The observer effect. At what point does the act of measuring a specimen make the measurements invalid?
X-ray fluorescence microscopy is an interesting and powerful technique. Trace elements can be imaged in astonishing detail from intact specimens. But every measured fluorescent X-ray photon had its origins in an electron forcibly removed from its resting place. Technically the very first photon we measure is a result of specimen damage, but how X-rays does it take before we observe the damage?
“Our dehydrated samples received a combined radiation dose of ~106 Gy with no apparent morphological changes in Compton scattering …” (Hare et al., Metallomics (8) 2016).
But what do radiation-induced morphological changes look like? And how can we be sure? We wanted to find out.
The experiment is simple. Keep on measuring a sample (a – left) until it breaks (b – left). At some point we see an obvious change in the bulk specimen and it is broken (c – left). Totally broken. Before the total annihilation of the specimen there are more subtle changes.
Comparing three different chemical-free specimen preparations, we find that cryofixation (b – right) provides the most protection. Although the sample is damaged, nothing can move and the specimen appears unchanged up to a radiation dose approaching 108 Gy. Upon thawing however the extent of the damage is revealed with the specimen being destroyed (below). Lyophilized (freeze dried) specimens (a – right) showed the next level of radiation protection. However, we observe that different elements display damage before others. Potassium begins to show signs of damage at 107 Gy, with other elements following the trend with the ultrastructure destruction (c – above) at 3 x 107 Gy.
In contrast, anesthetized specimens (c – above) show a different profile , with iron exhibiting signs of damage as early as 106 Gy before the bulk sample is destroyed (left). This result confirms previous assertions we’ve made, it’s nice when results agree.
What does all this mean? Firstly, we were correct in our previous assessment, and our quote above was correct (phew), we can safely irradiate a lyophilized c. elegans with ~106 Gy with no changes in morphological or elemental distribution at ~1um length scales. Secondly, we’ve set limits for X-ray fluorescence microscopy of c. elegans at ~1um length scales for a variety of preparations. Experiments can now be planned so they don’t exceed certain limits depending of what elements are of interest. This is a huge bonus, no longer are we flying dark and crossing our fingers. No longer do we have to mage vague statements about not seeing morphological changes. We can have confidence that what we observe is what we believe we observe. The observer effect is not in play within these limits.
Head over to Analytical Chemistry to see the article with all the gory details, and don’t forget to check out the supplementary material for extra information and some detail rich movies where you can see what the changes look like.