Serodiagnosis of environmental mycobacterial infections

Journal of Microbiological Methods 86 (2011) 283–290

Contents lists available at ScienceDirect

Journal of Microbiological Methods j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j m i c m e t h

Serodiagnosis of environmental mycobacterial infections Henriette Stavri a,⁎,⁎⁎⁎, Irina Ulea a,⁎⁎⁎, Dorel L. Radu a, e, Manuela Gheorghiu Branaru b, Olga Moldovan b, Miron A. Bogdan b, Cornelia Tudose b, Marinela Raileanu b, Dan Duiculescu c, Luminita Ene c, Viorel Olar d, Catalin Ionita d, Gabriela Loredana Popa e,⁎⁎⁎, Mircea I. Popa e,⁎⁎⁎, Patrick J. Brennan f,⁎⁎,⁎⁎⁎ a

Cantacuzino Institute, Mycobacterial Antigens Department, Bucharest, Romania “Marius Nasta” Pneumophtisiology Institute, Bucharest, Romania “Dr. Victor Babes” Hospital for Tropical and Infectious Diseases, Bucharest, Romania d “Carol Davila” Emergency Military Clinical Hospital, Bucharest, Romania e “Carol Davila” University of Medicine and Pharmacy, Bucharest, Romania f Colorado State University, Fort Collins, CO, USA b c

a r t i c l e

i n f o

Article history: Received 12 January 2011 Received in revised form 12 May 2011 Accepted 14 May 2011 Available online 20 May 2011 Keywords: ELISA Immunomagnetic Environmental mycobacteria Serodiagnosis Tuberculosis

a b s t r a c t To demonstrate the usefulness of enzyme-linked immunosorbent assay for serodiagnosis of mycobacterioses due to environmental mycobacteria we utilized a panel of glycolipid antigens selective for Mycobacterium avium–intracellulare, Mycobacterium kansasii, Mycobacterium xenopi, Mycobacterium scrofulaceum and Mycobacterium gordonae. The levels of circulating antibodies were determined against the environmental mycobacteria, and Mycobacterium tuberculosis in human immunodeficiency virus-negative and -positive patient sera. The method used immunomagnetic separation of the antigens, with covalent immobilization of antibodies to superparamagnetic amine and carboxyl terminated particles in solutions of the specific antigens. Enzyme-linked immunosorbent assay was performed on 195 patient sera: 34 with infections due to environmental mycobacteria, 114 with tuberculosis, 47 with other respiratory diseases. There were 46 human immunodeficiency virus-1 infected individuals. Among the 34 infections due to environmental mycobacteria, 9 patients were singularly infected with an environmental mycobacterium, and 25 co-infected with both M. tuberculosis and an environmental mycobacterium. Sensitivity, specificity and false positivity ranges were determined for each of the volunteer groups: tuberculosis positive, human immunodeficiency virus negative; tuberculosis positive, human immunodeficiency virus positive; those with infections due to individual environmental mycobacteria (such as M. scrofulaceum and M. kansasii); and those with other respiratory diseases. We demonstrate that such multiple assays, can be useful for the early diagnosis of diverse environmental mycobacterial infections to allow the start of treatment earlier than henceforth. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Environmental mycobacteria (EM) are emerging pathogens causing opportunistic infections in humans and animals (Primm et al., 2004) and include those Mycobacterium species that are not members of the Mycobacterium tuberculosis (M. tuberculosis) complex or M. leprae. Only some of these mycobacteria, natural inhabitants of

soil and waters, are pathogenic, for the healthy immunocompetent adult (Falkinham, 1996; Phillips and von Reyn, 2001). Diagnosis of EM infections is difficult, particularly as mixed infections with M. tuberculosis but important due to different treatment regimens (Nagai et al., 2001). The gold diagnostic standard is culture, since EM and M. tuberculosis exhibit distinctly different morphological and biochemical features (Heifets and Desmond, 2005); however, culture

Abbreviations: EM, environmental mycobacteria; M. tuberculosis, Mycobacterium tuberculosis; TB, tuberculosis; AFB, acid-fast bacillus; ELISA, enzyme-linked immunosorbent assay; GPL, glycopeptidolipid; LOS, lipooligosaccharides; PGL, phenolic glycolipids; MAI, M. avium/M.intracellulare; ATS, American Thoracic Society; NTM, non-tuberculous mycobacteria; ORD, other respiratory diseases; GL, glycolipid; TLC, thin layer chromatography; PBS, phosphate buffered saline; IgG, immunoglobulins G; BSA, Bovine Serum Albumin; MES, 2-(N-morpholino)ethanesulfonic acid; MP, magnetic particles; C18, MPG CPG-C18, (magnetic porous glass particles, with a lateral chain of 18 Carbon atoms; BM, BioMag® superparamagnetic Iron Oxide; HRP- SpA, Horseradish peroxidase- Staphylococcal protein A; PPV, positive predictive value; NPV, negative predictive value; FNR, false negative rate; FPR, false positive rate; GL-NH2, GL-BioMag ®Plus Amine- M. tuberculosis H37Rv IgG; GL-COOH, GL-BioMag ®Plus Carboxyl- M. tuberculosis H37Rv IgG; OD, optical density; IU, immunoenzymatic units; HIV, human immunodeficiency virus. ⁎ Correspondence to: H. Stavri, Cantacuzino Institute, Mycobacterial Antigens Department, Bucharest, Romania. ⁎⁎ Correspondence to: P.J. Brennan, Department of Microbiology, Immunology and Pathology, C321 Microbiology Building, 200 Lake Street, Colorado State University, Fort Collins, CO 80523-1682, USA. Tel.: + 1 970 491 6700; fax: + 1 970 491 1815. E-mail addresses: [email protected] (H. Stavri), [email protected] (P.J. Brennan). ⁎⁎⁎ Principal Authors. 0167-7012/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.mimet.2011.05.010

