Joseph Falkinham, III
- Non-tuberculous mycobacteria (NTM) are ubiquitous in the environment including water and soil.
- NTM preferentially inhabit surfaces and interfaces; biofilms have high NTM numbers.
- NTM have a lipid-rich, hydrophobic surface, a major determinant of their ecology and epidemiology.
- NTM are oligotrophic (grow above 50 µg AOC/mL) and can grow on a wide range of compounds.
- NTM are readily phagocytosed and grow in protozoa and amoebae.
- NTM are highly resistant to chemical disinfectants.
- Disinfection can select for NTM and thus lead to increased infectious risk.
- A number of Mycobacterium species, Mycobacterium bovis BCG and Mycobacterium marinum form spores ( Ghosh et al. 2009) and, thus, can survive long periods of time.
Although there is a necessary focus on the obligate pathogenic mycobacteria, Mycobacterium tuberculosis
, the vast majority of the over 140 species of Mycobacteriu
m are environmental opportunistic pathogens ( Theron and Cloete 2002
; Falkinham 2002
; Tortoli 2003
). The latter are referred to as nontuberculous mycobacteria (NTM), atypical mycobacteria, or environmental mycobacteria. NTM cause pulmonary disease or skin infections in immuno-competent individuals and disseminated disease in immuno-deficient individuals. NTM are ubiquitous in the environment, found in natural waters ( Falkinham et al. 1980
; Collins et al. 1984
; von Reyn et al. 1993
), drinking water distribution systems ( Covert et al. 1999
; Falkinham et al. 2001
; Le Dantec et al. 2002a
), aquaria ( Agbalika et al. 1984
), soils (Brooks et al. 1980; Kirschner et al. 1992
; Yajko et al. 1995
; Iivanainen et al. 1997
), dusts ( De Groote et al. 2006
), and aerosols ( Wendt et al. 1980
; Parker et al. 1983
). Biofilms are the preferred habitat for NTM ( Schulze-Röbbecke and Fischeder 1989
; Schulze-Röbbecke et al. 1992
; Falkinham et al. 2001
). Further, NTM colonize and grow in in-line filters ( Rodgers et al. 1999
) and cells of M. avium
grown in biofilms are more resistant to chlorine ( Steed and Falkinham 2006
) and antibiotics ( Falkinham 2007
). The recent discovery of spore formation by a number of different Mycobacterium species, including NTM ( Ghosh et al. 2009
), offers a radical reevaluation of the persistence of NTM in the environment and in patients.Risk Factors for NTM Disease
. Risk factors for NTM disease include chronic obstructive pulmonary disease (COPD), smoking, occupational lung diseases (black lung), changes in chest structure, immunodeficiency due to HIV infection or chemotherapy, deficiency in α-1-antitrypsin, and heterozygosity for mutations in the cystic fibrosis transductant regulator (CFTR) gene ( Wolinsky 1979
; Marras and Daley 2002
; Kim et al. 2005
). Recently it has been suggested that individuals suffering from gastro-esophageal disease are at risk for NTM pulmonary infection ( Koh et al. 2007
; Thomson et al. 2007
). There has been recognition that elderly individuals of low weight are at risk for NTM infection ( Prince et al. 1989
; Reich and Johnson 1991
; Kennedy and Weber 1994
). Young children are susceptible to cervical lymphadenitis caused by NTM ( Wolinsky 1995
). Hypersensitivity pneumonitis has been associated with exposure to envelope fractions isolated from NTM ( Richerson et al. 1982
; Huttunen et al. 2000
), to NTM in household water ( Marras et al. 2005
), or to hot tubs ( Rickman et al. 2002
), and to aerosols generated in machining and grinding operations in the automotive industry ( Bernstein et al. 1995
; Shelton et al. 1999
; Steinhauer and Goroncy-Bermes 2007
). Pathogenesis of NTM infection has been reviewed ( McGarvey and Bermudez 2002
) and guidelines for the diagnosis and treatment of NTM infections has been published by the American Thoracic Society ( Griffith et al. 2007
. The mycobacteria causing the highest incidence of infection are: Mycobacterium kansasii
, Mycobacterium marinum
, Mycobacterium ulcerans
(causative agent of Buruli Ulcer), Mycobacterium intracellulare
, the Mycobacterium avium
complex (MAC), Mycobacterium malmoense
, Mycobacterium xenopi
, Mycobacterium simiae
, and the rapidly growing species Mycobacterium abscessus
, Mycobacterium chelonae
, and Mycobacterium fortuitum
( Wolinsky 1979
; Wayne and Sramek et al. 1992; Falkinham 1996
; Marras and Daley 2002
, Vaerewijck et al. 2005
). M. kansasii
causes pulmonary disease and disseminated disease in AIDS patients ( Lillo et al. 1990
; Alcaide et al. 1997
). Historically, M. kansasii
disease has been associated with exposure to cattle and stockyards. Human infection with M. marinum
is associated with exposure to fish, aquaria, or swimming pools ( Aubry et al. 2002
). M. ulcerans
infection is manifested as a necrotizing disease affecting skin, subcutaneous tissue, and sometimes bone (“Buruli Ulcer”) and usually associated with residence or activities near water courses in tropical areas ( Portaels et al. 2008
). It is considered the third most common human mycobacterial disease after tuberculosis and leprosy ( Portaels et al. 2008
). The recent isolation of M. ulcerans
from the insect called the water strider collected in Benin supports the hypothesis that this NTM pathogen might be transmitted to humans from aquatic niches via insects ( Portaels et al. 2008
). The M. avium complex (MAC) comprises the subspecies M. avium
(MAA, infecting birds, some ruminants like deer, and humans); M. avium
(MAH, infecting pigs and humans), M. avium
