General comment about the Mentor feedback: I appreciate the feedback and the time invested to evaluate my review. I agree with all comments and I have changed my review accordingly.

Second version of the review:

The paper seeks to understand why the introgression of SbMATE into Al-sensitive sorghum genotypes does not necessarily increase Al tolerance. By performing QTL mapping (both biparental mapping and association mapping) the authors detect a region in chromosome 9 that harbors three transcription factors, two of which (SbWRKY1 and SbZNF1) were shown to activate the SbMATE expression. Various experiments were performed in order to show that the expression of these transcription factors was different between Al-tolerant and -sensitive germplasm and to demonstrate that greater number motifs in the SbMATE promoter (depending on the number of MITE repeats) recruited more transcription factors that eventually led to higher SbMATE expression and greater Al tolerance.

The paper is well written and the methodology is adequate to answer the main question. The conclusion fits the data. The paper provides important evidence to explain why some sorghum lines remain as Al-sensitive even after superior alleles of the major Al-tolerance gene (SbMATE) is introgressed. This information is highly significant to breed sorghum for greater tolerance to Al stress, which is a major stress for plant growth worldwide.

Major points that should be addressed by the authors:

1. The introduction is short and there is room to add papers where the role of cis-elements in Al tolerance was investigated (Yokosho et al 2016 Plant Physiology 172: 2327-2336 doi: 10.1104/pp.16.01214; Li et al 2018 New Phytologist 219:149-162 doi: 10.1111/nph.15143). These papers should also be mentioned in the discussion.

2. The qRT-PCR used only 18S rRNA as endogenous control. Authors did not provide information that the expression of that gene is similar in conditions with and without Al stress.

3. Figure 1C: If BR012 is Al-sensitive with low SbMATE expression, I would expect that SC566-NIL (which has the BR012 background) should also have a low SbMATE expression compared to BR012. Clearly the SbMATE expression is lower than in SC566 but still much higher than in BR012. Do the authors agree with this interpretation which indicates that this result appears inconsistent?

4. Figure 3D: As stated in the figure 3D, SbMATE promoters with four and one MITE repeat were compared. However, for SbZNF1 and SbWRKY1, the specific activity is only up to two-fold higher in promoter with four repeats (black bars) compared with promoter with one repeat (white bars). Shouldn't that activity be higher in promoter with four MITE repeats?

5. Figure 4C: It seems clear the impact of transcription factors in SbMATE expression. However, in Figure 4D, that impact does not seem to reflect in greater difference in Al tolerance. Genotype Z+/W+ has 2.5 times more SbMATE expression but only around 15% more RNRG. Why?

6. Figures 4C and 4D: Shouldn't the bars have error bars? Or perhaps the graphics could show the least significant difference (LSD).

7. Discussion, 2nd paragraph: "Our results indicate that SbMATE and SbWRKY1 are coregulated (r = 0.3; P = 0.08; SI Appendix,Fig.S6)". Isn't the P value too high (0.08)?

8. Discussion, 4th paragraph: "The cDNA sequences of SbWRKY1 and SbZNF1 alleles in Al-tolerant and Al-sensitive lines were found to be identical". What about the promoter of the SbWRKY1 and SbZNF1 genes in Al-tolerant and -sensitive genotypes? Do you have any information about variability in their promoter regions?

9. Discussion, 5th and 6th paragraphs: The saving of "cost to the plant" is explained in two different ways. In Al-sensitive lines, the transcription factors will be highly expressed in the absence of Al3+ but some repressor will block the SbMATE expression. In Al-tolerant lines, the expression of transcription factors is reduced in the absence of Al3+ which will result in lower citrate efflux. The idea of a repressor could be more substantiate if the sequence upstream the MITE repeats show some motif for that repressor.

10. The discussion focus mainly on explained that the number of MITE repeats will increase the number of cis-elements for the transcription factors. However, the message should be clearer that does not matter is the SbMATE promoter has higher number of MITE repeats if the genotypes has inferior alleles for both transcription factors. That is the difference between SC566 and BR012 that have the same four MITE repeats but show different Al tolerance levels.

Minor points:

1. It would be interesting if the title includes information about the transcription factors SbWRKY1 and SbZNF1 being important to protect sorghum roots from aluminum toxicity. The way it is, it seems that only the repeat variants for the SbMATE transporter are important.

2. The introduction is mainly focused in sorghum. Does the same problem (Al tolerance not being improved after the introgression of superior allele) happen in other species where organic acid transporters are important for Al tolerance? Or is this problem specific to sorghum?

3. The supplementary material is confusing starting with figures, then references and tables. Can the tables be added before the references? Can the legends be in the same pages as the figures?