Differential Left and Right Carotid Artery Blood Flow and Altered Hippocampal Mitochondrial Function After Transverse Aortic Constriction in Aging Rats

Abstract

Background

The hippocampus is a key brain structure that has been implicated in vascular dementia cause and is highly sensitive to changes in cerebral blood flow. Brain hypoperfusion in cardiovascular disease may facilitate neurodegeneration in the hippocampus by limiting substrate transport and metabolism. Although most animal studies have relied on artery occlusion to lower brain blood flow, brain hypoperfusion can also stem from mechanical damage resulting from high blood flow velocity and pulsatility. This study assessed, within the same rodent, whether high and low cerebral blood flow differentially affected hippocampal glucose transport protein expression, mitochondrial fuel oxidation, and expression of mitochondrial quality control proteins.

Methods

Four‐week‐old male and female Sprague–Dawley rats underwent transverse aortic constriction (n=13) or control (n=18) surgeries. Bilateral carotid artery diameter and blood flow were measured 20, 30, and 40 weeks postsurgery. Right and left hippocampal mitochondrial respiration and expression of glucose transporters and mitochondrial quality control proteins were measured 40 weeks postsurgery.

Results

Right carotid blood flow velocity and pulsatility were highest in the right and lowest in the left carotid of transverse aortic constriction animals (P<0.05). Complex I (P=0.057), complex I&II (P<0.05), and complex II uncoupled (P<0.05) respiration rates were lower in the right hippocampus of transverse aortic constriction animals when compared with the left, and markers of mitochondrial fusion were upregulated in transverse aortic constriction animals compared with controls (P<0.05).

Conclusions

Although both limited and pulsatile blood flow alter mitochondrial fusion markers, only pulsatile flow impairs mitochondrial respiration, suggesting turbulent hemodynamics may drive metabolic dysfunction in vascular dementia.

Nonstandard Abbreviations and Acronyms

GLUT‐1

glucose transporter 1

GLUT‐3

glucose transporter 3

HSP‐60

heat shock protein‐60

OPA1

optic atrophy 1

TAC

transverse aortic constriction

SHAM

sham‐operated control

Clinical Perspective.

What Is New?

• High carotid blood flow velocity and pulsatility impair hippocampal mitochondrial respiration, whereas reduced carotid blood flow does not.

• Hippocampal protein expression of mitochondrial quality control markers is upregulated in response to altered carotid blood flow, but protein expression of endothelial and neuronal glucose transporters is unaffected.

What Are the Clinical Implications?

• Glucose oxidation by brain mitochondria is negatively affected to a greater degree by high carotid blood flow velocity and pulsatility than by limited carotid blood flow, and alterations in mitochondrial structure and function may precede changes in glucose transport, demonstrating that early dysregulation of mitochondrial processes may play a role in the cause of vascular dementia.

Cardiovascular and neurodegenerative diseases represent 2 of the most common age‐related chronic diseases among older adults, ¹ , ² , ³ which cumulatively account for 36% of the disease burden in the older adult population. ⁴ Cardiovascular health and brain health are closely coupled, ⁵ and cardiovascular disease can increase the risk of developing neurodegenerative diseases ⁶ and exacerbate the rate of progression and clinical manifestations of dementia. ⁷ Vascular pathologies can independently give rise to a spectrum of deficits in brain function and structure, clinically known as vascular cognitive impairment and vascular dementia. ⁸ Nevertheless, the cellular mechanisms by which neurodegeneration can be initiated, influenced, or exacerbated by cardiovascular disease remain poorly understood. ⁹ , ¹⁰

Chronic brain hypoperfusion is widely seen as a key link between cardiovascular disease and neurodegeneration. ¹¹ Human studies show impaired brain blood flow ¹² and brain perfusion ¹³ , ¹⁴ may lower oxygen and glucose delivery to neural mitochondrial and facilitate severe energetic deficits. ¹⁵ Although brain hypoperfusion is seldomly operationally defined, it can be a consequence of 2 distinct hemodynamic phenotypes. ¹⁶ The first rises from conditions that chronically impair perfusion pressures such as those that lower the amount of blood pumped into circulation each heart beat (heart failure with reduced ejection fraction, dilated cardiomyopathy), ¹⁷ , ¹⁸ or those that negatively influence blood pressure (carotid sinus hypersensitivity). ⁵ , ¹⁶ , ¹⁹ , ²⁰ The second stems from conditions that lower arterial compliance and increase blood flow velocity and pulsatility (eg, hypertension, carotid stenosis), which ultimately lead to chronic mechanical strain and end‐organ damage. ²¹ Both phenotypes can result in lower brain perfusion and brain glucose uptake. ²² , ²³ Lower perfusion and glucose uptake could negatively impact neural mitochondrial adenosine triphosphate (ATP) production. ¹⁵ , ²⁴ Thus, the disruption of glucose transport and altered mitochondrial ATP production in chronic brain hypoperfusion may be an underlying mechanism linking cardiovascular disease to neurodegeneration.

A major limitation in studying brain hypoperfusion in animal models of cardiovascular disease and neurodegeneration is the reliance on unilateral or bilateral artery occlusion to obstruct blood flow. ²⁵ , ²⁶ These models fail to recognize that brain hypoperfusion can also arise from high blood flow velocity and pulsatility. ²¹ Although both mechanisms of brain hypoperfusion have been independently reported to negatively impact neuronal mitochondria, ²⁷ , ²⁸ the magnitude of tissue hypoperfusion and the severity of cerebrovascular dysfunction that follow are not equivalent between the 2. ²⁹ Therefore, although it is likely that these contrasting hypoperfusion mechanisms can exert distinct effects on the cellular metabolic processes linking cardiovascular disease and neurodegeneration (eg, neuronal glucose transport and mitochondrial metabolism), it remains a knowledge gap that continues to be unaddressed in the literature.

To investigate the metabolic consequences of the 2 hemodynamic phenotypes of brain hypoperfusion, we used transverse aortic constriction (TAC) in male and female rats, given that it can generate the 2 hemodynamic phenotypes within the same animal. TAC involves the partial ligation of the aorta between the right and left common carotid arteries, inducing high pulsatile blood flow through the right carotid while limiting blood flow through the left carotid, a response that is parallel in the cerebral hemispheres. ²⁶ , ³⁰ As such, there were 2 primary aims for this study. The first was to characterize the long‐term hemodynamic response to TAC (20, 30, and 40 weeks postsurgery) in the left and right carotid arteries of male and female rats. The second was to investigate whether mitochondrial respiration, content, and protein expression of glucose transporters and mitochondrial quality control markers differed in the left and right hippocampus in TAC rats, and if these differed from sham‐operated controls (SHAM).