(MAS, infecting wild animals), and M. avium
(MAP, causing Johne’s disease in cattle and possibly an etiological agent of Crohn’s disease in humans) ( Turenne et al. 2008
). MAC subspecies are important pathogens of animals ( Biet et al. 2005
). Although M. intracellulare
causes pulmonary disease in immunocompetent individuals (approximately half the cases of pulmonary disease in elderly, slender individuals is caused by M. intracellulare
), it is not found in AIDS patients who are almost exclusively infected with M. avium
( Falkinham 1996
; Marras and Daley 2002
). Amongst AIDS patients, genotypic fingerprints and glycolipid antigens of M. avium
from hospital’s hot and cold water systems were identical to the ones from AIDS patients exposed to contaminated water indicating water as a source of infection ( von Reyn et al. 1994
). Pulmonary infections can also result from showering with contaminated water and the resulting aerosols ( Falkinham et al. 2008
). A study performed in South Wales, suggested that the clustering pattern of Crohn’s disease (suspected be linked to M. avium
) in Cardiff might be linked to either drinking water or inhaling aerosols from the river Taff which flows through the city ( Pickup et al. 2005
). MAP, which can survive for long periods in water and sediment ( Whittington et al. 2005
), are likely washed into rivers by rain falling on upland hill pastures, which are grazed by endemically infected cattle and sheep ( Pickup et al. 2006
Pulmonary disease, but rarely disseminated disease in AIDS patients, caused by the difficult-to-isolate M. malmoense is more common in Europe than the United States ( Zaugg et al. 1993). Pulmonary infections caused by the relatively heat-resistant M. xenopi usually occur in outbreaks involving hot water systems ( Sniadack et al. 1992). Tap water, used to rinse surgical devices after disinfection was shown to be the source of M. xenopi hospital acquired infections ( Astagneau et al. 2001). M. simiae infection in the United States appears to be geographically restricted to the central southwestern United States, principally Texas ( Sahly et al. 2002). An outbreak of M. simiae infection was traced to a hospital hot water supply ( Conger et al. 2004). One species, Mycobacterium scrofulaceum, was commonly isolated from children with cervical lymphadenitis up until 1985 ( Wolinsky 1995) and from natural waters ( Falkinham et al. 1980), but has almost disappeared and is seldom recovered today ( Falkinham et al. 2001).
A high frequency of NTM nosocomial infections are caused by the rapidly growing mycobacteria ( Wallace 1994; Wallace et al. 1998). Pulmonary infection by M. abscessus appears to be an emerging problem amongst individuals with cystic fibrosis ( Jönsson et al. 2007). Rapidly growing NTM have also been associated with skin and soft tissue infections ( Uslan et al. 2006), particularly following exposure to footbaths at nail salons ( Winthrop et al. 2002).
The Mycobacterial Cell Envelope. A major determinant of the ecology and virulence of NTM is the thick, lipid-rich outer membrane ( Nikaido et al. 1993; Brennan and Nikaido 1995; Daffe and Draper 1998). NTM have a true outer membrane ( Hoffman et al. 2008) and thus have a periplasmic space. The envelope contributes to the slow growth of NTM and their resistance to antibiotics ( Rastogi et al. 1981; Jarlier and Nikaido 1994), disinfectants ( Taylor et al. 2000; Falkinham 2003), heavy metals ( Falkinham et al. 1984), and acid ( Bodmer et al. 2000). It also contains many compounds that are involved in virulence ( McGarvey and Bermudez 2002). In as much as the envelope lipids result in a hydrophobic surface and as hydrophobicity is a determinant of aerosolization from water ( Parker et al. 1983) and surface attachment ( Bendinger et al. 1993), the envelope is a major determinant of the ecology of NTM. NTM numbers are much higher on surfaces or interfaces compared to bulk suspension ( Schulze-Röbbecke and Fischeder 1989; Schulze-Röbbecke et al. 1992; Falkinham et al. 2001; Torvinen et al. 2004). NTM cells grown in biofilms have increased resistance to disinfection ( Torvinen et al. 2004; Steed and Falkinham 2006). Further, M. avium cells grown in biofilms are transiently more resistant to disinfection than cells grown in suspension ( Steed and Falkinham 2006). Evidently this is an adaptation, as the cells lose the increased resistance after 1 day’s growth in suspension. ( Steed and Falkinham 2006). Evidence has been presented that biofilm formation might be enhanced by oxidative stress ( Geier et al. 2008).
NTM Metabolism. NTM are oligotrophic, able to grow in waters at AOC levels above 50 µg/mL ( Carson et al. 1978; George et al. 1980; Norton et al. 2004). As a result of their hydrophobic envelope, NTM are capable of degrading a variety of hydrocarbons, including chlorinated hydrocarbons ( Heitkamp et al. 1988; Burback and Perry 1993; Buers et al. 1997; Krulwich and Pelliccione 1979). The recovery of NTM from polluted sites ( Wang et al. 2006) is undoubtedly linked with the ability of NTM to degrade hydrocarbons. Although the hydrophobic envelope is relatively impermeable to hydrophilic compounds, thus limited growth, rates of oxygen consumption and protein synthesis are as high as other bacteria at the same temperature ( Krulwich and Pelliccione 1979), NTM simply expend a great deal of energy on synthesis of long chain fatty acids (C60-C80) and not on the production of more cells.
