Mycolic acids as diagnostic markers for tuberculosis case detection in humans and drug efficacy in mice

Mycolic acids are attractive diagnostic markers for tuberculosis (TB) infection because they are bacteria-derived, contain information about bacterial species, modulate host–pathogen interactions and are chemically inert. Here, we present a novel approach based on mass spectrometry. Quantification of specific precursor → fragment transitions of approximately 2000 individual mycolic acids (MAs) resulted in high analytical sensitivity and specificity. We next used this tool in a retrospective case–control study of patients with pulmonary TB with varying disease burdens from South Korea, Vietnam, Uganda and South Africa. MAs were extracted from small volume sputum (200 µl) and analysed without the requirement for derivatization. Infected patients (70, 19 of whom were HIV+) could be separated from controls (40, 20 of whom were HIV+) with a sensitivity and specificity of 94 and 93%, respectively. Furthermore, we quantified MA species in lung tissue of TB-infected mice and demonstrated effective clearance of MA levels following curative rifampicin treatment. Thus, our results demonstrate for the first time the feasibility and clinical relevance of direct detection of mycobacterial lipids as biomarkers of TB infection.

Thank you for the submission of your manuscript " Mycolic acids as diagnostic markers for tuberculosis case detection in humans and drug efficacy in mice " to EMBO Molecular Medicine. We have now heard back from the three referees whom we asked to evaluate your manuscript. You will see that they find the topic of your manuscript potentially interesting. However, they also raise significant concerns on the study, which should be addressed in a major revision of the manuscript.
In particular, reviewer #1 feels strongly about the lack of detailed descriptions of the technological aspect but also regarding the form of the paper. This last concern was somehow shared by reviewers #2 and #3 who found the manuscript rather unstructured in parts, with missing or inadequate information. Importantly, reviewer #1 questions the choice of the BCG strain while reviewer #2 is unsure about how quantitative the measure of MAs is.
Given the balance of these evaluations, we feel that we can consider a revision of your manuscript if you can convincingly address all issues that have been raised in a detailed point-by-point letter and modify your manuscript accordingly, within the time constraint outlined below.
Revised manuscripts should be submitted within three months of a request for revision. They will otherwise be treated as new submissions, unless arranged otherwise with the editor.
I look forward to seeing a revised form of your manuscript as soon as possible. This paper is strong on basic science, but light on detail, background and choice of literature references. There are insufficient data supplied to allow proper assessment and reproduction of the results. For example, on Page 3, for sputum profiles it is stated that "four representative samples are shown in Fig. S4" but there are no such profiles at that location. There are a number of problems with the references; for example, the following are inappropriate in certain contexts: Tonge 2000, Butler 2001, Erht and Schnappinger 2007, Layre et al. 2009. It is not appropriate to provide detailed criticisms of minor points, but the following matters are important.
The claimed sensitivity and specificity of 94 and 93%, respectively, is excellent and there is no need to doubt its authenticity. However, to allow verification of the procedure, it is necessary to provide precise coherent data to enable others to repeat the work. Figure 1 is not ideal, as BCG (presumably Pasteur) (Fig. 1A) is a poor example, in that it lacks methoxymycolates; use a representative M. tuberculosis strain. Again, in Fig. 1B, C76 mycolic acid is a bad example; it would be better to show the C80 component, ensuring that the mycolic acid structure is correct. As noted above, no sputa mass spectra are included for inspection; ideally Fig. 1 should have included a positive and negative sputum example.
It is claimed that the procedure allows differentiation of a range of mycobacterial strains, using cluster analysis. However, it is difficult to clearly understand the profiles on which these results are based (Fig. S4). All these figures must be explained much more clearly, particularly pointing out which signals correspond to the different mycolic acid classes.
Referee #2 (Comments on Novelty/Model System): The authors convincingly demonstrate the utility of using a sub-class of mycolic acids that can be extracted directly from tubercle bacilli in the sputum of smear-positive TB patients and in the lungs of acutely infected mice as a marker of TB infection. Therefore, this novel Mtb-derived biomarker has potential applicability both in pre-clinical and in clinical studies of anti-tubercular drug efficacy.

