Referee 2's Review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
*********

The work of Winnen et al. investigates the mechanisms of translocation by the type III secretion system (T3SS) and addresses the open question of the temporally ordered activity of effector protein. For two of the effector proteins involved in Salmonella invasion (SopE and SptP), this problem has been addressed previously resulting in the conclusion that effector proteins are translocated simultaneous but degraded at different rate by the host cell proteasome. Here, the Hardt group challenges this model and provided data for the alternative model that SptP and SipA are translocated with different rates.

The approach follows earlier work of the group that allowed to follow the depletion of an effector pool after docking to a host cell and translocation.

The work provides and interesting combination of biological experimentation and mathematical modelling.

Specific comments

1. The study requires some experimental simplification, i.e. the overexpression of the regulator HilA resulting in uniform expression of the SPI1-T3SS and effector in the bacterial population but also in higher amounts of T3SS and effector proteins per bacterial cell. The second simplification is that resynthesis of effectors is blocked by addition of the protein biosynthesis inhibitor chloramphenicol. Although these manipulations are clearly required to allow the study in this model system, they also restrict the interpretation of the results towards an infection process in vivo. These restrictions should be clearly stated.

2. The use of different effector proteins in this study and the previous study by the Galan group makes a direct comparison of the models difficult. Both groups used SptP as a 'late effector'. However, the Galan work used SopE and this work SipA as an 'early effector'. One important difference is that SopE is equally distributed in the host cell cytoplasm after translocation while SipA forms foci at the point of translocation. This means that SopE and SptP travel/diffuse over similar distances after translocation while the distance is different for SipA and SptP. This difference probably adds another factor to the equation.

3. It would be of interest to characterize the mechanisms that determine the affinity for the T3SS in more detail. The authors hypothesize that the cognate chaperone for SipA and SptP govern different affinities of the effector /chaperone complex to the T3SS. This might be experimentally tested by swapping the domains required for the recognition by the chaperone. Will SptP chaperoned by InvB be translocated more rapidly, and vice versa?

5. One technical limitation of the approach is that translocation of SipA can be followed but that only depletion of SptP can be accessed. How safe is the detection of depletion by immunofluorescence? What is the proportion of bacteria that contain SptP that is not detected by immunostaining? Suitable controls should be shown in the supplements.

6. The data for translocation kinetics for SopE have not been generated together with SipA and SptP but result from a previous study. Therefore, care should be taken with the discussion.

7. The quantification of the numbers of effector proteins and T3SS should be described in more detail

Minor comments:

Fig.1D is difficult to read, I suggest to use color.