Model of mixed dementia (MD).

All studies involving vertebrate animals were approved by the University of Kentucky Institutional Animal Care and Use Committee (IACUC), under protocol 2012–0932. Isoflurane (1–2%) was used to maintain anesthesia during neuroimaging; 5% isoflurane was used during euthanasia by exsanguination.

The MD mice were generated by placing amyloid overexpressing mice on a transient HHcy-inducing diet. Male and female B6.Cg-Tg(APPswe,PSEN1dE9)85Dbo/Mmjax (MMRRC No: 34832-JAX) transgenic AD model mice (APPswe/PS1dE9) were used for this study, along with noncarrier wildtype (WT) littermates. All mice were maintained on a congenic C57BL/6J background. Carrier mice express a chimeric humanized amyloid precursor protein (APP) with the Swedish mutation, and a mutant human presenilin 1 (PS1) missing exon 9, resulting in progressive accumulation of A® peptide, plaque deposition, concomitant neuroinflammatory changes, and eventual behavioral impairment [25–27]. To induce HHcy, mice were placed on a custom diet (Envigo Teklad TD.97345) for 8 weeks beginning at about 34 weeks of age (mean = 33.8, standard deviation = 0.7). The diet lacks vitamins B₆, B₉ (folate), and B₁₂ and has excess methionine [20]. Control mice were given 8 weeks of a nutritionally matched custom control diet with normal methionine and B-vitamin levels (TD.01636). See S1 Table for details of both diets. All mice were subsequently returned to standard nutritionally complete rodent chow (TD.2918) for a 2-week recovery period, followed by 2-weeks of once daily intraperitoneal injections of either 5 mg/kg MW151 or an equivalent volume of saline vehicle. Throughout the experiments mice were housed 1–5 animals per cage (503.22 usable cm²), maintained in a room at 23°C ± 2°C with a 14/10 hour light/dark cycle beginning at 6:00 AM, and had ad libitum access to food and water. Environmental enrichment included a 5x5 cm cotton Nestlet (Ancare, Bellmore, NY, USA), paper shredding, a Backless Shack mouse house (Shepherd Specialty Papers), and ½ x ½ x 2” aspen chew sticks (Lomir Biomedical Inc).

Overall experimental design.

Two separate cohorts were generated for these studies. Cohort 1 was tested in a behavioral battery and cohort 2 underwent magnetic resonance imaging (MRI). Both cohorts consisted of the same three primary experimental groups: WT mice given control diet and administered IP saline injections (WT ctrl), APP/PS1 mice given HHcy diet and administered IP saline injections (MD saline), and APP/PS1 mice given HHcy diet and administered IP MW151 injections (MD MW151). An additional group was included in the behavioral cohort to validate the presence of persistent cognitive deficits despite reversal of HHcy: WT mice given HHcy diet and administered saline (WT HHcy) (see S1 Fig). Verification of vascular injury by Prussian blue (PB) staining in the MD model mice is contained in S2 Fig, consistent with what has previously been reported as a result of this diet [20, 23]. Mice in the behavioral cohort were tested during days 7–14 of dosing (see behavioral battery section below). Mice in the MRI group were imaged between day 12 and 14 of dosing. Injections were performed after the conclusion of the behavioral or neuroimaging assays for that day. All mice were sacrificed within 14–16 hours of receiving the final dose of MW151 at about 46 weeks of age. Tissue was harvested from the behavioral battery cohort for biochemical and histological analyses to avoid potential confounding effects of extended isoflurane administration in the neuroimaging group [28].

Behavioral battery design.

