Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

Contents lists available at ScienceDirect

Journal of Clinical Tuberculosis and Other

Mycobacterial Diseases
journal homepage: [Link]/locate/jctube

Effect of anti-tuberculosis treatment on the systemic levels of tissue

inhibitors of metalloproteinases in tuberculosis – Diabetes co-morbidity
Nathella Pavan Kumar a, c, *, Kadar Moideen a, Vijay Viswanathan b, Shanmugam Sivakumar c,
Syed Hissar c, Hardy Kornfeld d, Subash Babu a, e
a
National Institutes of Health – NIRT – International Center for Excellence in Research, Chennai, India
b
Prof. M. Viswanathan Diabetes Research Center, Chennai, India
c
National Institute for Research in Tuberculosis, Chennai, India
d
University of Massachusetts Medical School, Worcester, MA, USA
e
LPD, NIAID, NIH, MD, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Objectives: To study the association of Tissue inhibitors of matrix metalloproteinases (TIMP) levels with
Mycobacterium tuberculosis tuberculosis-diabetes comorbidity (TB-DM) comorbidity at baseline and in response to anti-TB treatment (ATT).
Diabetes mellitus Methods: We examined the levels of TIMP-1, -2, -3 and -4 in pulmonary tuberculosis alone (TB) or TB-DM at
Tissue inhibitors of metalloproteinases
baseline and after ATT.
Results: TIMP-1, -3 and -4 were significantly increased in TB-DM compared to TB at baseline and after ATT. ATT
resulted in a significant reduction in TIMP-2 and -3 levels and a significant increase in TIMP-1 in both TB and TB-
DM. TIMP-1, -3 and -4 were also significantly increased in TB-DM individuals with bilateral, cavitary disease and
also exhibited a positive relationship with bacterial burden in TB-DM and HbA1c in all TB individuals. Within the
TB-DM group, those known to be diabetic before incident TB (KDM) exhibited higher levels of TIMP-1, -2, -3 and
-4 at baseline and TIMP-2 at post-treatment compared to those newly diagnosed with DM (NDM). KDM in­
dividuals on metformin treatment exhibited lower levels of TIMP-1, -2 and -4 at baseline and of TIMP-4 at post-
treatment.
Conclusions: TIMP levels were elevated in TB-DM, associated with disease severity and bacterial burden, corre­
lated with HbA1c levels and modulated by duration of DM and metformin treatment.

1. Introduction remodeling and repair of tissue following destruction by matrix metal­

loproteinases (MMPs). Therefore, proteolytic balance between MMPs
The TIMP family consists of four members (TIMP-1, -2, -3 and -4) and TIMPs is vital in normal tissue remodeling and various pathological
with significant homology, that inhibit matrix metalloproteinases conditions [10].
(MMPs) with some specificity. TIMPs are endogenous inhibitors of Published studies have reported that TIMP levels were higher in
MMPs and regulate MMP response by forming 1:1 complexes with MMPs serum and pleural fluid of TB patients compared to serum of healthy
[1,2]. TIMP family is a biological inhibitor of several MMP enzymes, controls and non-TB pleural fluid [9]. We have also previously reported
which are involved in the process of the tumor and cell invasion through that TIMP-4 is a significant biomarker for the discrimination of TB-DM
the extracellular matrix and its expression is stimulated by several from TB [11]. However, a comprehensive analysis of the relationship
physiological triggers in various cell types [3,4]. TIMP-1 was previously of TIMPs with TB-DM and their association to disease pathology or
determined to be critical in the immune response to TB [5]. TIMPs may bacterial burdens has not been performed. We have previously demon­
also be crucial in the growth of fibrosis [6], which is characteristic of strated that the clinical and biochemical characteristics of newly diag­
healing TB infection [7]. TIMPs have been advocated as potential bio­ nosed DM individuals with TB are significantly different from those with
markers for TB with good sensitivity and specificity to discriminate TB TB and known DM [11]. Metformin is the most widely-used medication
from healthy individuals [8,9]. TIMPs (TIMP-1, -2 and -3) help in the for type 2 diabetes and published studies have reported that it may be a

* Corresponding author at: National Institute for Research in Tuberculosis, # 1 Mayor Sathyamoothy Road, Chetpet, Chennai, India.
E-mail address: pavankumarn@[Link] (N.P. Kumar).

