Intersectional genetics approach to overexpress TDP-43 in a single spinal motor neuron pool of two rhesus macaques.

(A) The FLEx system exploits Cre recombinase to trigger the inversion and excision of the gene of interest from between two loxP sites, thus ‘flipping’ the gene into an active configuration. (B) Inactive human wild-type TDP-43 DNA packaged in an AAV9 vector was injected intrathecally via lumbar puncture. The animals were placed in the Trendelenburg position for 10 minutes after this injection to facilitate rostral spread of the virus to the cervical segments [31]. In the same procedure, a second virus, rAAV2 retro, was injected into the right brachioradialis muscle. This vector encoded Cre recombinase. The TDP-43 gene will be switched into the active, readable state in any cells expressing both viruses. This system ensured that transduction was limited to the motor neurons innervating the right brachioradialis muscle.

Muscle signal intensity from T2-weighted Rapid Acquisition with Relaxation Enhancement (RARE) scans.

All panels show cross-sectional images of the left and right arms from two monkeys. For orientation, on a single scan for each animal the brachioradialis muscle is demarcated with a red boundary, the ulnar (U) and radius (R) bones are labelled, and a superficial blood vessel indicated with the filled arrowhead. In some scans, artifacts caused by the implanted EMG wires can be seen in both a flexor and extensor muscle of monkey Ma, indicated with an open arrowhead. (A, B) Baseline scans prior to virus injection. (C, D) 4-6 weeks after virus injection, no signal abnormalities are seen. (E, F) 8-14 weeks after virus injection, bands of hyperintense signal are seen localized to the brachioradialis muscle. (G, H) Signal abnormalities normalize after 8 weeks in both animals

Quantification of RARE-MRI signal intensity changes in the brachioradialis muscle.

(A-D) Change in signal intensity with time after the virus injection (time 0, vertical dashed line), for left and right sides in two monkeys. Intensity has been normalized by subtracting the mean and dividing by the standard deviation of a baseline period (grey shading). Each colored line presents data from a different proximo-distal slice through the forearm. Horizontal dotted lines mark the significance limits of P<0.05, 0.01, 0.005 and 0.001 after correction for multiple comparisons by the Benjamini-Hochberg procedure. Note the peaks in signal seen on the right side in both animals. (E, F) The same data as (A-D), replotted as color maps to show the spatiotemporal pattern. The color scale has been chosen such that black indicates non-significant changes, while red and blue represent significant increases and decreases, respectively. Significance was determined at a threshold of P<0.0017 after correcting for multiple comparisons. Both monkeys showed a focal, time-limited increase in signal in the right BR; in monkey Mi (C) this was followed by a signal decrease. A signal decrease was also recorded during this time in the left BR of monkey Mi. No signal decreases were recorded in the left or right BR of monkey Ma (D).

Evidence for motor neuron degeneration in spinal segments innervating the brachioradialis muscle.

(A) Image from a previous study in our group (Monkey W) involving retrograde viral transfection with rAAV2retro with a tdTomato genetic payload. Motor neurons labelled with the fluorescent protein could be clearly seen. (B) Example image from monkey Mi, showing that no motor neurons with tdTomato could be seen. (C) Image of ventral horn of the spinal cord from monkey Mi, stained for choline acetyl transferase (ChAT) to identify motor neurons. Scale bar for (A-C) = 200µm. (D) Motor neuron counts per section in a healthy control macaque (Monkey Z). There were no significant differences in cell numbers between left and right sides in any segment between C1 and T1. (E) Motor neuron counts for the two monkeys infected with TDP-43 constructs as in Figure 1. Asterisks mark significant differences between left and right side (*, P<0.05; **, P<0.01; ***, P<0.0001). Bars are the mean of 6-18 sections; error bars indicate standard errors of the mean.

Spinal distribution of TDP-43 and phosphorylated TDP-43 (pTDP-43).

(A-D) Enlarged view of spinal motor neurons within lamina IX stained for TDP-43 in brown, using DAB. (A), from a control animal without virus transfection (monkey Z); TDP-43 was restricted to the nucleus of motor neurons. (B) from the C7 segment of Monkey Mi. (C) from C6 segment of Monkey Ma. (D) from C5 segment of Monkey Mi. Note dense staining in the cytoplasm, with visible clumps of TDP-43, indicated by asterisks (*), and the less prominent nuclear staining in some cells (arrows) in segments which innervate the brachioradialis muscle. (E-H) Enlarged view of spinal motor neurons stained for pTDP-43(pS409/410). (E) no pTDP-43 staining was identified in a control macaque (monkey Z) without virus transfection. (F) from C7 segment of Monkey Ma. (G) C6 segment of Monkey Mi. (H) C5 of Monkey Ma. Nuclear staining was less prominent in motor neurons of segments innervating the brachioradialis muscle (arrows), than more distant segments. (I,J) Bar charts showing the counts of pTDP-43+ motor neurons in each spinal segment on the left and right side. There were no significant differences between sides. Asterisks mark segments with significantly different cell counts from the segments which had reduced ChAT+ motor neuron counts. Bars show the mean over 6-18 sections; error bars show the standard error of the mean.

TDP-43 and pTDP-43 in the sensorimotor cortex.

(A-F) Images show parasagittal sections of the cortex in the region surrounding the central sulcus (CS). Pre-central (primary motor cortex, M1) and post-central (primary somatosensory cortex, S1) cortices are labelled. Sections have been stained to label TDP-43 or pTDP-43 in brown, using DAB. (A) TDP-43 was found in the cytoplasm of the large Betz cells of layer V (indicated by arrows) in M1 of a healthy control macaque (Monkey W), without virus transfection (see enlarged view). (B, C) In the transfected animals, Monkey Mi and Monkey Ma, staining was more dense, with clear cytoplasmic staining of Betz cell cytoplasm and nuclei in the enlarged images. (D) No pTDP-43 expression was detected in M1 of the healthy control macaque (Monkey W). (E, F) Staining for pTDP-43 in the transfected animals identified dense staining in the cytoplasm of layer Vb cells (indicated by arrows). Nuclear staining was present but less prominent. (G, H) Counts of pTDP-43+ Betz cells in two subdivisions of M1, in two transfected animals. Examples of the division of New M1 and Old M1 are shown in (B). (L-O) the density of pTDP-43+ cell was calculated as cells per unit length of layer Vb. Plots show results for left and right M1, for each monkey separately. Scale bars: low power images: 1000µm; high power images: 100µm. Bars in (G-L) are the average over 15-22 sections sampled from the mediolateral extent of M1; error bars indicate standard error of the means. Asterisks indicate significant differences: *, P<0.05; **, P<0.005.