18
December 2023 ESRFnews
BIO-IMAGING
samples tagged with heavy metals with a laser, scan
them with high-energy X-rays, at about 70 keV. As
these do not interact with the light atoms that make
up biological tissue, there is no ablation and no tissue
destruction – the X-rays only cause the heavy isotopes
(mainly lanthanides) to fluoresce. And because of the
brilliance of ESRF X-rays, the MEZ-XRF scans can be
very quick – about 1000 Hz (pixels per second), versus
200 Hz for IMC. “The EBS has given us a gain in photon
flux at high energy by almost two orders of magnitude,”
says di Michiel.
The MEZ-XRF scans of human breast tumours,
tonsils and appendixes already gave data comparable
to that given by IMC. In the breast cancer images,
for example, Bodenmiller, Strotton and colleagues
could identify a host of different cell types at sub-
micron resolution in three different types of tumour,
and thereby determine which treatments would be
suitable – Herceptin if the “HER2” tags are visible, say,
or immunotherapy if the immune-cell tags are present
around an inflamed tumour (see fig. 1, above).
The benefit over IMC is not just speed and non-
destructiveness however MEZXRF could build on
the penetrating properties of Xrays to image whole
samples in 3D This could be faster than a similar IMC
demonstration two years ago by Bodenmillers group in
which the researchers reconstructed 3D imagery out of
serially sectioned 2D IMC images a process that took
one week for a human breastcancer sample Nat Cancer
3 122 According to Strotton 3D MEZXRF could
simplify and speed up the process to become the first 3D
multiplex method Thats where things really start to
get exciting he says
Jon Cartwright
regions for more sensitive scans that reveal finer cell
features, or low expression but important molecular
markers.”
Most bio-imaging methods have to make compromises
somewhere. For instance, electron microscopy can
image at sub-nanometre resolution, but at low speed
and throughput and with limited ability to “multiplex”
– that is, see many marked features. On the other hand,
such multiplexing is possible with fluorescent imaging,
where certain molecules in a sample are tagged with
molecular “fluorophores”, which fluoresce when
illuminated with light of the correct wavelength. In
optical fluorescent imaging, specific wavelengths of
visible light usually make the fluorophores shine, but
the narrowness of this visible spectrum typically means
that only four or five differently coloured tags can be
employed and still distinguished.
In recent years, a group led by Bernd Bodenmiller at
the University of Zurich and ETH Zurich developed
an alternative method – one that allowed imaging for
40 or more tags, at micron (cellular) resolution. Known
as imaging mass cytometry (IMC), it involves tagging
specific molecules with antibodies bound with heavy
metal isotopes A laser scans a 2D section of a sample
ablating one pixel at a time raster fashion meanwhile
a plasma torch directs the resultant plumes of isotopes to
a mass spectrometer where they can be identified The
commercialised method has become widely adopted
by core facilities particularly for clinical research
where it avoids the troublesome habit of certain samples
autofluorescing themselves but it is generally limited
to 2D and it is destructive precluding any subsequent
analyses on the same tissue
Then two years ago to avoid these IMC issues
Bodenmillers group had an idea instead of ablating
The EBS
has given
us a gain in
photon
flux at high
energy by
almost two
orders of
magnitude
M. S T R O T T O N
Figure 1. MEZ-XRF scans of three different types of human breast cancer: human epidermal growth factor receptor 2 positive (HER2+), luminal B
(LumB), and luminal B HER2 positive (LumB HER2+). In each, cell types can be identified based on the expression of multiple heavy-metal markers
mapped by X-ray fluorescence at ID15A. The overall composition of cell types determines which cancer treatment will be most beneficial.
stromal cells T-cells exhausted CD4 T-cells
HER2
+
PR
+
epithelia HER2
+
epithelia CK7
+
epithelia
PR
+
epithelia epithelial cells PD1
+
helper T-cells
ER
+
epithelia