[Link]

Available online 22 April 2021

2405-5794/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
([Link]
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

candidate for host-directed therapy for TB [12,13]. Retrospective Table 1

human studies indicate that metformin diminishes the risk of progres­ Demographics of the study groups and biochemical parameters in TB-DM and
sion to active TB disease [14,15]. Similarly, the association of TIMPs TB.
with TB individuals with KDM or NDM or of TB-KDM individuals with or Study Demographics TB-DM TB HC p Value
without metformin use has never been examined. No. of subjects 64 24 24 –
Therefore, the aim of this study was to examine the association of the recruited
systemic levels of TIMP-1, -2, -3 and -4 in TB-DM individuals and KDM/NDM 32/32 – – –
compare them with TB individuals without DM and healthy controls. We Metformin Rx Yes/ 16/16 – – –
No
demonstrate elevated levels of TIMPs in association with TB-DM in
Gender (Male/ 44/20 17/7 14/10 NS
comparison to TB and healthy controls. ATT resulted in a significant Female)
reduction in TIMP-2 and -3 levels and a significant increase in TIMP-1 in Median Age (Range) 52 (31–70) 43 (30–67) 35(27–62) NS
both TB and TB-DM. Median Height, cm 159 164 162 NS
(129–176) (121–181) (125–190)
Median Weight, kg 49 (31–64) 44 (30–90) 55 (45–90) NS
2. Materials and methods Smear Grade: 0/1+/ 0/22/24/18 0/9/9/6 NA –
2+/3+
2.1. Ethics statement Fasting Blood 158 93 (73–103) 88 (75–105) p<
Glucose, mg/dL (109–427) 0.0001
Post Prandial 220 112 110 p<
The Ethics Committees of the Prof. M. Viswanathan Diabetes
Glucose, mg/dL (183–448) (80–129) (78–120) 0.0001
Research Center and NIRT approved this study. Informed written con­ Glycated hemoglobin 10.3 5.6 5.5 p<
sent was obtained from all participants. level, % (7.3–15.6) (5.0–5.8) (5.0–5.7) 0.0001

The values represent the geometric mean (and the 95% confidence intervals)
2.2. Study population except for age where the median (and the range) are depicted.

All the study participants were prospectively recruited from ten

Spearman rank correlation was used to compare TIMPs concentrations
participating clinics (TB units) in and around Chennai. Study partici­
with HbA1c levels. Analyses were performed using GraphPad PRISM
pants were identified on the basis of being smear positive for acid-fast
Version 7.
bacilli and enrolled on being Mycobacterium tuberculosis culture posi­
tive on solid media. Study participants were 25–60 years of age and
3. Results
excluded if they had prior episode of TB disease, had received >7 days of
treatment for TB disease, had taken more than seven doses of a fluo­
3.1. Study population characteristics
roquinolone within the past 30 days, were pregnant or nursing, were
seropositive for HIV, or were receiving immunosuppressive therapy.
The baseline characteristics including demographic and biochemical
Plasma samples were collected from 64 individuals with TB-DM and
features of the study population are shown in Table 1. As shown, TB-DM
24 individuals with TB without DM and 24 healthy control individuals,
individuals had significantly higher levels of fasting and post-prandial
recruited in Chennai, India. This was the same set of individuals previ­
glucose as well as HbA1c compared to TB. No significant differences
ously used for studying the association of MMPs with TB-DM [16]. To
were observed in age, sex, smear or culture grades at baseline between
define cavitary disease as well as unilateral versus bilateral lung
the TB-DM and TB groups (Table 1).
involvement, chest X-rays were used. To define bacterial burdens smear
grades were used and they are as classified as 1+, 2+ and 3+. Glycemic
status (DM or normoglycemia) was diagnosed on the basis of oral 3.2. Heightened levels of circulating TIMPs in TB-DM and alterations
glucose tolerance test and/or HbA1c levels (for known diabetics), ac­ following ATT
cording to the WHO criteria. Amongst the 64 TB-DM individuals, 32
were KDM and 32 were NDM. Amongst the KDM individuals, 16 were on We examined the systemic levels of circulating TIMPs in TB-DM, TB
metformin containing anti-diabetic medication and 16 were not. The and HC individuals by measuring the circulating levels of TIMP-1, -2, -3
study groups were matched with regard to age and gender and the and -4 (Fig. 1). As shown, Fig. 1A, systemic levels of TIMP-1 (GM of 36.5
baseline characteristics of the study participants are shown in Table 1. ng/ml in TB-DM vs. 22.2 ng/ml in TB vs 16.97 ng/ml in HC), TIMP-2
Standard ATT was administered to TB-DM individuals using the directly (GM of 4.4 ng/ml in TB-DM vs 3.1 ng/ml in HC), TIMP-3 (GM of 2.2
observed treatment, short course (DOTS) strategy. At 6 months ng/ml in TB-DM vs. 1.1 ng/ml in TB vs 0.58 ng/ml in HC) and TIMP-4
following ATT initiation, fresh plasma samples were obtained. All TB- (GM of 2.7 ng/ml in TB-DM vs 1.6 ng/ml in HC) were significantly
DM and TB individuals were culture negative at the end of ATT. higher in TB-DM compared TB or HC individuals.
We also examined the effect of ATT on TIMP levels in TB-DM in­
2.3. ELISA dividuals. As shown in Fig. 1B, there were consistent and statistically
significant trends for a reduction in TIMP-2 and -3 in TB-DM. In marked
Circulating levels of TIMP-1, -2, -3 and -4 were estimated using a contrast to the other TIMPs measured, the levels of TIMP-1 were
multiplex luminex assay system (Bio-Rad Laboratories, Inc) in plasma consistently higher at TB treatment completion than at baseline in TB-
samples. The lowest detection limits were as follows: TIMP-1, 0.02 ng/ DM. Thus, treatment of TB results in alteration of circulating levels of
mL; TIMP-2, 0.067 ng/mL; TIMP-3, 0.059 ng/mL; TIMP-4, 0.0067 ng/ TIMPs, albeit with TIMP-1 trending in the opposite direction as the other
mL. TIMPs measured.

