Feeding and follow-up of TDP-43 (A315T)mice
The
TDP-43 (A315T) mice designed by Wegorzewska et al. [
15] contain the mutant human
TDP-43 gene, preceded by the Flag-tag, under control of the mouse Prnp, leading to the highest expression in spinal cord and brain, but also in other tissues. Originally, these mice were on a mixed C57BL/6 J and CBA background. We obtained these mice from the Jackson Laboratory, stock number 010700 (The Jackson Laboratory, Bar Harbor, Maine, USA). The mice were already backcrossed for 10 generations into a C57BL/6 J background, leading to a pure C57BL/6 J background. For genotyping, we used the primers and protocol as previously described [
20]. Mice were kept in a conventional facility and fed ad libitum with dry pellet food (Ssniff® R/M-H, Ssniff Spezialdiäten GmbH, Soest, Germany) for the non-transgenic (NTG) mice and the
TDP-43 (A315T) mice with normal food. The gel fed mice got DietGel®boost (ClearH20, Maine, USA) from 30 days on for male and 80 days on for female mice. This gel food contains all necessary nutrients, but is a soft, high calorie, easily digestible paste with hardly any fibers.
The mice on gel food were put in cages, without bedding to prevent that wood fibers would stick to their food. These cages have a grid on the bottom, which allows the stool and urine to pass, but does not cause difficulties walking. Because of the high grade of humidity of the gel food, the fur of the mice looked wet. In some cages mice with the earliest disease onset, lost (part) of their fur (and whiskers) because of grazing by the others. This phenomenon does not affect their well-being and was reversible within 2 weeks after they were put alone in a cage.
NTG C57BL/6 J mice of both genders were compared to TDP-43 (A315T) mice with normal food and gel food for motor performance, body appearance and survival once a week till onset, than twice a week. Onset of gait abnormalities was defined as the moment at which a clear swimming gait appeared, with the paws wide-based and the animal being unable to hold its lower body from the ground. End stage (ES) for the TDP-43 (A315T) mice with normal food was the moment at which the animal appeared immobile (even after being gently pushed) and lethargic, since we experienced that in this stage the death would occur within 24 h. For the gel fed mice, ES was the moment at which the mouse could not turn from a side anymore within 30 s. At this point the mice were euthanized using a lethal dose of 10% Nembutal and tissues were collected.
Histopathologic analyses
Scoring of the NMJ was done as previously described [
35]. Eight mice were analyzed in each group at 100–150 NMJs/mouse.
For the measurement of creatine kinase (CK) levels, blood was taken from the right atrium after euthanizing, but before perfusion of the mice (at ES for the TDP-43 (A315T) mice with normal or gel food). The sample was centrifuged for 10 minutes at 14000 rpm. The supernatant was pipetted into another eppendorf tube, which was again centrifuged for 10 minutes at 14000 rpm. This supernatant was stored at −80°C until all the samples were collected. CK levels were determined by the same technique as used for human samples at the laboratory of UZ Leuven, Belgium.
We performed immunostainings for total (mouse and human) TDP-43 and glial fibrillary acid protein (GFAP) on spinal cord of ES (normal or gel fed) TDP-43 (A315T) mice compared to NTG mice. To this end, mice were euthanized with 10% Nembutal and perfused with PBS and 4% PFA. Lumbar spinal cord was dissected and dehydrated overnight in 30% sucrose in PBS. The tissue was embedded in Tissue Tek (OCT Compound, 361603E, VWR International, Randor, Pennsylvania, USA) and frozen at −80°C. Slices of 20 μm were obtained by using a microtome (Slee cryostat, Mainz, Germany). For immunostainings, they were blocked with 10% normal donkey serum (Sigma Aldrich, St Louis, Missouri, USA) at room temperature, incubated overnight at 4°C with the C-terminal total (mouse and human) TDP-43 (polyclonal rabbit-anti-TDP43-Ab 12892-1-AP, Proteintech, 1/200), followed by incubation with NeuN-Ab (MAB377, Millipore, Billerica, Massachusetts, USA, 1/200 for 1 h at room temperature) or incubated for 2 hours at room temperature with GFAP-Ab (polyclonal rabbit anti-GFAP-Ab, DAKO, Glostrup, Denmark, 1/500) followed by 3 washes with PBS-T and incubation with the secondary Ab for 1 h at room temperature [1/500, Alexa Fluor 555 anti-rabbit (for TDP-43-Ab and GFAP-Ab) or Alexa Fluor 488 anti-mouse (for NeuN), Invitrogen Life Technologies, Carlsbad, CA, USA]. After 3 washes with PBS-T, slides were mounted with Vectashield with DAPI and analyzed using a Zeiss Imager M1 microscope (Zeiss, Oberkochen, Germany).