Nontuberculous mycobacteria are ubiquitous in the environment. Although the focus here is on aquatic environments, it should be noted that mycobacterial levels can be much higher in soils (e.g., 106/gm). However, they are present in virtually all natural waters sampled, including marine waters ( Falkinham et al. 1980
; Collins et al. 1984
; Iivanainen et al. 1993
; von Reyn et al. 1993
; Glover et al. 1994
; Iivanainen et al. 1999
) and water from drinking water distribution systems (du Moulin et al. 1986; Fischeder et al. 1991
; Covert et al. 1999
; Falkinham et al. 2001
) and from plumbing in buildings ( du Moulin et al. 1988
; Nishiuchi et al. 2007
). NTM species frequently isolated from drinking water and hospital water systems were M. avium
, M. intracellulare
, M. gordonae
, M. kansasii
, M. xenopi
, M. chelonae
, M. fortuitum
, and M. simiae
( Vaerewijck et al. 2005
; Carson et al. 1978
). A detailed overview over Mycobacterium
species isolated from water distribution systems can be found in a review by Vaerewijck et al. 2005
Over 20 NTM species have for example been isolated from the Rio Grande River including the pathogens M. avium and M. kansasii ( Bland et al. 2005). Members of the M. fortuitum complex group were most commonly isolated from this freshwater body. Total mycobacteria numbers correlated with coliform and E. coli counts. NTM have been isolated from cold and hot water taps, bottled water, ice machines, foot baths, heated nebulizers, shower heads, swimming pools, and hot tubs collected in the environs of Paris, France ( Le Dantec et al. 2002a). They have been reported to survive, persist, grow and colonize in drinking water supply systems ( Covert et al. 1999, Falkinham et al. 2001). NTM were isolated from 38% of 42 drinking water distribution systems, including both surface- and groundwater-fed systems, over 21 states in the continental United States ( Covert et al. 1999). Even higher numbers have been reported from water samples collected in the Los Angeles Metropolitan District; specifically members of the M. avium complex (MAC) were recovered from 92% of reservoirs, 95% of homes, and 100% of hospitals ( Glover et al. 1994). In natural waters like river water and lake catchment mycobacteria are frequently detected especially if the waters receive runoff from the surrounding soils and pastures. A study in South Wales reported 48 of 70 water samples (69%) from the River Tywi were positive for M. avium subsp. paratuberculosis as detected by an IS900-specific PCR method ( Pickup et al. 2006). The river catchment comprises 1,100 km2 surface area with more than a million cattle and more than 1.3 million sheep. Also most sediment samples from lakes receiving water inflow from the catchments tested positive.
Examples of reported mycobacterial concentrations in aqueous environments:
- Water from 18 cold water and 16 hot water taps from 34 sites on two temporarily vacant hospital floors in Boston, Massachusetts was examined for the presence of M. avium with 14 sites (41% of samples) testing positive ( du Moulin et al. 1988). Water from hot water taps (average temperature 55°C) yielded most positive samples with up to 500 CFU M. avium per 100 ml. Glycolipid characterization identified 7 of 11 analyzed strains as type 4 serovar, which was also the dominant serovar in AIDS patient in the area.
- Eight drinking water distribution systems across the United States were tested for NTM with special emphasis on M. avium and M. intracellulare ( Falkinham et al. 2001). The study examined 528 water and 55 biofilm samples taken over a 18-month period. NTM were recovered from 15% of the samples by cultivation with numbers ranging from 10 – 7 x 105 CFU per liter. Water treatment resulted in 2- to 4-log reductions in NTM counts. NTM were found to be on the average 25,000-fold more prevalent in the distribution system compared to the treatment plants suggesting growth (“re-growth) within the distribution system. Recovery was more frequent in fall and winter samples, but NTM seemed to be permanent residents in the systems without strong seasonality. Recovery and numbers of M. avium and M. intracellulare in drinking water samples were low with recovery frequencies of 3 and 1%, respectively (in biofilms recovery was more frequent with 5 and 24%). Numbers of M. avium in raw waters correlated with turbidity and higher overall microbial numbers. M. intracellulare was seldom found in raw or system water samples, but was frequently recovered from biofilms (average 600 CFU/cm2). The preference for this environmental niche would indicate that accurate monitoring of mycobacteria would require sampling of both water and biofilms.
- Thirty three groundwater samples from three drinking water treatment plants and 72 samples from drinking water distribution systems in Germany were tested for NTM ( Fischeder et al. 1991). Culture-positive results were found for 82% of all samples with typical concentrations between 102 and 103 CFU/L (maximum: 4.5 x 105 CFU/L). Species (identified biochemically and by thin layer chromatography of characteristic mycolic acids) included M. gordonae (most water samples), M. flavescens, M. kansasii, M. chelonae (domestic water systems), and M. fortuitum (drinking water treatment plant).