Referee #2 (Other Remarks):
There is an urgent need for new tools for detecting active TB infection and for use as biomarkers to monitor responses to new TB drugs in human disease and in animal models of TB infection. This manuscript makes some progress towards achieving these goals by establishing the potential utility of mycolic acids (MAs) as biomarkers for detecting TB in humans and for monitoring the effects of chemotherapy in infected mice. In this study, the authors carried out an exhaustive lipidomic analysis of MAs in the CMN family of Actinomycetes thereby producing an extremely valuable resource of information on this important class of lipids. They then apply this information to identify a subclass of MAs that can be extracted directly from TB patient sputum and mouse lung and used as a highly specific signature MA marker to detect acute TB infection in humans and mice. This is a well written manuscript that reports interesting and potentially significant findings. However, the authors should address the following points: 1. Pg. 4, paragraph 3, and Fig. 3B. What was the lung bacillary load in the "MTB-infected" animals shown in Fig. 3B? What was the limit of detection in terms of colony-forming units per lung? Moreover, how does the sensitivity of the MA-based assay compare in detecting M. tuberculosis in sputum vs. mouse lung tissue vs. in vitro culture? 2. MA analysis may provide a potentially useful bacterial biomarker for monitoring TB drug efficacy in human disease or animal models, thereby providing an alternative to the gold-standard CFU count currently used in pre-clinical and clinical studies. Although the authors convincingly demonstrate that the MA signal can be used to differentiate active TB from non-active disease in humans and acute infection from no infection or drug-cured infection in mice, it is not clear how quantitative of bacillary load this measure actually is. What is the limit of detection in mouse lung or human sputum? Moreover, is there a linear relationship between MA level and CFU count?
3. The sensitivity and specificity of an MA-based assay was shown to approach that of smear microscopy. However, the smear microscopy method for TB detection has been surpassed by other methods with significantly superior sensitivity, such as GeneXpert. These newer methods should be mentioned since any new diagnostic method will have to compare with these rather than smear microscopy in order to have an impact on TB diagnosis. Fig 2C, numbered dots correspond to patient IDs in table 1, but this doesn't become clear until the discussion. I suggest to include this information in the legend. A sputum profile is now shown in Fig. 1D.

In
There are a number of problems with the references; for example, the following are inappropriate in certain contexts: Tonge 2000, Butler 2001, Erht and Schnappinger 2007, Layre et al.2009.
We have carefully checked all referencing and have made adjustments to the cases mentioned above.
It is not appropriate to provide detailed criticisms of minor points, but the following matters are important.
The claimed sensitivity and specificity of 94 and 93%, respectively, is excellent and there is no need to doubt its authenticity. However, to allow verification of the procedure, it is necessary to provide precise coherent data to enable others to repeat the work.
We have added definitions for sensitivity and specificity. Page 3, end of second last paragraph now reads: "Using these data, we calculated a statistical sensitivity of 94% and a specificity of 93% (sensitivity: number of true positives divided by the sum of true positives and false negatives; specificity: number of true negatives divided by the sum of true negatives and false positives) and even slightly better values for the HIV positive individuals alone (Fig. 2F)." Our cutoff values used to calculate these values are now indicated in Fig. 2C and additional technical information for others to repeat these results are presented in Table S1. Figure 1 is not ideal, as BCG (presumably Pasteur) (Fig. 1A) is a poor example, in that it lacks methoxymycolates; use a representative M. tuberculosis strain.
We agree with the reviewer and have replaced the M. bovis BCG Pasteur strain profile with that obtained from M. tuberculosis (Fig. 1A).
Again, in Fig. 1B, C76 mycolic acid is a bad example; it would be better to show the C80 component, ensuring that the mycolic acid structure is correct.
Modified as suggested. A C80 mycolic acid species from M. tuberculosis (m/z 1,164; C80H156O3) is now presented in Fig. 1B.
As noted above, no sputa mass spectra are included for inspection; ideally Fig. 1 should have included a positive and negative sputum example.
Modified as suggested. A mass spectrum of sputum from a TB patient is shown in Figure 1D together with a negative control condition.
It is claimed that the procedure allows differentiation of a range of mycobacterial strains, using cluster analysis. However, it is difficult to clearly understand the profiles on which these results are based (Fig. S4). All these figures must be explained much more clearly, particularly pointing out which signals correspond to the different mycolic acid classes.
We have added more explanations to the figure legend as requested. Major MA signals are now labeled in this figure (Supplementary Fig. 6).