This experiment consisted of 14 WT ctrl mice (10F/4M), 13 MD saline mice (7F/6M), and 14 MD MW151 mice (7F/7M). There were 14 mice in the WT HHcy group included for verification (5F/9M) of persistent HHcy-induced cognitive deficits. Mice in all 4 groups were run and analyzed together statistically, but for purposes of clarity only the primary comparisons of interest (WT ctrl, MD saline, and MD MW151 groups) are shown in Figs 1 and 2. To facilitate acclimatization to the experimenter, mice were gently handled during weekly diet changes and weigh-ins using the cupping method [29], as well as for 4 non-consecutive days during the two-week diet recovery period. The battery was carried out over 7 days beginning on day 8 of IP injections. All testing was performed in designated behavioral testing rooms, with mice transported in their home cages and allowed to habituate for one hour each day prior to testing. On dosing day 8, mice underwent open field (OF) testing in the morning, followed by the frailty assay in the afternoon. On the morning of day 9, mice underwent the short-term object location test (OLT) in the same arena as open field. On days 11–13 mice underwent testing in the 6-arm radial arm water maze (RAWM). In the morning of day 14, mice were tested in the open pool (OP) followed by the marble burying task in the afternoon. In the evening, mice were separated and singly housed in a cage with a fresh nestlet for the overnight nesting assay. The following morning mice were humanely euthanized for tissue harvest and their nests were scored by blinded staff. Automated tracking was performed with Noldus EthoVision XT 10 (Noldus, Leesburg, Virginia, USA) for OF, OLT, and RAWM. All trials were recorded and monitored in real-time during all trials for all tests. In the event of tracking failure, the videos were manually re-scored by a blinded lab member.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. MD mice show altered non-cognitive behaviors versus WT control, unaffected by treatment with MW151.

(A) MD mice in both treatment groups lost significant weight on HHcy diet, which was recovered by time of sacrifice and unaffected by MW151 treatment. Results of two-way mixed measures ANOVA revealed a significant effect of time (F(1.693, 86.34) = 178.0, p < .0001), group (F(3, 51) = 13.02, p < .0001), and a time by group interaction (F(6, 102) = 24.02, p < .0001). ****p < .001, *p < .05 for the MD saline mice versus WT ctrl comparisons and ††††p < .001, ††p < .01 for MD MW151 versus WT ctrl, Sidak’s posthoc tests. (B) No significant differences between groups are observed in the nesting assay, despite overall significance in the Kruskal-Wallis test (H(4) = 9.833, p = 0.020). (C) No differences in the number of marbles buried are observed between any of the groups (F(3, 51) = 0.3992, p = 0.75). (D) There was a significant effect in the Frailty index scores (H(4) = 10.78, p = 0.0129), with the MD MW151 mice found to have a significantly higher score than WT ctrl, **p < .01, Dunn’s post hoc test. (E) Open field speed and time in center are shown, where the MD saline mice have a higher average speed than WT ctrl, and also spend less time in the center of the arena. MW151 does not alter either of these parameters. Results of one-way ANOVAs showed significant overall differences in average OF speed (F(3, 51) = 10.89, p < .0001) and OF time in center (F(3, 51) = 4.038, p = 0.0119), **p < .01, *p < .05 versus WT ctrl group, Holm-Sidak post-hoc comparisons.

https://doi.org/10.1371/journal.pone.0262474.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. MW151 treatment in MD mice has no effect on OLT performance but may reduce errors in the RAWM.

(A) Neither the MD saline nor the MD MW151 groups spent significantly more time on average exploring the object in the novel versus the familiar location, in contrast to the WT ctrl group which did show a significant preference for the object in the novel location: multiple paired t-tests with Holm-Sidak adjustment (T(13) = 3.522, p = 0.0149) *p < .05. (B) Learning curves of RAWM errors by block (3 consecutive trials) are shown for the WT ctrl and MD mice in the left panel. Errors are collapsed into average per day for statistical analysis in the right panel. Two-way mixed measures ANOVA showed a significant effect of day (F(1.955, 97.73) = 41.08, p < .0001) and group (F(3, 50) = 4.088, p = 0.0113). The MD saline mice (n = 12) committed significantly more errors than WT ctrl on both day 2 and 3 of testing, *p < .05 versus WT ctrl group, Tukey post-hoc comparisons. In contrast, MD MW151 mice did not commit statistically more errors than WT ctrl on any day. There was no significant difference between the MD MW151 and MD saline groups on any of the three days. (C) Results of one-way ANOVAs for average latency to platform in the final block (F(3, 50) = 1.076, p = .368) and average swim speed (F(3, 50) = 0.278, p = 0.841) indicated no statistical difference between any groups.

https://doi.org/10.1371/journal.pone.0262474.g002

Open field (OF).