- NTM concentrations and species compositions were studied in two drinking water plants (Orly and Ivry, near Paris, both receive surface water) and the corresponding distribution systems, as well as in groundwater that is directly transferred into the distribution system without entering a water treatment plant ( Le Dantec et al. 2002a). Maximum concentrations of NTM were found to be between 20-20,000 CFU per liter at the Orly plant and 20- 2,000 CFU per liter (Ivry). NTM numbers dropped dramatically during water treatment to 10-20 CFU per liter (Orly) and to a maximum of 60 CFU per liter (Ivry). Comparing NTM removal rates in the two plants, slow sand filtration was found more efficient than rapid sand filtration. The NTM species in the drinking water distribution system (DWDS) were found to be different from the ones isolated from the water leaving the plants. The overall NTM recovery frequency in the DWDS samples was 72% (n=144). Recovery numbers were similar for systems with either ground water or treated surface water. However, differences in the abundances of different NTM species were found in groundwater and treated surface water. For example, M. gordonae was found more frequently in treated surface water than in groundwater, whereas M. nonchromogenicum was only found in groundwater samples or samples containing both water types, but not in treated surface water samples. NTM concentrations between 1 - 50 CFU per liter and 51 – 500 CFU per liter were found in 78% and 21% of the positive samples, respectively. Only one sample contained more than 1,000 CFU NTM/L. Potentially pathogenic mycobacteria were found in 14-15% of the positive samples.
- NTM were found in 90% (n=50) biofilm samples from water treatment plants, domestic water supply systems and aquaria in Germany ( Schulze-Röbbecke et al. 1992). NTM concentrations typically ranged between 103 -104 CFU/cm2. The highest concentration was 5.6 x 106 CFU/cm2. Surfaces containing organic compounds (e.g., plastics or rubber) supported greater colonization and biofilm formations compared to surfaces composed of inorganic compounds (e.g., copper or glass).
- A study of 16 drinking water distribution systems in 8 Finnish localities revealed NTM concentrations between 15 and 140 CFU/L ( Torvinen et al. 2004). The water systems were located in boreal regions with high NTM numbers in natural waters and soils ( Iivanainen et al. 1997). Counts and isolation frequencies increased between the water treatment facilities (35% of samples positive) to the distal portions of the distribution systems (80% of samples positive). Systems using surface water and applying ozonation showed higher NTM numbers. Cell numbers and growth correlated with AOC concentrations in the water leaving the treatment plant. NTM concentrations in old and young deposits in distribution systems showed medians of 1.8 x 105 and 3.9 x 105 CFU/g [dry weight] and exceeded those in water samples by a factor of 2.5 x 104. Highest densities were found in developing biofilms at the distal sites of systems using surface water.
- NTM numbers in brook water samples collected in Finland from 53 drainage areas ranged from 10 to 2,200 CFU/L ( Iivanainen et al. 1993). The frequency of isolation from the 53 samples was 100%. There was a positive correlation (P < 0.001) between NTM numbers and the number of heterotrophic bacteria, the presence of peatlands, chemical oxygen demand, water color, and concentrations of some metals, and a negative correlation with pH. Peatlands can be a reservoir from which runoff water disperses bacteria into the surrounding environment (Iivanainen et al. 1997). Rain led to an increase in water color, COD, acidity, and NTM counts. High NTM concentrations were reported in areas with acidic sulfide soils. The study suggested that acidic waters with a high organic content are potential sources of NTM.
NTM are quite resistant to chemical disinfectants typically employed for water treatment such as chlorine, chloramine, chlorine dioxide, or ozone ( Taylor et al. 2000
). This is most likely due to their lipid rich, hydrophobic cell envelope ( Rastogi et al. 1981
; Brennan and Nikaido 1995
). A complicating factor in measurement of NTM susceptibility to antimicrobial agents is their tendency to aggregate. Thus, without precautions to ensure one is working with a single cell suspension, it is not clear whether one is measuring the susceptibility of single cells or aggregates ( Taylor et al. 2000
). NTM are more resistant to commonly used disinfectants than are bacterial indicators, such as Escherichia coli
( Pelletier et al. 1988
; Health Canada 2006
). For example, MAC resistance to chlorine is 700-3,000 times greater than that for E. coli
, resistance to chlorine dioxide and ozone at least 100- and 50-fold greater ( Taylor et al. 2000
). In a typical drinking water system, the chlorine is unlikely to be effective in controlling MAC ( Health Canada 2006
; AWWA 1999). M. chelonae
and M. fortuitum
strains (commonly found in drinking water treatment plants) were reported to survive 60 min of exposure to 0.3 ppm (60% survivors) and 0.7 ppm (2% survivors) at pH 7 ( Carson et al. 1978
). The high resistance to disinfection led to the hypothesis that reduction in NTM numbers by almost 2 log units (range 1.96 to 4 log units) during drinking water treatment must be due to other mechanism such as filtration and removal of turbidity ( Falkinham et al. 2001
). In fact, disinfection is likely to select for NTM, allowing these slowly growing pathogens to proliferate in the distribution systems and in household plumbing ( Falkinham et al. 2001
Due to the high resistance to commonly used chlorine concentrations, inactivation with copper and silver ions was examined. NTM are relatively resistant to heavy metals and oxyanions ( Falkinham et al. 1984) and M. avium was found more resistant to copper and silver ions than Legionella pneumophila ( Lin et al. 1998; Kusnetsov et al. 2001). Exposure times had to be 100-fold longer for M. avium to achieve comparable killing in vitro ( Lin et al. 1998; Kusnetsov et al. 2001). The high tolerance could contribute to their persistence in drinking water ( Taylor et al. 2000) and is suspected to lead to selection of NTM in water settings like DWDS, industrial water reservoirs, hot tubs, and home spas where disinfectants are widely used ( WHO 2001). Metal and chlorine resistance is likely the reason for the survival and growth of M. avium in in-line drinking water filters ( Rodgers et al. 1999). Slow growth also contributes to the resistance of M. avium, M. intracellulare, and M. scrofulaceum to disinfectants ( Falkinham 2003). Survival might also be enhanced by entering the dormant, spore state ( Ghosh et al. 2009). Many NTM outbreaks have been associated to water originating from disinfected sources leading to the worrisome conclusion that disinfection can lead NTM predominance and proliferation and, ultimately, to NTM infection and disease ( WHO 2001).
