Basic and Translational Science

Growth of Mycobacteria in Urine

Determined by Isothermal Microcalorimetry:
Implications for Urogenital Tuberculosis and
Other Mycobacterial Infections
Gernot Bonkat, Alexander Bachmann, Anna Solokhina, Andreas F. Widmer, Reno Frei,
Thomas C. Gasser, and Olivier Braissant
OBJECTIVE To overcome the limitations of current urine-based diagnostic assays of urogenital tuberculosis,
we used isothermal microcalorimetry to detect the metabolic activity of Mycobacterium tubercu-
losis and other commonly neglected pathogenic mycobacteria in urine and accurately determine
their growth parameters.
METHODS A microcalorimeter equipped with 48 channels was used. Detection was accomplished, and
growth was monitored for 4 different Mycobacterium species in sterilized and modified urine at
37°C by measuring metabolic heat flow (␮W ⫽ ␮J/s) as a function of time. These strains were
M. smegmatis, M. phlei, M. kansasii, and M. tuberculosis. The data were integrated to perform
curve fitting and extract the growth parameter from the raw data.
RESULTS In sterilized urine, M. smegmatis showed the fastest growth rate (0.089 ⫾ 0.017 [h⫺1]), followed
by M. phlei (0.072 ⫾ 0.016 [h⫺1]) and M. kansasii (0.007 ⫾ 0.001 [h⫺1]). No growth of M.
tuberculosis was detected in sterilized urine. However, in serum-supplemented urine, growth of M.
tuberculosis was observed within 3 weeks at a growth rate of 0.008 ⫾ 0.001 [h⫺1]. Biofilm
formation was enhanced in the serum supplemented urine.
CONCLUSION Isothermal microcalorimetry allows rapid and accurate detection of mycobacterial growth in
urine. Given the absence of data on the mycobacterial growth in urine, isothermal microcalo-
rimetry could be used to unravel key aspects of Mycobacterium physiology in the urinary tract and
potentially contribute to improvement in the diagnosis and treatment of urogenital
tuberculosis. UROLOGY 80: 1163.e9 –1163.e12, 2012. © 2012 Elsevier Inc.

U
rogenital tuberculosis (UGTB) is among the
most common manifestations of extrapulmonary
tuberculosis worldwide.1,2 Because of its insidi-
ous evolution and late onset of symptoms, the diagnosis
and treatment are notoriously delayed, resulting in sig-
nificant morbidity (eg, end-stage renal failure, shrunken
bladder, and testicular destruction).3,4 UGTB is mainly
caused by members of the Mycobacterium tuberculosis
complex. Nevertheless, several mycobacteria other than
tuberculosis are also associated with urogenital infec-
tions.5,6 Many efforts have been made to develop faster
diagnostic assays for UGTB.7-9 However, the optimal
treatment cannot be initiated on the basis of rapid tests
Financial Disclosure: The authors declare that they have no relevant financial
interests.
From the Department of Urology, Division of Infectious Diseases and Hospital
Epidemiology, and Clinical Microbiology Laboratory, University Hospital Basel, Basel,
Switzerland; and Laboratory of Biomechanics and Biocalorimetry, University of Basel
Faculty of Medicine, Basel, Switzerland
Reprint requests: Gernot Bonkat, M.D., Department of Urology, University Hos-
pital Basel, Spitalstrasse 21, Basel 4031 Switzerland. E-mail: bonkatg@uhbs.ch
Submitted: January 16, 2012, accepted (with revisions): April 27, 2012
© 2012 Elsevier Inc.
All Rights Reserved
alone, because they fail to assess mycobacterial viability
(ie, growth) or provide affordable phenotypic drug sus-
ceptibility testing. Similarly, fast diagnostic methods,
such as matrix-assisted laser desorption ionization-time of
flight-mass spectrometry or genotype, which are useful
in tuberculosis lung infections, have the same draw-
backs.10-12 Therefore, urine or tissue culture remains es-
sential in the diagnosis of UGTB and its treatment. In
addition, culture is still the only method that provides
insights into mycobacterial growth and physiology but at
a workload that most laboratories can no longer afford.
Thus, although improvement in therapy and efficient
surveillance rely on this basic knowledge, such basic
microbiologic data about the growth of mycobacteria in
urine are extremely scarce, if not unavailable. In this
context, measurement of mycobacterial metabolic heat
production might provide this missing key information.
Isothermal microcalorimetry (IMC), a nonspecific ana-
lytic tool for the measurement of heat production rates in
the ␮W range, can achieve such measurements. This
technique has been used in material science, chemistry,
0090-4295/12/$36.00 1163.e9
http://dx.doi.org/10.1016/j.urology.2012.04.050
and weapons surveillance system research13,14; however,
but only rare attempts have been made to use it in the
biomedical setting15-18 and, especially, in the field of
urology.19 Using IMC, the purpose of the present study
was to measure the growth of different mycobacteria in
urine and quantitatively determine their growth param-
eters.

