Mycoplasma contamination in human leukemia cell lines

Journal of Immunological Methods, 149 (1992) 43-53

43

© 1992 Elsevier Science Publishers B.V. All rights reserved 0022-1759/92/$05.00

JIM 06244

Mycoplasma contamination in human leukemia cell lines I. Comparison of various detection methods Cord C. Uphoff, Suzanne M. Gignac and Hans G. Drexler German Collection of Microorganisms and Cell Cultures, Human and Animal Cell Cultures Collection, Braunschweig, Germany (Received 1 July 1991, accepted 29 November 1991)

The sensitivity and reliability of seven assays for mycoplasma detection were tested on a panel of leukemia cell lines. The assays used were: microbiological cultivation on broth and agar, immunofluorescent visualization of mycoplasmal DNA using DAPI (both direct staining and after multiplication of the contaminants on an indicator cell line), a nucleic acid hybridization assay with a radioactive probe specific for mycoplasmal rRNA, and ELISA with mycoplasma-specific polyclonal antibodies, a biochemical method utilizing 6-MPDR, and a mycoplasma-specific monoclonal antibody in immunofluorescence staining. The broth-agar method, the two DAPI tests and the RNA hybridization assay produced the highest detection rates; a number of false-negative cases were recorded by the other tests. The detection rates, costs, requirement for specialized equipment and other characteristics were evaluated for each method. Since each technique also has disadvantages and certain limitations and since no method can be regarded as the 'gold standard', at least two procedures should be used in routine screening for mycoplasma in cell cultures. K£y

words: Mycoplasma; Cell line; Leukemia; Detection

Introduction Over the past decades, cell culture has become an indispensable tool of modern research. But the more frequent use of cell culture is accompanied by increased contamination, i.e., cellular contamination (cross-contamination between cell cultures) and contamination with microorganisms (microbial contamination). While bacterial and fungal infections of cultures are relatively easy to

CO"espomience to: H.G. Drexler, DSM, German Collec~ tion of Microorganisms and Cell Cultures, Mascheroder Weg 1B, D-3300 Braunschweig, Germany (Tel.: 49-531-618760; Fax 49-531-618718).

prevent, to detect and to treat, contamination with mycoplasmas represents a much bigger problem in terms of incidence, detect ability, prevention and eradication. It has been estimated that between 5 and 35% of cell cultures in current use are infected with mycoplasmas (Del Giudice et aI., 1984; McGarrity et aI., 1985; Mowles, 1988; Hay et aI., 1989). Mycoplasmas belong to the class Mollicutes, order Mycoplasmatales; the latter is divided into three families (Mycoplasmataceae, Acholeplasmataceae, Spiroplasmataceae); each family contains different genera. For instance, the genus Mycoplasma comprises more than 50 known species which are further subdivided into a multitude of strains. Mycoplasmas are the smallest

44

free-living, self-replicating organisms (0.2-2 ~m in diameter). Mycoplasmas lack a cell wall so that antibiotics interfering with the murein formation of cell walls (such as penicillin which is often added to cell cultures) are not effective at the standard concentrations used. Mycoplasmas are extracellular parasites usually attached to the external surface of the cell membrane. Mycoplasma contamination is not without serious consequences since virtually every cellular process can be altered by the infection. In the worst cases, contamination leads to diminished cell growth and eventually to the loss of cultures. Mycoplasmas from human, bovine and porcine sources are the most prevalent groups with Mycoplasma orale (human source), M. hyorhinis (porcine), M. arginini (bovine) and Acholeplasma laidlawii (bovine) being the single most common isolates (Del Giudice et aI., 1984; McGarrity et aI., 1985). While establishing our cell culture collection at the German Collection of Microorganisms and Cell Cultures (DSM), we obtained a large number of continuous cell lines from various scientific laboratories. 44 out of 95 cell lines (45%) turned out to be mycoplasma-positive. The 95 cell lines were received from 34 laboratories all over the world. 16 of these 34 laboratories (47%) submitted at least one mycoplasma-contaminated culture. Numerous methods for detecting mycoplasma infection have been described (Del Giudice et aI., 1984; McGarrity et aI., 1985). While some tests are rather cumbersome and time-consuming, other less fastidious techniques lack sensitivity and specificity. We attempted to evaluate and compare the usefulness, practicality and costs of seven different methods for their daily application in the routine setting of a cell culture bank. In the present study we concentrated on a large panel of continuous human leukemia cell lines in suspension.