284

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

is a protracted process due to the notoriously slow growth rates of all species of Mycobacterium. Accordingly, rapid, sensitive and inexpensive assays to confirm EM infections, as distinct from tuberculosis (TB) are required. Molecular techniques such as PCR and the use of DNA probes, or mycolic acid analysis by high pressure liquid chromatography (Heifets, 2004) are unaffordable for the developing world or for small clinical settings (Tonjum et al., 1998). A variety of immunological, particularly serological, tests have been developed to identify those infected with M. tuberculosis (Amicosante et al., 1999; Andersen et al., 2000; Daniel, 1996; Garg et al., 2003; Hoff et al., 2007; Kitada et al., 2005; Stavri et al., 2005). The search for a convenient serological test for TB and EM diseases is challenging because it requires balancing sensitivity against specificity (Abebe et al., 2007; Julian et al., 2004). The results of M. tuberculosis serologic diagnosis showed a high variability due to the enzyme-linked immunosorbent assay protocols used (different antigens, such as culture filtrate antigens (Laal et al., 1997), purified extracts of glycolipid (Escamilla et al., 1996), mycobaterial sonicates (Rosen, 1990), or more defined mycobacterial antigens such as 38-kDa (Rv0934), MPT59 (Rv1886c), MPT51 (Rv3803c) and LAM (Garg et al., 2003; Daniel, 1987). The sensitivities and specificities of the 38 kDa antigen in ELISA-based serological tests vary with the population, from respectively 89%/93% for Chinese (Cole et al., 1996), 55%/98% for Polish (Demkow et al., 2004), to 29%/96% for South Africans (Lodes et al., 2001). These subjects had all pulmonary TB and smear/culture positive status. When glycolipid antigens were utilized in ELISA, the sensitivities and specificities, based on positive acid-fast bacillus and culture of M. tuberculosis, vary from 60 to 71%/100% (Julian et al., 2004), to 75.2%/91,8% (Niculescu et al., 1995). The use of purified, species specific antigens, particularly in simple ELISA formats, could fulfill the criteria of high sensitivity and specificity, and affordability for diagnosis of EM infections. However, to date, no rapid antibody detection systems for diagnosis of mycobacterial infections other than TB, have been reported. Yet, the wide diversity of mycobacteria within the environment is matched by an equally diverse range of species and serotype-specific glycolipid antigens, variously known as glycopeptidolipids (GPLs), lipooligosaccharides (LOSs) and phenolic glycolipids (PGLs) (Brennan et al., 1982). In this paper, we developed a diagnostic immunoenzymatic assay based on some of these specific antigens for individuals infected with the most widely encountered species i.e. M. avium/M. intracellulare (MAI) complex, M. kansasii, M. xenopi, M. scrofulaceum, and M. gordonae.

Dr. Victor Babes Infectious and Tropical Diseases Hospital; and Carol Davila Emergency Military Clinical Hospital. Selection criteria included: (a) patients with suspected EM infection or EM/M. tuberculosis coinfection, based on clinical, radiological data supported by bacteriological tests for EM; (b) patients with EM infections (positive controls who met clinical, radiological, and bacteriological criteria for mycobacterioses and were classified according to the American Thoracic Society (ATS) guidelines for diagnosis of non-tuberculous (NTM) related diseases (American Journal of Respiratory and Critical Care Medicine, 1997)); (c) patients with TB diagnosed according to the World Health Organization guidelines (WHO, 2009); (d) patients with other respiratory diseases (ORD) or without any EM infections (negative controls) including: lung infections (acute respiratory tract infections, i.e. lower respiratory tract infections and pneumonia but not nosocomial forms); acute or chronic bronchitis; and asthma. Bacteriological diagnosis (direct examination by microscopy, culture and resistance testing) was performed on all patients. All diagnoses of mycobacteriosis or/and TB were confirmed by identification of bacteria by culture according to standard methodology, using Lowenstein-Jensen medium. The acid fast smears were prepared from sputum, by the modified Petroff technique, described by Kent and Kubica (1985). Sputum specimens for Ziehl–Neelsen staining were obtained from some patients in the first three days after admission to hospital, usually before the beginning of specific treatment (Kubica, 1984), while others have been admitted to hospital for control, during and after their treatment. The sera were collected in the same way, either at the start of the treatment, for suspected infection with EM or/and M. tuberculosis or ORD, or when the patients have been observed, during or after therapy. All samples were tested without knowledge of the clinical diagnosis of each patient, but the study groups were not randomly selected, as samples were intentionally collected for the development of the ELISA EM from patients strongly suspected of having mycobacteriosis or tuberculosis and from those with other respiratory diseases. In addition to positive AFB and culture of M. tuberculosis, radiographic evidence and clinical symptoms were used as the basis for diagnosis to estimate sensitivities and specificities. The study was approved by local ethical committee (IRB #0002508) and the patients provided informed consents to participate in the study. All sera samples were blinded and stored at −70 °C. All of the subjects in this study had been vaccinated at birth with BCG according to the Romanian National Immunization Program. The subjects recruited from the three centers and their bacteriological diagnoses are presented in Table 1.

2. Materials and methods 2.1. Study groups

2.2. Growth of mycobacteria

Serum samples were obtained from 195 patients (12–72 years old) from three centers in Bucharest: M. Nasta Pneumophtysiology Institute;

The mycobacteria reference species used were obtained from: Pasteur Institute Paris, France (M. tuberculosis H37Rv, M. kansasii,

Table 1 Distribution of subjects according to culture results and HIV status. Culture results

M. tuberculosis EM MAI M. kansasii M. xenopi M. gordonae M. scrofulaceum Negative Total

No. of subjects

114 34 12 5 11 4 2 47 195

Hospital 1

Hospital 2

Hospital 3

HIV status positive/negative

HIV status positive/negative

HIV status positive/negative

0/57 0/27 0/8 0/5 0/9 0/3 0/0 0/21 0/103

27/0 3/0 2/0 0/0 0/0 1/0 0/2 14/0 44/0

0/48 0/4 0/2 0/0 0/2 0/0 0/0 0/12 0/48

Hospital 1 — “Marius Nasta” Pneumophtisiology Institute, Bucharest; Hospital 2 —“Victor Babes” Hospital for Tropical and Infectious Diseases; and Hospital 3 — “Carol Davila” Emergency Military Clinical Hospital.