METHODS

This study was approved by the Institutional Animal Care and Use Committee at the University of Maryland College Park. Male and female Sprague–Dawley rats underwent either TAC surgery to mimic chronic hypertension and induced heart failure with preserved ejection fraction or a SHAM surgery, which was identical to TAC, barring the constriction of the aorta. Following surgical procedures and wound healing, animals were paired‐housed, fed water and food ad libitum, and aged for 40 weeks.

TAC Surgery

Male and female Sprague–Dawley rats (4 weeks old) underwent either the TAC or SHAM surgical procedures as described previously. ³¹ Briefly, each animal was weighed and anesthetized via inhalation of isoflurane (≈2% isoflurane supplemented with 100% oxygen [400–500 mL/min]). Once unresponsive, a single dose of buprenorphine (0.05–0.1 mg/kg) was injected subcutaneously, and the surgical area was clipped clean and sterilized. A horizontal skin incision was made at the suprasternal notch region, a longitudinal ≈1 cm cut was made in the sternum, and the left and right carotids were visualized. In both procedures, a 4‐0 silk suture attached to a blunted needle was passed around the aortic arch between the origin of the right innominate and left common carotid arteries. A bent, blunted 20‐gauge needle was then placed next to the aortic arch, the suture was tightened around the needle, and the needle was then promptly removed. The sternum was closed with a single silk suture, and the skin was sutured shut with 5‐0 monofilament suture and treated with betadine. SHAM animals received the same procedure as TAC animals but did not have the suture tied around the aortic arch.

Cardiac and Carotid Doppler Ultrasound

Carotid blood flow and artery diameter were measured with a Vevo3100 (FUJIFILM VisualSonics) animal Doppler ultrasound imaging system at 20, 30, and 40 weeks postsurgery. All animals were placed under anesthesia in an induction chamber with 2% to 5% isoflurane supplemented with 100% oxygen (400–500 mL/min). Once animals became unresponsive, the fur on the chest was removed, ophthalmic gel was applied to both eyes, and animals were placed on an imaging platform and administered isoflurane via nose cone (≈2% isoflurane supplemented with 100% oxygen [400–500 mL/min]). Ultrasound imaging of the aortic arch and cardiac tissue was used to confirm aortic arch binding and obtain cardiac parameters such as heart rate, cardiac output, ejection fraction, and stroke volume under anesthesia 40 weeks postsurgery. Carotid artery blood flow was determined with the use of B‐mode and pulsed wave (PW) Doppler imaging of the bilateral carotid arteries (Figure 1A). Ultrasound files (PW Doppler and B‐mode) were analyzed in triplicates with the Vevolab software (FUJIFILM VisualSonics), and carotid arterial blood flow velocity (mean and peak velocities), blood flow pulsatility (pulsatility index), and artery diameter during systole and diastole were determined.

Figure 1. TAC carotid hemodynamic phenotype is characterized by greater right and lower left carotid blood flow that does not worsen with time.

A, Doppler ultrasound waveforms of the left and right carotid arteries of TAC animals. B, Peak carotid blood flow velocity in the right and left carotids of TAC and SHAM animals 20, 30, and 40 weeks postsurgery. C, Calculated right and left carotid flow rate in TAC and SHAM animals 20, 30, and 40 weeks postsurgery. D, Blood flow pulsatility in the right and left carotid of TAC and SHAM animals 20, 30, and 40 weeks postsurgery. *P<0.05, **P<0.01, ***P<0.001. Data are presented as mean+SEM. SHAM indicates sham‐operated controls; and TAC, transverse aortic constriction.

Euthanasia and Tissue Isolation

Rats were first anesthetized (~2% isoflurane supplemented with 100% oxygen [400–500 mL/min]). Once unresponsive, the heart was removed via bilateral thoracotomy, the head was removed, and the brain was dissected out. Whole brain samples were placed in ice‐cold 1x PBS, and the bilateral hippocampi were isolated and weighed. The hippocampus was selected as a structure of interest because it is a subcortical structure particularly sensitive to neurodegenerative diseases ³² , ³³ and is one that is heavily targeted in the animal literature of brain hypoperfusion. ²⁶ , ²⁸ , ²⁹ Following bilateral hippocampal isolation, ≈11 mg of the right and left hippocampus were homogenized in 4.4 mL of ice‐cold respiration media (50 mmol/L 3‐(N‐morpholino)propanesulfonic acid [MOPS], 20 mmol/L glucose, 1 mmol/L EGTA, 100 mmol/L KCl, 10 mmol/L MgCl₂, 0.2% BSA) using separate ice‐cold Dounce homogenizers. The remaining hippocampal samples were placed in separate conical tubes with ice‐cold 1x RIPA lysing buffer containing protease and phosphatase inhibitors (ThermoFisher Scientific, Waltham, MA) and homogenized using a hand‐held motorized microtube homogenizer (VWR, Radnor, PA). Samples were spun at 1500 x g for 15 minutes to obtain purified protein from each sample, which were then stored at −80 °C for further analysis.

Hippocampal Respiration

Hippocampal mitochondrial oxygen consumption was obtained via liquid‐phase respiration in a temperature‐controlled respiration chamber fitted with a Clark‐type electrode (Oxytherm⁺, Hansatech Instruments, Norfolk, United Kingdom). Next, 1 mL of hippocampal homogenate was placed into the respiration chamber with the temperature preset at 37 °C with constant stirring at 60 rpm, and a previously published substrate‐uncoupler‐inhibitor‐titration respiration protocol was used to assess mitochondrially fueled metabolism. ³⁴ While continually monitoring oxygen consumption, non‐ADP‐phosphorylating leak respiration was induced by adding the CI (complex I)‐linked substrates pyruvate (5 mmol/L), malate (0.5 mmol/L), and glutamate (10 mM). Then ADP (2.5 mmol/L) was added to measure maximal CI‐linked, ADP‐phosphorylating oxygen consumption, and then succinate (10 mmol/L) was added, to assess the combined CI&CII (complex II) phosphorylating oxygen consumption rate. Maximal uncoupled respiration was obtained with the stepwise titration of protonophor carbonyl cyanide p‐trifluromethoxyphenylhydrazone (FCCP) (0.5 μmol/L additions separated by 60 seconds), which dissipates the mitochondrial protonmotive force and allows for an estimation of maximal electron transport system electron transportation capacity. Rotenone (0.5 μmol/L) was then used to inhibit CI and obtain CII‐linked uncoupled respiration. Last, CII and CIII were inhibited with malonate (2 mmol/L) and antimycin A (2.5 mmol/L) to assess residual nonmitochondrial oxygen consumption.