Materials and metbods Cell cultures The cell lines listed in Table I were used for this analysis. We selected cell lines known to be

TABLE I CONTINUOUS CELL LINES USED IN PRESENT STUDY. Cell line

Cell type

Original leukemia

DSMno.

CML·T1 CTV·I EHEB HDLM·2

Tcell Monocyte Bcell Hogdkin cell Erythroid cell T cell Bcell Erythroid cell Tcell Bcell Tcell Hodgkin cell Plasma cell Hodgkin cell Plasma cell Tcell Monocyte Tcell Bcell

CML AMoL B-CLL Hodgkin's lymphoma Erythroleukemia ALL B-CLL Erythro· leukemia T·NHL B-NHL ALL Hodgkin's lymphoma Plasma cell leukemia Hodgkin's lymphoma Myeloma AML AML ATL Burkitt's lymphoma Myeloma

ACC7 ACC13 ACC67 ACC17

HEL JURKAT JVM·3 K·562 KARPAS·299 KARPAS·422 KE·37 KM-H2 L-363 L-428 LP·I MKB-l ML·2 MT·I NAMALWA OPM-2 PI2/ICHlKAWA RC-2A REH SPI·SOl SPI·802 THP·l U-266

Plasma cell T cell Myeloid cell Pre Bcell Erythroid cell Erythroid cell Monocyte Plasma cell

ACCll ACC 149 ACCl8 ACClO ACC3l ACC32 ACC46 ACCS ACC49

ACC41 ACC48 ACC IS ACC28 ACC24 ACC50

ALL AML

ACC34 ACC6

ALL ALL

ACC22 ACC86

ALL

ACC92

AMoL Myeloma

ACCl6 ACC9

Key: ALL - acute lymphoblastic leukemia; AML - acute myeloid leukemia; AMoL - acute monocytic leukemia; ATL - adult T cell leukemia; CLL - chronic lymphocytic leukemia; CML - chronic myeloid leukemia; NHL - non· Hodgkin's lymphoma.

mycoplasma-infected and included negative controls. All cultures were continuous human leukemia-lymphoma cell lines and were grown in suspension at 37 0 C in a 5% CO 2-in air humidified atmosphere using the following culture me-

45

dia (all from Gibco, Eggenstein, Germany): RPMI 1640 (cat. no. 041-01875), Iscove's modified Dulbecco's medium (only cell line LP-1) (MOM; cat. no. 041-01980), 5-20% heat-inactivated (for 45 min at 56 0 C) fetal calf serum (FCS; cat no. 011-06290). Freedom from mycoplasma infection of the FCS used was ascertained microbiologically with the broth-agar assay (see below). The adherent green monkey kidney cell line VERO-B4 (DSM ACC 33) grown in RPMI 1640 with 10% FCS was used for indirect DAPI DNA staining (see below). Mycoplasma detection methods Direct DAPI DNA staining. Cytochemical staining with DAPI or Hoechst 33258 is a sensitive technique for visualizing any DNA present in the preparation (Russell et aI., 1975; Chen, 1977). Cytospin slide preparations were air-dried and stained for 15 min with the DAPI solution. The stock solution containing 5 mg 4' ,6-diamidino-2phenylindole-dihydrochloride (DAPI; Sigma, Deisenhofen, Germany; cat. no. 0-1388) dissolved in 50 ml phosphate-buffered saline (PBS) was diluted 1/100 with methanol. Mter rinsing with water, slides were mounted using the following solution: 50% McIlwain's buffer (57 ml of 0.2 M citric acid + 43 ml of 0.1 M Na 2 HP04 ) and 50% glycerine. Slides were analysed under an immunofluorescence microscope (Nikon Labophot) fitted with an ultraviolet filter set (excitation 365/10 nm; beam splitter DM 400 nm; barrier filter BA 400 nm). Nuclei of fluorescent cells emit blue light. Mycoplasmas could be detected as a thin rim around the unstained cytoplasm of the cells (but not intracytoplasmatically) or as free clusters between cells. Indirect DAPI DNA staining. Incubation of culture supernatant from the cell line to be analysed with an indicator cell line can amplify low level mycoplasma infections (Hay, 1986). The cell line VERO 0.5 x 104 cells/mt) was grown overnight as a monolayer on microscope coverslips in petri dishes. Culture supernatant was pipetted onto the slide and incubated for 3-7 days (before complete confluence was reached). After washing, cells were fIXed in Carnoy's fixa-