MATERIAL AND METHODS

Using IMC, we investigated metabolic heat production during
the growth of 4 different mycobacteria. Two fast-growing my-
cobacteria, M. smegmatis (DSM 43465, Deutsche Sammlung
von Mikroorganismen und Zellkulturen, Braunschweig, Ger-
many) and M. phlei (DSM 49239), were chosen for their ease of
cultivation. In addition, M. tuberculosis H37Ra (American
Type Culture Collection 25177) was chosen because of its
attenuated nature. Finally, M. kansasii (clinical isolates pro-
vided by Dr. Jaton, University Hospital Lausanne, Lausanne,
Switzerland) was chosen, because several reports have men-
tioned urinary infection by this Mycobacterium species. Before
microcalorimetric measurement, the mycobacteria were main-
tained frozen at ⫺80°C in Middlebrook 7H9 medium with 10%
Oleic Albumin Dextrose Catalase supplement and 20% glyc-
erol. Precultures were performed at 37°C in Middlebrook 7H9
medium containing 10% Oleic Albumin Dextrose Catalase and
0.5% Tween 80 to avoid clump formation. Urine from a healthy
donor was collected in a sterile container and rapidly filtered
through a 0.2-␮m pore size Stericup filter (Millipore, Billerica,
MA). The sterilized urine was stored at 4°C until use (within 72
hours) or alternatively at ⫺80°C for longer storage. For the
microcalorimetric measurements, sterilized 4-mL glass micro-
Figure 1. (Upper) From left to right, M. smegmatis (plain
calorimetric ampules were prepared with 2.9 mL of filtered,
gray curve), M. phlei (dashed gray curve), M. kansasii
sterilized urine and 100 ␮L of prepared inoculum culture. To
(dashed black curve), and M. tuberculosis (plain black
further enhance the growth of slow-growing mycobacteria (M.
curve). (Lower) Representative heat flow curves for M. kan-
tuberculosis and M. kansasii), similar measurements were con-
sasii and M. tuberculosis in urine supplemented with 10%
ducted with the addition of calf serum (10% final concentra-
calf serum. From left to right, M. kansasii (plain gray curve),
tion) to simulate pathologic conditions (ie, proteinuria). In
M. kansasii forming a biofilm (dashed gray curve), M. tuber-
these cases, the ampules were filled with 2.5 mL urine, 300 ␮L
culosis (plain black curve), and M. tuberculosis forming a
calf serum, and 200 ␮L inoculum. The sealed ampules were
biofilm (dashed black line).
introduced into the microcalorimeter (TAM48 Waters/TA,
New Castle, DE) according to the manufacturer’s instructions.
The heat production rate (ie, heat flow) was measured contin- All calculations were performed using the R statistical software
uously for ⱕ1000 hours (⬃42 days). Experiments were con- (R Foundation for Statistical Computing, Vienna, Austria) and
ducted in triplicate and repeated twice for fast-growing myco- the grofit package.21,22
bacteria and 3 times for slow-growing mycobacteria.