Hippocampal Mitochondrial Content

Citrate synthase activity was assessed as an estimate of mitochondrial content and was performed following previously published protocols. ³⁵ Hippocampal homogenates were added to a cuvette containing, 100 mmol/L Tris HCl, 0.1 mmol 5,5‐dithiobis‐(2‐nitrobenzoic acid), 0.25 mmol/L acetyl‐CoA, and 0.1% Triton, and background was read. Substrate dependent activity was initiated by addition of 0.5 mmol/L oxaloacetate into a final volume of 1.0 mL, and absorbance was followed at 412 nm for 180 seconds. Activity was calculated using a millimolar extinction coefficient of 13.6 for the mercaptide ion.

Western Blotting

Right and left hippocampal protein concentrations were first determined via Pierce bicinchoninic acid (BCA) protein assay (ThermoFisher Scientific). Equal amounts of protein (30 μg/lane) were loaded and separated by SDS‐PAGE using Mini‐PROTEAN® stain‐free tris‐glycine eXtended (TGX) precast gels (2% SDS, 25% glycerol, 0.01% bromophenol blue) (Bio‐Rad, Hercules, CA) and transferred to polyvinylidene fluoride (PVDF) membranes (Bio‐Rad). Membranes were then blocked in 5% to 6% nonfat dry milk in tris‐buffered saline with 0.1% Tween® 20 detergent (TBST) for 1 hour and then incubated at 4 °C overnight in 3% nonfat dry milk in TBST with primary antibodies for target proteins. The primary antibodies for this study included: GLUT‐1 (glucose transporter 1) (Cell Signaling number12939, 1:900), GLUT‐3 (glucose transporter 3) (Invitrogen number MA532697, 1:900), DRP1 (dynamin‐related protein 1) (Cell Signaling number 8570, 1:1000), FIS1 (fission 1) (Millipore Sigma number HPA017430, 1:900), MFN2 (mitofusin 2) (Cell Signaling number 83667, 1:1500), OPA1 (optic atrophy 1 isoform) (Cell Signaling number 80471, 1:1000), and HSP‐60 (heat shock protein‐60) (Cell Signaling number 12165, 1:1600). A horseradish peroxidaze (HRP)‐conjugated secondary antibody was used (Cell Signaling number 7074), and the SuperSignal West Pico Chemiluminescent Substrate detection kit was used to initiate the immunoreactive protein reaction (ThermoFisher Scientific). Protein bands were detected and imaged with the Bio‐Rad ChemiDoc system and software (Bio‐Rad), and ImageJ software was used to quantify bands using densitometry. For all Western blot experiments, total protein was used as the normalizing control and was obtained using stain‐free imaging technology (Bio‐Rad).

Statistical Analysis

To test any main effects of condition (TAC versus SHAM), hemisphere (right versus left), or condition by hemisphere interactions on carotid blood flow, pulsatility index, and artery diameter, a series of independently run 2×2 ANOVA were used, and significant condition by hemisphere interactions were followed up with a post hoc analysis using the Tukey honestly signficant difference (HSD) correction for multiple comparisons. Additionally, repeated 1‐way ANOVA models were independently run and used to determine whether there were any main effects of time (20, 30, and 40 weeks) on carotid blood flow, pulsatility index, and artery diameter measures. Separate 2×2 ANOVA models were also used to test whether there were any main effects of condition (TAC versus SHAM), hemisphere (right versus left), or condition by hemisphere interactions on hippocampal respiration, mitochondrial content, and protein expression of markers associated with mitochondrial quality control. Because previous literature has found hemispheric differences in brain perfusion within TAC, and because we also hypothesized these hemispheric differences would also apply for mitochondrial respiration and protein expression of target markers, all 2×2 ANOVA models for mitochondrial outcomes were followed up with planned comparisons to specifically test whether there were any hemispheric differences within TAC. Pearson product–moment correlations were used to determine if carotid artery blood flow velocities and pulsatility 40 weeks postsurgery were related to mitochondrial respiration. Last, separately run, independent samples t tests or Mann–Whitney U tests were used as exploratory analyses to address any possible sex differences on carotid blood flow, pulsatility index, artery diameter measures, hippocampal respiration, mitochondrial content, and protein expression within either the TAC or SHAM conditions. All analyses were conducted using the statistical software JASP (version 0.17.3) and GraphPad Prism (version 10.2.0), and statistical significance was set at P<0.05. Data that support the findings of this study are available from the corresponding author upon reasonable request.

Excluded and Missing Data

Thirty‐one animals were used in the investigations (7 TAC male, 6 TAC female, 8 SHAM male, and 10 SHAM female). However, for the assessment of carotid hemodynamics, a total of 1 TAC (1 female) and 8 SHAM (4 male, 4 female) animals were excluded from the analysis due to poor ultrasound quality or missing data. Furthermore, Doppler ultrasound time series revealed 3 TAC animals had left carotid blood flow obstructed to such degree that blood flow wave forms could not be analyzed (eg, flat wave forms) despite good image contrast and measurable flows on the right carotid (Figures S1–S6A). Instead of assigning a blood flow value of 0 for these animals, the lowest measured left carotid values within the TAC group were assigned as proxy left carotid values to provide a realistic conservative flow value. Last, 2 male TAC animals with complete data at 20 and 30 weeks, but incomplete data at 40 weeks (due to poor ultrasound image resolution), were included in the analyses for which the respective data were available (eg, 20 and 30 weeks 2×2 ANOVA) but excluded from the time points for which no data were available (eg, 40 weeks 2×2 ANOVA). As such, for the carotid hemodynamic assessments only, a total of 12 TAC (7 male, 5 female) and 10 SHAM (4 male, 6 female) animals with a complete set of data for most time points were used. For all other study outcomes, data were obtained from 30 animals, because respiration data for 1 TAC (female) and Western blot data for 1 SHAM (male) could not be obtained due to an unstable oxygen trace and protein degradation, respectively. Animal characteristics for the full (n=31) cohort can be found in the Table 1, and animal characteristics by condition and sex can be found in Table S1.

Table 1.

Animal Characteristics

	TAC (n=13), mean (SEM)	SHAM (n=18), mean (SEM)
Sex
Male	7	8
Female	6	10
Age, wk	45.5 (0.3)	45.7 (0.3)
Body mass, g	470 (32.9)	464 (37.9)
Heart mass, g	1.4 (0.1)	1.3 (0.09)
Heart mass to body mass, mg/g	3.1 (0.14)	3.0 (0.16)
Heart rate, bpm	318.7 (8.8)	327.2 (8.0)
Left ventricular width, mm	3.6 (0.14)	3.5 (0.10)
Left ventricular width to body mass, mm/mg	8.0 (0.47)	8.4 (0.61)
Septal width, mm	3.2 (0.11)	3.0 (0.08)
Septal width to body mass, mm/mg	7.1 (0.49)	7.0 (0.43)
Cardiac output, mL/min	76.5 (3.7)	82.4 (4.0)
Stroke volume, μL	240 (10.0)	253 (12.7)
Ejection fraction, %	72 (3.6)	77 (2.0)
Tibialis anterior mass, g	1.6 (0.11)	1.5 (0.11)
Total hippocampus wet weight, g	0.11 (0.003)	0.11 (0.002)
Right hippocampus wet weight, g	0.057 (0.001)	0.056 (0.001)
Left hippocampus wet weight, g	0.058 (0.001)	0.057 (0.001)
Right hippocampal citrate synthase activity, μmol/g per min	31.5 (5.0)	31.1 (4.3)
Left hippocampal citrate synthase activity, μmol/g per min	36.4 (4.3)	28.1 (3.0)

Data are presented as mean (SEM). SHAM indicates sham‐operated controls; and TAC, transverse aortic constriction.