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

M. scrofulaceum, M. gordonae); Dr. Joseph Lam, Department of Microbiology, College of Biological Sciences, University of Guelph, Ontario, Canada (MAI); and American Type Culture Collection (M. xenopi). All species were cultivated on Lowenstein-Jensen medium (Metchock et al., 1999), subcultured for four weeks, on potato-synthetic liquid Sauton medium (Osborn, 1979). Then the inocula were cultured on the surface of liquid Sauton medium (Allen, 1998) for 4–6 weeks at 37 °C in order to obtain sufficient bacterial mass as a surface pellicle for antigen isolation; the exception was M. xenopi which was grown at 42 °C. 2.3. Extraction procedure and purification of glycolipid antigens The characteristic glycolipid (GL) antigens were extracted from all species according to Brennan et al. (1978). Thin layer chromatography (TLC) of the glycolipids was performed according to Brennan et al. (1982), using TLC aluminum sheets 20 × 20 cm Silica gel 60 F254 Merck with a concentrating zone of 20 × 2.5 cm. The solvent system used was CHCl3:CH3OH:H2O (60:35:8), and the chromatograms were visualized after spraying and charring with 0.1% orcinol in 40% aqueous H2SO4. 2.4. Preparation of anti-sera Positive control sera developed for ELISA EM were obtained by a modification of the method of Janicki et al. (1971). Suspensions of each heat inactivated bacillus (MAI, M. kansasii, M. xenopi, M. gordonae, M. scrofulaceum and M. tuberculosis H37Rv), 40 mg/ml in 0.2 M phosphate-buffered saline (PBS), pH 7.4, were disrupted by sheer stress (French Pressure Cell Apparatus), 16,000 psi, and mixed with equal volumes of incomplete Freund adjuvant. Rabbits (3 animals/mycobacterial species) were injected with the above suspension, 0.1 ml per animal, s.c., on four limbs, weekly, for 4 weeks. The animals were bled two weeks after the last inoculation and the sera were pooled. The immunoglobulins G (IgG) were separated from the sera according to Harboe and Ingild (1973). The production of anti-M. tuberculosis H37Rv serum was similar to that for positive controls, except that heat inactivated cells were mixed with 6 weekold unheated culture filtrate of M. tuberculosis H37Rv (filtered through a 0.2 μ Millipore filter), to 40 mg/ml concentration. The mixture was injected into three goats, each animal receiving 1 ml suspension. The animals were bled after 6 weeks, the sera were pooled and IgG was separated as described and the protein concentration was determined, 36 mg/ml (Lowry et al., 1951). The goats were used for M. tuberculosis antisera production due to the higher yield that could be obtained. We had used these sera for different scientific purposes. The anti-mycobacterial goat antisera had been used in many studies (Daniel et al., 1984; Shende et al., 2007). 2.5. Activation and immunomagnetic purification of the antigens Activation of functionalized superparamagnetic microparticles (BioMag ®Plus Amine terminated and BioMag ®Plus Carboxyl terminated; Polysciences Inc., Eppelheim, Germany), was performed according to manufacturer's protocol. For each mg of activated BioMag ®Plus Amine particles, 0.2 mg/5.6 μl M. tuberculosis H37Rv IgG and 14.4 μl Pyridine Washing Buffer, 0.01 M, pH 6 were utilized. For each mg of activated BioMag ®Plus Carboxyl particles, 0.5 mg/15 μl M. tuberculosis H37Rv IgG and 0.05 mg of a carrier protein, Bovine Serum Albumin (BSA) Fraction V, were utilized for blocking. The BSA was dissolved in 10 μl, 0.05 M 2-(N-morpholino) ethanesulfonic acid (MES) buffer, pH 5.2. The coupling efficiencies when using BioMag ®Plus Amine and BioMag ®Plus Carboxyl–M. tuberculosis H37Rv IgG were calculated according to the manufacturer data sheets (Polysciences, Inc.). To determine the coupling efficiency, the absorbance of pre- and post-coupling solutions, with Pyridine

285

Washing Buffer as control, was measured at 280 nm. The efficiency was expressed as the % protein uptake, using the following formula:  A280

D

Pre Coupling Solution

  × D − A280

 A280Pre

Post Coupling Solution

Coupling Solution

 ×D

 × D × 100

dilution.