Selected studies on disinfection are summarized in the following:
- Free chlorine concentrations of 1.0 ppm were shown to eliminate cultivability of 105 CFU of NTM isolates within 8 hours of exposure, whereas 0.15 ppm had virtually no effect ( Pelletier et al. 1988). The data suggested that free chlorine residual levels of 0.1 ppm or less may be an inadequate control strategy to eliminate NTM from in water distribution systems.
- Le Dantec et al. studied the chlorine susceptibility of different NTM isolates from a water distribution system in Paris. Disinfection at a concentration and contact time typically used in water treatment lines (60 mg x min x liter-1), resulted in inactivation of >4 log10 unit of M. gordonae, and 1.5 log units of M. fortuitum or M. chelonae. Even the most susceptible NTM species, M. aurum and M. gordonae, were 100 and 330 times more resistant to chlorine under the chosen conditions than E. coli. Chlorine resistance of M. gordonae was higher under low-nutrient conditions such as encountered in water than under high nutrient conditions. Chlorine inactivation was reported more efficient when the temperature was increased from 4°C to 25°C and when the pH decreased ( Le Dantec et al. 2002b).
- Taylor et al. 2000 measured the efficacy of different chemical disinfectants for killing M. avium, M. intracellulare, and M. scrofulaceum strains. Data were presented as CT99.9% values (product of disinfectant concentration in ppm and time in min to achieve a 3 log reduction in cultivatable cells). Results were compared to those for an E. coli control. Mean CT99.9% values ranged from 51 to 204 for chlorine (E. coli control: 0.09), from 91 to 1,710 for monochloramine (E. coli control: 73), from 2 to 11 for chlorine dioxide (E. coli control: 0.02), and from 0.10 to 0.17 for ozone (E. coli control: 0.002). Cells from more slowly growing strains were more resistant to chlorine than more rapidly growing strains. This correlated with the observation that water-grown cells were 10-fold more resistant to chlorine than medium-grown cells ( Taylor et al. 2000; Falkinham 2003). M. scrofulaceum strains were significantly more susceptible to the disinfectants suggesting that one reason for its disappearance is uniform application of disinfection standards across the United States.
- Lin et al. measured the susceptibility of M. avium to copper and silver ions. Bacteria in 100 ml buffered water at a concentration of 3 x 106 CFU ml-1 were exposed to different metal concentrations (copper/silver ions: 0.1/0.01, 0.2/0.02, 0.4/0.04, 0.8/0.08 mg per liter). Colony formation on Middlebrook 7H10 agar and grow index values from the BACTEC system showed that the following contact times were required for a 2 and 3 log reduction: 3d and 5 d (0.1 mg l-1 Cu2+, 0.01 mg l-1 Ag+), 3 d and 5 d (0.2 mg l-1 Cu2+, 0.02 mg l-1 Ag+), 2d and 4 d (0.4 mg l-1 Cu2+, 0.04 mg l-1 Ag+), and 1 d and 2 d (0.8 mg l-1 Cu2+, 0.08 mg l-1 Ag+), respectively ( Lin et al. 1998).
- Bland et al. compared the chlorine susceptibility of an M. fortuitum strain isolated from the Rio Grande River with the ‘hardy’ E. coli O157:H7 strain Sakai. Exponentially growing cells were mixed into deionized water supplemented with chlorine. Initial cell concentrations were around 107 CFU per ml. E. coli cultivability dropped by 6 log units after 1 hour in 1 ppm free chlorine, while M. fortuitum counts were still at 55% of the initial value under the same conditions. Above 1 ppm free chlorine, no E. coli counts were obtained, whereas M. fortuitum cultivability dropped only 3.5 log units after exposure to 8 ppm free chlorine for 1 h ( Bland et al. 2005).
- Williams et al. characterized the population diversity of model potable water biofilms grown in annular reactors in the absence of disinfectant or receiving chlorine (free residual chlorine: 0.24 ± 0.11 ppm), chloramine (free residual chlorine: 0.048 ± 0.073 ppm), or inactivated disinfectant residual. Sequence diversity of 16S rRNA clone libraries suggested that biofilms harvested on days 56 and 69 incubation in the presence of chloramine showed little species diversity and contained mainly Mycobacterium and Dechloromonas sequences. The other biofilms showed more phylogenetic diversity and were dominated by a variety of alpha- and additional beta-proteobacteria. The results suggested that chloramine might not able to prevent NTM species from inhabiting drinking water biofilms and might even select for NTM ( Williams et al. 2005).