2.4. Statistical analysis 3.3. Circulating TIMPs are markers of radiographic TB disease severity
and bacterial burdens in TB-DM
Geometric means (GM) were used for measurements of central ten­
dency. Statistically significant differences between the two groups were Since the circulating TIMP levels were significantly enhanced in TB-
analysed using the Mann Whitney test with Holm’s correction for mul­ DM individuals, we wanted to determine the association between the
tiple comparisons. Linear trend post-test was used to compare TIMPs systemic levels of TIMPs and disease severity in TB-DM. To this end, we
concentrations with smear grades (reflecting bacterial burdens) and measured the circulating levels of TIMPs in TB-DM individuals with

2
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

Fig. 1. Elevated circulating levels of TIMPs in TB-DM individuals. (A) The plasma levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were measured in TB-DM (n = 64),
TB (n = 24) and HC (n = 24) individuals at baseline. The data are represented as scatter plots with each circle representing a single individual. P values were
calculated using the Kruskal-Wallis test with Dunn’s post-hoc for multiple comparisons. (B) The plasma levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were measured
in TB-DM individuals at baseline (pre-T) and at 6 months of ATT (post-T). The data are presented as line graphs with each line representing a single individual. P
values were calculated using the Wilcoxon signed rank test.

cavitary versus non-cavitary disease and unilateral versus bilateral dis­ exhibited a significant positive association with HbA1c levels in TB in­
ease at baseline. As shown in Fig. 2A, the circulating levels of TIMP-1 dividuals, with or without DM at baseline showing a significant associ­
(GM of 42.962 ng/ml in cavitary vs. 33.993 ng/ml in non-cavitary dis­ ation of these factors with poor glycemic control. To determine whether
ease), TIMP-3 (GM of 2.927 ng/ml in cavitary vs. 1.663 ng/ml in non- TIMP levels differ based on the duration of diabetes in TB-DM, we
cavitary disease) and TIMP-4 (GM of 3.841 ng/ml in cavitary vs. estimated the systemic levels of TIMPs in KDM (Median HbA1c 10.5%)
2.557 ng/ml in non-cavitary disease) were higher in TB-DM individuals (n = 32) and NDM (Median HbA1c 6.8%) (n = 32) individuals. As shown
with cavitary disease compared to those without. Similarly, as shown in in Fig. 3B, systemic levels of TIMP-1 (GM of 41.716 ng/ml in KDM vs.
Fig. 2B, the circulating levels of TIMP-1 (GM of 42.817 ng/ml in bilateral 31.369 ng/ml in NDM), TIMP-2 (GM of 6.359 ng/ml in KDM vs. 3.864
vs. 33.993 ng/ml in unilateral disease), TIMP-2 (GM of 62.412 ng/ml in ng/ml in NDM), TIMP-3 (GM of 2.600 ng/ml in KDM vs. 1.437 ng/ml in
bilateral vs. 1.692 ng/ml in unilateral disease) and TIMP-4 (GM of 3.489 NDM) and TIMP-4 (GM of 3.274 ng/ml in KDM vs. 2.479 ng/ml in NDM)
ng/ml in bilateral vs. 2.522 ng/ml in unilateral disease) were higher in were significantly higher in KDM compared to NDM individuals at
TB-DM individuals with bilateral disease compared to those with uni­ baseline. As shown in Fig. 3C, systemic levels of TIMP-2 (GM of 3.552
lateral disease. To determine the association of circulating TIMPs and ng/ml in KDM vs. 2.051 ng/ml in NDM) alone were significantly
bacterial burdens, we performed a correlation of the circulating levels of increased in KDM compared to NDM individuals upon completion of
TIMP family in TB-DM individuals with smear grades. As shown in ATT. Thus, KDM is associated with elevated systemic levels of circu­
