This study used aging F344/BN rats to temporally relate CSF volume, production and turnover with brain Aβ40 and Aβ42 accumulation. A linear increase in ventricular and total cranial CSF volumes with age and a non-linear relationship between CSF production and age were found. The CSF turnover rate estimates how many times the entire CSF volume is replaced per day and was found to increase from 3 to 12 months and decrease from 12 to 30 months. The significant increase in CSF production from 3 to 12 months and its subsequent decrease to 30 months highly influenced the shape of the turnover curve. There was a rapid increase of cortical and hippocampal Aβ40 and Aβ42 to about 12 months, after which the rate of increase in Aβ40 slowed. Aβ42 continued on the same rising slope to 20 months after which the slope decreased. Both Aβ40 and Aβ42 approached a steady state concentration in the brain by 30 months. The reversal of the Aβ40/Aβ42 ratio in the brain with advancing age was an interesting finding.
The cranial CSF spaces were imaged by MRI. Fiducial points were placed within these spaces to construct a 3D model of the lateral, third, and fourth ventricles, and the subarachnoid space around the brain to calculate total CSF volume. Some measurement error may have been introduced by the observer; however, it would have been consistent throughout the age groups, and our results are within the range of measurements reported by others for ventricular and total cranial CSF volumes [32–35].
Determining the causes of Aβ accumulation in the brain with aging may be crucial to the development of effective AD therapies. Aβ accumulates as a result of an imbalance between production and clearance . Studies suggest that overall brain Aβ accumulation in aging and in AD is not due to an increase in Aβ production, but rather a decrease in Aβ clearance from the brain [22, 31, 36], although some increase in production cannot be completely excluded by these studies. Several pathways for Aβ clearance exist: i) active transport via receptors at the blood-brain barrier (BBB) and blood-CSF barrier (BCSFB), ii) in situ enzymatic degradation and iii) CSF bulk flow and turnover.
Previous reports have shown Aβ clearance pathway alterations with age. There is a significant decrease in Aβ efflux transporter expression with age and an increase in influx transporter expression at the BBB [31, 36]. Interestingly, the opposite trend is true for Aβ transporter expression at the BCSFB, perhaps a compensatory response to the overall decrease in BBB Aβ clearance . Studies have also shown decreases in Aβ-degrading enzymes with age . CSF turnover, therefore, may become increasingly more important for metabolite clearance in the aging brain .
CSF turnover is a modulator of CSF and brain protein concentrations [9, 10]. Although it appears that there is no association between Aβ accumulation in brain parenchyma and CSF turnover rates in the young rats, there may be an inverse relationship between the two in the elderly rats, as part of an age-related multi-factorial decrease in Aβ clearance via the BBB and CSF systems. Clearly, Aβ concentrations in brain increase while CSF production and turnover increase up to 12 months of age. Preston (2001) also found that Aβ begins to accumulate before there is any decrease in CSF production and turnover . However, CSF turnover in association with other age-dependent Aβ clearance defects may contribute to the overall increase in Aβ brain concentrations with advancing age. This temporal association between Aβ and CSF dynamics in the older rats does not prove a cause and effect relationship, but one may argue that the decrease in CSF turnover is a contributing factor. Improving CSF turnover by shunting these rats may show a causal relationship between the two if shunting lowers the rate of Aβ accumulation in aging. Such a study would require measurement of brain, CSF and blood Aβ concentrations in both shunted and sham-operated age-matched controls.
The procedure for measuring CSF production, V-C perfusion, is extremely invasive, and the rats experienced high mortality rates, particularly in the elderly. However, the number of rats at 3 and 12 months was sufficient for showing a significant rise in CSF production, and the decrease from 12 to 30 months fits with existing studies [3, 9, 10]. In a study by Preston (2001) using Fischer 344 inbred rats, CSF production was shown to increase from 3 to 19 months (1.21 ± 0.27 to 1.48 ± 0.53 μL/min) and then significantly decrease to 30 months (0.65 ± 0.16 μL/min). A 12 month measurement was not included in this study, but judging from our data (see Figure 3A) it would likely have been higher than the 19 month production rate . Because this study and the Preston study used similar rat strains and found similar production results, we believe that the technical difficulties associated with our methods are not substantive.