The open field test was performed as described previously [30]. The testing apparatus consisted of four adjacent 40 cm x 40 cm arenas with overhead lighting set to approximately 20 lux. An individual mouse was placed into each arena to allow simultaneous testing of up to four mice at a time, with each open field trial lasting 15 minutes. Infrared cameras were used to enable body center-point tracking in the low light, and the measurement of total distance traveled, average speed of movement, and percent time in the center zone of the arena (20 x 20 cm area) was performed automatically.

Frailty assay.

Mice were tested according to [31]. Briefly, each mouse was assessed by a blinded observer across 27 aging-related parameters and graded based on presence and severity of each. A score of 0 indicated normal/absence of age-associated changes, 0.5 indicated modest degree of the trait, and 1 indicated severe degree. The final scores for each trait were then summed to generate a frailty index score, with a higher score indicating increased age-associated changes. Core body temperature was also measured with a rectal probe. See S2 Table for traits and grading criteria used.

5-minute object location test (OLT).

This test was performed in the same apparatus as the open field assay, but with two identical Lego® pyramids placed in the top left and top right corners of the box, 5 cm distant from the edges of the arena. Mice underwent a familiarization phase wherein they were given 10 minutes to freely explore both objects, followed by a 5-minute intertrial interval spent in a clean and empty holding cage. During this time, both objects were replaced with identical but clean objects, but the upper right corner object moved instead to the lower right corner. Mice were then returned to their assigned arena for the testing phase, wherein they were given 5 minutes to freely explore both objects. Detection of nose-point within 1 cm of the object counted as exploration of that object.

Six-arm radial arm water maze (RAWM).

RAWM was run according to the protocol previously described [19]. Briefly, the maze consisted of six arms in a circular tank filled with water containing non-toxic white tempera paint. The tank was enclosed by black curtains arranged in a square, with geometric extra-maze visual cues affixed to three of four curtain sides. A circular platform was placed at the end of one of the arms, approximately 1–2 cm below the water’s surface to provide an escape. Animals were split into two groups and tested in a back-to-back 3-day paradigm. Each day consisted of 15 trials (45 trials total), with the first 12 trials on day 1 alternating between hidden and visible (flagged) platform. The platform was hidden for all trials after the first 12. Trials were run in two blocks of 6 and one block of 3 each day. Mice were run in groups of 5–8 at a time, alternating by block to provide time to rest. The goal arm was fixed for each mouse but varied between animals, and each mouse started from a different arm between trials. The start arm sequence was randomized in reference to the goal arm, but with a pattern consistent for all mice. Errors were defined as entry from the pool center into any non-goal arm, or 15 seconds in any location without movement to a new zone. Entry to a new zone was defined in the software as movement of the center point of the mouse ≥ 5 cm past the dividing line between zones. To visualize learning curves, groups of 3 consecutive trials were averaged into blocks, for a total of 5 blocks per day. For statistical analysis by ANOVA, total errors were averaged by day.

Open pool (OP).

The day following RAWM, all mice underwent testing for swimming/motivational or visual impairments in the open pool [32]. The arms and visual cues were removed from the testing arena, and the platform raised to ~0.5 cm above the water surface. Mice received 10 trials each, with the platform raised and flagged for all trials. Each mouse was assigned a starting quadrant and the platform position alternated between the 3 remaining quadrants between trials. Time to find the platform from the starting quadrant was measured for each trial. Criterion for impairment was defined as an average latency to platform of > 20 seconds for the final block (last 3 trials). One female mouse from the MD saline group reached the impairment criterion (49.6 s) and was excluded from the RAWM and OP analyses.

Marble burying assay.

The marble burying test was performed with a slight modification of the protocol previously published [33]. Each mouse was tested in a new housing cage without food or water, filled to a height of 5 cm with fresh bedding material and 12 glass marbles (1.4 cm diameter) evenly spaced in a grid pattern. After exploring the cage for 30 minutes the mouse was carefully removed and a picture of the cage bottom taken from directly overhead. Two blinded scorers counted the number of buried marbles (defined as < 50% of the marble surface still visible) and their scores were averaged.

Nesting assay.