Fig. 2C, TIMP-1, -3 and -4 exhibited a significant positive correlation lating TIMPs at baseline and TIMP-2 following standard ATT.
with smear grades in TB-DM individuals, indicating a positive associa­
tion of these factors with bacterial burdens. Thus, disease severity and 3.5. Metformin treatment is associated with diminished circulating TIMPs
bacterial burden in TB-DM are associated with elevated systemic levels
of circulating TIMPs at baseline. Use of the anti-diabetic drug metformin has been associated with
lower risk for TB infection, progression from TB infection to active TB
3.4. Circulating TIMPs exhibit a positive relationship with HbA1c in TB disease and for mortality in TB-DM. To test whether this protective effect
individuals and are increased in individuals with KDM of metformin was reflected by differences in circulating TIMPs, we
compared plasma TIMP levels in KDM individuals who reported use of
To elucidate the association between systemic levels of circulating metformin at baseline (n = 16) compared to those on non-metformin
TIMPs and glycemic control in TB patients with or without DM at antidiabetic regimens (n = 16). Importantly, no significant differences
baseline, we determined the relationship between the circulating levels were observed in HbA1c levels between KDM individuals on metformin
of TIMPs in TB individuals with or without DM with HbA1c levels (Median HbA1c 11.3%) compared to KDM individuals not on metformin
(Fig. 3A). As shown, the circulating levels of TIMP-3 and TIMP-4 (Median HbA1c 10.2%). As shown in Fig. 4A, systemic levels of TIMP-1

3
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

Fig. 2. Elevated circulating levels of TIMP 1, 3 and 4 in cavitary and bilateral disease in TB-DM individuals and relationship to bacterial burdens (A) The plasma
levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were measured in TB-DM individuals with cavitary versus non-cavitary disease. (B) The plasma levels of TIMP-1, TIMP-
2, TIMP-3 and TIMP-4 were measured in TB-DM individuals with bilateral versus unilateral disease. (C) The relationship between the plasma levels of TIMP-1, TIMP-
2, TIMP-3 and TIMP-4 and smear grades as estimated by sputum smears was examined in TB-DM individuals. The data are represented as scatter plots with each
circle representing a single individual. For bilateral and cavitary disease P values were calculated using the Mann-Whitney test with Holm’s correction for multiple
comparisons. For bacterial burden relationship P values were calculated using the Linear trend post – test.

4
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

Fig. 3. Elevated TIMPs exhibit a positive relationship with HbA1c and are elevated in individuals with KDM (A) The relationship between the plasma levels of TIMP-
1, TIMP-2, TIMP-3 and TIMP-4 and HbA1c levels was examined in all TB individuals with and without DM. The data are represented as scatter plots with each circle
representing a single individual. (B) The plasma levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were measured in TB-DM individuals with known diabetes. (KDM)
versus newly diagnosed diabetes (NDM) at baseline. (C) The plasma levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were measured in TB-DM individuals with known
diabetes (KDM) versus newly diagnosed diabetes (NDM) at 6 months of ATT. The data are represented as scatter plots with each circle representing a single indi­
vidual. For HbA1c P values were calculated using the Spearman Rank Correlation. For KDM, P values were calculated using the Mann-Whitney test with Holm’s
correction for multiple comparisons.