Statistical Analysis RESULTS

The raw data (ie, the heat flow signal) were used to determine
the maximal number of active bacteria in urine, assuming that
a single bacterial cell produces approximately 2 pW.20
In addition, the cumulative heat produced over time was
obtained by integrating the heat flow data. Using the heat over
time curve, the maximal growth rate was determined by fitting
the Richards growth model to the heat over time curve. Simi-
larly, the net growth rate was estimated by fitting a simple
exponential growth model (Nt ⫽ N0 ⫻ e␮t) to the exponential
part of the growth curve. For these calculations, the heat flow
curves of samples producing biofilm or not reaching sufficient
thermal stability were discarded, because the growth model
cannot be applied to such data. Using the net growth rate, the
generation time was calculated using the relation Tg ⫽ ln2/␮.
The growth of each Mycobacterium species in the steril-
ized urine was easily recorded and resulted in clearly
different heat flow patterns (Fig. 1, Upper). The produc-
tion of heat by M. smegmatis and M. phlei was immedi-
ately detected. These 2 species produced a sharp peak in
the heat flow record that reached 80 ␮W for M. smeg-
matis and 47 ␮W for M. phlei and lasted 40-60 hours. In
contrast, M. kansasii showed lower activity with a max-
imal value of 8 ␮W, resulting in a broader peak appearing
at 200-360 hours. Finally, M. tuberculosis did not produce
any increasing signal in these conditions. However, de-
creasing activity of the inoculum was observed during the
first 12 hours before returning to baseline. This likely