RESULTS

Animal Characteristics

Thirty‐one animals were used (7 TAC male, 6 TAC female, 8 SHAM male, and 10 SHAM female) in this study. Animal characteristics for the full (n=31) cohort can be found in the Table. Overall, no significant differences between TAC and SHAM were observed in any of the characteristics, including hippocampal mitochondrial content, heart mass, septal width, cardiac output, ejection fraction, and stroke volume (Table 1).

TAC Carotid Blood Flow Velocities and Pulsatility Were Chronically Lowest on the Left Hemisphere and Highest on the Right Hemisphere

At all time points measured, 2×2 ANOVA revealed there were significant main effects and interactions on peak carotid blood flow velocity (Figure 1B, Table S2) and pulsatility index (Figure 1C, Table S3). At 20 weeks, peak artery velocity was lower in the left carotid of TAC (mean [M]=902.1±99.4) when compared with the right (M=1716.8±84.9 mm/s, P<0.001) of TAC, as well as the right (M=1450.5±59.1 mm/s, P<0.001) and left (M=1376.3 mm/s±73.9, P<0.01) of SHAM animals. These patterns held steady at 30 and 40 weeks postsurgery, and peak blood flow velocity remained lowest in the left carotid of TAC (30 weeks M=898.5±129.2 mm/s; 40 weeks, M=794.9±103.6 mm/s) when compared with the right of TAC (30 weeks M=1692.0±85.2 mm/s, P<0.001; 40 weeks M=1573.7±81.9 mm/s, P<0.001) (Figure 1B), as well as the right (30 weeks M=1490.8±81.2 mm/s, P<0.01; 40 weeks M=1551.7±131.2 mm/s, P<0.01) and left (30 weeks M=1505.0±92.9 mm/s, P<0.001; 40 weeks M=1442.0±85.7 mm/s; P<0.001) of SHAM animals (Figure 1B).

Blood flow pulsatility was consistently highest on the right carotid of TAC animals, whereas it was simultaneously lowest on the left carotid. At 20 weeks postsurgery, blood flow pulsatility was significantly higher in the right carotid of TAC (M=1.43±0.02 pulsatility index (P.I) animals when compared with the left (M=1.05±0.04 P.I, P<0.001) of TAC, as well as the right (M=1.30±0.03 P.I, P<0.05) and left (M=1.23±0.02 P.I, P<0.001) of SHAM animals (Figure 1C). Carotid blood flow pulsatility remained significantly higher on the right carotid of TAC at all other time points (30 weeks M=1.39±0.03 P.I, 40 weeks M=1.43±0.03 P.I) when compared with the left (30 weeks M=0.98±0.05 P.I, P<0.001; 40 weeks M=0.86±0.04 P.I, P<0.001) of TAC, as well as the left (30 weeks M=1.22±0.02 P.I, P<0.01; 40 weeks M=1.22±0.02 P.I, P<0.001) carotid of SHAM controls (Figure 1C). Outside of 20 weeks postprocedure, carotid blood flow pulsatility on the right of TAC was only significantly higher than the right of SHAM animals at 40 weeks postprocedure (TAC M=1.43±0.03 P.I, SHAM M=1.29±0.03 P.I; P<0.05) (Figure 1C).

Last, carotid flow rate was also calculated as previously described ³⁶ and was found to be lower in the left (20 weeks M=0.58±0.07, 30 weeks M=0.59±0.1, 40 weeks M=0.51±0.07) of TAC animals at all time points when compared with the right (20 weeks M=1.06±0.07, P<0.001; 30 weeks M=1.07±0.05, P<0.001; 40 weeks M=1.02±0.05, P<0.001) of TAC animals, as well as the right (20 weeks M=0.90±0.04, P<0.01; 30 weeks M=0.91±0.04, P<0.05; 40 weeks M=0.98±0.06, P<0.001) and left (20 weeks M=0.84±0.04, P<0.05; 30 weeks M=0.91±0.06, P<0.05; 40 weeks M=0.94±0.05, P<0.001) of SHAM animals (Figure 1D, Table S4). A significant main effect of time was observed for blood flow pulsatility in the left carotid of TAC animals (Table S5). Post hoc comparisons revealed that blood flow pulsatility was significantly lower at 40 weeks when compared with 20 (20 weeks M=1.05±0.04, 40 weeks M=0.86±0.04; P<0.05), and 30 weeks when compared with 40 (30 weeks M=0.98±0.05, 40 weeks M=0.86±0.04; P<0.01). No other significant main effects of time were observed on carotid artery velocities, blood flow pulsatility (Table S5), or carotid flow rate.

Sex differences in TAC animals were only observed on the right side, and females tended to showcase significantly greater peak velocities at 30 and 40 weeks than males (Figure S2A). No significant sex differences were seen in right carotid blood flow pulsatility in TAC (Figure S2A). In SHAM animals, females displayed significantly greater peak blood flow velocities at 40 weeks than males (Figure S2B). When compared with male SHAM animals, blood flow pulsatility was also found to be significantly higher in female SHAM animals at 20, 30, and 40 weeks postsurgery (Figure S2B).

Carotid Artery Cross‐Sectional Area and Diameter Are Influenced by TAC Surgery and Sex

Carotid artery cross‐sectional area in systole was calculated as cross‐sectional area=D²×0.785, ³⁷ and no significant main effects or interactions on systolic cross‐sectional area were detected (Figure S3A). Independently run 2×2 ANOVA models revealed no significant main effects or condition by hemisphere interactions on either systolic or diastolic carotid artery diameter at any of time points measured (Figures S4A and S5A). A significant main effect of time was observed for left systolic artery diameter in SHAM animals, but post hoc comparisons indicated none of the differences survived multiple comparison correction (Table S6). No other significant main effects of time on either systolic carotid artery diameter, diastolic artery diameter (Table S6), or systolic cross‐sectional area in either TAC or SHAM animals were observed.

