a. Look at Figure 2. Why would the photos in part B be insufficient by themselves to form a convincing argument that CHIP28 is a water channel? Why is the graph important? b. The data in figure 2 is necessary to support the argument that CHIP28 is a water channel. Why is it not sufficient by itself to convince your fellow scientists that CHIP28 is a water channel? A 1.5 1.4 1.2 1.0 3. Time (min) 0.5 1.5 2.5 Time (mln) Flg. 2. Increased osmotic water permeability of CHIP28 RNA-injected Xenopus oocytes. After 72 hours, control-injected and CHIP28 RNA injected (10 ng) oocytes were transferred from 200 mosM to 70 mosM modified Barth's buffer, and changes in size were observed by vid- eomicroscopy (19). (A) Osmotic swelling of representative control-injected (open circles) and CHIP28 RNA-injected (tilled squares) 0ocytes. Time of rupture is denoted (X). (B) Photos of injected oocytes at indicated times. Oocytes injected with CHIP28 RNA (3 min) or control (5 min) are denoted 3/5. SCIENCE. VOL. 256 17 APRIL 1992 Relative volume Control stdIHO inhibition resulted from incubation in 3 mM HgCl, (Table 1). When the CHIP28 RNA- injected oocytes were incubated first with HgCl, and then with B-mercaptoethanol, inhibition was reversed, and the oocytes swelled and ruptured. These reagents did not alter osmotic water permeability in control- injected oocytes (Table 1). We measured the osmotic water perme- ability in oocytes injected with increasing amounts of in vitro-transcribed CHIP28 RNA. The quantity of RNA injected pro- duced corresponding increases in osmotic water permeability that reflected the quanti- ty of CHIP28 protein expressed (Table 2). However, the relative osmotic water perme- ability calculated per CHIP28 molecule in normal human RBCS (P cm'/s per molecule of CHIP28) is greater than that in oocytes expressing CHIP28 (P, 10-14 cm'/s per molecule of CHIP28) (24). This suggests that either some of the CHIP28 protein expressed in oocytes may be trapped within intracellular compartments, the organization of CHIP28 in oocyte extra- cellular membranes may be suboptimal, reg- ulatory factors may be required, or the oocyte cytoskeleton may impede more rapid swelling. The characteristics of CHIP28 are also consistent with other known features of water channels. For example, (i) the total number of CHIP28 monomers in human red cells (2 x 10) (11) is similar to the number of water channels estimated from biophysical analyses of human RBCS (2.7 x 10) (2); (ii) the 28.5-kD molecular size of a CHIP28 monomer (11) is similar to the 30-kD func- = 1.6 x 10-13 !!