Nesting assay was performed as previously described [19, 34]. Briefly, mice were singly housed overnight in a new cage with no enrichment save for bedding and a new Nestlet. Two observers blinded to the experimental conditions scored the quality of the nests according to a semi-quantitative 5-point scale: (1)–nestlet not noticeably touched, >90% intact; (2)–nestlet partially torn up, 50–90% remaining intact; (3)–shredded >50% but without identifiable nesting site and bedding spread around cage; (4)–identifiable but flat nest, >90% of the nestlet torn up with the material gathered within one quarter of the cage floor area; (5)–a near perfect nest where >90% of the nestlet is torn up, the nest is crater shaped, and walls are higher than the mouse body for at least half of the circumference. Scores were assigned for each nest in half-point increments and averaged between the observers to generate an average nesting score.

Brain tissue harvest and plasma homocysteine measurement.

To avoid potential confounding effects of extended isoflurane administration during neuroimaging on endpoints of interest, tissue was harvested from the mice that underwent behavioral testing as previously reported [19, 21]. Briefly, mice were anesthetized with 5% isoflurane, blood was collected via cardiac puncture and mice were transcardially perfused over 5 minutes with 50 mL cold 1x PBS before brain removal and dissection. Whole blood was separated by centrifugation in EDTA tubes (Greiner Bio-One, #454428) at 2000xg for 20 minutes at room temperature. Plasma was collected, flash frozen in liquid nitrogen and stored at –80°C until processing for homocysteine measurement. Plasma samples from a randomly chosen subset of 5 mice from each group were diluted 1:5 in ARCHITECT Multi-Assay Manual Diluent and analyzed on an ARCHITECT i2000SR instrument at the University of Kentucky Clinical Laboratory. The brain was cut down the midline and the right hemisphere immediately post-fixed in 4% paraformaldehyde for 18-24h at 4°C and cryopreserved in 30% sucrose for a minimum of 48 h at 4°C. Fixed samples were cut into 30 μm sections and stored in cryoprotectant solution (50% 1X PBS, 25% ethylene glycol, 25% glycerol) at -20°C until staining. The hippocampus was removed from the left hemisphere, flash frozen in liquid nitrogen, and stored at -80°C until processing for cytokine measurement.

Immunofluorescent measurement of GFAP and Thioflavin S-positive amyloid plaques.

Six to eight free-floating sections spaced approximately 300 microns apart through the dorsal hippocampus were used for each animal. Sections were blocked with 10% goat serum (Lampire Biological Laboratories, #7332500) and 0.2% Triton X-100 in PBS. Rat anti-GFAP antibody (Invitrogen #13–0300) was diluted 1:3000 in PBS with 3% normal goat serum and 0.2% Triton X-100. Sections were incubated in primary antibody for 14–16 hours at 4°C before a 2-hour room temperature incubation with an Alexa633-conjugated anti-rat secondary antibody (Invitrogen #A-21094, 1:1000). Sections were mounted, partially dried, and treated for 5 minutes with a 1% Thioflavin S (Sigma #1892) solution in water, followed by two 5-minute washes with 80% ethanol. They then underwent a 5-min incubation with 1x TrueBlack (VWR, #10119–144) in 70% ethanol to reduce autofluorescence before drying and coverslipping in Vectashield mounting medium with DAPI (Vector Laboratories, #H-1200). Entire slides were imaged on a Zeiss Axio Scan Z1 digital slide scanner at 20x magnification. Hippocampus was outlined manually in HALO (Indica Labs) by a blinded investigator. Algorithm intensity settings were manually thresholded based upon negative control tissue, and the positive pixel algorithm (Area Quantification FL v1.2) was applied to the outlined regions for the GFAP and Thioflavin S channels. For 2 WT ctrl, 3 MD saline, and 4 MD MW151 animals the automated scanner was unable to achieve consistent autofocus across sections, and these slides were therefore not included in the analyses.

MesoScale discovery cytokine enzyme-linked immunosorbent assay (ELISA).