tive (75 ml methanol and 25 ml galcial acetic acid) followed by a methanol wash. Finally, slides were stained for 15 min with the above DAPI solution. A positive reaction resulted in a hazy blue veil (composed of tiny dots) covering the entire cell. Microbiological broth-agar cultures. In this detection method mycoplasmas were first enriched in a liquid medium and then distributed on agar plates to allow for formation of the characteristic 'fried-egg'-style colonies (McGarrity et aI., 1985). Two types of broth/ agar were prepared. Friis medium (Friis, 1975): Bacto brain heart infusion (Difco, Hamburg, Germany; cat. no. 0037-01-6), Bacto PPLO broth (Difco; cat. no. 0554-01-1), Hanks' BSS (Gibco; cat. no. 076-01200), porcine serum (Gibco; cat. no. 036-06250), yeast extract (Oxoid, Wesel, Germany; cat. no. L21), and phenol red solution (Merck, Darmstadt, Germany; cat. no. 7241); for the production of Friis agar plates agar no. 1 (Oxoid; cat. no. Ltn and dextran solution (Sigma; cat. no. 0-9885) were added. PH medium (modified Haytlick's medium) (Hayfiick, 1965): Bacto PPLO broth, horse serum (Gibco; cat. no. 034-06050), yeast extract, and DNA solution (Difco; cat. no. 3231-12-1); for the preparation of PH agar plates Bacto PPLO agar (Difco; cat. no. 0412-01-1) was used instead of Bacto PPLO broth. 200 ILl cell culture supernatant were added to 2 ml of both Friis and PH broth and incubated for 4-5 days. Then, 100 ILl of the latter broths were pipetted onto Friis and PH agar plates, respectively, and allowed to spread. During the incubation period (up to four weeks) the formation of typical colonies was scored under an inverted microscope. Biochemical 6-MPDR assay. This assay principle is based on the biochemical difference between mammalian cells and mycoplasmas in respect of the presence of the enzyme adenosine phosphorylase (McGarrity, 1982). This technique utilizes 6-methyl purine deoxyriboside (6-MPDR), a non-toxic analogue of adenosine, which is converted by adenosine phosphorylase into 6-methyl purine and 6-methyl purine riboside. Both metabolites are toxic to mammalian cells and partially or completely destroy the indicator cells.

46

This assay is available commercially as a test kit ('MycoTect', Gibco; cat. no. 062-05672). To perform the test the indicator cell line VERO was cultivated in 24-well plates together with a sample of supernatant (200 ~I) from the culture to be analysed. On the next day the 6-MPDR solution was added and after an incubation period of 3-4 days (confluence of the cells) the viability of the cells was monitored by staining with crystal violet. Monolayers with contaminated cells remained unstained due to lysis of the indicator cells caused by the mycoplasma-induced cytotoxicity of 6MPDR. Uncontaminated cells were not affected by 6-MPDR and showed a blue stained layer of viable cells. RNA hybridization assay. This test takes advantage of the fact that mycoplasmal ribosomal RNA (rRNA) is different from mammalian rRNA and employs the principle of nucleic acid hybridization in solution. The commercially available test kit (Gen-Probe, San Diego, CA, USA) contains an 3H-labelled DNA probe homologous to mycoplasmal rRNA. The test kit is supposed to detect at least 17 species of mycoplasma and acholeplasma including those most commonly seen in infected cultures (see above). The culture material to be examined was centrifuged at 300 x g. The resulting supernatant was pelleted at 15,000 x g. In order to increase the sensitivity of the assay, we extended the incubation period of the mycoplasma-containing pellet with the radioactive probe from 3 to 18-20 h. After the separation and washing steps, scintillation fluid was added and the samples examined in a l3-scintillation counter (LKB Rackbeta 1209, Pharmacia LKB, Freiburg, Germany). The results were calculated as a certain percentage of input counts hybridized. In an attempt to decrease the costs of the assay, we compared the application of 100%, 50% and 25% of the recommended quantity of reagents. ELISA. In this enzyme-linked immunosorbent assay (ELISA), polyclonal antibodies raised against specific mycoplasma antigens on mycoplasmas were used. The commercially distributed test kit (Mycoplasma detection kit, Boehringer Mannheim, Mannheim, Germany; cat. no. 1296 744; kindly provided by Dr. W. Eberle,