- Whan et al. 2001 measured the chlorine susceptibility of two M. avium subsp. paratuberculosis strains (bovine and human strain) exposed to 0.5, 1.0, and 2.0 ppm chlorine for 15 and 30 min. Cells (initial concentration 106 CFU/ml) were suspended in sterile deionized water (subjected to reverse osmosis) or the same water containing MgCl2, CaCl2, NaHCO3 and BSA (to mimic commercial water treatment operations). Log10 reductions in the range between 1.32 and 2.82 were reported. A later study with the same chlorine concentrations and contact times produced similar log reductions for free cells (0.78 – 1.73 log units), but much lower reduction values (0.16 – 0.94 log units) when the cells were ingested by Acanthamoeba polyphaga ( Whan et al. 2006).
- Increased chlorine resistance of NTM phagocytozed Acanthamoeba polyphaga was also found in a study by Adékambi et al. 2006 who found surviving culturable that were exposed to 15 ppm free chlorine for 24 hours. No culturable NTM were found under the same conditions in the absence of amoeba.
- Lagunas-Solar et al. 2005 investigated the use of radiofrequency power (RF) to inactivate waterborne pathogens. RF is a physical method leading to uniform heating of a water volume. Water samples were seeded with app. 3 x 106 CFU M. avium subsp. paratuberculosis per ml and subjected to a frequency of 10 to 14 MHz. Growth was completely and rapidly halted in less than 1 min when temperatures between 60-65°C were reached. That result agrees with the temperature limit for growth of M. avium complex (MAC) strains ( Schulze-Röbbecke and Bucholtz 1992).
The recent discovery that NTM form spores (Ghorse et al. 2009) will challenge the contemporary view of the persistence of NTM in both the environment and in patients. A large number of growth and persistence measurements of M. avium
complex (MAC) strains have been reported as MAC are major pathogens. M. avium complex strains survive in water and sediments for months up to years and grow under varied conditions ( Sniadack et al. 1992
; Health Canada 2006
). M. avium
and M. intracellulare
can grow over a relatively wide temperature range (15°C - 42°C; M. avium
to 45°C) and grow over a wide pH range ( George et al. 1980
). High NTM numbers are found in waters and soils of low pH and high organic content; namely the coastal swamps of the eastern United States ( Kirschner et al. 1992
) and the boreal forest soils ( Iivanainen et al. 1997
) and drainage waters of Finland ( Iivanainen et al. 1999
). M. avium
and M. intracellulare
strains grew at pH levels as low as 4.0 and at low oxygen levels ( Kirschner et al. 1992
). However, a great diversity of NTM species was also recovered from the alkaline water of the Rio Grande River (mean pH of 8.30; Bland et al. 2005
). The number and growth of NTM in drinking water distribution system waters correlated strongly with the concentration of assimilable organic carbon (AOC) in the water leaving the waterworks ( Torvinen et al. 2004
) and in a pilot drinking water distribution system ( Norton et al. 2004
). In the pilot system, growth of M. avium
occurred at AOC concentrations above 50 µg/L ( Norton et al. 2004
). Assimilable organic carbon might be more readily available in systems using ozonation ( Torvinen et al. 2004
). A reduction of AOC and biodegradable organic carbon is believed to reduce mycobacterial numbers by slowing growth ( Falkinham et al. 2001
). Growth of M. avium
in natural waters was stimulated by the addition of humic and fulvic acids ( Kirschner et al. 1999
). This correlates with high numbers of mycobacteria in boreal waters and brown water swamps which represent humic- and fulvic-rich, acidic, microaerobic environments ( Iivanainen et al. 1993
; Kirschner et al. 1992
NTM Growth in Protozoa and Amoebae. NTM are readily phagocytosed and grow in some species of amoeba ( Cirillo et al. 1997; Steinert et al. 1998) and protozoa ( Strahl et al. 2001). Adékambi et al. 2006 showed that 26 NTM species could grow in Acanthamoeba polyphaga trophozoites and cysts. Intracellular growth in amoeba also leads to increased virulence ( Cirillo et al. 1997). Studying a hospital water network, NTM were more frequently recovered from water samples containing an amoeba species ( Thomas et al. 2006).
Examples of studies on mycobacterial persistence are given in the following:
- M. avium cells were shown to survive substantially longer than E. coli in drinking water biofilms (weeks compared to days) making the latter a bad indicator for this pathogen in biofilms and its release ( Lehtola et al. 2007). Biofilms were grown in tap water-fed Propella reactors on polyvinyl coupons. The tap water contained 0.15 ppm chlorine. After 1 month, biofilms that had developed were seeded by running water with M. avium through the reactors with a cell concentration of approximately 105 CFU/ml for 2 hours (retention time 12.6 hours). Numbers of culturable M. avium cells in biofilms grown under high-shear turbulent flow decreased from 155 to 50 CFU/cm2 on average within 4 weeks. The number of M. avium cells visualized by FISH was on average 870- and 5,000-fold higher than numbers obtained by cultivation for biofilm and outlet water samples, respectively. Data indicated that microbes were released from the biofilms into the reactor outflow water as planktonic cells were not washed out at the expected rate ( Lehtola et al. 2007).