Hippocampus dissected from the left hemisphere was homogenized using an Omni Bead Ruptor 24 (Omni International) at a 1:20 weight to volume ratio in PBS lysis buffer: PBS with 1 mM phenylmethylsulfonyl fluoride, 0.5 mM EDTA and 0.2X Halt Protease Inhibitor Cocktail (Thermo Scientific, #87786). Homogenates were centrifuged at 12,000×g for 20 min at 4°C. Supernatants were collected for cytokine measurement using MesoScale Discovery custom mouse V-Plex ELISA kits, with IL-1β, IL-6, TNFα, and CXCL1 run multiplexed as part of the Proinflammatory Panel 1 Mouse Kit (K15048) and run according to manufacturer’s instructions on the QuickPlex SQ 120 reader. Only IL-1β is displayed in the results section as significantly altered in the disease model, but the other cytokines can be viewed in S3 Fig and S1 File. All samples were incubated for 16–18 hours in the plate at 4°C, shaking at 1000 rpm. Cytokine levels were normalized to the total milligrams (mg) of protein loaded in the sample as determined by BCA Protein Assay (ThermoFisher #23225). Cytokine concentrations were calculated based on standard curves using Discovery Workbench software (Meso Scale Discovery, v4.0). Final values are expressed as the concentration of cytokine measured in the PBS-soluble fraction per mg of total protein loaded per well. One WT ctrl sample was below the limit of detection and one MD saline sample was lost.

Cerebral blood flow (CBF) measurement by pseudo-continuous arterial spin labeling (pCASL).

Mice in the second experiment underwent magnetic resonance imaging at the University of Kentucky Magnetic Resonance Imaging and Spectroscopy Center, with scans performed on a 7 Tesla ClinScan system (Bruker BioSpin, Germany). There were 10 mice per group: 5F/5M WT, 5F/5M MD saline, and 6F/4M MD MW151 group. All mice were anesthetized with 5% isoflurane and transferred to the scanner bed with heating pad, where anesthetic depth was maintained in a ~1.2% isoflurane and air mixture using a nose cone. Imaging was performed after mice reached a stable temperature (35–36°C) and respiratory rate (70–100 breaths/min), which was monitored continuously. A whole-body volume coil was used for transmission and a mouse brain surface coil used for receiving. T2-weighted structural images were acquired with a field of view (FoV) = 23 mm, matrix = 256 x 256; slice thickness = 0.6 mm, 24 slices, repetition time (TR) = 3810 ms, echo time (TE) = 41 ms. For pCASL, interleaved control and labeled images were acquired with a train of 200 Hanning window-shaped radiofrequency pulses of duration = 200 μs, spacing = 200 μs, 25 deg flip angle, max gradient (G_z) = 90 mT/m, mean G_z = 7 mT/m. Two-dimensional multislice spin-echo planar imaging was used with a FoV = 18 mm, matrix = 64 x 64, slice thickness = 1.0 mm, 6 slices, TR = 4,000 ms, TE = 24 ms, with 120 repetitions. A separate normalization image was subsequently acquired with TR = 10,000 ms and 10 repetitions (M₀). Data processing was performed in a Linux terminal where the tagged signal was subtracted from the untagged and normalized by dividing by M₀. A single slice at the dorsal hippocampal level was outlined in FSLeyes for all mice and average cerebral blood flow pixel value calculated.

Metabolite concentration measurement by proton magnetic resonance spectroscopy (MRS).

After pCASL, a 5.2 x 2.0 x 1.3 mm³ voxel was placed around the bilateral hippocampus for water-suppressed MRS measurement: TR = 1500 ms, TE = 135 ms, 400 scans averaged. This was followed by a scan without water suppression, 10 averages. Shim was manually adjusted to get the FWHM below 30 Hz for all animals. Water-suppressed and unsuppressed scans were processed using LCModel with the LCMgui interface [35]. Analyzed metabolite peak values with %SD > 20% were excluded from analyses. If a metabolite was excluded from more than 2 animals of a single group it was not analyzed across groups. Eight metabolites met inclusion criteria: creatine (Cr), phosphocreatine (PCr), lactate (Lac), glutamate (Glu), glutamate and glutamine (Glu+Gln), taurine (Tau), total choline (GPC+PCh), N-acetylaspartate (NAA), and N-acetylaspartate and N-acetylaspartylglutamate (NAA+NAAG). Values for metabolites are expressed as a ratio against the total creatine (Cr+PCr) concentration.