1163.e10 UROLOGY 80 (5), 2012

Table 1. Growth parameter of M. tuberculosis and mycobacteria other than tuberculosis in urine and modified urine
Maximal Growth Rate
Microorganism Growth Media ␮ [h⫺1]
M. smegmatis Sterilized urine 0.311 ⫾ 0.017
M. phlei Sterilized urine 0.178 ⫾ 0.011
M. kansasii Sterilized urine 0.019 ⫾ 0.001
M. kansasii Modified urine* 0.038 ⫾ 0.008
M. tuberculosis Sterilized urine 0†
M. tuberculosis Modified urine* 0.015 ⫾ 0.001
* Added with 10% calf serum.
†
No growth detected and decreasing signal only observed, indicating some remaining activity of inoculum; decreasing signal indicated
remaining activity of inoculum lasted ⬍12 h.
resulted from survival without additional growth of the
inoculum.23 In addition, the ureolytic activity of M.
tuberculosis could also explain such a signal. The maximal
activity peaks observed corresponded to the concentra-
tion of metabolically active cells of about 50 ⫻ 106
cells/mL for M. smegmatis, 25 ⫻ 106 for M. phlei and 2 ⫻
106 for M. kansasii (Table 1). These results are consistent
with the growth rates and generation times observed
(Table 1). In the modified urine supplemented with calf
serum, both M. tuberculosis and M. kansasii grew and
showed only a single activity peak. M. tuberculosis pro-
duced a peak that reached approximately 7 ␮W and
lasted 250 hours compared with M. kansasii, which
reached a peak of approximately 13 ␮W and lasted about
200 hours (Fig. 1, Lower). The maximal number of active
cells was about 3 ⫻ 106 cells/mL for both strains (Table
1). In these conditions (ie, urine plus 10% calf serum),
approximately one fourth of the cultures in the micro-
calorimetric ampules formed clearly visible biofilms at the
air-liquid interface. Biofilm formation resulted in up to
⬃2.5 times greater and prolonged activity (Fig. 1,
Lower), sometimes lasting ⬎1 month after the nonbio-
film-forming cultures had returned to baseline.
COMMENT
UGTB is the second most common form of extrapulmo-
nary tuberculosis in countries with a severe epidemic and
the third most common form in regions with a low
incidence of tuberculosis. The diagnosis is generally de-
layed, chiefly in developing countries, because of its in-
sidious evolution with symptom paucity and inspecificity,
but also because of the very long period required to
confirm a diagnosis using conventional culture ap-
proaches (ⱕ8 weeks).
Our study has provided the first quantitative measure-
ments of the growth rate of mycobacteria in urine and
emphasizes such insidious evolution. IMC, a technique that
is extremely sensitive and nondestructive, allowed precise
monitoring of the growth of the 4 different mycobacteria
tested in urine. To the best of our knowledge, such long-
term measurements and precise determination of the growth
parameters and activity of mycobacteria in urine have never
before been achieved. Similar measurements using classic
culture methods have often been impaired by the formation
UROLOGY 80 (5), 2012
Net Growth Rate
␮ [h⫺1] Generation Time [h] Samples (n)
0.089 ⫾ 0.017 8.0 ⫾ 1.6 6
0.072 ⫾ 0.016 9.9 ⫾ 2.7 6
0.007 ⫾ 0.001 99.0 ⫾ 14.1 5
0.015 ⫾ 0.001 46.2 ⫾ 3.1 6
0† — 9
0.008 ⫾ 0.001 86.6 ⫾ 10.8 6
of clumps or biofilms by the mycobacteria. In addition, the
large culture volumes and regular sampling for 1 month
carries significant risks of contamination and raises biosafety
concerns. Therefore, IMC appears to be a valuable tool for
performing studies of UGTB. Previous studies have empha-
sized the potential of microcalorimetry in the detection of
M. tuberculosis and mycobacteria other than tuberculosis
and in phenotypic drug susceptibility testing.17,18 Therefore,
we believe that with few methodologic modifications, this
technique could be applied for the diagnosis of UGTB and
the detection of resistant mycobacterial strains. In our study,
using a very poor medium (ie, urine), detection was
achieved within 10 days for M. kansasii and 20 days for M.
tuberculosis, which is still within the range of the current
reference standard (ie, BACTEC Mycobacteria Growth In-
dicator Tube, Franklin Lakes, NJ).8,24 To make such detec-
tion of UGTB faster, one could use media such as Middle-
brook (solid or liquid) or Lowenstein-Jensen in the
microcalorimetric vials, such as was proposed in other stud-
ies.15,17,18 Those media could then be inoculated with urine
samples previously concentrated by centrifugation and de-
contaminated using the N-acetyl-L-cysteine-sodium hydrox-
ide procedure to inactivate contaminants and avoid mixed
bacterial growth.
In this context, previous studies have shown that in a
specific medium, the growth curve of an organism can be
used as a signature and that numeric analysis of the
growth-related heat flow pattern can be performed to
achieve identification,25,26 allowing discrimination be-
tween contaminants and mycobacteria. This certainly
warrants additional study investigating the potential of
IMC in the field of UGTB.
Another noteworthy aspect of our study was in the
observation of the metabolic activity of the biofilms
formed. Because IMC only measures heat, it was possible
to study, intact, in situ biofilms. In our context, the slow
growth rates of M. tuberculosis and M. kansasii, combined
with biofilm formation, contributed to maintaining a
rather high activity for a period of ⬎1 month. Biofilm
formation certainly contributes to enhancing the overall
pathogenicity of mycobacteria, because it allows them to
stay active in the urogenital system. Similarly, it is likely