Immunomagnetic purification of rabbit IgGs was performed with two types of particles: MPG CPG-C18 (magnetic porous glass particles, with a lateral chain of 18 Carbon atoms (beads with Fe 2+, 500 Å in size)), obtained from CPG Inc., NJ, USA (C18) and BioMag® superparamagnetic Iron Oxide (BM), Polysciences Inc., a suspension of particles, 10 μm in size, bound to M. tuberculosis H37Rv cells. The MP– cells complexes were utilized for immunomagnetic separation of the M. tuberculosis H37Rv common antibodies from each of the control sera raised on rabbits against the mycobacterium species of interest. The specificity of these IgGs was tested in ELISA EM. Using a magnetic separation device, the purified antigens and antibodies were disassociated from the magnetic particles (MP) complexes, before being used in ELISA. 2.6. ELISA assay ELISA BK, is an in-house assay developed by “Cantacuzino” Institute, using GL antigens isolated from M. tuberculosis H37Rv (Niculescu et al., 1995) and Horse Radish Peroxidase - Staphylococcal Protein A (HRP-SpA) conjugate as secondary antibody. Briefly, ELISA plates were coated with 0.1 ml of 10 μg/ml n-hexane soluble fraction of antigens isolated from M. tuberculosis H37Rv. The plates were allowed to dry overnight at room temperature and were either used immediately or sealed in aluminum foil and stored at 4 °C. The plates were blocked with 1% skim milk in PBS, then incubated with 0.1 ml patient serum, previously diluted 1/25 in PBS, for 30 min at 37 °C. The plates were washed and the bound antibodies were detected by incubation with HRP-SpA conjugate as secondary antibody, 30 min at 37 °C. After the plates were washed, orto-phenylene di-amine substrate was added for 30 min to develop color, and the reaction was read at 450 nm. The multiple EM ELISA method was performed similarly, except the fact that each EM magnetically purified GL antigens, diluted in nhexane (60 μg/ml), 0.1 ml per well, was coated on a different microplate. The polystyrene microplates with lids (F16, Maxisorp, Nunc-Immuno Module; Nalgen Nunc International) contain 6 separate modules (16 wells) in a frame. For one multiple EM assay, 6 modules were used, one for each of 5 different EM GL antigens, and one for the M. tuberculosis antigen. The absorption was measured in a microplate reader (dual wavelength operation mode at 450 nm with the background absorption measured at a wavelength of 630 nm; the final absorbance value was calculated as A450–A630). 2.7. Data analysis Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), false negative rate (FNR), false positive rate (FPR), accuracy and error were calculated according to Altman and Bland (1994) and standard methods (Chinn, 2001). Sensitivity was defined as ability of ELISA EM to detect cases of EM infection (or TB, for ELISA BK). Specificity was defined as the ability of ELISA EM results to be negative for TB and ORD (or ORD, in ELISA BK) patients. PPV was defined as EM infected patients vs. EM ELISA positive results ratio. NPV was defined as the ratio between patients non infected with EM (TB plus ORD patients) and EM ELISA negatives. FPR = 1 − specificity; FNR = 1 − sensitivity; Accuracy = ratio between EM patients,

286

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

correctly diagnosed by ELISA EM and the overall number of EM patients. Error = ratio between EM patients incorrectly diagnosed by ELISA EM and the overall number of EM patients. 3. Results 3.1. Immunomagnetic purification of antibodies and antigens M. tuberculosis H37Rv IgG coupling efficiency was calculated according to manufacturer, expressed as the% protein uptake Typical values of protein uptake, according to the manufacturer, should be N60%. Our results were higher for BioMag ®Plus Carboxyl– M. tuberculosis H37Rv IgG, i.e. 96.5% and lower for BioMag ®Plus Amine–M. tuberculosis H37Rv IgG GL-BioMag®PlusAmine– M. tuberculosis H37Rv IgG (GL–NH2), i.e. 40.5%. The coupling efficiency of mycobacterial GL-BioMag®PlusAmine - M. tuberculosis H37Rv (GL-NH2) and mycobacterial GL-BioMag ®Plus Carboxyl M. tuberculosis H37Rv (GL -COOH) was calculated using the same formula as that for M. tuberculosis IgG. For GL-NH2, the efficiency ranged between 82.8% (M. gordonae) and 8.7% (M. xenopi), while for GL– COOH, the highest efficiency was 56% (M. kansasii) and the lowest 14% (M. xenopi). Our results proved that, although the mycobacterial GLs may not possess an available amine group, they were stably adsorbed to BioMag particles. TLC was performed on all GLs before and after MP purification (BioMag ®Plus Amine and BioMag ®Plus Carboxyl) in order to visualize the differences. The bands were numbered based on their mobility on TLC plates. When the GLs were immunomagnetically purified, the differences in mobility were evident (Figs. 1 and 2 are but Fig. 2. TLC of immunomagnetically purified M. scrofulaceum. Solvent: chloroform/ methanol/water (60:35:8). The bands were numbered based on their mobility on TLC plate. M. scrofulaceum: II, IV, VI and VII; M. scrofulaceum–COOH: VI and VII; and M. scrofulaceum–NH2: II, IV, VI and VII.

two examples). TLC of GLs before and after magnetic purification, corroborated with ELISA results, shows that COOH-terminated particles allowed better purification of the M. kansasii, M. xenopi and M. scrofulaceum glycolipid antigens whereas NH2-terminated particles act better for those from MAI and M. gordonae. C-18 and BM superparamagnetic iron oxide particles were bound to M. tuberculosis H37Rv cells, in order to perform the immunomagnetic separation of the common antibodies from each of the control sera raised on rabbits against the mycobacterium species of interest (MAI, M. kansasii, M. xenopi, M. gordonae, M. scrofulaceum). We have compared ELISA results obtained with three types of positive controls, mycobacterial IgGs and IgGs additionally immunomagnetically purified with C18 and BM superparamagnetic iron oxide (Fig. 3). The results showed that the controls immunomagnetically purified with BM, had a higher specificity than those obtained with CPG-C18, i.e. the positive controls had higher values for the homologous species and lower for the others (Table 2).

3.2. ELISA results

Fig. 1. TLC of immunomagnetically purified MAI. Solvent: chloroform/methanol/water (60:35:8). The bands were numbered based on their mobility on TLC plate. MAI: I, IV, VI and VII; MAI–COOH: I, IV, VI and VII; and MAI–NH2: VI and VII.

Therefore, in the ELISA assays of patient sera, we used the controls purified with BM particles. The purified controls were used at different dilutions, depending on the quality of each antibody. The cut-off value separating positive from negative sera was calculated for each test: Cut off = Control OD × 0.18. The 0.18 value was chosen so that the cut off could represent 2 standard deviations above optical density (OD) average values for patients with ORD, known not to be infected with M. tuberculosis or EM. For ELISA BK, the cut-off provided a 75.2% sensitivity and a 91.8% specificity, when 1072 sera were tested

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

287

Fig. 3. Multiple EM and BK ELISA with BM particles purified IgG controls. Antigens: M. tuberculosis H37Rv; M. kansasii–COOH; MAI–NH2; M. xenopi–COOH; M. scrofulaceum–COOH; and M. gordonae–NH2. EM IgG controls purified with BM: C+ M. xenopi, C+ M. kansasii, C+ MAI. C+ M. scrofulaceum, and C+ M. gordonae. BK IgG control: C+ M. tuberculosis. C plate: PBS.

