A Transmembrane Segment Determines the Steady-state Localization of an Ion-transporting Adenosine Triphosphatase
Lisa A. Dunbar, Paul Aronson, and Michael J. Caplan


    The H, K-ATPase is found in the apical membrane of resting stomach parietal cells, while the Na, K-ATPase, whose alpha subunit is 63% homologous to the H, K-ATPase, is located in the basolateral membrane.  Previous studies have shown that Na, K-ATPase alpha subunit chimeras, composed of the first 519 amino acids from the H, K-ATPase, are localized to the apical membrane.  In this experiment they created more H, K-ATPase/Na, K-ATPase chimeras to try and isolate the specific sequence of the H, K-ATPase alpha subunit that contains the sorting information.  They also tested whether the TM4 (fourth transmembrane region) domain of the alpha subunit interacted with the GSL-rich domain of the apical membrane to achieve localization.  Lastly the authors performed a functional study on the enzymatic activities of the chimeric ion-transporter.

Data Presented

Figure 1.  Here we see that cells containing chimera I and II have a similar pattern of staining as the negative control Na, K-ATPase.  Cells infected with chimera III has a very different pattern of staining then the control and the other chimeras and the negative control.  Figure 1I has a thicker ringed staining pattern.  Figure 1K shows a staining pattern that is concentrated in the upper portion of the photo.  Unfortunately this figure has no positive control so I am reluctant to say that this is the staining pattern of the apical membrane.  Had the authors included a photo of the staining pattern of endogenous H, K-ATPase I would be able to say more about this figure.  As it stands, the only conclusion I feel comfortable making about this figure is that chimera III stains differently than I, II and Na, K-ATPase and that it is possible that this is because it is localized in the apical membrane.  Chimera III was composed of the 324-519 amino acid region of the H, K-ATPase (along with the 85 amino acid epitope tag).  This region was unique to chimera III and is responsible for the difference in staining.

Figure 2.  Cells containing chimeras IV and VI had the same staining pattern as the negative control, while chimeras V and VII showed staining patterns that differed from the negative control.  Chimera IV had 356-519 amino acids from H, K-ATPase and chimera VI had ~324/325-329* of H, K-ATPase.  It would seem that these portions of the H, K-ATPase due not cause a difference in staining from the negative control.  Again a positive control was not included, which becomes more worrisome because the staining pattern seen in chimeras V and VII is different from the staining pattern seen in chimera III.  Chimera V contains a 324-356 amino acid region of H, K-ATPase and chimera VII contains ~329-356*.  This figure leads me to conclude that the region of 329-356 amino acids (H, K-ATPase) is causing the different staining pattern from the negative control.

*Chimeras VI and VII have ambiguous sequences.   The oligo sequences that were used to make these chimeras are not well defined.

Figure 3. The Na, K-ATPase beta subunit shows both the staining patterns of the chimera V and Na, K-ATPase.  I think that the authors should have included a figure of just the Na, K-ATPase alpha subunit, so I could compare that staining pattern to the beta subunit, but from the other figures I can see that they are similar. From this figure I would say that the beta subunit is associating with both the chimera and the endogenous Na, K-ATPase.

Figure 4.  This figure is a comparison of the sequences of the Na, K-ATPase and the H, K-ATPase.  There is a lot of homology between the two sequences, which makes me wonder how the few differences in amino acid sequence could lead to the different localization of the proteins.  Figure A could be improved by adding numbers to the amino acids. Chimeras I-V were all described by number and it would be nice to see numbers, just for reference.  Also in the figure caption it states "The amino acids corresponding to the junction points of the chimeras are shown by arrowheads."  It is not clear which chimeras they are talking about.  If they are talking about chimeras VI and VII they should say that, or maybe add to the figure by showing the exact sequence used for these chimeras.  Figure B does not add much to the paper.

Figure 5.  This figure shows the detergent solubility of chimera H519N.  Alkaline phosphate is a positive control, showing the fractions that would contain a GPI-linked protein, and Na, K-ATPase is a negative control.  The chimera H519N more closely resembles Na, K-ATPase.  These results indicate that the H, K-ATPase ion transporter does not associate with GPI as a mechanism for localization.

Figure 6.  Chimera VIII shows a different staining pattern than the negative control, once again no positive control was used.  This is a similar staining pattern to chimera V.  This is an interesting result because chimera VIII does not have the TM4 domain that seemed to be responsible for the staining pattern for chimera VII (figure 2m and 2o).  Chimera VIII contained ~324-329 a.a.(the second ectoregion) and ~356-519 a.a.(part of the cytoplasmic loop).

Figure 7.  Cells containing chimeras III, V, and VII all died, but these were all apically located.  Chimera VIII, which is also apically located, survived.  The results of chimera VIII seem to indicate that location is not a factor for the lack of ouabain resistance in the three chimeras.  The chimeras that did not confer resistance to ouabain were all missing the 329-356 a.a. domain of the Na, K-ATPase.  It seems likely that this region is important for conferring ouabain resistance.  In figure 7b the pH is lowered on apical region by the transfection with chimera III.  I don't really understand this result.  The only explanation seems to be that chimera III acts as an H+ pump and the increase in H+ pumps causes a decrease in pH.  But in the next graph the addition of ouabain causes the pH to rise to average levels.  I guess this result would be more clear if I knew the mechanism of ouabain's inhibition of Na, K-ATPase.  If chimera III is functioning as an H, K-ATPase, why does ouabain, which inhibits Na, K-ATPase, inhibit H, K-ATPase.  What is it's mechanism of inhibition?  Does it rely on a domain that is closer to the COOH-term?

The researchers asked what specific sequence of the H, K-ATPase manifests sorting information.  The result of the this experiment was that the region of ~329-356 a.a and having both the ~324-329 a.a. and 356-519 a.a regions will cause the chimeric protein to be localized apically.  As I stated before, because there was no positive control I'm hesitant to use the word "apically."  This result shows that there is not one single amino acid sequence that functions as a localization sequence.  It is possible that the chimeras that were found in the apical region were misfolded and this is the reason that more than one signal directs them to the apical region.  Functional tests were performed, but not to an extent that proved that the apically located protein with the ~329-356 a.a. of the H, K-ATPase was functioning.  A good experiment might be to see how long the chimeric proteins stay in the apical membrane as opposed to a wildtype H, K-ATPase.  This could be done using pulse-chase.  If they are quickly degraded, it is likely that they are misfolded.  Also, if chimera III, V and VII had no Na, K-ATPase activity, then they should be tested to see if they have H, K-ATPase activity as well, especially if there is a drug similar to ouabain for H transporters.  Another area of study would be to see if there is any interaction with the H, K-ATPase and other membrane entities (similar to the detergent-solubility fractionation).  This experiment could use immunopercipitation or a yeast two-hybrid system to see if the ion transporter is complexed with any other proteins.  But I think the most promising experiments are the ones they proposed of "determining the specific resides of the TM4 and its flanking regions, which are responsible for apical localization."  I would use this same chimeric method and to make even more involved chimeras I would repeat the oligo ligation method.