J. Rousseau et al. / Bioorganic & Medicinal Chemistry Letters 27 (2017) 708–712
711
89Zr-2-Trastuzumab were very efficient (>90%) as confirmed by
iTLC-SG. After purification by PD-10 column, the radiochemical
purities were >99% for all three radioimmunoconjugates and their
specific activities were in the range of 0.3 – 0.6 MBq/lg. Compared
to unmodified Trastuzumab, no significant change in immunoreac-
tivity of all three radioimmunoconjugates was observed
(96, 104 and 101% immunoreactivity for 89Zr-DFO-Trastuzumab,
89Zr-1-Trastuzumab, and 89Zr-2-Trastuzumab, respectively). This
indicates that chelator conjugation and 89Zr labeling did not reduce
the binding capability of Trastuzumab to the HER2 antigen.
To assess the stability of 89Zr complexes of 1 and 2, in vitro
stability assays were conducted for 89Zr-1-Trastuzumab,
89Zr-2-Trastuzumab and 89Zr-DFO-Trastuzumab. These three
radioimmunoconjugates were incubated in mouse plasma at
37 °C, and the extent of demetalation was monitored over time
and quantified by iTLC. As shown in Table 1, progressive demet-
alation was observed for all three radioimmunoconjugates in
mouse plasma. 89Zr-DFO-Trastuzumab displayed the best plasma
stability. While 77% of 89Zr-DFO-Trastuzumab remained intact
after one day, only 25% of 89Zr-1-Trastuzumab and 46%
89Zr-2-Trastuzumab were still intact. On Day 3, there were 3,
16 and 55% of 89Zr-1-Trastuzumab, 89Zr-2-Trastuzumab and
89Zr-DFO-Trastuzumab, respectively, remained intact.
In addition to in vitro stability assays, PET imaging was also
conducted to assess the in vivo behaviour of these three radioim-
munoconjugates. Mice were injected with 8.4 2.0 lg of radioim-
Fig. 3. Representative maximum intensity projection PET images of 89Zr-labeled
Trastuzumab conjugates obtained at 1 and 3 days post-injection. Spleen is indicated
by arrows.
munoconjugates and PET/CT images were acquired at one and
three days post-injection. As shown in Fig. 3, spleen was clearly
visualized for all three radioimmunoconjugates on both images
acquired on Day 1 and 3. Spleen is the organ that is typically
observed in antibody imaging, and was used here as a surrogate
target. Clear spleen visualization indicated that significant por-
tions of radioimmunoconjugates remained intact and reached this
target organ. Bone uptake (especially in the knees) was also
observed and increased over time, suggesting there was progres-
sive demetalation for all three radioimmunoconjugates. The bone
uptake of 89Zr-1-Trastuzumab and 89Zr-2-Trastuzumab was higher
than that of 89Zr-DFO-Trastuzumab as observed from both images
taken on Day 1 and 3. The observed inferior in vivo stability of
complexes of 89Zr-1 and 89Zr-2 was consistent with the observa-
tions from the in vitro plasma stability assays.
At Day 3, all mice were euthanized and tissues/organs of inter-
est were harvested and counted to obtain comprehensive biodistri-
bution data of all three radioimmunoconjugates. Compressive
tissue uptake values were provided in Supplementary Table 1. As
shown in Fig. 4, significantly higher spleen uptake (p < 0.001)
was observed using 89Zr-DFO-Trastuzumab (48.9 7.92% ID/g)
than 89Zr-1-Trastuzumab (24.6 4.08% ID/g) and 89Zr-2-Tras-
tuzumab (27.64 4.22% ID/g). In addition, bone uptake was also
significantly lower (p < 0.001) for 89Zr-DFO-Trastuzumab
(7.58 0.36% ID/g) than 89Zr-1-Trastuzumab (19.5 3.59% ID/g)
and 89Zr-2-Trastuzumab (18.3 2.91% ID/g). These biodistribution
data were consistent with observations from PET images, and con-
firmed that 89Zr-1-Trastuzumab and 89Zr-2-Trastuzumab had a
higher extent of in vivo demetalation leading to higher bone uptake
and lower spleen uptake.
Fig. 4. Biodistribution of 89Zr-labeled Trastuzumab conjugates at three days post-
injection. Data are expressed as mean SD (n = 6). For statistics analysis, * (p < 0.05)
and *** (p < 0.001) indicate the difference is significant when compared to the
uptake of 89Zr-DFO-Trastuzumab.
Both in vitro and in vivo studies of 89Zr-1-Trastuzumab and 89Zr-
2-Trastuzumab suggested that chelators 1 and 2 did not form a
stable complex with 89Zr. This could be due to for example (1) the
Zr complexes of 1 and 2 were not the expected eight-coordinate
complexes despite the presence of four hydroxamate groups per
chelator; or (2) the eight-coordinate Zr complexes of 1 and 2 were
not in an optimized coordination arrangement due to the steric con-
straints imposed by the 2,20-iminodiacetamide backbone and the
shorter arm (X = CH2 for 1 and X = CH2CH2 for 2, Fig. 1) of hydroxa-
mate moieties. A similar finding was reported by Boros et al.19 on
the comparison of stability between 89Zr complexes of two
hydroxamate chelators L3 and L4 (Fig. 1) that shared a same cyclam
backbone. L3 was an octadentate tetrahydroxamate chelator with
an (N-methyl-N-hydroxy)carboxamidomethyl group connected to
all four amino groups of the cyclam. For L4, it was a hexadentate tri-
hydroxamate chelator with substitution of a hydroxamate group
with a longer arm ((N-methyl-N-hydroxy)carboxamidoethylcar-
Table 1
In vitro plasma stability of 89Zr-DFO-Trastuzumab, 89Zr-1-Trastuzumab and
89Zr-2-Trastuzumab. Data are presented as the percentage of the total activity
corresponding to the intact radioimmunoconjugates.
89Zr-DFO-Trastuzumab 89Zr-1-Trastuzumab 89Zr-2-Trastuzumab
Day 0 >99
Day 1 77
Day 2 69
Day 3 55
>99
25
13
3
>99
46
18
16