English subtitles

← To detect diseases earlier, let's speak bacteria's secret language

Get Embed Code
28 Languages

Showing Revision 12 created 03/27/2019 by Oliver Friedman.

  1. You don't know them.
  2. You don't see them.
  3. But they're always around,
  4. whispering,
  5. making secret plans,
  6. building armies with millions of soldiers.
  7. And when they decide to attack,
  8. they all attack at the same time.
  9. I'm talking about bacteria.
  10. (Laughter)

  11. Who did you think I was talking about?

  12. Bacteria live in communities
    just like humans.

  13. They have families,
  14. they talk,
  15. and they plan their activities.
  16. And just like humans, they trick, deceive,
  17. and some might even cheat on each other.
  18. What if I tell you that we can listen
    to bacterial conversations
  19. and translate their confidential
    information into human language?
  20. And what if I tell you that translating
    bacterial conversations can save lives?
  21. I hold a PhD in nanophysics,
  22. and I've used nanotechnology
    to develop a real-time translation tool
  23. that can spy on bacterial communities
  24. and give us recordings
    of what bacteria are up to.
  25. Bacteria live everywhere.

  26. They're in the soil, on our furniture
  27. and inside our bodies.
  28. In fact, 90 percent of all the live cells
    in this theater are bacterial.
  29. Some bacteria are good for us;
  30. they help us digest food
    or produce antibiotics.
  31. And some bacteria are bad for us;
  32. they cause diseases and death.
  33. To coordinate all
    the functions bacteria have,
  34. they have to be able to organize,
  35. and they do that just like us humans --
  36. by communicating.
  37. But instead of using words,
  38. they use signaling molecules
    to communicate with each other.
  39. When bacteria are few,
  40. the signaling molecules just flow away,
  41. like the screams of a man
    alone in the desert.
  42. But when there are many bacteria,
    the signaling molecules accumulate,
  43. and the bacteria start sensing
    that they're not alone.
  44. They listen to each other.
  45. In this way, they keep track
    of how many they are
  46. and when they're many enough
    to initiate a new action.
  47. And when the signaling molecules
    have reached a certain threshold,
  48. all the bacteria sense at once
    that they need to act
  49. with the same action.
  50. So bacterial conversation consists
    of an initiative and a reaction,

  51. a production of a molecule
    and the response to it.
  52. In my research, I focused on spying
    on bacterial communities
  53. inside the human body.
  54. How does it work?
  55. We have a sample from a patient.
  56. It could be a blood or spit sample.
  57. We shoot electrons into the sample,
  58. the electrons will interact with any
    communication molecules present,
  59. and this interaction
    will give us information
  60. on the identity of the bacteria,
  61. the type of communication
  62. and how much the bacteria are talking.
  63. But what is it like
    when bacteria communicate?

  64. Before I developed the translation tool,
  65. my first assumption was that bacteria
    would have a primitive language,
  66. like infants that haven't developed
    words and sentences yet.
  67. When they laugh, they're happy;
    when they cry, they're sad.
  68. Simple as that.
  69. But bacteria turned out to be nowhere
    as primitive as I thought they would be.
  70. A molecule is not just a molecule.
  71. It can mean different things
    depending on the context,
  72. just like the crying of babies
    can mean different things:
  73. sometimes the baby is hungry,
  74. sometimes it's wet,
  75. sometimes it's hurt or afraid.
  76. Parents know how to decode those cries.
  77. And to be a real translation tool,
  78. it had to be able to decode
    the signaling molecules
  79. and translate them
    depending on the context.
  80. And who knows?
  81. Maybe Google Translate
    will adopt this soon.
  82. (Laughter)

  83. Let me give you an example.

  84. I've brought some bacterial data
    that can be a bit tricky to understand
  85. if you're not trained,
  86. but try to take a look.
  87. (Laughter)

  88. Here's a happy bacterial family
    that has infected a patient.

  89. Let's call them the Montague family.
  90. They share resources,
    they reproduce, and they grow.
  91. One day, they get a new neighbor,
  92. bacterial family Capulet.
  93. (Laughter)

  94. Everything is fine,
    as long as they're working together.

  95. But then something unplanned happens.
  96. Romeo from Montague has a relationship
    with Juliet from Capulet.
  97. (Laughter)

  98. And yes, they share genetic material.

  99. (Laughter)

  100. Now, this gene transfer
    can be dangerous to the Montagues

  101. that have the ambition to be the only
    family in the patient they have infected,
  102. and sharing genes contributes
  103. to the Capulets developing
    resistance to antibiotics.
  104. So the Montagues start talking internally
    to get rid of this other family
  105. by releasing this molecule.
  106. (Laughter)

  107. And with subtitles:

  108. [Let us coordinate an attack.]

  109. (Laughter)

  110. Let's coordinate an attack.

  111. And then everybody at once responds
  112. by releasing a poison
    that will kill the other family.
  113. [Eliminate!]

  114. (Laughter)

  115. The Capulets respond
    by calling for a counterattack.

  116. [Counterattack!]