5
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

Fig. 4. Diminished circulating levels of TIMPs in KDM individuals on metformin treatment (A) The plasma levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were
measured in KDM individuals on metformin treatment versus no metformin treatment at baseline. (B) The plasma levels of TIMP-1, TIMP-2, TIMP-3 and TIMP-4 were
measured in KDM individuals on metformin treatment versus no metformin treatment at 6 months of ATT. The data are represented as scatter plots with each circle
representing a single individual. P values were calculated using the Mann-Whitney test with Holm’s correction for multiple comparisons.

(GM of 24.97 ng/ml in Metformin vs. 39.408 ng/ml in Non-Metformin), comorbidity have poorer ATT outcomes with longer times to sputum
TIMP-2 (GM of 2.988 ng/ml in Metformin vs. 4.997 ng/ml in Non- culture conversion, which in in turn leads to higher risk of death or
Metformin) and TIMP-4 (GM of 2.414 ng/ml in Metformin vs. 3.506 treatment failure, and increased risk of relapse after successful
ng/ml in Non-Metformin) were significantly diminished in KDM in­ completion of anti-TB treatment [22,23].
dividuals on metformin compared to KDM individuals not on metformin. TIMPs are known to be important inhibitors of MMPs, and they are
As shown in Fig. 4B, TIMP-4 (GM of 4.993 ng/ml in Metformin vs. 5.756 also gradually recognized to have impending roles in inflammatory
ng/ml in Non-Metformin) alone was significantly diminished in KDM response [24]. Published studies clearly report that presence of metal­
individuals on metformin compared to KDM individuals not on met­ loproteinases and their inhibitors play an key role in integrity and
formin upon completion of ATT. Thus, metformin therapy in KDM in­ remodeling of extra cellular matrix components in inflammatory con­
dividuals is associated with diminished systemic levels of circulating ditions [25]. The imbalance of TIMP and MMP activities are linked to TB
TIMPs. severity but this has not previously been explored in the context of TB-
DM comorbidity. We, therefore hypothesized that alterations in TIMP
4. Discussion levels would reflect disease pathogenesis, extent and severity of disease
and response to treatment. Our existing analysis revealed that TB-DM
Many epidemiological and clinical studies have revealed that DM is patients exhibit significantly enhanced systemic levels of TIMP-1, -3
one of the major risk factors for TB infection and DM is allied with a two and -4 compared to TB individuals without DM and healthy controls.
to four-fold increased risk of active TB. Evidence from the published Other published studies have also reported that TIMP-1 concentrations
studies also reports that DM patients with uncontrolled blood glucose were significantly elevated in TB patients in comparison to controls and
are at advanced risk to active TB than individual with controlled DM also associated with disease severity [8]. In addition, a recently pub­
[17,18]. The interfaces amongst DM and TB are multidimensional and lished study reported that TIMP-1 is a key biomarker for the diagnosis of
poorly known, though changes have been observed in innate and TB [5]. It is also been well described and reported that TIMP-1 has been
adaptive immune responses [19]. The detrimental effects of DM on TB significantly elevated in the active TB disease in comparison to other
incidence and consequences are now broadly accepted. More than a few pulmonary disorders like pneumonia [5]. Although the role of TIMPs in
studies have shown higher susceptibility to TB in animal models of TB- TB remain unclear, [Link] has been implicated to aggressively dysregulate
DM co-morbidity [20,21]. The actual mechanisms causing this suscep­ the balance between MMPs and TIMPs [26]. Our study is one of the first
tibility to TB are still vague and are in need of comprehensive evalua­ to report on the systemic levels of TIMP expression following ATT. Our
tion. In addition to the heightened risk for TB, persons with TB-diabetes data suggest that while TIMP-1 levels are elevated, other TIMP levels