As it pertains to sex differences, TAC females had on average significantly lower cross‐sectional area, as well as systolic and diastolic artery diameters than males, but these differences were only significant on the right side at 20 weeks (Figures S3B, S4B, and S5B). In SHAM animals, females displayed significantly lower systolic cross‐sectional area at all time points when compared with males (Figure S3C). Furthermore, sex differences in carotid artery diameter were extensive, and females displayed lower systolic and diastolic diameters than males at all time points (Figures S4C and S5C).

Hippocampal Mitochondrial Respiration Is Impaired in the Right, But Not the Left, Hemisphere of TAC Animals

At 40 weeks post‐TAC, a significant main effect of hemisphere (F (1, 56)=5.08, P<0.05, ηp²=0.08) in CI‐linked coupled respiration was observed, and follow‐up contrast revealed that within condition hemispheric differences were greatest in TAC (t (56)=1.94, P=0.057) than SHAM (t (56)=1.18, P=0.24) animals, but these hemispheric differences within TAC animals did not reach statistical significance (P=0.057) (Figure 2A). There was also a significant condition by hemisphere interaction (F (1, 56)=4.69, P<0.05, ηp²=0.077) for CII‐linked uncoupled respiration, and follow‐up contrasts within TAC animals revealed that the right hippocampus had significantly lower CII‐linked uncoupled respiration when compared with the left hippocampus (t (56)=−2.60, P=0.012) (Figure 2A). Last, because there was a main effect of hemisphere that approached significance (F (1, 56)=3.75, P=0.058, ηp²=0.063) for CI&II‐linked coupled respiration, this was followed up with a contrast analysis, and the right hippocampus of TAC animals had significantly lower respiration rates than the left (t (56)=−2.29, P=0.025) (Figure 2A). No main effects of condition (F (1, 56)=1.06, P=0.30, ηp²=0.01), hemisphere (F (1, 56)=0.055, P=0.81, ηp²=0.0009), or condition by hemisphere interaction (F (1, 56)=0.88, P=0.35, ηp²=0.01) were observed for mitochondrial content as measured via citrate synthase activity (Table 1).

Figure 2. Hippocampal mitochondrial respiration is impaired in the right hemisphere in TAC animals, and SHAM animals display sex dimorphisms not present in TAC.

A, Hippocampal mitochondrial respiration in right and left hemisphere of TAC and SHAM animals 40 weeks postsurgery. B, Right and left hemispheric hippocampal. Mitochondrial respiration in male and female TAC animals 40 weeks postsurgery. C, Right and left hemispheric hippocampal. Mitochondrial respiration in male and female SHAM animals 40 weeks postsurgery. *P<0.05, ***P<0.001. Data are presented as mean+SEM. CI indicates complex I; CII, complex II; ETS, electron transfer system; F, female; FCCP, carbonyl cyanide p‐trifluoromethoxyphenylhydrazone; G, glutamte; J_O, oxygen consumption; LEAK, leak respiration; M, malate; M, male; P, pyruvate; ROX, residual nonmitochondrial oxygen consumption; SHAM, sham‐operated controls; and TAC, transverse aortic constriction.

No sex differences in mitochondrial respiration were found in TAC animals (Figure 2B). However, in SHAM controls, CI‐linked coupled respiration was lower in SHAM females when compared with male counterparts (M=3.30±0.18 versus M=4.1±0.12, P<0.001; r=−0.794) (Figure 2C). Female SHAM animals also displayed lower CI&II‐linked coupled respiration when compared with male counterparts (M=5.82±0.24 versus M=7.08±0.16, P<0.001; r=−0.688) (Figure 2C).

Pearson product–moment correlations were used as an exploratory analysis to determine if carotid artery blood flow velocities and pulsatility 40 weeks postsurgery were related to mitochondrial respiration. For this analysis, only SHAM and TAC animals with a complete set of hemodynamic data were included. That is, the 3 TAC animals that were found to have poor left carotid time series were excluded from this analysis. Overall, carotid artery peak velocity (r=−0.453, P=0.004 [−0.67 to −0.16]) (Figure 3A), blood flow pulsatility (r=−0.429, P=0.006 [−0.65 to −0.13]) (Figure 3B), and mean velocity (r=−0.448, P=0.004 [−0.66 to −0.15]) (Figure S6A) were all negatively associated to CI&II‐linked coupled respiration, and higher velocities and pulsatility were associated with impaired respiration rates. Similarly, carotid artery peak velocity (r=−0.473, P=0.002 [−0.68 to −0.18]) (Figure 3C), and blood flow pulsatility (r=−0.481, P=0.002 [−0.69 to −0.19]) (Figure 3D), and mean velocity (r=−0.45, P=0.003 [−0.67 to −0.16]) (Figure S6B) were also negatively associated to CII‐linked uncoupled respiration, and higher velocities and pulsatility were associated with lower uncoupled respiration rates.

Figure 3. High carotid blood flow velocity and pulsatility are negatively associated with hippocampal mitochondrial respiratory states.

A, Scatterplot depicting the negative association between carotid artery velocity and pulsatility to CI&II coupled hippocampal respiration 40 weeks postsurgery. B, Scatterplot depicting the negative association between carotid artery velocity and pulsatility to CII uncoupled hippocampal respiration 40 weeks postsurgery. CI indicates complex I; CII, complex II; ETS, electron transfer system; and J_O, oxygen consumption.

Alterations in Carotid Hemodynamics Are Coupled to an Upregulation in Hippocampal Protein Expression of Mitochondrial Fusion Proteins

There were significant main effects of condition for both the long (F (1, 56)=4.92, P<0.05, ηp²=0.081) and short (F (1, 56)=5.01, P<0.05, ηp²=0.082) (Figure 4A) OPA1 isoforms, as well as total OPA1 (F (1, 56)=5.09, P<0.05, ηp²=0.083) (calculated as the sum of the density signal of the short and long isoforms), with TAC animals displaying higher protein expression than SHAM animals 40 weeks post‐TAC surgery. Furthermore, a significant main effect of condition (F (1, 56)=6.53, P<0.05, ηp²=0.10), where TAC animals showcased higher protein expression than SHAM, was also detected for HSP‐60 (Figure 4A).

Figure 4. Hippocampal protein expression of mitochondrial fusion markers is greater in TAC animals and among females in TAC and SHAM.

A, Protein expression of mitochondrial fusion markers in the right and left hippocampus of TAC and SHAM animals 40 weeks postsurgery. B, Protein expression of mitochondrial fusion markers in male and female TAC animals 40 weeks postsurgery. C, Protein expression of mitochondrial fusion markers in male and female SHAM animals 40 weeks postsurgery. *P<0.05, **P<0.01. Data are presented as mean+SEM. A.U indicates arbitrary units; F, female; HSP‐60, heat shock protein‐60; L, left; L‐OPA1, OPA1 long isoform; M, male; MFN, mitofusin 2; OPA1, optic atrophy 1; R, right; SHAM, sham‐operated controls; S‐OPA1, OPA1 short isoform; and TAC, transverse aortic constriction.

