Treatment with the copper compound CuATSM has no significant effect on motor neuronal pathology in patients with ALS

Abstract Aims Although the orally available brain‐penetrant copper compound CuATSM has demonstrated promising effects in SOD1‐linked mouse models, the impact of CuATSM on disease pathology in patients with amyotrophic lateral sclerosis (ALS) remains unknown. Methods The present study set out to address this deficit by performing the first pilot comparative analysis of ALS pathology in patients that had been administered CuATSM and riluzole [N = 6 cases composed of ALS‐TDP (n = 5) and ALS‐SOD1 (n = 1)] versus riluzole only [N = 6 cases composed of ALS‐TDP (n = 4) and ALS‐SOD1 (n = 2)]. Results Our results revealed no significant difference in neuron density or TDP‐43 burden in the motor cortex and spinal cord of patients that had received CuATSM compared with patients that had not. In patients that had received CuATSM, p62‐immunoreactive astrocytes were observed in the motor cortex and reduced Iba1 density was found in the spinal cord. However, no significant difference in measures of astrocytic activity and SOD1 immunoreactivity was found with CuATSM treatment. Discussion These findings, in this first postmortem investigation of patients with ALS in CuATSM trials, demonstrate that in contrast to that seen in preclinical models of disease, CuATSM does not significantly alleviate neuronal pathology or astrogliosis in patients with ALS.

However, SOD1-ALS accounts for only $2% of all patients with ALS [9,10], and the translational relevance of CuATSM in treating the $90% of patients with sporadic disease is less clear. Importantly, whether CuATSM alleviates neuronal pathology in patients with sporadic ALS is not known. The present study set out to address this question in patients that had been administered CuATSM. Given that all patients were also prescribed the glutamatergic modulator riluzole, which is the only approved ALS therapy currently available in Australia, patients with only riluzole treatment were also included in this analysis.

Case selection
Human brain tissue was obtained from the New South Wales Brain Bank, which holds a neuropathologic series collected with informed consent through regional brain donor programs. All cases with pathological confirmation of clinical ALS that had been administered CuATSM (NCT04082832) were selected for this study (n = 6) ( Table 1). Given that these patients had also been administered riluzole [11], an age-matched ALS cohort that had been on a similar daily dosage of riluzole was selected (n = 6). All cases had previously been staged for topographical progression of TDP-43 [12,13] and assessed for genetic mutations in the C9ORF72, TARDBP and SOD1 genes. A SOD1 mutation was identified in three cases (p.I114T in CuATSMriluzole, p.I114T and p.V149G in riluzole [14] counterstained with 0.5% cresyl violet as previously described [15]. Immunofluorescence staining was carried out on 10-μm-thick sections. Following microwave antigen retrieval (0.01 M citrate buffer, pH 6.0), formaldehyde quenching was carried out using 0.1% sodium borohydride, followed by protein blocking. The sections were then

Quantitation of pathologies
Sections were quantified as previously described [15]. In the motor cortex, 2Â 500-μm-wide strips through the entire cortical thickness from the pial surface to white matter were sampled in each cortical section, and neurons with and without pTDP-43 and p62 inclusions were counted at 200Â magnification using a 10 Â 10 eyepiece graticule (500 μm Â 500 μm) with standard inclusion (lower and left) and exclusion (upper and right) borders in contiguous, non-overlapping fields. In the spinal cord, both anterior horns were identified, and neurons with and without pTDP-43 and p62 inclusions were also counted at 200Â magnification using the same 10 Â 10 eyepiece graticule.
The density of neurons within each region of interest was calculated for each case, and the proportion with pTDP-43 or p62 inclusions was expressed as a percentage of these. The density of Iba1-positive microglia was also quantified in this same manner. The areal fraction occupied by GFAP immunopositive astroglia in each region of interest was assessed using a point-counting method on 200Â magnification as previously described [16]. Given that a proportion of astrocytes had obvious p62 immunoreactivity (described in the results), the areal fraction occupied by p62 in each region of interest was also assessed.
The proportion of glial cells with p62 or pTDP-43 was graded on a four-point scale: 0 = no detectable pathology across the entire section; 1 = mild (some pathology observed in most fields of view at 100Â magnification); 2 = moderate; 3 = severe as previously described [17]. Consistent with a recent report [18], diffuse SOD1 immunoreactivity was observed in all ALS cases, and the intensity of these was graded in neurons and glia on a four-point severity scale of 0-3. Quantitation was performed by two raters blind to case details and treatment group with an inter-and intra-rater variance of <5%.