(367 from active TB patients and 705 from ORD patients), leading to a good balance of sensitivity and specificity in diagnosing TB. We have considered the limits of positive control OD to range between 0.800 and 0.1200. The test was positive if the determined OD was higher than the calculated cut off. The results could be also

evaluated in immunoenzymatic units (IU), according to the formula:

Patient serum OD−Cut off : Control OD−Cut off

Table 2 ELISA specific controls purified with C18 and BM vs. native controls (OD450). GL M. tb C+ M tb C+ MAI C+ MAI C18 C+ MAI BM C+ Kans C+ Kans C18 C+ Kans BM C+ Xen C+ Xen C18 C+ Xen BM C+ Scrf C+ Scrf C18 C+ Scrf BM C+ Gord C+ GordC18 C+ Gord BM Cplate

GL MAI

GL MAI COOH

GL MAI NH2

GL Kans

GL Kans COOH

GL Kans NH2

GL Xen

GL Xen COOH

GL Xen NH2

GL Scrf

GL Scrf COOH

GL Scrf NH2

GL Gord

GL Gord COOH

GL Gord NH2

1200

765

430

450

530

345

505

925

410

405

520

360

430

320

340

300

920

955

970

1075

430

430

630

1000

980

510

780

580

755

480

470

340

900

1075

1010

1090

870

830

390

960

885

460

750

650

680

490

410

380

820

1085

1090

1170

835

220

300

690

345

305

705

495

590

580

500

455

560

235

180

290

1090

1285

1120

410

255

200

455

340

330

340

350

320

360

315

175

160

1050

1305

1165

220

200

185

320

300

295

280

285

235

460

225

150

140

1110

1350

1205

270

170

120

290

270

255

280

265

230

455

505

475

560

605

455

565

955

1045

605

1095

790

950

860

785

660

525

480

465

520

580

450

530

990

1190

1010

940

610

780

840

780

605

250

430

425

480

550

420

480

1080

1320

1100

765

605

980

800

750

720

660

605

570

490

545

345

420

780

590

670

1090

1210

1120

515

420

370

650

620

545

600

510

405

460

400

340

370

960

1090

1020

490

415

390

595

460

390

325

305

290

295

370

305

360

1100

1320

1150

395

290

330

465

380

290

310

655

500

575

650

540

585

685

700

770

820

925

880

450

375

300

320

600

520

575

585

470

500

610

490

535

805

920

900

325

300

275

305

510

395

420

510

400

455

580

500

485

910

1090

1165

15

15

15

15

15

15

15

15

10

15

15

10

15

15

15

15

M. tb: M. tuberculosis; MAI: M. avium–intracellulare; Kans: M. kansasii; Xen: M. xenopi; Scrf: M. scrofulaceum; Gord: M. gordonae; GL: glycolipid; C+: positive control; BM: BioMag® superparamagnetic Iron Oxide; C18: MPG CPG-C18, magnetic porous glass particles, with a lateral chain of 18 Carbon atoms; and Cplate: plate control.

288

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

Table 3 ELISA EM results on culture positive TB patients. Patients ELISA EM ELISA EM Specificity % False positive no. negative false positive results % MAI M. kansasii M. xenopi M. gordonae M. scrofulaceum

114 114 114 114 114

94 101 92 95 93

20 13 22 19 21

85.1 89.8 83.8 85.7 84.4

14.9 10.2 16.2 14.3 15.6

Table 4 ELISA EM results on culture negative patients with other respiratory diseases.

MAI M. kansasii M. xenopi M. gordonae M. scrofulaceum

Patients no.

ELISA EM negative

ELISA EM false positive

Specificity %

FPR %

47 47 47 47 47

45 47 44 43 44

2 0 3 4 3

95.9 100 94 92.1 94

4.1 0 6 7.8 6

FPR: false positive results.

The results ranged between 0 and 100 IU. Positive: 10 IU; low positive: 1–9 IU; and negative: 0 UI. If the result was lower than 10 IU, it is recommended to repeat the assay. Taking into consideration the results of coupling efficiency of GLs purified with BioMag ®Plus Amine–M. tuberculosis H37Rv IgG and BioMag ®Plus Carboxyl–M. tuberculosis H37Rv and the percentage of specific glycolipids we used the following GLs in EM ELISA: MAI GL– NH2, M. kansasii and M. xenopi GL–COOH, M. gordonae GL–NH2 and M. scrofulaceum GL–COOH. EM ELISA was performed on 195 patient sera, each serum being simultaneously tested with 5 GLs, immunomagnetically purified with GL–NH2 or with GL–COOH, depending of the species. The statistical data concerning multiple EM ELISA results were calculated. 26 out of 34 patients with EM infections, culture positives, proved to be positives in EM ELISA, i.e. 80.9% sensitivity for all EM. The highest sensitivity was for M. kansasii, i.e. 100% and the lowest for M. scrofulaceum, i.e. 66.6%. The FNR of EM ELISA was calculated as 19.1%. Only 8 patients were infected with a single EM; 26 out of 34 EM patients were co-infected with TB. The specificity of EM ELISA was calculated for culture positive TB patients (Table 3), for each of the EM species. Therefore, we have determined that it ranged between 83.8% (M. xenopi) and 89.8% (M. kansasii).We have determined also the

Fig. 4. ELISA EM and ELISA BK quantitative results (IU) in sera from EM (n = 8) and EM–M. tuberculosis co-infected (n = 26) patients. Each dot indicates an individual serum sample. The Ox axis indicate cut-off values separating positive from negative sera and was calculated for each test: Cut off = Control OD × 0.18. The 0.18 value was chosen so that the cut-off could represent 2 standard deviations above 450 nm optical density (OD) average values for patients with other respiratory diseases, known not to be infected with M. tuberculosis or EM. EM, environmental mycobacteria; M. tb, M. tuberculosis. ELISA BK, in-house assay, using GL antigens isolated from M. tuberculosis H37Rv and HRP-SpA conjugate as secondary antibody. IU, imunoenzimatic units.