In TAC animals, sex differences in protein expression were only found within the right hippocampus, whereas females were found to have significantly higher protein expression of the short (M=1.33±0.12 versus M=0.89±0.12, P<0.05; d=1.36) and long (M=1.88±0.13 versus M=1.32±0.11, P<0.05; d=1.76) OPA1 isoforms compared with males (Figure 4B). No significant sex differences in hippocampal protein expression were found in the left hemisphere of TAC animals. In SHAM animals, females had higher protein expression than males of MFN2 (M=1.31±0.07 versus M=0.99±0.11, P<0.05; d=0.82) (Figure 4C), and DRP1 (M=1.18±0.07 versus M=0.88±0.09, P<0.05; d=0.88) (Figure 5C). No other sex differences were evident in SHAM controls.

Figure 5. Hippocampal protein expression of mitochondrial fission markers are not altered in TAC, but sex differences are present in SHAM controls.

A, Protein expression of mitochondrial fission markers in TAC and SHAM. B, Protein expression of mitochondrial fission markers in male and female TAC animals 40 weeks postsurgery. C, Protein expression of mitochondrial fission markers in male and female SHAM animals 40 weeks postsurgery. *P<0.05. Data are presented as mean+SEM. A.U indicates arbitrary units; DRP1, dynamin‐related protein 1; F, female; FIS1, fission 1; L, left; M, male; R, right; SHAM, sham‐operated controls; and TAC, transverse aortic constriction.

Protein Expression of Glucose Transporters Unaltered in TAC Animals

Overall, 40 weeks post‐TAC, we found no significant main effects of condition (F (1, 56)=0.424, P=0.51, ηp²=0.008), hemisphere (F (1, 56)=0.33, P=0.56, ηp²=0.006), or condition by hemisphere interaction (F (1, 56)=0.11, P=0.73, ηp²=0.002) in hippocampal protein expression of GLUT‐1 (Figure 6). Similarly, no significant main effects of condition (F (1, 56)=0.38, P=0.53, ηp²=0.007), hemisphere (F (1, 56)=0.22, P=0.63, ηp²=0.004), or a condition by hemisphere interaction (F (1, 56)=0.09, P=0.75, ηp²=0.002) for hippocampal protein expression of GLUT‐3 (Figure 6) were observed 40 weeks post‐TAC. Furthermore, no sex differences were found in the protein expression of glucose transporters in either TAC or SHAM animals.

Figure 6. Protein expression of glucose transporters is not altered in TAC 40 weeks postsurgery.

Data are presented as mean+SEM. A.U. indicates arbitrary units; GLUT‐1, glucose transporter 1; GLUT‐3, glucose transporter 3; L, left; R, right; SHAM, sham‐operated controls; and TAC, transverse aortic constriction.

DISCUSSION

Although primarily used to cause cardiac pressure overload, TAC also alters blood flow within the carotid arteries so that right carotid artery (upstream of the constriction) blood flow velocities and pulsatility are greater than the left (downstream of the constriction). Thus, the transverse aortic constriction model is uniquely suited to simultaneously study how the 2 primary hemodynamic mechanisms linked to brain hypoperfusion ¹⁶ , ³⁸ influence cellular processes associated with neurodegeneration. Overall, we report that distinct carotid hemodynamics are coupled to hemispheric differences in coupled and uncoupled mitochondrial respiration, where the right hippocampus shows deficits in mitochondrial respiration and the left hippocampus does not. We also report that TAC animals have higher expression of mitochondrial fusion and uncoupled protein response (UPR^MT) proteins in both hemispheres, compared with SHAM animals. Together, these findings suggest the molecular mechanisms linking cardiovascular disease to neurodegeneration could be uniquely influenced by peripheral hemodynamic patterns and could inform future studies on flow‐ and perfusion‐dependent processes.

Carotid Artery Hemodynamics

Carotid blood flow velocities are higher in the right carotid of TAC animals when compared with the left carotid at every time point following constriction, and carotid blood flow pulsatility is higher in the right of TAC animals when compared with not only the left carotid within TAC, but also the right and left carotids of SHAM animals. This is in line with existing literature showing arterial pulse pressure is highest in the right carotid of TAC animals when compared with the right and left of SHAM animals ²⁹ and may indicate losses in arterial compliance are greatest in the right carotid of TAC. It has been previously reported that the divergent blood flow phenotype in TAC can develop as soon as 1 day postsurgery and remain largely unchanged 7 days postsurgery. ³⁰ A similar pattern has been reported in a mouse model of Alzheimer disease with TAC‐induced brain hypoperfusion, where the bilateral hemodynamic phenotype is present 3 weeks post‐TAC, and largely unchanged 8 weeks post‐TAC. ³⁹ Because most studies assessing carotid artery hemodynamics in TAC have used acute (1 day and 7 days post‐TAC), ³⁰ or pre‐ and postmodels, ²⁹ , ³⁹ , ⁴⁰ there are little data on characterizing carotid blood flow in TAC from a chronic point of view. Furthermore, these rats were 20 weeks old at the first ultrasound assessment, and therefore growth was unlikely to contribute to worsening hemodynamics. As such, these results build on the existing literature and show TAC carotid flow characteristics are stable over time and can be therefore adapted for chronic studies.

Recent work by De Montgolfier and colleagues provides compelling evidence showing that 6 weeks post‐TAC surgery there were 2 distinct hemodynamic phenotypes that induce hypoperfusion, yet the pulsatility associated with the right carotid leads to a more pronounced degree of hypoperfusion in the right cerebral hemisphere when compared with the left. ²⁹ Moreover, these TAC‐induced hemispheric differences in brain perfusion are accompanied by other hemispheric differences such as greater deficits in blood–brain barrier integrity, higher adverse microvascular events (eg, microbleeds), and greater losses in microvascular density in the right hemisphere when compared with the left. ²⁹ Thus, although blood flow velocity in the right carotid of TAC does not substantially differ with SHAM controls in this study, the higher sustained pulsatility seen in the right carotid of TAC may lead to sustained hypoperfusion, cerebral microvascular stress, and neurovascular damage. Overall, this longitudinal characterization of carotid blood flow velocity and pulsatility resulting from TAC demonstrates this model may be suitable to chronically study how, and to what extent, the 2 primary hemodynamic mechanisms linked to brain hypoperfusion ²⁹ influence discrete neurophysiological processes linked to neurodegenerative diseases such as vascular dementia.

