococcus epidermidis on attractiveness.



Vota questo articolo
(0 Voti)

Lascia un commento

Assicurati di aver digitato tutte le informazioni richieste, evidenziate da un asterisco (*). Non è consentito codice HTML.

 

Scienzaonline con sottotitolo Sciencenew  - Periodico
Autorizzazioni del Tribunale di Roma – diffusioni:
telematica quotidiana 229/2006 del 08/06/2006
mensile per mezzo stampa 293/2003 del 07/07/2003
Scienceonline, Autorizzazione del Tribunale di Roma 228/2006 del 29/05/06
Pubblicato a Roma – Via A. De Viti de Marco, 50 – Direttore Responsabile Guido Donati

Photo Gallery