6
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

decrease significantly or do not change. A previous report on TIMP-1 Funding statement

levels in sputum showed decrease in TIMP-1 at 2 months post-
treatment, which is different from our data on circulating levels albeit This project has been funded in whole or in part with Federal funds
at a later time point [8]. Notably, Hwang et al [27] reported higher from the Government of India’s (GOI) Department of Biotechnology
pleural fluid TIMP-1 levels in in TB pleuritis patients who went on to (DBT), the Indian Council of Medical Research (ICMR), the United States
have residual pleural thickening. Our data also disclosed a significant National Institutes of Health (NIH), National Institute of Allergy and
relationship of TIMP levels with the severity of TB disease (as estimated Infectious Diseases (NIAID), Office of AIDS Research (OAR), and
by the bilateral and cavitary disease) and increasing bacterial burdens, distributed in part by CRDF Global [grant USB1-31149-XX-13]. The
showing that comorbid DM increases this response, which could reflect contents of this publication are solely the responsibility of the authors
elevated bacterial load and/or a specific perturbation of immune func­ and do not represent the official views of the DBT, the ICMR, the NIH, or
tion. Of further interest are the findings that TIMP levels are positively CRDF Global. This work was also funded in part by the Division of
correlated with HbA1c, showing a relationship with poor glycemic Intramural Research, NIAID, NIH.
control, which drives diabetic complications in all tissues [28,29].
Previously, we have reported that there was a bimodal distribution of CRediT authorship contribution statement
baseline HbA1c between KDM and NDM individuals in our cohort, with
significantly greater baseline A1c in the KDM group [30]. Our current NPK: Conceptualization, Formal analysis, Investigation, Methodol­
study adds to this clear heterogeneity in the appearance of TB-DM co­ ogy, Writing - original draft. HK: Conceptualization, Data curation,
morbidity. We determined that systemic TIMPs were significantly Funding acquisition, Resources, Writing - review & editing. SB:
heightened in KDM in comparison to NDM at baseline and after Conceptualization, Funding acquisition, Resources, Writing - original
completion anti-TB treatment, indicating the increased severity of TB draft, Writing - review & editing. VV: Data curation, Project adminis­
disease in KDM individuals. Metformin is an approved antidiabetic drug tration, Resources. SH: Data curation, Project administration. SS:
in routine clinical use and has drawn attention as an impending Formal analysis, Investigation. KM: Investigation, Methodology.
adjunctive, host-directed therapy (HDT) for TB independent of its
glucose-lowering activity [12,15,31]. Studies in mice reported that upon
metformin treatment there is a lowering of bacterial burdens [31]. Declaration of Competing Interest
Enhanced immune control of TB in metformin-treated mice was asso­
ciated with reduced systemic inflammation, which matches our finding The authors declare that they have no known competing financial
that TIMP levels are lower in KDM individuals treated with metformin. interests or personal relationships that could have appeared to influence
Use of metformin has been linked to reduced risk for TB infection, for the work reported in this paper.
progression from TB infection to TB disease and for mortality during TB
treatment in diabetic individuals [32]. Our results deliver new confir­ Acknowledgments
mation for a host-directed role for metformin in that individuals on
metformin treatment revealed decreased systemic TIMP levels, recom­ We thank the staff of Department of Clinical Research and the
mending a host-protective effect of metformin in TB-DM with potential Department of Bacteriology, NIRT for valuable assistance in bacterial
implications for its use in TB without DM. cultures and radiology and the staff of MVDRC, RNTCP, especially Dr.