Although our TAC animals tended to display a phenotype suggestive of impaired cardiac function and cardiac hypertrophy, these were not significantly different than our SHAM controls. The model was intended to produce a modest cardiac phenotype and includes both male and female animals, increasing variability in size‐dependent parameters. To our knowledge, we are the first to extend the TAC model in rats to 40 weeks postsurgery, and the use of a 20‐gauge needle facilitated the survivability of our animals to the intended end point. Moreover, we note that although we do not have a robust heart failure phenotype, our TAC animals displayed the same carotid hemodynamic phenotype reported in previous work with aggressive ligation aproaches. ²⁶ , ²⁹ , ³⁰ , ³⁹ Therefore, the blunted cardiac function and cardiac hypertrophy phenotypes did not translate to a loss in the carotid and, by extension, the brain hemodynamics that are central to this study.

Mitochondrial Respiration and Quality Control

Mitochondria are key organelles that supply the vast majority of ATP used by neurons ¹⁵ , ⁴¹ , ⁴² , ⁴³ and may be key organelles compromised in chronic brain hypoperfusion. The primary finding of this study is hemispheric differences in mitochondrial respiration following TAC. Specifically, when compared with the left hippocampus, the right hippocampus of TAC animals displays significantly lower CI&II‐linked coupled respiration, as well as CII‐linked uncoupled respiration. Furthermore, despite not reaching statistical significance (P=0.057), we note that the right hippocampus in TAC also displays lower CI‐linked coupled respiration when compared with the left hippocampus (M=3.46 ± 0.23 versus M=4.10 ± 0.24), which may be of physiological relevance. Because coupled respiration represents the maximal respiratory capacity of the electron transfer pathways in the presence of saturating substrates and ADP, ⁴⁴ CI&II‐linked coupled respiration represents the upper limits in coupling the transfer of electrons along the electron transport system to the phosphorylation of ADP to ATP. ⁴⁵ Moreover, because uncoupled respiration used protonophores to disrupt the inner mitochondrial membrane and dissipate the protonmotive force, CII‐uncoupled represents the maximal electron transfer through complex II, III, and IV. ⁴⁴ , ⁴⁶ Thus, the deficits in coupled and uncoupled respiration in the right hippocampus may suggest that high pulsatility within brain‐feeding arteries may have effects that propagate into cellular organelles such as the mitochondria and disrupt the structural and functional properties of the electron transfer system. This intriguing possibility could be further explored by quantifying mitochondrial complex/super complex formation via blue native gel electrophoresis, and this may be natural to next expand on the observations of the present study. Nevertheless, because previous studies show TAC induces a more robust hypoperfusive response in the right hemisphere than the left, ²⁹ our results are complimentary to these observations and suggest it is possible that the metabolic consequences of hypoperfusion are also more pronounced on the right hemisphere than the left.

Meanwhile, despite no hemispheric differences in the protein expression mitochondrial quality control markers in TAC, markers associated with mitochondrial fusion and the UPR^MT were upregulated in both hemispheres of TAC animals when compared with SHAM controls. Both the long‐ and short‐OPA1 isoforms, as well as HSP‐60, are significantly higher in hippocampus of TAC compared with SHAM. It is worth noting that: (1) the UPR^MT and heat shock proteins (HSPs) are integral for the regulation of electron transport chain (ETC) stoichiometries during mitochondrial remodeling and complex formation, ⁴² , ⁴⁷ (2) inner‐membrane‐bound OPA1 is integral for the regulation of mitochondrial cristae shape, ⁴⁸ and (3) whole‐brain mitochondrial complex assembly formation is impaired in chronic hypertension. ²⁷ Therefore, it is possible that the bilateral upregulation of OPA1 in TAC may be an adaptive response to increase mitochondrial cristae complexity to improve metabolism in the presence of altered carotid hemodynamics, and possibly low tissue perfusion, ²⁹ whereas the dual upregulation of HSP‐60 may serve to ensure mitochondrial stoichiometries and ETC complex formation are sustained, and proteotoxic stress is minimized. However, this possibility should be studied in future investigations through a robust combination of additional mitochondrial quality control marker quantification, and advanced imaging of mitochondrial morphology. Because mitochondrial quality control is a complex and closely related set of cellular responses that involve mitochondrial dynamics (eg, fission and fusion), mitophagy, and the unfolded protein response (UPR^MT), ⁴⁷ , ⁴⁹ , ⁵⁰ , ⁵¹ , ⁵² quantifying other markers of dynamics (eg, phosphorylated DRP1, mitochondrial fission factor [MFF], mitochondrial dynamics of 51 kDa protein [MiD51]), mitophagy (eg, Parkin, PTEN‐induced putative kinase 1 [PINK1], BCL2/adenovirus E1B 19‐kDa‐interacting protein 3 [BNIP3], phospho‐ubiquitin) and UPR^MT (eg, mitochondrial 70kDa heat shock protein [mtHSP70], mitochondrial 90kDa heat shock protein [mtHSP90], 40kDa heat shock protein [HSP40]) would help further characterize whether adaptive mechanisms are engaged in the presence of altered hemodynamics. Furthermore, high‐resolution imaging of mitochondria cristae through transmission electron microscopy or immunofluorescence could also provide meaningful context as to how mitochondria structurally adapt to adverse cardio‐/cerebrovascular hemodynamics. Nevertheless, taken collectively, these results suggest neural mitochondria may be able to upregulate quality control mechanisms in the presence of altered carotid hemodynamics linked to brain hypoperfusion, but these adaptive responses may be insufficient when the hemodynamic phenotype involves high blood flow pulsatility.

Glucose Transporter Protein Expression

Surprisingly, there are no hemispheric differences in the expression of either GLUT‐1 or GLUT‐3 within TAC, nor any differences between TAC and SHAM. Poulet and colleagues report hippocampal GLUT‐1 (but not GLUT‐3) mRNA is downregulated in TAC animals when compared with SHAM controls 3 weeks and 4 weeks postsurgery. ³⁹ Intriguingly, when looking at the cortex, GLUT‐1 was only downregulated 4 weeks postsurgery. ³⁹ Because ischemia reperfusion studies have shown acute increases in blood flow are accompanied by an initial upregulation in GLUT mRNA levels, ⁵³ it is possible that downregulation of GLUT mRNA in acute models of TAC is also a temporary response. ⁵⁴ , ⁵⁵

