The assessment of cortical hemodynamic responses induced by tubuloglomerular feedback using in vivo imaging

Figures

• 

  FIGURE 1. Experimental overview assessing the influence of loop diuretic (Furosemide) and Phlorizin on renal hemodynamics. (a) Imaging setup for laser speckle contrast imaging for real‐time high‐resolution tissue blood flow measurements. (b) Time‐averaged blood flow map of the renal surface acquired with the LSCI. The yellow star‐like shaped regions are microvasculatures with relatively high blood flow compared to the surrounding renal tissue. (c) The microvasculatures are segmented (demonstrated by black borders) for blood flow time series extraction. (d) The timeline for the whole experiment. The recording takes place during the drug infusion.

• 

  FIGURE 2. Illustration of the analysis from segmentation to a time‐frequency spectrogram. (a) A close‐up image of a star vessel that can be imaged at the renal surface. The images of vessels undergo a segmentation process, from which (b) blood flow index time series can be extracted. (c, d) A closer look at the B reveals that the TGF oscillations during the control period (black) disappear with the furosemide infusion (red). (e) Power spectrum shows a prominent peak at around 0.03 Hz associated with TGF during the control period. (f) The spectrogram of superlet analysis for the control period (left) shows a sharp localization of the TGF frequency around 0.03 Hz sustained for 30 min that disappears completely with the furosemide infusion (right).

• 

  FIGURE 3. Three types of TGF‐induced responses in different vessels. The top panel represents time‐averaged power spectra: One peak around 0.03 Hz corresponds to a single frequency of TGF oscillation (a); a TGF peak cannot be detected (b); Multiple peaks within the narrow frequency band correspond to TGF oscillations with different frequencies (c). The bottom panel represents the time‐frequency spectrogram matching to power spectra on (a–c): A stable TGF frequency over the observation time (d), No observable TGF frequency (e); Multiple TGF frequencies within the observation time (f).

• 

  FIGURE 4. Eliminating TGF with furosemide administration in normotensive Sprague–Dawley rats. (a) An example time series taken from one vessel shows that the TGF oscillations present during the control period (black) disappear after furosemide infusion (red). (b) Carpet plots show TGF‐mediated blood flow oscillations across many vessels, with the color range representing filtered blood flow index. Oscillations are strong in animals 1–3 and weak in animals 4 and 5 in control (first column). Regardless, the TGF oscillations are eliminated in all animals, as demonstrated by the lack of oscillatory patterns (second column). The plots only show 600 s of 30‐min recordings to visualize the oscillatory patterns better.

• 

  FIGURE 5. Increased local blood flow and the TGF dynamic changes induced by furosemide administration. Top panel: Box plots represent the median (central red line), the 25th (bottom edge), and 75th (top edge) percentiles of the analyzed TGF metrics: (a) mean blood flow index (BFI), (b) standard deviation of filtered TGF time series (Sigma), and (c) area under the curve (AUC). Reduced Sigma and AUC metrics represent the elimination of TGF oscillations. (d) Three metrics are plotted for every observed microvessel. Top row: 3 animals show a good separation in metrics between the control and the furosemide condition. Bottom row: Metrics of control and furosemide are clustered together. *p < 0.001; **p < 0.005. The units for BFI, Sigma, and AUC are arbitrary. μMean values for control and furosemide data points.

• 

  FIGURE 6. TGF oscillations in segmented vessels under phlorizin administration in normotensive Sprague–Dawley rats. (a) Exemplary blood flow time series demonstrating sustained TGF oscillations for 600 s in both control and phlorizin. Note the different y‐axis in the blood flow index (a.u.). (b) First column (control): TGF band‐filtered signals show consistent TGF oscillations over the observation time. In the second column (phlorizin): TGF oscillations are still well pronounced. The plots only show 600 s of 30‐min recordings to visualize the oscillatory patterns better. μMean values for control and furosemide data points.

• 

  FIGURE 7. Different vessels can exhibit different amplitudes of the TGF frequency peak. (a) The time‐averaged power spectra of 40+ vessels show that vessels can have heterogeneous TGF peak amplitudes during the control period. After the phlorizin infusion, the dominant TGF frequency peak around 0.03 Hz remains well‐visible but exhibits variations in power across vessels. (b) Vessels show a dominant TGF frequency between 0.025 and 0.03 Hz in varying degrees of power in both conditions.

• 

  FIGURE 8. Decreased nephron blood flow and TGF dynamic changes induced with phlorizin administration. Top panel: Box plots represent the median (central red line), the 25th (bottom edge), and 75th (top edge) percentiles of the analyzed TGF metrics: (a) mean blood flow index (BFI), (b) standard deviation of filtered TGF time series (Sigma), and (c) area under the curve (AUC). (d) Three‐dimensional scatter plot of the three metrics for each microvessels. (d) Top row: 3 animals show a good spatial separation across all three dimensions. Bottom row: control and phlorizin clusters show similar metrics in 2 animals. **p < 0.0005. The units for BFI, Sigma, and AUC are arbitrary. μMean values for control and phlorizin data points.