Boehringer Mannheim) contains antibodies to M. arginini, M. hyorhinis, A. laidlawii, and M. orale. This assay was used both for the detection of mycoplasma infection and for the determination of the species of mycoplasma involved. The detection antibodies were conjugated with biotin to which a streptavidin-alkaline phosphatase complex bound in the next step. The alkaline phosphatase hydrolysed the substrate 4-nitrophenyl phosphate to a yellow nitrophenol product. Positivity was evaluated visually and with a microplate-reader (Titertek Multiskan, Flow). Monoclonal antibody CCM-2. The elongation factor TU (EF-TU) is a key enzyme in the assembly of amino acid chains during protein biosynthesis and is located in large quantities in the cytoplasm of bacteria and related microorganisms. The monoclonal antibody (McAb) CCM-2 specifically binds the EF-TU of mycoplasmas and can thus be used to indicate mycoplasma infection in cell cultures (Blazek et aI., 1990)., Cytospin slide preparations were fIXed with 70% ethanol in order to perforate the mycoplasmal cell membrane and then air-dried. The biotinylated McAb CCM-2 (kindly provided by Dr. V. Kamla, Dusseldorf, Germany) was diluted 1/1000 in PBS. Slides were stained for 60 min at 37 0 C. After washing, streptavidin-FITC (Dianova, Hamburg, Germany; cat. no. 016-090-084) was applied to the slides at a dilution of 1/25 for 60 min. Slides were washed and mounted using a 50% PBS-50% glycerol solution. Positivity was evaluated under epi-immunofluorescence microscopy (Nikon Labophot) at 1000 X magnification. Positive staining was localized at the cell membrane and around the cells but had to be clearly discerned from background staining in the eukaryotic nucleus and cytoplasm.

Identification of mycoplasma species The respective contaminating mycoplasma species were identified by an immunobinding assay using rabbit antisera (Kotani et at., 1986). Nitrocellulose filters (Sartorius, Gottingen, Germany; cat. no. 11306-013N) were blotted onto mycoplasma colonies growing on the agar plates. The filters were incubated with 400 1'1 rabbit antisera at a dilution of 1/1000 in 24-well plates for 60 min. After washing with PBS, 400 ILl of a

47

1/1000 goat anti-rabbit peroxidase-conjugated antiserum (Nordic-Biogenzia Lemania, Bochum, Germany; cat. no. GAR/lgG(H + L)/PO) were added for another 60 min. Following further washes with PBS, filters were stained with the substrate solution containing 4-chlor-1-naphthol (3 mg dissolved in 1 ml methanol and diluted in 5 ml PBS) for 5 min and air-dried. Positive identification resulted in deep-blue spots on the filter which were scored under an inverted microscope at 100 x magnification. The following mycoplasma preparations were used as reference strains in preliminary tests and for reference purposes; rabbit antisera against

these seven mycoplasma species were employed in the immunobinding assay (both reference preparations and antisera kindly provided by Prof. H. Kirchhoff, Hannover, Germany): M. arginini, M. bovis, M. fermentans, M. hominis, M. hyorhinis, M. orale, A. laidlawii.

Results and discussion

Identification of mycoplasma species The species of the contaminating mycoplasmas could be identified by the immunobinding assay using rabbit antisera and by the ELISA proce-

TABLE II RESULTS OF MYCOPLASMA DETECflON AND IDENTIFICATION Cell line

Agar

OAPI direct

OAPI indirect

RNA

ELISA

6·MPOR

McAb

Identification of species b

CIV·l HDLM·2 HEL lURKAT JVM·3 K·562 KARPAS-299 KE·37 KM·H2 L-363 L·428 MKB-l ML·2 NAMALWA P12/ICHIKAWA RC·2A REH SPI·801 SPI·802 CML·T1 EHEB KARPAS-422