that their resistance to antimicrobial agents also increases
when biofilms have formed. In the future, such a meth-
1163.e11
odologic approach could be used, not only to determine
the growth parameters, but also to rapidly and quantita-
tively assess the drug susceptibility of the pathogenic
mycobacteria.17
Finally, with respect to the cost, isothermal microcalo-
rimeters are available at a cost of US$1000 to ⬎US$500
000, depending on the instrument sensitivity and
throughput. Assuming the use of a mid-grade instrument,
the cost would be US$4.5 and US$6.5 per sample, in-
cluding US$2.5 for consumables (ie, vials filled with
medium) and US$2.0 and US$4.0 for calorimeter use,
maintenance, and amortization. In industrialized coun-
tries, the personnel cost represents a substantial addition
to the calorimetry cost, at approximately US$10.5 per
sample. However, the personnel costs are rather low in
developing countries, about US$1 per sample. Therefore,
the overall cost of 1 sample would be US$5.5-US$17.
CONCLUSION
The results of our study have shown that IMC provides a
rapid, quantitative, accurate, and safe method to gain
insight into the mycobacterial growth in urine at a min-
imal workload and cost. Owing to the lack of such data in
published reports, this technique will undoubtedly con-
tribute to the general understanding of UGTB, resulting
in better therapy.
References
1. Figueiredo AA, Lucon AM, Junior RF, et al. Epidemiology of
urogenital tuberculosis worldwide. Int J Urol. 2008;15:827-832.
2. Jacob JT, Nguyen TM, Ray SM. Male genital tuberculosis. Lancet
Infect Dis. 2008;8:335-342.
3. Zarrabi AD, Heyns CF. Clinical features of confirmed versus sus-
pected urogenital tuberculosis in region with extremely high prev-
alence of pulmonary tuberculosis. Urology. 2009;74:41-45.
4. Wise GJ, Shteynshlyuger A. An update on lower urinary tract
tuberculosis. Curr Urol Rep. 2008;9:305-313.
5. Hepper NG, Karlson AG, Leary FJ, et al. Genitourinary infection
due to Mycobacterium kansasii. Mayo Clin Proc. 1971;46:387-390.
6. Brooker WJ, Aufderheide AC. Genitourinary tract infections due
to atypical mycobacteria. J Urol. 1980;124:242-244.
7. van Vollenhoven P, Heyns CF, de Beer PM, et al. Polymerase
chain reaction in the diagnosis of urinary tract tuberculosis. Urol
Res. 1996;24:107-111.
8. Hillemann D, Richter E, Rüsch-Gerdes S. Use of the BACTEC
mycobacteria growth indicator tube 960 automated system for
1163.e12
recovery of mycobacteria from 9,558 extrapulmonary specimens,
including urine samples. J Clin Microbiol. 2006;44:4014-4017.
9. Hemal AK, Gupta NP, Rajeev TP, et al. Polymerase chain reaction
in clinically suspected genitourinary tuberculosis: comparison with
intravenous urography, bladder biopsy, and urine acid fast bacilli
culture. Urology. 2000;56:570-574.
10. Boehme CC, Nabeta P, Hillemann D, et al. Rapid molecular
detection of tuberculosis and rifampin resistance. N Engl J Med.
2010;363:1005-1015.
11. Lotz A, Ferroni A, Beretti JL, et al. Rapid identification of myco-
bacterial whole cells in solid and liquid culture media by matrix-
assisted laser desorption ionization-time of flight mass spectrometry.
J Clin Microbiol. 2010;48:4481-4486.
12. Saleeb PG, Drake SK, Murray PR, et al. Identification of myco-
bacteria in solid-culture media by matrix-assisted laser desorption
ionization-time of flight mass spectrometry. J Clin Microbiol. 2011;
49:1790-1794.
13. Bunyan P, Baker C, Turner N. Application of heat conduction
calorimetry to high explosives. Thermochim Acta. 2003;401:9-16.
14. Wadso I. Isothermal microcalorimetry in applied biology. Ther-
mochim Acta. 2002;394:305-311.
15. Braissant O, Wirz D, Göpfert B, et al. “The heat is on”: rapid
microcalorimetric detection of mycobacteria in culture. Tuberculo-
sis (Edinb). 2010;90:57-59.
16. Braissant O, Wirz D, Göpfert B, et al. Use of isothermal micro-
calorimetry to monitor microbial activities. FEMS Microbiol Lett.
2010;303:1-8.
17. Howell M, Wirz D, Daniels AU, et al. Application of a microcalo-
rimetric method for determining drug susceptibility in mycobacte-
rium species. J Clin Microbiol. 2012;50:16-20.
18. Rodríguez D, Daniels AU, Urrusti JL, et al. Evaluation of a low-cost
calorimetric approach for rapid detection of tuberculosis and other
mycobacteria in culture. J Appl Microbiol. 2011;111:1016-1024.
19. Bonkat G, Braissant O, Widmer AF, et al. Rapid detection of
urinary tract pathogens using microcalorimetry: principle, tech-
nique and first results. BJU Int. Epub 2012 Feb 7.
20. James AM. Thermal and Energetic Studies of Cellular Biological Sys-
tems. Bristol: IOP; 1987:1-13.
21. R Development Core Team. R: A Language and Environment for
Statistical Computing. Vienna, Austria: R Foundation for Statistical
Computing; 2011.
22. Kahm M, Hasenbrink G, Lichtenberg-Frate H, et al. Grofit: fitting
biological growth curves with R. J Stat Softw. 2010;33:1-21.
23. Vinnerås B, Bölske G, Wahlström H, et al. Survival of mycobac-
terium tuberculosis and Mycobacterium bovis in human urine. Wa-
ter Sci Technol. 2011;63:1075-1080.
24. Dinnes J, Deeks J, Kunst H, et al. A systematic review of rapid
diagnostic tests for the detection of tuberculosis infection. Health
Technol Assess. 2007;11:1-196.
25. Bermúdez J, López D, Sanahuja A, et al. Discriminant analysis of
microcalorimetric data of bacterial growth. Can J Microbiol. 1988;
34:1058-1062.
26. López D, Viñas M, Lorén JG, et al. Analysis of microcalorimetric
curves for bacterial identification. Can J Microbiol. 1987;33:6-11.
UROLOGY 80 (5), 2012