To analyze the intestinal tract by immunostainings, segments of intestine (ileum and colon) were collected, opened along the mesenteric border and pinned flat in a sylgard lined dissection dish. Using fine forceps the mucosal and submucosal layers were removed prior to fixation in 4% paraformaldehyde (30 min). After rinsing in PBS, circular muscle layers were peeled in case of the small intestine; while for the colon the longitudinal muscle was removed. Tissues were treated in permeabilizing (0.5% triton-x) and 4% goat/donkey serum, prior to a 24 h (4°C) incubation in primary antibodies: specific human TDP-43 antibody (monoclonal mouse anti-hTDP-43-Ab 60019-2-Ig, Proteintech, Chicago, USA, 1/200), total TDP-43 (polyclonal rabbit-anti-TDP43-Ab 12892-1-AP, Proteintech, 1/200), GFAP (Abcam, Cambridge, UK, 1/5000), NO-synthetase (Santa Cruz biotechnologies, Santa Cruz, USA, 1/400), ChAT (Chemicon International, 1/500) and HuCD (Invitrogen Life Technologies, 1/500). After rinsing, fluorescently labeled appropriate secondary antibodies were added for 2 h. Immunohistochemical staining was visualized under an epifluorescence microscope (BX 41 Olympus, Belgium) with specific filter cubes (EX/DM/EM in nm) for blue (325-375/400/435-485), green (460-495/505/510-550) and red fluorescent probes (570-590/595/600-660). Images were recorded using Cell^F software on an XM10 (Olympus) camera.
For counting of motor neurons, the lumbar spinal cord was cut in 20 μm thick slices, with each 6th slice placed on a slide (10 in total along the whole lumbar region). These were stained by Cresylviolet (Sigma). After taking photos of the ventral horns at 40x (10/mouse) with a Zeiss Imager M1 microscope (Zeiss), manual counting and analyses of the size of motor neurons was done with use of AxioVision (Zeiss). Six NTG mice, 5 ES TDP-43 (A315T) mice with normal food and 5 ES TDP-43 (A315T) mice with gel food were analyzed.
Staining and quantification of the dorsolateral corticospinal tract axons in the distal thoracic spinal cord (of 4 NTG, 4 ES
TDP-43 (A315T) mice on normal food and 4 ES mice on gel food) was done as described by Wegorzewska et al. [
15].
Western blot and qPCR
Spinal cord and brain from presymptomatic TDP-43 (A315T) or NTG mice (n = 3 for each genotype) were collected in Tripure isolation reagent (Roche Diagnostics, IN, USA) or tissue protein extraction reagents (Thermo Scientific, Rockford, IL, USA) with Complete (Complete EDTA-free, Roche Diagnostics) for qPCR and Western blot respectively.
RNA extraction was performed as described previously [
36]. qPCR assays were run in triplicate with n = 3 for each condition. The mRNA expression in spinal cord and brain of mouse TDP-43 (Mm.PT.51.5553804 Tardp exon 2–3, IDT, Coralville, Iowa, USA) was compared to the expression of several housekeeping genes, using taqman assays: beta-actin (VIC-MGB 4352341E-0905010, Applied Biosystems, Life Technologies, CA, USA), mouse hypoxanthine-guanine phosphoribosyltransferase (Mm01545399_m1 Hprt, Applied Biosystems) and mouse glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Mm.PT.39a.1 GAPDH exon 2–3, IDT, Leuven, Belgium). Expression of mouse Sort1 + Ex17b mRNA in the brain of
TDP-43 (A315T) and NTG mice was determined using Cyber green. The primers were: mouse Sort1 + Ex17b: 5′-AAATCCCAGGAGACAAATGC-3′ and 5′-GAGCTGGATTCTGGGACAAG-3′; mouse GAPDH: 5′- TGGCCTTCCGTGTTCCTAC -3′ and 5′- GAGTTGCTGTTGAAGTCGCA -3′.
For Western blot, 40 μg of spinal cord of
TDP-43 (A315T) or NTG mice (n = 3 for each condition) were blotted using a precast gel (NuPAGE®Bis-Tris gel IM-8042, Novex, Life technologies, Carlsbad, California, USA) and following the protocol described in Van Hoecke et al. [
37], allowing separation of the mouse TDP-43 (of 43 kDa) and the exogenous human TDP-43, with the Flag-tag running at a slightly higher molecular weight as described in Wegorzewska et al. [
15]. The membranes were probed with polyclonal rabbit-anti-TDP43-Ab (12892-1-AP, Proteintech, 1/500). The mouse TDP-43 levels were compared between
TDP-43 (A315T) mice and NTG mice. Alfa-tubulin (mouse anti-alfa-tubulin, T6199, Sigma, 1/5000) was used as a loading control. Also, the total TDP-43 expression between 90 days old
TDP-43 (A315T) males and females was compared by using the polyclonal rabbit-anti-TDP43-Ab (12892-1-AP, Proteintech, 1/500) and normalized against mouse GAPDH (mouse anti-GAPDH AM4300, Ambion, Life Technologies, 1/5000).