+a + + + + + + + + + + + + + + + + + +

+ + +1 + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + + + +

+ +

+

+/+c +/-

M. arginini M. orale

+/+ -/+/+ +/-

Morale + M. hyorhinis A. laidlawii M. arginini + M. hyorhinis M arginini M fermentans + M arginini M. fermenlans M. orale Morale Morale M arginini A. laidlawii + M. hyorhinis M. fermentans

LP·l MT·l OPM·2

THP·l

+ + + + + + + + + + + +

+

+

U·266

+ + + + + + + + + + + + +

-/-

-/-/-

+/+/+ +/+ +/+/-/-

-/+ +/+ +/-

-/-/-/-/-/-/-/-/-/-

Not identified d

Not identified M. hominis

M. orale Not identified

a + / _ = positive or negative detection of mycoplasma infection. As identified by immunobinding assay using rabbit antisera and by ELISA. Assessment of positivity or negativity by two independent observers. d Species not identified with the available antisera (M. arginin~ M. hovis, M fermentans. M hominis. M hyorhinis, M. orale, A. laidlawii).

b C

4X

dure for 16 of the 19 infected cell1ines (Table 11). Cell lines HEL and SPI-802 presumably harboured mycoplasmas not included in the panel of seven antisera used. The mycoplasma(s) contaminating cell line RC-2A reacted with all seven antisera, possibly due to some unknwon cross-reactivity. We found the following distribution of mycoplasma species: M. orale (n = 6); M. arginini (n = 5); M. hyorhinis (n = 3); M. Jermentans (n = 3); A. laidlawii (n = 2); M. hominis (n = 1). Four of 16 cell lines had double-infections. The cell lines were received from five different laboratories in England, Germany and Japan. The occurrence of the different mycoplasma species was similar to those reported previously except for M. Jermentans (Levine et aI., 1974; TaylorRobinson et aI., 1977; McGarrity, 1982; Del Giudice et aI., 1984).

Microbiological broth-agar cultures Using the broth-agar assay, 19 of the 27 cell lines tested were unequivocally found to be infected with mycoplasmas (Table 11). In all but one (cell line KM-H2 with M. Jermentans) of the 19 positive cell lines both the Friis medium and the PH medium agar plates displayed mycoplasma colonies. Characteristic •fried-egg' -type colonies were seen for 12/19 cell lines (Fig. 1). However, there were wide variations in size, morphology, and speed of growth between the various positive cell lines and also between the Friisand PH-agar plates. Possible sources for misinterpretation were 'pseudocolonies' stemming firstly from transplanted cells which might multiply on the agar for a limited period of time or aggregate in clumps, and secondly from artefacts in the production of

Fig. I. Microbiological broth·agar detection assay: typical ·fried·egg'-type mycoplasma colonies on a Friis medium agar plate (cell line ML-2; M. arllinini; phase contrast X 125).

49

the agar (e.g., crystals, bubbles, etc.). The interpretation of mycoplasma colonies on agar plates is clearly subjective and dependent on experience. However, in the present study there was 100% concordance between the three observers regarding identification of myocplasma colonies or pseudo-colonies. Mycoplasma colonies were detectable after an average of 3-4 days' incubation on agar (in some cases after 6 days at the latest). Agar plates found to be negative after 1 week developed no new colonies in the second or third week. Several mycoplasma species were described which cannot be cultured at all or which do not grow well on agar plates (Hopps et aI., 1973; Barile et aI., 1983). Here, tht: seven mycoplasma reference species available and mycoplasmas isolated from cell lines crv-l, P12/ICHIKAWA,

MKB-l and SPI-H02 were grown under aerobic and anaerobic conditions. In the direct comparison of the respective pairs, all mycoplasma species grew equally well and formed colonies after about the same incubation period (including M. hyorhinis).

The microbiological broth-agar culture method is commonly regarded as the most sensitive and most specific detection assay for mycoplasma infections in cell cultures (Hay et aI., 1989). In most comparative studies, this technique was used as the standard method (Del Giudice et aI., 1984; McGarrity et aI., 1985). Direct DAPI DNA staining 18 of the 19 positive cell lines were recognized by the direct DAPI staining (Fig. 2); one cell line (HEU showed equivocal staining (Table 2). Two

Fig. 2. Direct immunofluorescent DAP! DNA staining of a massively mycoplasma-infected cell line (on a cytospin slide preparation). Note the staining around and between the cells. The eukaryotic cell nucleus is strongly stained (cell line KE-37; magnification X 525).