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

289

Table 5 Data analysis of ELISA EM results on EM culture positive patients. ELISA EM

Patients no.

ELISA EM positive

ELISA EM false negative

Sensitivity

FNR

Accuracy

Error

MAI M. kansasii M. xenopi M. gordonae M. scrofulaceum Total

12 5 11 4 2 34

9 5 8 3 1 26

3 – 3 1 1 8

80 100 78.5 80 66.6 80.9

20 0 21.5 20 33.4 19.1

75 100 72.7 75 50 76.5

25 0 27.3 25 50 23.5

specificity of ELISA EM on culture negative patients (ORD), which ranged between 92.1% for M. gordonae and 100% for M. kansasii (Table 4). ELISA BK, performed simultaneously with ELISA EM on 114 culture positive, immunocompentent and immunosuppressed, TB patients had an overall sensitivity of 77% and specificity of 95.9%. Immunosuppressed patients' sera usually did not respond strongly in ELISA. The ELISA BK sensitivity for HIV positive patients, fell to 46.8%, while it rose to 88.5% for 85 immunocompetent TB patients. The specificity of ELISA BK was high, but probably this value may be biased due to severe immunosuppression (data not shown). ELISA EM and ELISA BK quantitative results (IU) on EM and EM–M. tuberculosis coinfected patients are graphically presented in Fig.4. The sensitivity and FNR of ELISA EM on EM culture positive patients are presented in Table 5. ELISA EM had an overall 80.9% sensitivity, ranging from 66.6% (M. scrofulaceum) to 100% (M. kansasii), specificity ranging from 83.8 to 89.8% (TB) and 92.1–100% (ORD), PPV: 61.5–72.7% (TB) and 88.8–100% (ORD); negative predictive values (NPV: 98.1– 100%; FPR: 7.5–11.6%; FNR ranging from 0 (M. kansasii) to 33.4 (M. scrofulaceum), overall accuracy 76.5%, ranging between 50 (M. scrofulaceum) and 100% (M. kansasii) and error 23.5% ranging between 0 (M. kansasii) and 50% (M. scrofulaceum). 4. Discussion The incidence of EM infections in the general population of Romania was only 2.9 cases per 100,000 population (Marica, 2008), but as well as in other countries, mixed infections with TB and EM bacilli are also reported (Fujita et al., 2006). The global incidence of TB in Romania was 109.8 cases per 100,000 population in 2007 (Marica, 2008). It is very important to differentiate between TB and EM infections, since the treatments for these two diseases are different. The individualization is based on species of mycobacteria recovered, site and severity of the infection, antimicrobial drug susceptibility test results, concurrent diseases and the patient general condition. We used TLC to decide which immunomagnetically GLs to use for a specific EM in ELISA. By applying the simultaneous EM ELISA, our results showed that some cross reactivity was observed but the OD/IU for the homologous species was significantly higher when using immunomagnetically purified GLs than crude GLs. The ELISA EM and ELISA BK assays were performed on all patient sera.The simultaneous ELISA, comprising 5 different EM and M. tuberculosis GL antigens, allowed a combined result concerning the presence of EM and M. tuberculosis antibodies. We obtained about the same results as Wayne et al. (1988); 26 out of 34 sera of the patients infected with EM proved to be positive in ELISA BK as well as in ELISA EM. Eight patients (4 infected with MAI, 1 with M. kansasii, 2 with M. xenopi, and one with M. gordonae) were positive only in ELISA EM. It is very difficult to find a higher number of patients infected only with EM, because of high prevalence of TB in Romania. Most of the patients infected with EM, had been previously infected with M. tuberculosis; thus the presence of EM usually represents a re-infection. The low number of patients with EM nonM. tuberculosis infected is not enough for a statistical interpretation of

the assay data. However, the results on TB patients showed that there is a high specificity of the assay: 95.6% for MAI, 91.3% for M. xenopi, 100% for M. kansasii, 84.4% for M. scrofulaceum and 85.7% for M. gordonae. Thus we proved that the patients were not randomly chosen. They were patients with a confirmed EM and M. tuberculosis infection, in an area with high TB prevalence. The co-infection with M. tuberculosis and some other EM was previously described (Fujita et al., 2006; Tsukamura et al., 1981). We consider that immunomagnetically purified mycobacterial GLs can be used as antigens in ELISA, for increasing the sensitivity of the assay. The GLs could be purified with amine and carboxyl terminated superparamagnetic BioMag particles. The coupling efficiency of mycobacterial GL–NH2 and GL–COOH were approximately similar. But TLC of GL before and after magnetic purification, corroborated with ELISA results, showed that COOH-terminated particles allow better purification for M. kansasii, M. xenopi and M. xenopi while NH2terminated particles act better on MAI and M. gordonae GL. The specificity of the assay was increased by utilizing controls, M. tuberculosis antibodies-free, immunomagnetically purified with BM superparamagnetic Iron Oxide bounded with M. tuberculosis H37Rv cells. The use of such purified antigens in ELISA, can increase both sensitivity and specificity of serologic assays in EM diagnosis. The combination of multiple-EM antigen ELISA using immunomagnetically purified GLs, gives good discrimination between healthy controls and sera from patients with TB or EM disease. For a more precise and differential diagnosis of TB and EM diseases, further combinations of highly specific antigens reactive only to TB and/or EM patient sera would be necessary. We have shown that EM infection with or without TB can be recognized serologically using EM immunomagnetically purified GL. Furthermore, we suggest that co-infection with EM may be quite frequent in Romanian TB patients. The multiple assay on sera collected from patients investigated for TB and/or EM diagnosis, could be performed in parallel with direct microscopic smear examination and could be utilized in as an important tool for early laboratory diagnosis of the disease, with pulmonary or extrapulmonary localization, even in non-bacillary cases. After completion and evaluation of the EM assay, in the currently presented context, we intend to continue the studies in partnership with clinicians in order to obtain a better correlation between the ELISA EM data and the clinical manifestations, studied in a higher number of patients infected with organisms from the Mycobacterium genus, taking into consideration the moment of serum collection according the stage of the disease: acute, convalescent, taken years after infection. The EM infection represents a disease whose diagnosis can prove quite a big challenge. In these conditions, even if immunoenzymatic diagnosis cannot alone clarify the diagnosis, we consider ELISA EM useful and even recommendable, as an aid in the diagnosis in the presence of other epidemiologic, clinical, paraclinical and laboratory data. Also, the utilization of a species specific ELISA EM could be helpful in seroepidemiologic studies, both in countries with high incidence and prevalence of TB and EM, or in countries with low TB rate and higher EM incidence.