Statistics
Statistical analysis was performed using SPSS (Version 25) with a pvalue of <0.05 taken as significant. Demographic differences among groups were determined using one-way ANOVA for age and postmortem delay, and c 2 -test for gender and presence of SOD1 mutation.
Group differences were assessed using multivariate analysis. Correlation analyses were performed with Spearman rank correlation analyses. Consistent with previous reports, SOD1 cases did not demonstrate pTDP-43 immunoreactivity [19] and, as such, were excluded from analyses of pTDP-43.

RESULTS
P62-positive inclusions were seen in all cases, whereas pathological pTDP-43 aggregates were only observed in ALS cases without a SOD1 mutation. No pathological brain changes to suggest prolonged hypoxia were observed in the present series. Co-existing neurodegenerative pathologies were absent.

Group demographics and daily drug dosage
There were no significant differences in age at death between p62-positive astrocytes in the CuATSM treatment group p62-immunoreactive astrocytes ( Figure 1A) were observed in the motor cortex of all ALS cases that had received CuATSM but not in ALS cases that had not ( Figure 1D). These astrocytes had typical astrocyte morphology ( Figure 1B) and were not immunoreactive for TDP-43 ( Figure 1C). Immunofluorescent triple labelling for p62, GFAP and SOD1 ( Figure 1G-K) confirmed these to be p62-positive astrocytes.   bioavailability has also been found in patients with sporadic ALS, the therapeutic benefits of CuATSM seen in SOD1 models have been proposed to be translatable to patients with the sporadic disease [21].

Group differences
The present study performs the first postmortem investigation into patients that had been treated with CuATSM during life. Our results revealed no significant difference in neuron density or TDP burden in patients that had received CuATSM compared with patients that had not. In contrast to that seen in preclinical models [3-6, 8, 22, 23], P62 is an autophagy substrate that plays a critical role in aggregate degeneration [24] and was observed here in motor cortical astrocytes of CuATSM-treated patients only, where it filled the astrocyte cytoplasm in a similar way to GFAP (Figure 1). Although p62-immunopositive astrocytes have not been described in postmortem tissue of patients with ALS, elevated p62 levels have been identified in reprogrammed skin-derived astrocytes from patients with sporadic, SOD1 and C9ORF72 ALS [23,25]. Importantly, no relationship between astroglial sequestosome activity with CuATSM treatment was found in patient cell lines [23], suggesting that the p62-astrocytes observed here do not reflect a therapeutic response to CuATSM. Further to this, in contrast to the reduced astrocytic activity seen in SOD1 rodent models [3][4][5][6][7][8], no significant change in measures of GFAP was found with CuATSM here. Instead, lower levels of activated microglia were observed in the spinal cord of patients with CuATSM treatment. Interestingly, a negative correlation between disease duration and activated microglia has been reported in the spinal cord of ALS cases [26], but this was not observed in the present cohort. As is the case for most quantitative pathological studies, the main methodological issue warranting consideration is the relatively small sample sizes, and future replication of these results in a larger sample is needed. Nevertheless, the assessment of tissue from six patients that had been on the CuATSM trial is significant, and the striking and consistent findings provide strong support to suggest our results are representative.
In summary, this first postmortem analysis of patients treated with CuATSM demonstrates p62-immunoreactive astrocytes in the motor cortex and lower Iba1-positive microglia in the spinal cord.
Importantly, however, no significant difference in neuronal density and the pathological burden was found with CuATSM treatment, T A B L E 2 Mean (±SD) neuronal and glia pathology in patients treated with CuATSM and riluzole or riluzole only.

CONFLICT OF INTEREST STATEMENT
The authors declare that they have no competing interests.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.