50

out of eight negative cell lines were judged to be positive. These false-positive results were probably caused by staining artefa"cts or cell detritus. It is conceivable that free cellular DNA from broken cells might attach to the cell surface of viable cells. This test is dependent on the intensity of the infection since only massively contaminated cultuers can be detected. While differences in interpretation are possible and borderline decisions occur to a certain degree, good cytospin preparations are also mandatory. However, the test is easily and quickly performed making it useful for regular monitoring and first-line screening. An alternative fluorochrome would be Hoechst 33258 which gives identical results.

Indirect DAPI DNA staining In this assay a mycoplasma infection is transferred to an otherwise well-established cell line adherent to the staining slide, thus circumventing the problem of dead cells and insufficient cell preparations encountered with the direct DAPI test. The reading of the fluorescent staining test is also facilitated by the spreading of the typical staining pattern over the entire cell due to its adherence to the glass slide. There was complete concordance in the results of the indirect DAPI DNA staining and the microbiological broth-agar technique (Table 11). Another commonly used indicator cell line is the Swiss mouse embryo fibroblast cell line 3T6 (Del Giudice et aI., 1978). Previous reports also documented high positive detection rates: 96100% (Del Giudice et al., 1978, 1984; McGarrity et a\., 1986; Yoshida et aI., 1988). RNA hybridization assay In order to enhance the sensitivity and decrease the costs of the assay, the incubation period was extended to 18-20 h and only a quarter of the recommended quantity of test reagents was used. A preliminary comparative test series confirmed that these alterations did not adversely influence the results. There was 100% concordance between the data from the agar plate cultures and the RNA hybridization test with regard to positivity and negativity for mycoplasma infection (Table 11). 0.4%

of the total activity as enumerated by f3-counting is recommended by the supplier as the cut-off value for positivity. However, there were some definitely negative cell lines in the range of 0.4% ± 0.05% wheras all positive cell lines showed clearly higher values (mean 5.23%, range 1.0614.23%). Some bacterial species may also be detected by this test thereby falsely indicating a positive mycoplasma contamination. However, this type of infection would presumably be easily recognized by inverted microscopy. More important, RNA from eukaryotic cells will not hybridize with the DNA probe. Using this RNA hybridization assay, Douma et al. (1989) reported a positive detection rate of 77% (13 cell lines identified as mycoplasma-positive out of 17 infected cell lines). These differences in detection rates might be explained by the facts that (j) we used longer incubation periods, thereby significantly increasing the sensitivity; and Gi) we employed the improved, newer version of the commercial test kit. ELISA The visual evaluation was completely in accordance with that obtained by an ELISA reader, suggesting that automated recording of the staining reaction is optional. A major advantage of this assay is the feasibility of simultaneously determining the contaminating species of mycoplasma and detecting the infection. A limiting factor is the fact that only four mycoplasma species can be detected. According to the literature, these supposedly account for 86-99% of the contaminating species (McGarrity, 1982; Del Giudice et aI., 1984; McGarrity et al., 1986). In the present study, however, five out of 19 mycoplasma infections were not detected by this test, while all eight negative cell lines were correctly scored as non-infected (Table 11): HEL and SPI-802 with unidentified species, KM-H2 and PI2/ICHlKAWA with M. termentans, REH with M. hominis. It is clear that 100% of those mycoplasma infections which should be identified by the assay were indeed found, but it is also obvious that the panel of contaminating mycoplasma species may differ between various cell culture studies, situations and over time.

51

Biochemical 6-MPDR assay

Monoclonal antibody CCM-2

Using this test system, 14 out of 27 cell lines score<.! as mycoplasma-positive and the remaining 13 as mycoplasma-negative (Table II). The positive results coincided with those found by other tests suggesting no false-positive data. Five of the 13 cell lines classified as mycoplasma-negative appear to be false-negative cases: CfV-l and ML-2 with M arginini, L-428 with M. orale, SPI802 and HEL with unidentified mycoplasma species. Any detrimental effects of the indicator cells or the culture conditions are excluded since these cell lines were found to be positive in the indirect DAPI assay (see above) using the same VERO cells. We conclude that these particular mycoplasmas must have either no adenosine phosphorylase activity or activities below the sensitivity level of the assay. Widely divergent positive detection rates have been reported in the literature: 56%, 76%, 97% and 100% (Hatanaka et aI., 1975; McGarrity, 1982; McGarrity et aI., 1986; Yoshida et aI., 1988; Douma et aI., 1989). Again, these discrepancies between the various studies can be explained by the variable occurrence of adenosine phosphorylase-negative species and strains. It might be possible to reduce the costs of the commercially available test kit by obtaining its various components separately.