  117. And they have a battle.

  118. This is a video of real bacteria
    dueling with swordlike organelles,

  119. where they try to kill each other
  120. by literally stabbing
    and rupturing each other.
  121. Whoever's family wins this battle
    becomes the dominant bacteria.
  122. So what I can do is to detect
    bacterial conversations

  123. that lead to different
    collective behaviors
  124. like the fight you just saw.
  125. And what I did was to spy
    on bacterial communities
  126. inside the human body
  127. in patients at a hospital.
  128. I followed 62 patients in an experiment,
  129. where I tested the patient samples
    for one particular infection,
  130. without knowing the results
    of the traditional diagnostic test.
  131. Now, in bacterial diagnostics,

  132. a sample is smeared out on a plate,
  133. and if the bacteria grow within five days,
  134. the patient is diagnosed as infected.
  135. When I finished the study
    and I compared the tool results
  136. to the traditional diagnostic test
    and the validation test,
  137. I was shocked.
  138. It was far more astonishing
    than I had ever anticipated.
  139. But before I tell you
    what the tool revealed,

  140. I would like to tell you about
    a specific patient I followed,
  141. a young girl.
  142. She had cystic fibrosis,
  143. a genetic disease that made her lungs
    susceptible to bacterial infections.
  144. This girl wasn't a part
    of the clinical trial.
  145. I followed her because I knew
    from her medical record
  146. that she had never had
    an infection before.
  147. Once a month, this girl
    went to the hospital
  148. to cough up a sputum sample
    that she spit in a cup.
  149. This sample was transferred
    for bacterial analysis
  150. at the central laboratory
  151. so the doctors could act quickly
    if they discovered an infection.
  152. And it allowed me to test my device
    on her samples as well.
  153. The first two months I measured
    on her samples, there was nothing.

  154. But the third month,
  155. I discovered some bacterial
    chatter in her sample.
  156. The bacteria were coordinating
    to damage her lung tissue.
  157. But the traditional diagnostics
    showed no bacteria at all.
  158. I measured again the next month,
  159. and I could see that the bacterial
    conversations became even more aggressive.
  160. Still, the traditional
    diagnostics showed nothing.
  161. My study ended, but a half a year later,
    I followed up on her status
  162. to see if the bacteria
    only I knew about had disappeared
  163. without medical intervention.
  164. They hadn't.
  165. But the girl was now diagnosed
    with a severe infection
  166. of deadly bacteria.
  167. It was the very same bacteria
    my tool discovered earlier.
  168. And despite aggressive
    antibiotic treatment,
  169. it was impossible
    to eradicate the infection.
  170. Doctors deemed that she would not
    survive her 20s.
  171. When I measured on this girl's samples,

  172. my tool was still in the initial stage.
  173. I didn't even know
    if my method worked at all,
  174. therefore I had an agreement
    with the doctors
  175. not to tell them what my tool revealed
  176. in order not to compromise
    their treatment.
  177. So when I saw these results
    that weren't even validated,
  178. I didn't dare to tell
  179. because treating a patient
    without an actual infection
  180. also has negative
    consequences for the patient.
  181. But now we know better,
  182. and there are many young boys
    and girls that still can be saved
  183. because, unfortunately,
    this scenario happens very often.
  184. Patients get infected,
  185. the bacteria somehow don't show
    on the traditional diagnostic test,
  186. and suddenly, the infection breaks out
    in the patient with severe symptoms.
  187. And at that point, it's already too late.
  188. The surprising result
    of the 62 patients I followed

  189. was that my device
    caught bacterial conversations
  190. in more than half of the patient samples
  191. that were diagnosed as negative
    by traditional methods.
  192. In other words, more than half
    of these patients went home thinking
  193. they were free from infection,
  194. although they actually carried
    dangerous bacteria.
  195. Inside these wrongly diagnosed patients,
  196. bacteria were coordinating
    a synchronized attack.
  197. They were whispering to each other.
  198. What I call "whispering bacteria"
  199. are bacteria that traditional
    methods cannot diagnose.
  200. So far, it's only the translation tool
    that can catch those whispers.
  201. I believe that the time frame
    in which bacteria are still whispering
  202. is a window of opportunity
    for targeted treatment.
  203. If the girl had been treated
    during this window of opportunity,
  204. it might have been possible
    to kill the bacteria
  205. in their initial stage,
  206. before the infection got out of hand.
  207. What I experienced with this young girl
    made me decide to do everything I can

  208. to push this technology into the hospital.
  209. Together with doctors,
  210. I'm already working
    on implementing this tool in clinics
  211. to diagnose early infections.
  212. Although it's still not known
    how doctors should treat patients

  213. during the whispering phase,
  214. this tool can help doctors
    keep a closer eye on patients in risk.
  215. It could help them confirm
    if a treatment had worked or not,
  216. and it could help answer simple questions:
  217. Is the patient infected?
  218. And what are the bacteria up to?
  219. Bacteria talk,

  220. they make secret plans,
  221. and they send confidential
    information to each other.
  222. But not only can we catch them whispering,
  223. we can all learn their secret language
  224. and become ourselves bacterial whisperers.
  225. And, as bacteria would say,
  226. "3-oxo-C12-aniline."
  227. (Laughter)

  228. (Applause)

  229. Thank you.