- The prevalence of NTM was studied in two surface drinking water systems was studied to assess the effect of filtration ( Hilborn et al. 2006). Water samples were taken from plant effluent, distributed water, and at point-of-use (POU) sites (i.e. cold-water taps). NTM were isolated from 38 out of a total of 139 samples (corresponding to 27% of samples). Thirty isolates were identified as M. avium, 5 as M. kansasii, 2 as M. fortuitum complex, and one as a member of the M. avium complex. The concentration in these samples ranged between 1 and >500 CFU/500 ml (median 161 CFU/500 ml). NTM were isolated from all water samples collected from POU sites yielding the majority of NTM positive samples. No statistically significant correlation was observed between the CFU of NTM and the chlorine concentration or the temperature in the distribution system. The implementation of filtration and ozonation in one of the plants did not result in lower NTM detection rates within the study period. When studying the presence of M. avium groups at POU sites, electrophoretic groups (defined by pulsed-field electrophoresis), were found to persist for up to 26 months despite the change in water treatment in one plant ( Hilborn et al. 2006)
- M. avium subsp. paratuberculosis was reported to remain culturable in lake water microcosms for 632 days ( Pickup et al. 2005). Microcosms with sterile water were seeded to an initial density of app. 103 to 104 CFU per ml. Quantitative PCR detected the presence of genomes until the end of the experiment of 841 days. PCR detectable units (PDUs) did not fall below 103 genome copies per ml ( Pickup et al. 2005).
- A clinical isolate of M. intracellulare (strain LM1) was subjected to starvation in deionized water and survival was monitored for more than two years ( Archuleta et al. 2005). Loss of viability was described biphasic. In the first phase viability decreased rapidly over a 5 – 10 day period (adaptation phase), whereas in the second phase loss of viability was very slow (persistence phase). Cultivability dropped to app. 50% within 7 days and to 27% after 89 days. Thereafter, cultivability remained rather constant during the study period of 2.2 years and was at no point lower than 15%. Metabolic activity (as measured by the reduction of a tetrazolium dye) dropped by 60% by day 10. After that, loss of metabolic activity slowed down and was 80% by day 53. After 1.7 years the metabolic rate was at 2.0% of the initial value (during growth in media) and was similar to a metabolic rate of 1.5% of heat-killed cells. When supplementing medium after 2.2 years of starvation, metabolic activity was reported to increase to app. 60% of the initial value after only 1 hour at 37°C and to 100% after 7.7 h. Starved cells were completely resistant to the antibiotics pyrazinamide and isoniazid, however not to kanamycin. Supported also by other data, the authors suggested an organized metabolic shutdown to enter a dormant state as part of a survival strategy ( Archuleta et al. 2005).
- An Australian study examined the survival of M. avium subsp. paratuberculosis in dam water and sediment ( Whittington et al. 2005). Water troughs were artificially contaminated with a mixture of chaff, water, and feces containing the target bacterium to a final concentration of app. 2 x 106 viable M. avium subsp. paratuberculosis. The dam water was untreated and contained numerous unidentified invertebrates and protozoa. Cultivability in water and/or sediment was found up to 16 weeks in a location semi-exposed to sunlight and up to 36 weeks in a shaded location. Survival in sediment was 12 to 20 weeks longer than survival in the water column ( Whittington et al. 2005).
- Whan et al. reported phagocytosis of M. avium subsp. paratuberculosis by both Acanthamoeba castellanii and Acanthamoeba polyphaga ( Whan et al. 2006). Whereas bacterial numbers (determined by culture) did not change significantly during the first 7 days, a 1-1.5 log increase in bacterial numbers was seen by day 24 with both amoeba species.
- Mura et al. used qPCR to examine survival of a bovine and a human strain of M. avium subsp. paratuberculosis in the presence of A. polyphaga ( Mura et al. 2006). Approximately 106 amoebal trophozoites were infected with mycobacteria. In a short time experiment, bacterial numbers dropped slightly by day 8 until as measured by qPCR. This was followed by a recovery until day 16 with final numbers equaling or exceeding the initial numbers. In a long term experiment, encystment of amoeba after 4 weeks resulted in a 2-log reduction in mycobacteria. Numbers recovered, however, until the end of the study period of 24 weeks.
There is almost no data available on infective doses of NTM. Barriers, such as the in ability to identify “virulent” types of M. avium
, listed on EPA’s Candidate Contaminant List, do not allow performance of risk assessment. NTM infections occur in both immunocompetent (primarily pulmonary) and in immunocompromised individuals (primarily AIDS patients and those with cancer or undergoing chemotherapy). In immunocompetent individuals, a number of risk factors have been identified (above) and it is likely that disease is a consequence of the interaction of multiple factors including the immune status of the host and the number and relative virulence of the individual mycobacterium ( Health Canada 2006
Molecular identification of waterborne mycobacteria is normally performed after isolation by cultivation ( Vaerewijck et al. 2005
). Molecular methods have mainly been developed for mycobacterial detection and strain differentiation in clinical samples.Selected molecular detection methods are summarized in the following:
- PNA-FISH: Lehtola et al. (2006) described a high-affinity peptide nucleic acid (PNA) probe for detecting different M. avium species including M. avium subsp. avium, M. avium subsp. paratuberculosis, and M. avium subsp. silvaticum. The method was successfully used to detect NTM in spiked water samples and lab-grown potable water biofilms ( Lehtola et al. 2006).
- Species-specific DNA probes are commercially available from a number of companies including:
AccuProbe® (from Gen-Probe Inc.): HPA with luminometer detection for culture identification.
GenoType® Mycobacterium (from Hain Lifesciences): 23S rRNA regions are PCR amplified, reverse hybridized to a strip with immobilized species specific probes ( Richter et al. 2006).