290

H. Stavri et al. / Journal of Microbiological Methods 86 (2011) 283–290

Acknowledgement We dedicate this paper to the memory of Professor Ronald J. Doyle, University of Louisville, KY, who came up with the idea of utilizing superparamagnetic particles in diagnosis of mycobacterial infections. We thank Kathy Revesz, University of Medicine and Dentistry of New Jersey, Newark, NJ, for her helpful comments and critical comments on the manuscript. This work was supported by Romanian Ministry of Education, Research and Youth, under the 35/2005 and 61-026/2007 grants. All of the authors had participated in this research. References Abebe, F., Holm-Hansen, C., Wiker, H.G., Bjune, G., 2007. Progress in serodiagnosis of Mycobacterium tuberculosis infection. Scand. J. Immunol. 66, 176–191. Mycobacteria: General Culture Methodology and Safety Considerations. In: Allen, B.W. (Ed.), Humana Press, Totowa, N.J. Altman, D.G., Bland, J.M., 1994. Diagnostic tests 2: predictive values. BMJ 309, 102. American Journal of Respiratory and Critical Care Medicine, 1997. Diagnosis and treatment of disease caused by nontuberculous mycobacteria. This Official Statement of the American Thoracic Society was approved by the Board of Directors, March 1997: Medical Section of the American Lung Association, 156, pp. S1–S25. Amicosante, M., Houde, M., Guaraldi, G., Saltini, C., 1999. Sensitivity and specificity of a multi-antigen ELISA test for the serological diagnosis of tuberculosis. Int. J. Tuberc. Lung Dis. 3, 736–740. Andersen, P., Munk, M.E., Pollock, J.M., Doherty, T.M., 2000. Specific immune-based diagnosis of tuberculosis. Lancet 356, 1099–1104. Brennan, P.J., Heifets, M., Ullom, B.P., 1982. Thin-layer chromatography of lipid antigens as a means of identifying nontuberculous mycobacteria. J. Clin. Microbiol. 15 (447), 55. Brennan, P.J., Souhrada, M., Ullom, B., McClatchy, J.K., Goren, M.B., 1978. Identification of atypical mycobacteria by thin-layer chromatography of their surface antigens. J. Clin. Microbiol. 8, 374–379. Chinn, S., 2001. Statistics for the European Respiratory Journal. Eur. Respir. J. 18, 393–401. Cole, R.A., Lu, H.M., Shi, Y.Z., Wang, J., De-Hua, T., Zhou, A.T., 1996. Clinical evaluation of a rapid immunochromatographic assay based on the 38 kDa antigen of Mycobacterium tuberculosis on patients with pulmonary tuberculosis in China. Tuber. Lung Dis. 77, 363–368. Daniel, T.M., Gonchoroff, N.J., Katzmann, J.A., Olds, G.R., 1984. Specificity of Mycobacterium tuberculosis antigen 5 determined with mouse monoclonal antibodies. Infect. Immun. 45 (1), 52–55. Daniel, T.M., 1996. Immunodiagnosis of tuberculosis. In: Rom WM, G.S. (Ed.), Tuberculosis. Little, Brown, Boston, pp. 223–231. Daniel, T.M., Debane, S.M., 1987. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay. Am. Rev. Resp. Dis. 135, 1137. Demkow, U., Ziółkowski, J., Filewska, M., Białas-Chromiec, B., Zielonka, T., MichałowskaMitczuk, D., Kuś, J., Augustynowicz, E., Zwolska, Z., Skopińska-Rózewska, E., Rowińska-Zakrzewska, E., 2004. Diagnostic value of different serological tests for tuberculosis in Poland. J. Physiol. Pharmacol. 55 (Suppl. 3), 57–66. Escamilla, L., Mancilla, R., Glender, W., Lopez-Marin, L.M., 1996. Mycobacterium fortuitum glycolipids for the serodiagnosis of pulmonary tuberculosis. Am. J. Respir. Crit. Care Med. 154, 1864–1867. Falkinham 3rd., J.O., 1996. Epidemiology of infection by nontuberculous mycobacteria. Clin. Microbiol. Rev. 9, 177–215. Fujita, Y., Doi, T., Maekura, R., Ito, M., Yano, I., 2006. Differences in serological responses to specific glycopeptidolipid-core and common lipid antigens in patients with pulmonary disease due to Mycobacterium tuberculosis and Mycobacterium avium complex. J. Med. Microbiol. 55, 189–199. Garg, S.K., Tiwari, R.P., Tiwari, D., Singh, R., Malhotra, D., Ramnani, V.K., Prasad, G.B., Chandra, R., Fraziano, M., Colizzi, V., Bisen, P.S., 2003. Diagnosis of tuberculosis: available technologies, limitations, and possibilities. J. Clin. Lab. Anal. 17, 155–163. Harboe, N., Ingild, A., 1973. Immunization, isolation of immunoglobulins, estimation of antibody titre. Scand. J. Immunol. Suppl. 1, 161–164. Heifets, L., 2004. Mycobacterial infections caused by nontuberculous mycobacteria. Semin. Respir. Crit. Care Med. 25, 283–295.