In theory, the approach using a specific McAb with fluorescence staining might have considerable potential, as it might be possible to detect single cells harbouring few mycoplasmas. However, the differentiation of specific staining of mycoplasmas along the contours of cells from non-specific background staining in the cytoplasm and on the cell membrane of the eukaryotic cell proved to be very difficult and often not reproducible: in assessing in positivity or negativity, two independent observers showed concordance for only 11/19 (58%) of the mycoplasma-infected cell lines (Table II). Thus, there was agreement on mycoplasma contamination in only 6/19 cases (32%). Unequjvocal staining was found in strongly infected cultures. Using this test, all eight uninfected cell lines were judged to be negative. Using incubation on agar as the reference method, Blazek et al. (1990) reported complete concordance between the immunofluorescence assay using the McAb CCM-2 and the microbiological test on 16 positive and 4 negative cell lines. The discrepancies between our and these results are not due to rare mycoplasma species in our series, since all mycoplasmas identified here belong to the panel of species previously recognized by the McAb (Blazek et aI., 1990). Possibly, antibodies of higher avidity would lead to stronger

TABLE III COMPARISON OF SEVEN MYCOPlASMA DETECTION TESTS Assay

Agar-broth Direct DAPI Indirect DAPI RNA hybridization EUSA 6-MPDR Monoclonal antibody

Positive a

Total b

Working time (h)

100 95 100 100

100 89 100 100

1 0.5 2 3

Detection rate (%)

Test period 3 weeks Ih 6-7 days 1.5 days

Costs (OEM) 1.90 0.10 2.60

26.30 d

Sample volume (mJ) 0.4 0.1 0.3 1.5

Equipment •

IF microscope IF microscope ~-counter,

radioactive facilities 74 74 32

81 81 52

2 2 4

1.5 days 4-5 days 0.5 days

• Cell cultures correctly identified as mycoplasma-positive. b Cell cultures correctly identified as either mycoplasma-negative or -positive. • Besides equipment of a standard cell culture lab. d When used as recommended by the supplier.

17.45 d 15.40 d 4.70

2.0 1.0 0.1

Optional EUSA-reader IF microscope

52 TABLE IV SYNOPSIS OF FEATURES OF SEVEN MYCOPLASMA DETECTION ASSAYS Assay

Advantages

Disadvantages

Agar-broth

Long incubation, subjective, no growth of some species,

6-MPDR

Inexpensive, high sensitivity, standard/reference method, high detection rate Inexpensive, Quick Inexpensive, high detection rate, high sensitivity Objective, high detection rate, Objective, specific, species identification Traditional method

Monoclonal antibody

Specific

Direct DAPI Indirect DAPI RNA hybridization ELISA

staining results and hence to a more reproducible application. Alternatively, an indicator cell line with little cross-reactivity with the McAb could be used. Using another monoclonal antibody raised against a mycoplasmal membrane antigen, Douma et al. (1989) detected 100% of infected cultures in a dot-enzyme immuoassay. Summary

Seven methods for the detection of mycoplasma infections in cell culture suspensions were compared with regard to sensitivity (detection rates), simplicity, cost and other parameters (Table III). The major advantages and limitations of the assays are summarized in Table IV. Overall, the best results were obtained with the classical methods of microbiological broth-agar and DAPI DNA staining, and RNA hybridization. The costs of some tests could substantially be decreased. It is advisable to use at least two different procedures for mycoplasma detection (Hay et aI., 1989). Although the broth-agar assay detected all infections in this study, it cannot be regarded as the 'gold standard' since some mycoplasma strains are difficult or impossible to cultivate in artificial media. Other authors have described lower positive detection rates for this microbiological method (McGarrity et aI., 1986; Yoshida et aI., 1988).