INNO-LiPA Mycobacteria (from Innogenetics N.V.): also a DNA strip assay for detection of 16S-to 23S rRNA ITS detection of PCR products ( Richter et al. 2006).
PCR assays targeting 16S rRNA, hsp65, 32-kDa protein genes or ITS sequences have been described for detecting and differentiating different mycobacterial species ( Telenti et al. 1993; De Beenhouwer et al. 1995; Kox et al. 1995; Steingrube et al. 1995; Park et al. 2000; Haefner et al. 2004; Prammananan et al. 2005). Identification is achieved by sequencing, RFLP, or hybridization to specific probes (OSCPH-EIA, reverse cross blot hybridization). Sensitivities as low as approximately 2-3 NTM cells have been reported. Confirmation that the amplified PCR product is often required ensuring specificity. A PCR-based method designed for detection of NTM in drinking water biofilms showed a sensitivity of app. 100 M. avium cells/sample ( Schwartz et al. 1998). Genus-specific primers targeted a 16S rRNA gene region with PCR products being subsequently hybridized with a second mycobacteria-specific probe. The two-step approach was reported to reduce the probability of false-positive detection.
IMS-PCR: Two assays were reported amplifying a gene coding for a mycobacterial antigen (for M. ulcerans detection) and a multicopy DNA insertion element IS900 (for M. avium subsp. paratuberculosis detection) in water after IMS purification of target cells. Detection limits app. 100 M. ulcerans/ml (Roberts et al. 1997) and 10 CFU/ml M. avium subsp. paratuberculosis cells were reported ( Whan et al. 2005).
qPCR: The assay targeting 16S rRNA genes was designed for detecting and differentiating clinically relevant mycobacteria in clinical specimens. Different species were differentiated by melting point analysis ( Shrestha et al. 2003). A qPCR assay for detecting M. avium subsp. paratuberculosis was developed targeting the IS900 insertion element specific for this bacterium. The assay included an internal amplification control could consistently and quantifiably detect 102 target cells in 20 ml of artificially contaminated drinking water (Rodriguez-Lazaro et al. 2005).
Rapid-cycle PCR: The assay amplifies a 1,000 bp 16S rRNA gene segment and differentiates between Mycobacterium tuberculosis complex, Mycobacterium avium, and other nontuberculous mycobacteria using different specific FRET probes. It utilizes the LightCycler platform, includes an IAC and was reported detect as few as 5-10 genome copies per reaction. Evaluation was performed with cultured isolates ( Lachnik et al. 2002).
PCR-microarray: The method uses a genus-specific probe and 20 species-specific probes for detection and genotyping of medically important species. Polymorphic regions of the ITS sequences were PCR amplified using labeled primers and hybridized to the microarray. The time demand for complete procedure was reported to be 4.5 hours. Validation was performed with cultured clinical isolates and specimens. An initial DNA concentration of 5 pg/µl could be detected, corresponding to about 1,000 genome copies ( Park et al. 2005). The mycobacterial ITS-region was identified in an earlier report as a suitable target for differentiating epidemiologically relevant species and subspecies using multiplex PCR ( Park et al. 2000). Tobler et al. utilized the gene encoding the 65kDa heat shock protein for detecting and identifying different Mycobacterium species. The PCR-diagnostic microarray approach was designed for differentiating up to 37 different species ( Tobler et al. 2006).
NASBA: A NASBA-based assay (GenoType Mycobacteria Direct, GTMD; Hain Lifescience GmbH) is commercially available for the detection of Mycobacterium tuberculosis complex, M. avium, M. intracellulare, M. kansasii, and M. malmoense. Although the test is developed for detection from clinical specimens, it might be an interesting technology for environmental applications.
GC-MS: A gas chromatography-mass spectrometry method was developed for detection of M. xenopi in drinking water based on detection of trace levels of the secondary alcohol 2-docosanol that is characteristic for M. xenopi. This compound was detected in 7 of 10 drinking water samples that were culture-positive for this species. The three GC-MS negative samples contained very low M. xenopi numbers. The method cannot differentiate between live and dead bacteria, but results were in agreement with culturing results. Although the detection limit was not determined, the assay was considered very sensitive. Results could be obtained within 2 days of receipt of samples ( Alugupalli et al. 1992).
The following pictures are part of the Public Health Image Library (PHIL) from the Centers for Disease Control and Prevention (CDC). The photos are in the public domain and thus free of copyright restrictions. We would like to express our appreciation for providing these images.
This colorized scanning electron micrograph (SEM) depicts some of the ultrastructural details seen in the cell wall configuration of a number of Gram-positive Mycobacterium tuberculosis bacteria. TB bacteria become active, and begin to multiply, if the immune system can't stop them from growing. The bacteria attack the body and destroy tissue. If in the lungs, the bacteria can actually create a hole in the lung tissue. Some people develop active TB disease soon after becoming infected, before their immune system can fight off the bacteria. Other people may get sick later, when their immune system becomes weak for another reason.
Photo ID: 9997
Content provider(s): Centers for Disease Control and Prevention/Dr. Ray Butler
Photo credit: Janice Carr
Under a magnification of 3841X, this scanning electron micrograph SEM) reveals some of the ultrastructural morphologic details exhibited by a number of Gram-positive bacilli, or “rod-shaped”, Mycobacterium fortuitum bacteria.
Photo ID: 11033
Content provider(s): Centers for Disease Control/ Margaret M. Williams; Janice Haney Carr
Credit: Janice Haney Carr