My Research

    Renal autoregulation is the process by which blood flow in the kidney is kept constant despite changes in systemic blood pressure. This process is mediated by two mechanisms: the  tubuloglomerular feedback (TGF) and the faster myogenic reaction (MYO). TGF resonates at the frequency range of 0.02-0.05Hz, while the MYO resonates at 0.1-0.3Hz. The susceptibility to renal injury in hypertension and the progression of chronic renal failure is exacerbated by autoregulatory dysfunction. Hence, it is necessary to characterize the effects of drugs and disease functional state of the TGF and myogenic mechanisms in order to help prevent this type of damage. Generally, the progression of disease is subtle and transient. Traditional time invariant signal processing techniques that are often used to address aforementioned aims are often not able to discern these subtle changes in dynamics. Unfortunately, until recently, these time invariant techniques are the only ones available. In order fully characterize the subtle changes in dynamics of renal autoregulation, time-varying signal processing techniques must be used. For example, time-varying spectrums can be used to determine how the dynamics of autoregulatory mechanisms in the frequency domain changes with time. Therefore, these techniques will be able to detect the subtle and transient nature of renal autoregulation and may potentially be used as an early diagnosis of kidney related diseases. An example of this is shown below:

    The figure to the left shows kidney blood flow data from a rat model. In this experiment, systemic blood flow was reduced via an aorta clamp at approximately 80 seconds. This is shown on the top plot with an initial reduced kidney blood flow. Due to autoregulation, the blood flow recovered after approximately 20 seconds back to the baseline. At approximately 150 seconds, the aorta clamp was released suddenly, simulating a step increase in systemic blood pressure. This was seen on the kidney blood flow as a sudden increase. Due to the step increase input, the system does not have enough time to compensate for the increase and thus will resonate. The blood flow plot shows this oscillation as a slow wave overlay with a faster frequency. This seems to indicate the presence of the slow TGF and the faster MYO mechanisms. If one was to use traditional time invariant techniques, such as the fast Fourier transform, the result will be the plot on the left. Although this type of analysis will give great information on the frequency components of the signal to be analyzed, it will not provide any information on the time varying dynamics of the signal. The plot shown on the center is a time varying spectrum calculated using the Smoothed Pseudo Wigner-Ville technique. This type of time varying method can not only provide information on the frequency components of the system, but also time information. For example, this plot will show that during the time that the aorta is clamped (approximately 80 to 180 seconds), the faster mechanism is somewhat attenuated. Although one can see this qualitatively on the blood flow plot, the spectrum will show this in a more quantitative way.

    In the literature, there is no consensus on the use and effects of anesthetics on the dynamics of renal autoregulation. To this end, we compared the time varying dynamics of autoregulation between rats using 3 different anesthetics. Both Sprague-Dawley (SDR) and spontaneously hypertensive rats (SHR) were used, as these are two models commonly used in the study of renal hemodynamics. Furthermore, nitro-L-argining-methyl ester (L-NAME) was also used in this study, as this drug can increase the gain of the TGF and MYO mechanisms. Data from this experiment was analyzed using the complex-demodulation method (CDM). This recently developed method has the advantage of being able to calculate instantaneous spectral amplitude at all points in time. Using this method, one can see precisely how various frequency components of a signal can change with time. An example of this is shown below:

The figure to the left is the same trace as the one shown above. The top panel of the figure shows systemic blood pressure. One can seen exactly when the aorta clamp was applied in relation to the kidney blood flow shown on the middle panel. The last panel shows the result from the CDM calculation. The blue line in this case is the instantaneous spectral amplitude for TGF, while the red line is for MYO. As can be seen, the aorta clamp attenuated the MYO response, but upon the clamp's release the MYO response will be returned to normal. Furthermore, the clamping and its release caused a transient increase in the TGF response. However, after some time, this reponse will be reduced back to the baseline levels.

This experiment was performed on animals anesthetized using halothane, inactin, or isoflurane. We calculated the average power of the two renal autoregulatory mechanisms 150 seconds immediately after the release of the aorta clamp. This was done because the resonance oscillations tend to be transient in nature. If the entire trace was used to calculate the average, this average will be distorted by the signal that does not contain the oscillations we are interested in.