Referee #2 (Comments on Novelty/Model System):
The authors convincingly demonstrate the utility of using a sub-class of mycolic acids that can be extracted directly from tubercle bacilli in the sputum of smear-positive TB patients and in the lungs of acutely infected mice as a marker of TB infection. Therefore, this novel Mtb-derived biomarker has potential applicability both in pre-clinical and in clinical studies of anti-tubercular drug efficacy.

There is an urgent need for new tools for detecting active TB infection and for use as biomarkers to monitor responses to new TB drugs in human disease and in animal models of TB infection. This manuscript makes some progress towards achieving these goals by establishing the potential utility of mycolic acids (MAs) as biomarkers for detecting TB in humans and for monitoring the effects of chemotherapy in infected mice. In this study, the authors carried out an exhaustive lipidomic analysis of MAs in the CMN family of Actinomycetes thereby producing an extremely valuable resource of information on this important class of lipids. They then apply this information to identify a subclass of MAs that can be extracted directly from TB patient sputum and mouse lung and used as a highly specific signature MA marker to detect acute TB infection in humans and mice.
This is a well written manuscript that reports interesting and potentially significant findings. However, the authors should address the following points: 1. Pg. 4, paragraph 3, and Fig. 3B. What was the lung bacillary load in the "MTB-infected" animals shown in Fig. 3B?

What was the limit of detection in terms of colony-forming units per lung?
Bacillary load for the experiment shown in Fig. 3 was measured in 5 out of the 10 animals used in this experiment. The average value was 15.2E6 cfu/lung. We have included this information in the legend to Figure 3.

Moreover, how does the sensitivity of the MA-based assay compare in detecting M. tuberculosis in sputum vs. mouse lung tissue vs. in vitro culture?
We thank the reviewer for this valuable comment, which we have addressed in two ways: 1. An additional spiking experiment was performed to determine the minimum number of bacteria needed for MA detection in sputum and culture medium.
Increasing numbers (0 -2.5E6 cfu) of bacilli (M. tuberculosis H37Rv) were added to negative sputum as well as culture medium, which was used as an in vitro control for comparison. The results from this approximation experiment have been included as a new panel (Fig. 1E): -A linear response is observed over 3 orders of magnitude (R2>0.998).
-The minimum number of bacteria needed for detection by our mass spectrometry method was approximately 10,000 cfu for a signal to noise ratio of 3 (S/N=3) in medium and sputum.
-The complexity of the body fluid sputum results in matrix effects leading with slightly lowered sensitivities (see also below). This is mentioned in text on page 3, 3rd paragraph, which now reads: " We further determined the minimum number of bacterial cells necessary for our MA detection approach, by performing a spiking experiment with a serial dilution of M. tuberculosis cells added to non-TB sputum as well as culture medium. In sputum, around 10,000 cfu were sufficient to detect bacteria based on their MA signal (signal to noise ratio, S/N=3). A linear increase in MA levels was observed with increasing numbers of bacterial cells, for both, sputum (R2=0.998) and medium (R2=0.999). In extracts from medium, MA signals were slightly higher than in extracts from sputum, which was probably due to MA extraction efficiency or ion suppression effects caused by the complex matrix of this body fluid (Fig. 1E)." 2. Comparison of sensitivities between sputum, lung tissue and in vitro cultures -medium: 2.5xE6 cells give a signal corresponding to 0.84xE-6 g of MA (Fig. 1E). Thus, the ratio of 0.84xE-6/2.5xE6 = 0.33 can be used a measure for the sensitivity under these conditions where MA are detected from cells extracted from medium.
-sputum: in the case of sputum this value is reduced to approximately 0.2 (compare slopes in Fig.  1E).
-lung tissue: If we assume comparable linearity also in the case of lung tissue, this value is approximately 0.13 (2 ug MA/lung and approx. 15.2xE6 cfu/lung, Fig. 3B).
Thus, sensitivity for MA extracted and analyzed from culture medium is more than double that of extracts from lung tissue; not surprising given the much more complex nature of tissue as opposed to culture medium with respect to extraction and matrix effects.