Our results on TIMPs largely propose that heightened systemic levels Jayagopal Lavanya and Chennai corporation, especially Dr. Senthilna­
of TIMPs is a typical characteristic of TB-DM co-morbidity. Our results than for valuable assistance in recruiting the patients for this study. Data
suggest that specific TIMP-1, -3 and -4 may be a useful component of a in this manuscript were collected as part of the Regional Prospective
biomarker panel for active TB-DM comorbidity with other clinical and Observational Research for Tuberculosis (RePORT) India Consortium.
immunological parameters. However, our study suffers from the limi­
tation of a small sample size. Therefore, further validation of these
References
biomarkers in different endemic populations and different geographical
regions could serve as the basis to develop a biomarker test for active [1] Anand SP, Selvaraj P. Effect of 1, 25 dihydroxyvitamin D(3) on matrix
tuberculosis. At the end of successful anti-TB treatment, TIMP levels metalloproteinases MMP-7, MMP-9 and the inhibitor TIMP-1 in pulmonary
were found to remain increased in TB-DM compared to TB, which may tuberculosis. Clin Immunol 2009;133(1):126–31.
[2] Murphy G. Tissue inhibitors of metalloproteinases. Genome Biol 2011;12(11):233.
be key in tissue remodelling and scarring that lead to long-term fibrosis. [3] Nakano Y, Niida S, Dote K, Takenaka S, Hirao H, Miura F, et al. Matrix
Our data reveal two non-mutually exclusive mechanisms for the eleva­ metalloproteinase-9 contributes to human atrial remodeling during atrial
tion of TIMPs in TB and TB-DM. The first possibility is that TIMPs are fibrillation. J Am Coll Cardiol 2004;43(5):818–25.
[4] Khokha R, Murthy A, Weiss A. Metalloproteinases and their natural inhibitors in
directly driving pathological fibrosis and are therefore increased in these inflammation and immunity. Nat Rev Immunol 2013;13(9):649–65.
conditions [27,33]. The second possibility is that the elevations in the [5] Chen Y, Wang J, Ge P, Cao D, Miao B, Robertson I, et al. Tissue inhibitor of
levels of TIMPs are merely a reflection of elevated levels of MMPs in metalloproteinases 1, a novel biomarker of tuberculosis. Mol Med Rep 2017;15(1):
483–7.
these individuals. We plan on determining these different mechanistic [6] Nagase H, Visse R, Murphy G. Structure and function of matrix metalloproteinases
possibilities in future studies. In addition, these markers could be and TIMPs. Cardiovasc Res 2006;69(3):562–73.
beneficial as targets of host adjuncts to reduce the duration of chemo­ [7] Hunter RL. Pathology of post primary tuberculosis of the lung: an illustrated
critical review. Tuberculosis (Edinb) 2011;91(6):497–509.
therapy as well as agents reducing the incidence of relapse or recur­
[8] Ugarte-Gil CA, Elkington P, Gilman RH, Coronel J, Tezera LB, Bernabe-Ortiz A,
rence. Finally, our study reveals the complex network interlinking the et al. Induced sputum MMP-1, -3 & -8 concentrations during treatment of
pathogenesis of TB-DM to inflammatory pathology and possibly poor tuberculosis. PLoS ONE 2013;8(4):e61333.
[9] Sundararajan S, Babu S, Das SD. Comparison of localized versus systemic levels of
outcomes in TB-DM co-morbidity.
Matrix metalloproteinases (MMPs), its tissue inhibitors (TIMPs) and cytokines in
tuberculous and non-tuberculous pleuritis patients. Hum Immunol 2012;73(10):
5. Conclusions 985–91.
[10] Thrailkill KM, Clay Bunn R, Fowlkes JL. Matrix metalloproteinases: their potential
role in the pathogenesis of diabetic nephropathy. Endocrine 2009;35(1):1–10.
Our data reveal that heightened systemic levels of TIMPs are a [11] Andrade BB, Kumar NP, Sridhar R, Banurekha VV, Jawahar MS, Nutman TB, et al.
typical characteristic of TB-DM co-morbidity. TIMP levels are correlated Heightened plasma levels of heme oxygenase-1 and tissue inhibitor of
with the severity of pulmonary TB disease and with glycemic control. metalloproteinase-4 as well as elevated peripheral neutrophil counts are associated
with TB-diabetes comorbidity. Chest 2014;145(6):1244–54.
[12] Wallis RS, Hafner R. Advancing host-directed therapy for tuberculosis. Nat Rev
Immunol 2015;15(4):255–63.