Low sensitivity, subjective Subjective, adherent cell line needed Expensive (can be reduced), radioactive facilities Expensive, only four mycoplasma species Expensive (can be reduced), only certain strains Very subjective

We now employ the following three assays routinely in daily cell culture and quality control work: direct DAPI (weekly on all growing cell lines as a quick, siple and inexpensive screening method), RNA hybridization and broth-agar culturing (in our hands the most reliable and sensitive assays). All detection methods should include positive and negative controls .

.

References Barile, M.F. and McGarrity, G.I. (1983) In: J.G. Tully aDd S. Razin (Eds,), Methods in Mycoplasmology, Vol. 2. Academic Press, New York, p. 159. Blazek, R., Schmitt, K., Krafft, U. and Hadding, U. (1990) Fast and simple procedure for the detection of cell culture mycoplasmas using a single monoclonal antibody. I. Immunol. Methods 131,203. Chen, T.T. (1977) In situ detection of mycoplasma contamination in cell cultures by fluorescent Hoecbst 33258 stain. Exp. Cell. Res. 104, 255. Del Giudice, R.A. and Hopps, H.E. (1978) In: GJ. McGarrity, D.G. Murphy and W.W. Nichols (Eds.), Mycoplasma Infection of Cell Cultures. Plenum, New York, p. 57. Del Giudice, R.A. and Gardella, R.S. (1984) In: E.M. Levine (Ed.), Uses and Standardization of Vertebrate Cell Cultures. Tissue Culture Association, In Vitro, Monograpb no. 5, Gaithersburg, p. 104. Douma, R., Dular, R., Brodeur, B.R. and Kasatiya, S.S. (1989) Development and application of monoclonal antibodies for the detection of cell infectin, mycoplasmas. I. Immunol. Methods 124, 197. Friis, N.F. (1975) Some recommendations concerning primary isolation of Mycoplasma sui pneumoniae and Mycop/asmll flocculare. Nord. Vet. Med. 27, 337.

53 Hatanaka, M.R., Del Giudice, R. and Long, C. (1975) Adenine formation from adenosine by mycoplasmas: Adenosine phosphorylase activity. Proc. Natl. Acad. Sci. USA 72, 1401. Hay, R. (1986) In: R.I. Freshney (Ed.), Animal Cell Culture, A Practical Approach. IRL Press, Oxford, p. 71. Hay, R.J., Macy, M.L. and Chen, T.R. (1989) Mycoplasma infection of cultured cells. Nature 339, 487. Hayflick, L. (1 %5) Tissue cultures and mycoplasmas. Tex. Rep. BioI. Med. 23, 285. Hopps, H.E., Meyer, B.C., Barile, M.F. and Del Giudice, R.A. (1973) Problems concerning non-cultivable mycoplasma contamination in tissue cultures. Ann. N.Y. Acad. Sci. 225, 265. Kotani, H. and McGarrity, GJ. (1986) Identification of mycoplasma colonies by immunobinding. J. ain. Microbiol. 23,783. Levine, E.M. (1974) In: D.M. Prescott (Ed.), Methods in Cell Biology, Vol. 8. Academic Press, New York, p. 229. McGarrity, GJ. (1982) In: K. Maramorosch (Ed.), Advances in Cell Culture, Vol. 2. Academic Press, New York, p. 99.

McGarrity, G.J., Sarama, J. and Vanaman, V. (1985) Cell culture techniques. Am. Soc. Microbiol. 51, 170. McGarrity, G.J., Kotani, H. and Carson, D. (1986) Comparative studies to determine the efficiency of 6-methylpurine deoxyribose to detect cell culture mycoplasmas. In Vitro Cell. Dev. BioI. 22,301. Mowles, J.M. (1988) The use of ciprofloxacin for the elimination of mycoplasma from naturally infected cell lines. Cytotechnology 1, 355. Russell, W.C., Newman, C. and Williamson, D.H. (1975) A simple cytochemical technique for demonstration of DNA in cells infected with mycoplasmas and viruses. Nature 253,461. Taylor-Robinson, D. (1977) In: OJ. McGarrity, D.O. Murphy and W.W. Nichols (Eds.), Mycoplasma Infections of Cell Cultures. Plenum, New York, p. 47. Yoshida, T., Kawase, M., Sasaki, K., Mizusawa, H. and Ishidate, M. (1988) Comparative studies on the sensitivity of three methods for detecting mycoplasmal contamination in cell cultures. Bull JFCC 4, 9.