MA analysis may provide a potentially useful bacterial biomarker for monitoring TB drug efficacy in human disease or animal models, thereby providing an alternative to the gold-standard CFU count currently used in pre-clinical and clinical studies. Although the authors convincingly demonstrate that the MA signal can be used to differentiate active TB from non-active disease in humans and acute infection from no infection or drug-cured infection in mice, it is not clear how quantitative of bacillary load this measure actually is. What is the limit of detection in mouse lung or human sputum? Moreover, is there a linear relationship between MA level and CFU count?
Yes, there is a linear relationship between MA levels and CFU counts (see above and Fig. 1E).
well performed and is a valuable contribution to the field of molecular medicine. I have only minor comments on the manuscript, listed below.
1. Introduction, third paragraph, first line. Fig S1 does not contain molecular structures. Table S1 would probably be a good place to include this information.
We have included elemental compositions and information on the alpha branch chain in Table S1. This information is now included in Table S1. Modified as suggested.

Legends of supplementary figures are difficult to understand, in particular
Modified and we hope this is clearer now. Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed reports from the referees that were asked to re-assess it. As you will see the reviewers are globally supportive and I am pleased to inform you that we will be able to accept your manuscript pending the following final amendments: -Referee #2's recommendation of using a non-linear regression analysis to determine correlation coefficients in Figure 1E should be followed, and Figure 1E changed accordingly.
-Make sure that entry of sample 18 is correct.
-In the Material and Methods section p7, some text/embedded image is missing under "Isotopic Correction", please change accordingly.
-Please add a 5th key word.
Please submit your revised manuscript within two weeks.
I look forward to reading a new revised version of your manuscript as soon as possible.
Yours sincerely, Editor EMBO Molecular Medicine REFEREE REPORTS: Referee #1: The authors have made the necessary changes.
Referee #3 (Comments on Novelty/Model System): The reason not to qualify the novelty as "high" lies in the fact that bacterial typing on MS profiles is not novel. However, using MA for this purpose, is.
Referee #3 (Other Remarks): I have only two minor comments to the revised manuscript.
The first concerns Fig 1E. The authors appear to have performed a linear regression, but this gives too much weight to the data point at high(-er) 'number of cells'. A non-parametric regression analysis would be a better way to determine the correlation coefficient.
As a second remark, sample number 18 is listed as coming from a patient with AIDS, but a negative HIV status (table 1). Is this correct?
2nd Revision -Authors' Response 11 October 2011 We have addressed the points you list below (see my comments in this email in CAPS) and I have submitted the modified version via the online system.
Please let know if you have additional questions or if you need more information. I hope this revised version will now be acceptable for publication in this final form.
Thank you once again.
Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed reports from the referees that were asked to re-assess it. As you will see the reviewers are globally supportive and I am pleased to inform you that we will be able to accept your manuscript pending the following final amendments: -Referee #2's recommendation of using a non-linear regression analysis to determine correlation coefficients in Figure 1E should followed, and Figure 1E