Heifets, L., Desmond, E., 2005. Clinical mycobacteriology (tuberculosis) laboratory: services and methods. In: Cole, Stewart T., et al. (Ed.), Tuberculosis and the Tubercle Bacillus. ASM Press, Washington, DC. Hoff, S.T., Abebe, M., Ravn, P., Range, N., Malenganisho, W., Rodriques, D.S., Kallas, E.G., Soborg, C., Mark Doherty, T., Andersen, P., Weldingh, K., 2007. Evaluation of Mycobacterium tuberculosis-specific antibody responses in populations with different levels of exposure from Tanzania, Ethiopia, Brazil, and Denmark. Clin. Infect. Dis. 45, 575–582. Janicki, B.W., Chaparas, S.D., Daniel, T.M., Kubica, G.P., Wright, G.L., Yee, G.S., 1971. A reference system for antigens of Mycobacterium tuberculosis. Am. Rev. Respir. Dis. 104, 602–604. Julian, E., Matas, L., Alcaide, J., Luquin, M., 2004. Comparison of antibody responses to a potential combination of specific glycolipids and proteins for test sensitivity improvement in tuberculosis serodiagnosis. Clin. Diagn. Lab. Immunol. 11, 70–76. Kent, P.T., Kubica, G.P., 1985. Public health mycobacteriology, a guide for the level III laboratory. In: P. H. S. C. C. M. US Department of Health and Human Services (Ed.), American Society for Microbiology, Washington, DC, pp. 1–257. Kitada, S., Maekura, R., Toyoshima, N., Naka, T., Fujiwara, N., Kobayashi, M., Yano, I., Ito, M., Kobayashi, K., 2005. Use of glycopeptidolipid core antigen for serodiagnosis of mycobacterium avium complex pulmonary disease in immunocompetent patients. Clin. Diagn. Lab. Immunol. 12, 44–51. Kubica, G.P., 1984. Clinical microbiology. The Mycobacteria. A Source Book, Part A, pp. 133–175. Laal, S., Samanich, K.M., Sonnenberg, M.G., Belisle, J.T., O'Leary, J., Simberkoff, M.S., Zolla-Pazner, S., 1997. Surrogate marker of preclinical tuberculosis in human immunodeficiency virus infection: antibodies to an 88-kDa secreted antigen of Mycobacterium tuberculosis. J. Infect. Dis. 176, 133–143. Lodes, M.J., Dillon, D.C., Mohamath, R., Day, C.H., Benson, D.R., Reynolds, L.D., McNeill, P., Sampaio, D.P., Skeiky, Y.A., Badaro, R., Persing, D.H., Reed, S.G., Houghton, R.L., 2001. Serological expression cloning and immunological evaluation of MTB48, a novel Mycobacterium tuberculosis antigen. J. Clin. Microbiol. 39, 2485–2493. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265–275. Marica, C., 2008. Current Data Concerning the Epidemiology of Tuberculosis in Romania, Global Fight Day against Tuberculosis, Bucharest, Romania. Mycobacterium, In: Metchock, B.G., Nolte, F.S., Wallace Jr., R.J. (Eds.), 7th edition ed. ASM Press, Washington, DC. Nagai, H., Kawabe, Y., Nagayama, N., Tanaka, Y., Nishiyama, M., Suzuki, M., Masuda, K., Baba, M., Hori, A., Tamura, A., Akagawa, S., Machida, K., Kurashima, A., Yotsumoto, H., Mori, M., Kimura, S., 2001. HIV seroprevalence in patients with tuberculosis. Kekkaku 76, 679–684. Niculescu, D., Stavri, H., Teodor, I., Popa, M., Stavri, D., Ciomu, E., Homos, M., Popescu, L., 1995. Clinical usefulness of rapid serodiagnosis in pulmonary tuberculosis by ELISA with glycolipidic antigens extracted from Mycobacterium tuberculosis and with whole BCG suspension. Roum. Arch. Microbiol. Immunol. 54, 277–284. Osborn, T.W., 1979. Serial subculture of BCG on solid and liquid media. Tubercle 60, 83–90. Phillips, M.S., von Reyn, C.F., 2001. Nosocomial infections due to nontuberculous mycobacteria. Clin. Infect. Dis. 33, 1363–1374. Primm, T.P., Lucero, C.A., Falkinham 3rd., J.O., 2004. Health impacts of environmental mycobacteria. Clin. Microbiol. Rev. 17, 98–106. Rosen, E.U., 1990. The diagnostic value of an enzyme-linked immune sorbent assay using adsorbed mycobacterial sonicates in children. Tubercle, 71, pp. 127–130. Shende, N., Gupta, S., Upadhye, V., Kumar, S., Harinath, B.C., 2007. Isolation and analysis of circulating tuberculous antigens in Mycobacterium tuberculosis. Indian J. Tuberc. 54, 125–129. Stavri, H., Branaru-Gheorghiu, M., Moldovan, O., Raileanu, M., Popa, M.I., Popa, L., Ene, L., 2005. Rapid immunochromatographic serum assay of nontuberculous mycobacterial infections. Rom. Arch. Microbiol. Immunol. 64, 42–49. Tonjum, T., Welty, D.B., Jantzen, E., Small, P.L., 1998. Differentiation of Mycobacterium ulcerans, M. marinum, and M. haemophilum: mapping of their relationships to M. tuberculosis by fatty acid profile analysis, DNA-DNA hybridization, and 16S rRNA gene sequence analysis. J. Clin. Microbiol. 36, 918–925. Tsukamura, M., Mizuno, S., Murata, H., 1981. Occurrence of Mycobacterium tuberculosis and strains of the Mycobacterium avium–M. intracellulare complex together in the sputum of patients with pulmonary tuberculosis. Tubercle 62, 43–46. Wayne, L.G., Anderson, B., Chetty, K., Light, R.W., 1988. Antibodies to mycobacterial peptidoglycolipid and to crude protein antigens in sera from different categories of human subjects. J. Clin. Microbiol. 26, 2300–2306. WHO, 2009. Treatment of Tuberculosis: Guidelines.