Ll 0001 1111111p11

Time (days)

FIGURE 4. Plasma retinol kinetics in rats with various liver vitamin A stores. Model-simulated plasma tracer response curves after intravenous administration of [3H]retinol-labeled plasma to rats with low (---; liver vitamin A, 8 nmol), marginal ( ; liver vitamin A,

150 nmol), or high vitamin A status (—; liver vitamin A, 3450 nmol). Based on data from Green et al. (1987).

in the tracer response curves as well as in the final slope for rats with high, marginal, or low liver vitamin A levels. These features of the curves are related to the size of liver stores, the recycling of labeled vitamin A, and the fractional catabolic rate. As predicted by the three whole-body models presented above, the 1987 analysis revealed significant differences in vitamin A disposal rate among the three groups. Disposal rate ranged from 4.2 nmol/day in the rats with low vitamin A status to 41.3 nmol/day in those with high liver reserves; also the total time an average retinol molecule spent in plasma (mean residence time; Table I) was significantly lower in rats with high vitamin A status. Statistical analysis indicated that 90% of the variance in vitamin A disposal rate could be accounted for by variation in the plasma retinol pool size. We also found that there was a significant negative correlation between the fraction of the labeled dose in plasma at 5 days and the natural log of total liver vitamin A. This observation was extended in subsequent studies (Adams and Green, 1994; Duncan et al., 1993) that became useful for later work on the application of isotope dilution analysis to the prediction of vitamin A status in humans (Furr et al., 2005).

In two later studies, we further investigated the determinants of vitamin A utilization in rats (Green and Green, 1994a; Kelley and Green, 1998), following our observation in the 1987 paper that there is a significant effect of plasma retinol pool size on vitamin A utilization. In the two subsequent studies, data were collected on plasma retinol kinetics versus time after administration of [3H]retinol-labeled plasma as described earlier. Then model-based compartmental analysis was applied to describe whole-body vitamin A kinetics as viewed from the plasma space. In this approach, processes with similar kinetics are lumped into the same compartment, as opposed to the whole-body modeling approach described earlier in which data are obtained for various anatomical compartments. Here, the resulting model is simpler than the whole-body models. However, the same kinetic parameters may be estimated, making this method extremely useful. In the 1994 work (Green and Green, 1994a), rats were fed different levels of dietary vitamin A to affect vitamin A stores and balance. After a 41-day kinetic study, plasma tracer data were fit to a three-compartment, mammillary model (Fig. 2). The central plasma compartment, which is the site of dietary and tracer input, exchanges retinol with two extravascular compartments. One of these is a slower turning-over pool that includes retinyl ester stores, and the other is a faster turning-over pool of (presumably) retinol. After data for each rat in each of the four dietary groups were modeled, data from each group were analyzed using the multiple studies feature in SAAM (Lyne et al., 1992). With this tool, information from different subjects or animals that have been treated similarly is analyzed as one data set. Using the multiple studies feature, mean model parameters were estimated as a function of dietary intake. The results showed that the number of recyclings of retinol to plasma (recycling number; Table I) was not affected by vitamin A status and averaged 12-13 in all groups. In rats with low or marginal vitamin A status, vitamin A intake, vitamin A reserves, and plasma retinol concentration all influenced vitamin A kinetics. In rats with marginal vitamin A status (liver vitamin A, ^500 nmol), only 40% of the slow turning-over pool (compartment 3) could be accounted for by liver vitamin A. In the group with depleted liver vitamin A stores (<10 nmol), liver contained 3% of the vitamin A in the slow turning-over pool (275 nmol). See subsequent discussion about the slowly turning-over extrahepatic pool of vitamin A.

In the other study (Kelley and Green, 1998), a similar approach was used to investigate factors that influence vitamin A utilization rate in vitamin A-adequate rats under conditions of low vitamin A intake. After compart-mental analysis, multiple linear regression analysis was used to examine the impact of plasma retinol pool size, vitamin A intake, and liver vitamin A levels on vitamin A utilization. We found that, if liver stores are adequate, vitamin A disposal is not decreased to compensate for low vitamin A intake as long as plasma retinol concentration is normal. We also found that there are appreciable extrahepatic pools of vitamin A in rats, especially when liver levels are low (Green and Green, 1996) (Fig. 5). This extrahepatic pool appears to deplete even more slowly than the liver in response to lowered dietary vitamin A input. In contrast, when liver vitamin A levels are very high, the model-predicted total traced mass underestimates measured liver vitamin A levels (Fig. 5). This has implications when isotope dilution techniques (Furr et al., 2005) are used to assess very high vitamin A stores that will not be traced.

Overall, our modeling studies lead us to hypothesize that both hepatic and extrahepatic pools of vitamin A help maintain normal plasma retinol

0 0

Post a comment