The results from this analysis is given on the two plots below:

    The two bar graphs above shows the average MYO and TGF response immediately after the release of the aorta clamp using the CDM method. Some of the important findings is that after the administration of LNAME, isoflurane anesthetized rats shows a large increase in MYO as compared to the other two anesthetics. In SHRs, the TGF response of halothane anesthetized animals is much greater than in other anesthetics. Administration of LNAME lead to a decrease in TGF in halothane anesthetized rats, while giving an increase in MYO. In general, these results show that the renal autoregulatory hemodynamics from animals under different anesthetics are very different from each other.

    In addition to these responses, some animals also exhibited a large, MYO predominant response. An example of this is shown below:

    As can be seen here, very little of the typical slow TGF oscillations was seen. However, the faster oscillations shown here has some periodicity to them, as seen as some sort of envelope in the amplitude of the signal. It is unclear as to exactly what this change in amplitude of the oscillations is. It is important to note here that although TGF was not observed here, it does not mean that it is inhibited. Anatomically, the TGF mechanism is located downstream of the faster MYO response. If the MYO response is upregulated, it will dampen any change in systemic blood pressure sensed by the TGF. In other words, if the MYO response is able compensate for the changes in systemic blood pressure, there is TGF will not be stimulated to turn on.

    The frequency of occurrance of these MYO predominant responses are given below:

    Of interesting note here is that administration of LNAME on inactin anesthetized animals resulted in similar number of occurrences of the MYO predominant response as isoflurane anesthetized animals without LNAME. Furthermore, inactin anesthetized SHR without LNAME also had a similar number in occurrence as inactin anesthetized SDR with LNAME. This could be explained with by free radical molecules. LNAME provides its action by inactivating the production of nitric oxide, which is a vasodilator. Therefore, its inhibition will lead to more vasoactive vasculature. The current thought is that isoflurane and SHR results in an increase in systemic levels of free radicals. In general, these free radical molecules tend to quench the effects of nitric oxide, therefore simulating the effects of LNAME. Further evidence of this is shown in isoflurane anesthetized animals that are treated with LNAME. It could be seen that this type of response is seen very often in such a combination. This could be due to LNAME reducing an already reduced amount of nitric oxide, therefore leading to extremely vasoactive vessels.

    In conclusion, these studies show that steady state autoregulation was not affected, as all of the cases show that the blood flow in the kidney recovered back to baseline. However, as seen with the results using CDM calculations, the relative contribution of the two renal autoregulatory mechanisms are changed. Therefore, one must take the choice of anesthetic into careful consideration when comparing studies on renal autoregulatory hemodynamics. It is important to note that although a difference in these dynamics were found, one can not say for sure which is the best anesthetic to use, as the dynamic state of renal autoreagultion on unanesthetized animals has yet been characterized.

Future Work:

    As explained above, the choice of anesthetics has a profound effect on the dynamics of renal autoregulation. Therefore, the state of these dynamics under unanesthetized animals needs to be characterized. There are two general methods to go about doing this. First is through the use of telemetry methods. In this method, flow sensors are attached to the renal artery of an animal while anesthetized. After the recovery of the animal, the blood flow measurements can then be taken from the animal while it is awake. Although this type of technique has been used in the past to characterize steady state renal hemodynamics, this type of experimental setup is less useful for the detection of step response dynamics of renal autoregulation. This is because it is difficult to control systemic blood pressure in a step-wise fashion using this method. Therefore, one must turn to the second technique, which is via the use of de-cerebrate animals. In this experimental setup, the brain stem of the animal is cut in such a way so that the animal will no longer feel any pain, but still be alive. Using these conditions, anesthesia will not be necessary. Therefore, renal blood flow measurements taken from this setup will not be corrupted by the effects of anesthetics.

    The second line of research involves the detection of free radicals. As stated above, it was proposed that isoflurane exhibits its action through the accumulation of free radicals in the body. To test this idea, we can use an assay to check for the amount of free radicals in the kidney of rats under different anesthetics. Isoflurane would be expected to have a higher amount of free radicals.