7
N.P. Kumar et al. Journal of Clinical Tuberculosis and Other Mycobacterial Diseases 23 (2021) 100237

[13] Zumla A, Rao M, Wallis RS, Kaufmann SH, Rustomjee R, Mwaba P, et al. Host- [23] Jimenez-Corona ME, Cruz-Hervert LP, Garcia-Garcia L, Ferreyra-Reyes L, Delgado-
directed therapies for infectious diseases: current status, recent progress, and Sanchez G, Bobadilla-Del-Valle M, et al. Association of diabetes and tuberculosis:
future prospects. Lancet Infect Dis 2016;16(4):e47–63. impact on treatment and post-treatment outcomes. Thorax 2013;68(3):214–20.
[14] Vashisht R, Brahmachari SK. Metformin as a potential combination therapy with [24] Brew K, Dinakarpandian D, Nagase H. Tissue inhibitors of metalloproteinases:
existing front-line antibiotics for Tuberculosis. J Transl Med 2015;13:83. evolution, structure and function. Biochim Biophys Acta 2000;1477(1–2):267–83.
[15] Marupuru S, Senapati P, Pathadka S, Miraj SS, Unnikrishnan MK, Manu MK. [25] Woessner Jr JF. Matrix metalloproteinases and their inhibitors in connective tissue
Protective effect of metformin against tuberculosis infections in diabetic patients: remodeling. FASEB J 1991;5(8):2145–54.
an observational study of south Indian tertiary healthcare facility. Braz J Infect Dis [26] Kubler A, Luna B, Larsson C, Ammerman NC, Andrade BB, Orandle M, et al.
2017;21(3):312–6. Mycobacterium tuberculosis dysregulates MMP/TIMP balance to drive rapid
[16] Kumar NP, Moideen K, Viswanathan V, Shruthi BS, Sivakumar S, Menon PA, et al. cavitation and unrestrained bacterial proliferation. J Pathol 2015;235(3):431–44.
Elevated levels of matrix metalloproteinases reflect severity and extent of disease [27] Hwang KE, Shon YJ, Cha BK, Park MJ, Chu MS, Kim YJ, et al. Tissue inhibitor of
in tuberculosis-diabetes co-morbidity and are predominantly reversed following metalloproteinase-1 is responsible for residual pleural thickening in pleural
standard anti-tuberculosis or metformin treatment. BMC Infect Dis 2018;18(1): tuberculosis. Tohoku J Exp Med 2015;235(4):327–33.
345. [28] Lee SW, Song KE, Shin DS, Ahn SM, Ha ES, Kim DJ, et al. Alterations in peripheral
[17] Al-Rifai RH, Pearson F, Critchley JA, Abu-Raddad LJ. Association between diabetes blood levels of TIMP-1, MMP-2, and MMP-9 in patients with type-2 diabetes.
mellitus and active tuberculosis: a systematic review and meta-analysis. PLoS ONE Diabetes Res Clin Pract 2005;69(2):175–9.
2017;12(11):e0187967. [29] Dominguez-Rodriguez A, Abreu-Gonzalez P, Garcia-Gonzalez MJ, Kaski JC. High
[18] Jeon CY, Murray MB. Diabetes mellitus increases the risk of active tuberculosis: a serum matrix metalloproteinase-9 level predict increased risk of in-hospital cardiac
systematic review of 13 observational studies. PLoS Med 2008;5(7):e152. events in patients with type 2 diabetes and ST segment elevation myocardial
[19] Restrepo BI, Schlesinger LS. Impact of diabetes on the natural history of infarction. Atherosclerosis 2008;196(1):365–71.
tuberculosis. Diabetes Res Clin Pract 2014;106(2):191–9. [30] Kornfeld H, West K, Kane K, Kumpatla S, Zacharias RR, Martinez-Balzano C, et al.
[20] Ronacher K, van Crevel R, Critchley JA, Bremer AA, Schlesinger LS, Kapur A, et al. High prevalence and heterogeneity of diabetes in patients with TB in South India: a
Defining a research agenda to address the converging epidemics of tuberculosis report from the effects of diabetes on tuberculosis severity (EDOTS) study. Chest
and diabetes: part 2: underlying biologic mechanisms. Chest 2017;152(1):174–80. 2016;149(6):1501–8.
[21] Vallerskog T, Martens GW, Kornfeld H. Diabetic mice display a delayed adaptive [31] Singhal A, Jie L, Kumar P, Hong GS, Leow MK, Paleja B, et al. Metformin as adjunct
immune response to Mycobacterium tuberculosis. J Immunol 2010;184(11): antituberculosis therapy. Sci Transl Med 2014;6(263). 263ra159.
6275–82. [32] Degner NR, Wang JY, Golub JE, Karakousis PC. Metformin use reverses the
[22] Baker MA, Harries AD, Jeon CY, Hart JE, Kapur A, Lonnroth K, et al. The impact of increased mortality associated with diabetes mellitus during tuberculosis
diabetes on tuberculosis treatment outcomes: a systematic review. BMC Med 2011; treatment. Clin Infect Dis 2018;66(2):198–205.
9:81. [33] El-Din DSS, Amin AI, Egiza AO. Utility of tissue inhibitor metalloproteinase-1 and
osteopontin as prospective biomarkers of early cardiovascular complications in
type 2 diabetes. Open Access Maced J Med Sci 2018;6(2):314–9.