Sex Differences

Historically, cardiovascular disease has been studied in the context of male physiology and disease, ⁵⁶ despite the fact that cardiovascular disease is also the leading cause of death for women across the globe. ⁵⁷ , ⁵⁸ Additionally, emerging frameworks suggest that sex‐related differences in cardiovascular disease coupled with greater longevity may be key in understanding why the prevalence of cognitive decline and dementia is higher among women than men. ⁵⁹ We therefore performed an exploratory analysis to determine whether there are sex differences in right and left carotid hemodynamics, mitochondrial respiration, and protein expression of target markers. Carotid blood flow velocity is greater in female compared with male SHAM animals at the 40‐week time point, and these sex differences occur earlier in TAC (30 and 40 weeks), with higher carotid blood flow velocities in the right side of females. Because the right side of TAC showcased the sex differences in carotid blood flow velocities 10 weeks earlier than SHAM controls, this may suggest that higher peripheral blood flow velocities and associated pulsatility may speed up the development of sex dimorphisms in vascular health that may also be present at the cerebrovascular level. In support of this interpretation, recent frameworks show the cerebral vasculature of nondiseased female Sprague–Dawley rats has different structural properties than their male counterparts such as lower content of vascular smooth muscle cells, higher collagen deposits, and thicker internal elastic lamina. ⁶⁰ These structural differences, in turn, seem to impair the contractile capabilities of large cerebral arteries, lower arterial distensibility, elevate the myogenic response, and have negative effects on the vascular functional capabilities in females when compared with males. ⁶⁰ As such, female animals may have a reduced ability to cope with sustained higher pulsatile blood flow velocities over time, and the early manifestation of sex dimorphisms in blood flow seen in TAC could signify an early failing vasculature. However, future studies should aim to further characterize the peripheral and central vascular dimorphisms that may be present in TAC.

As it pertains to hippocampal mitochondrial physiology, SHAM females have significantly lower CI‐linked and CI&II‐linked coupled respiration, concomitant with higher protein expression of fusion (MFN2) and fission (DRP1) markers compared with SHAM males. These observations contribute to existing animal data where sex‐based differences in mitochondrial function and quality control are reported. For example, CI‐linked respiration is higher in cortical mitochondrial fractions obtained from 13‐month‐old female mice compared with males. ⁶¹ Hormones, such as estradiol, may also influence brain mitochondria, ⁶² because complete removal of estrogens in ovariectomized macaque monkeys is reported to alter brain mitochondrial shape from elongated tubules to a more donut‐shaped phenotype, ⁶³ which in this context is associated with mitochondrial stress. ⁶² Thus, the presence of sex dimorphisms in respiration and markers of mitochondrial dynamics in our SHAM animals could suggest that circulating hormone levels may be partially involved, but this should be addressed further in future studies. Nevertheless, our results contribute to a growing field of research and may aid in better understanding how sex dimorphisms in brain mitochondrial function and dynamics change throughout the lifespan and how these relate to cardiovascular health and disease.

Limitations

Although the present study provides evidence that opposing peripheral vascular hemodynamics linked to cerebral hypoperfusion can be coupled to hemispheric differences in hippocampal mitochondrial function, it is not without limitations. We acknowledge that our statistical approach on most of our experimental outcomes deviates from the traditional linear mixed‐model approach that treats hemisphere as a repeated measure within‐subject factor in the model. However, given that the explicit purpose of TAC surgery is to produce differential left and right peripheral and hemispheric hemodynamic phenotypes, there is an evidence‐based biological rationale to treat left and right hemispheres as independent. ²⁹ Treating each hemisphere as independent removes consideration of the within‐subject correlation between the 2 hemispheres, and so this warrants consideration in the interpretation of the results. Nevertheless, this approach is similar to previously published studies that have used TAC. ³⁰ , ³⁹

Moreover, although we were able to characterize the carotid hemodynamic changes in TAC and SHAM, our study did not determine if these were coupled to structural and functional changes within the carotid arteries themselves. Additionally, despite an assessment of carotid blood flow velocity and pulsatility, we did not quantify hemispheric nor hippocampal tissue perfusion. Thus, the data from this study cannot directly show that altered carotid blood flow in TAC translated to structural changes within the carotid arteries or hemispheric differences in brain perfusion. However, previous work has established that TAC not only leads to whole brain hypoperfusion, but that this hypoperfusion is not the same between the 2 cerebral hemispheres. ²⁶ , ³⁰ Thus, because our animals displayed the same carotid phenotype observed in prior TAC studies with brain hypoperfusion, it is also likely that our TAC animals did develop some degree of brain hypoperfusion.

It is also worth noting that although a recent meta‐analysis found no significant relation between constriction diameter and end‐systolic and end‐diastolic volumes in TAC, ⁶⁴ the degree of cardiac remodeling may be highly dependent on the gauge of the needle. ⁶⁵ A recent study in mice found that even seemingly close gauges can overtly affect the severity of the TAC phenotype, and using a 27‐gauge needle leads to significantly more pronounced manifestations of heart failure than 26‐gauge and 25‐gauge needles. ⁶⁵ Specifically, left ventricular mass, atrial weight, septal width, and cardiac function are all significantly worse in mice whose TAC was performed using 27‐gauge needles than in those whose TAC was done using 25‐gauge needles. ⁶⁵ Moreover, although most animals whose TAC procedure involved 25‐gauge and 26‐gauge needles showed full survivability, 15% of the animals whose TAC procedure involved 27‐guage needles died within 4 days postoperation. ⁶⁵ In this context, selecting a mild constriction for our rats through a 20‐gauge needle facilitated the survivability of our animals to the intended 40‐week end point. Moreover, as noted in the discussion, although we do not have a robust heart failure phenotype, our TAC animals displayed the same carotid hemodynamic phenotype seen in mice with aggressive ligation. ²⁶ , ²⁹ , ³⁰ , ³⁹ Therefore, the blunted cardiac function and cardiac hypertrophy phenotypes did not translate to a loss in the hemodynamic phenotypes that were central to this study.

Last, this study did not include the use of behavioral testing assays (such as the Morris Water Maze or Y‐Maze) that are normally used to link adverse physiological changes to cognitive performance. Molecular data and cognitive batteries are key in the establishment of animal models to study neurodegeneration, and although there is limited evidence showing behavioral deficits in TAC, ²⁹ future work should aim to include cognitive assays to fully characterize the behavioral consequences of TAC and how they relate to the molecular changes that happen within cortical and subcortical structures. Given that chronic blood flow pulsatility and low brain perfusion may be a key causative bridge linking cardiovascular disease to neurodgeneration, ⁵ , ²¹ the dual incorporation of brain perfusion metrics and behavioral assays in TAC must be a crucial component of future studies.

CONCLUSIONS

Although transverse aortic constriction is traditionally used to model cardiac pressure overload, the present study highlights its usefulness for investigating peripheral flow‐dependent mechanisms linked to brain hypoperfusion and their impact on neurophysiological processes associated with neurodegeneration. Specifically, TAC leads to hemispheric differences in mitochondrial respiration and overall changes in quality control markers in the hippocampus. These findings underscore how distinct perfusion patterns, such as high blood flow velocities and pulsatility, affect mitochondrial function differently, offering new insights into the cellular mechanisms linking cardiovascular pathology and neurodegeneration.

Sources of Funding

This work was supported by the American Heart Association grant 16SDG30770015/Sarah Kuzmiak‐Glancy and 23PRE1020728/Gabriel S. Pena.

Disclosures

None.

Supporting information

Tables S1–S6

Figures S1–S6