In the manuscript “Cab45 controls lysosomal hydrolase targeting”, Tran et al applied a novel hiSPECS technique to uncover a set of lysosomal cargo molecular that are hyper-secreted by HeLa cells depleted for Cab45/SDF4 protein. Western blot and RUSH assay were used to confirm the abnormal sorting of PSAP in Cab45 KD cells. The authors also claimed that depletion of Cab45 causes defects in lysosomal positioning and that this novel function of Cab45 is independent of the proteins' ability to bind Ca2+. For the most part, the data is quantified and the manuscript is well written. Unfortunately, in most of the quantification, there is no significant difference between WT and Cab45 KD (particularly Figs 2 and 3), which questions the authors claim whether “Cab45 is critical for lysosomal transport”. At this point, the data presented in the manuscript does not fully substantiate this claim. The authors do acknowledge that this subtle effect could be due to the longer secretion period for mass spectrometry and a shorter one for WB. In which case, they should use the same time course for WB analysis as well. The effect on lysosomal positioning is exciting and novel. Unfortunately, quantification presented in Figure 5 (B and D) indicated that this defect might not be rescuable by the expression of Cab45wt, questioning the connection between Cab45 depletion and lysosome positioning.

Specific points:
1. Authors should tone down their claim that Cab45 is critical for the packaging of lysosomal cargo. Presented data indicate a rather minor effect. What fraction of lysosomal cargo is actually missorted in Cab45 KD or KO cells?
2. hiSPECS is an interesting and sensitive technique, but it is not clear if this technique is providing unbiased results for all secreted glycoproteins. Specifically, ConA precipitation could be more efficient for the precipitation of glycoproteins with higher content of mannose residue. Which proteins are less secreted in Cab45 KD cells? Authors previously identified LyzC and COMP as Cab45 clients. COMP is a glycoprotein with two N-glycosylated chains. Is it significantly decreased in Cab45 KD analyzed by hiSPECS? Providing complete MS data may clarify some of these questions. Description of hiSPECS method applied to HeLa cells should be more detailed.
3. To avoid off-target defects, the MS/KD data should be controlled by re-expression of siRNA-resistant Cab45.
4. From Figure 4, the authors interpret that PSAP exits the Golgi faster compared to its exit in WT cells. However, PSAP appears to arrive at the Golgi twice as fast and in Fig 4F, there is no significant difference between WT and Cab45 KO! So the interpretation that Golgi exit is faster when Cab45 is depleted is not very accurate. GalT is a trans-Golgi, not TGN marker; therefore the RUSH results should be reworded as the exit from the Golgi.
5. The claim that “Cab45 is critical for lysosomal transport” can be bolstered by showing whether Cab45 depletion affect the function of lysosomes. A suggested experiment is to test the functionality of the lysosomal degradation using a substrate like DQ-BSA
6. Figures 5 B and D should be combined to assess the effect of Cab45wt rescue on the lysosome positioning.

The authors addressed and explained the reviewer's comments well. Validation of the secretion phenotype of Cab45 deficient cells pulldown experiment has been performed using PSAP antibody, which is impressive. Visualization of final localization of RUSH PSAP-mCherry by immunofluorescence microscopy in cells expressing Rab7-GFP after 30 and 180 min biotin addition is a very good approach. The manuscript will be suitable for publication after several important points (see below) are addressed in the discussion.

Comments:

1. The authors have shown that PSAP and PGRN are secreted in higher amounts when Cab45 or SORT, or M6PR are depleted. While the secretion of PSAP is significantly increased, all the quantification data for PGRN show that its secretion is not statistically significant. The one place where PGRN levels should have shown no statistical difference is S1A, and there it is depleted! This should be addressed in the text.
2. In Figures 3B and 3C the authors should not use the term "modest increase" for statistically insignificant data. Only PGRN has increased in SORT KO, and Cab45 depletion in CI-M6PR KO increases PSAP secretion. This result section should be rewritten to state the data as it is.
3. The WB data of Figure 3D, 3F is not very clean. PSAP and PGRN band intensity is very low. It would be better to provide a better WB image.
4. It is very interesting that Figure 5D shows a significant colocalization difference of PSAP with LAMP1 between Cab45KO and Cab45wt-rescue cells but no significant difference between Cab45KO and Control HeLa cell. This should be addressed in the text.
5. One of the points noted by the authors is the abundance in the number of lysosomal hydrolases secreted in Cab45-depleted cells. Yet, this does not appear to affect lysosomal degradative activity. The authors conclude that Cab45 is critical for lysosomal transport and positioning. However, the data suggests that Cab45 is dispensable for lysosomal function in HeLa cells! This should be acknowledged in the discussion.