Michael F. Barile

PROBLEMS CONCERNING “NONCULTIVABLE” MYCOPLASMA CONTAMINANTS IN TISSUE CULTURES

The inadvertent contamination of cell cultures by a variety of agents-bacteria, fungi, bacteriophages, viruses, and mycoplasmas-has long plagued the cell biologist.’,* Contamination of cultures with mycoplasmas is a particularly difficult problem because many of these agents exert no obvious effect on the appearance of the culture and hence go undetected unless the investigator makes a special effort to look for them. Indeed, experience has taught many that good cell husbandry can be effective in the elimination of these ubiquitous pests.3 Propagation of cell cultures in the absence of antibiotics, good handling techniques, and frequent monitoring are three of the major practices that can be used to eliminate contamination. Even under these conditions, however, one occasionally encounters mycoplasma contaminants that cannot be readily isolated by conventional techniques.4~5 The characterization and subsequent isolation of one such contaminant constitutes the subject of this report.

THE IDENTIFICATION AND SOURCES OF MYCOPLASMAS ISOLATED FROM CONTAMINATED CELL CULTURES

Mycoplasmas have been established as common contaminants of cell cult u r e ~ . ~ In 1962, the Bureau of Biologics* instituted a requirement to test for the presence of mycoplasmas in viral vaccines for human use that were produccd in cell cultures. Accordingly, we have maintained a continuing study in order to examine and establish the sources of mycoplasma contamination of cell cultures and of biologics for clinical use. The results of this continuing study will be summarized in this report.

Nucleotide sequence of the MgPa (mgp) operon of Mycoplasma genitalium and comparison to the P1 (mpp) operon of Mycoplasma pneumoniae

The attachment of Mycoplasma genitalium and Mycoplasma pneumoniae to ciliated epithelium involves two surface proteins designated MgPa and P1, respectively. We have previously cloned and sequenced the P1 (mpp) operon of M. pneumoniae, and report here the use of P1-derived probes to clone and sequence a 10.4-kb region of M. genitalium DNA that, by analogy to the P1 operon, contains the MgPa (mgp) operon. The deduced amino acid sequences of the 29-kDa (ORF-1), MgPa (160-kDa) and 114-kDa (ORF-3) proteins of the MgPa operon show extensive homologies with those of the 28-kDa, P1 (170-kDa) and 130-kDa proteins, respectively, encoded by the P1 operon. The common features and homology of these operons are consistent with previous observations that the MgPa and P1 proteins share cross-reactive epitopes, as well as similar biological function. The gene order of the MgPa operon is ORF-1, MgPa, ORF-3, with intervening regions of 6 and 1 nt, respectively. A consensus ribosome-binding site (RBS) sequence is found before ORF-1 and a sequence indicative of a transcription terminator is located beyond ORF-3; the absence of such sequences adjacent to the MgPa gene suggests that the operon is transcribed as a polycistronic message. The RBS sequence is followed by sequences of dyad symmetry that have the potential to form two alternative stem-and-loop structures, which could be involved in controlling initiation of translation.

Mycoplasmas in Cell Culture

This report will briefly review our current knowledge of: 1) the incidence, prevalence and sources of mycoplasma contamination in cell cultures; 2) the procedures for isolation, detection and identification of mycoplasmas, including procedures recommended for testing biological products produced for human use; 3) the effects of mycoplasma contamination and /or infection on the function and activities of various infected cell cultures; and 4) the recommended procedures for prevention and elimination of mycoplasma contamination. Extensive reviews on mycoplasma contamination in cell culture have been reported earlier (Barile, 1979; DelGiudice and Hopps, 1978; McGarrity and Kotani, 1985).

Wall-less prokaryotes from fall flowers in central United States and Maryland

Four spiroplasma strains and eleven isolates tentatively identified as acholeplasmas were obtained from fall flowers in Colorado, Nebraska, Illinois, and Maryland. Although the acholeplasma isolates were heterogeneous, all showed antigenic sharing with a group of unnamed organisms (L1 and related strains) isolated in othe studies from flowers in Florida. The W20 and W24 isolates from Nebraska were partially related to the L1 group by DNA-DNA homology and polyacrylamide gel electrophoresis (PAGE) analyses. A Colorado spiroplasma (W13) was identifed as a new strain of group IV complex. Three spiroplasma strains from flowers in Maryland “old fields” represent a new serovar with closest affinity to subgroup I-4 and to the LB12 and N525 serovars of group I. Widespread occurrence of acholeplasmas on flowers in this study, and on plant surfaces in general, suggests that, like spiroplasmas they probably will be found to reside in arthropods.

Incidence and Sources of Mycoplasma Contamination: A Brief Review

In 1956, Robinson and colleagues (1) reported the first isolation of a mycoplasma from a contaminated cell culture. Subsequently, mycoplasmas have been shown to be common and bothersome contaminants capable of altering the activity of cells and affecting the results of study. Because many of the vaccines prepared for human use are produced in cell cultures and are subject to mycoplasma contamination, the Bureau of Biologics has maintained a continuing study for the past 18 years to examine various aspects of mycoplasma contamination. This report will review some of our findings and present a brief, updated status report on the incidence, prevalence and sources of mycoplasma contamination. The subject has been reviewed in detail elsewhere (2–5).

The search for a viral agent in Hodgkin's disease

Methods previously employed to implicate viral agents and mycoplasma in the etiology of leukemia and Burkitts lymphoma were employed in a search for the cause of Hodgkins disease. In a prospective study of 26 patients, 34% had evidence of “C‐type” particles in tissues or plasma pellets. No budding virus was seen. Antibody to the herpes‐like virus (HLV or EBV) was as prevalent in normal controls as in patients with Hodgkins disease and the frequency of elevated anti‐HLV titers was similar. No mycoplasma was found in 41 specimens from 27 patients examined. There is no firm evidence to date linking Hodgkins disease to the viruses known to be responsible for animal leukemias or to HLV which is suspected to be the cause of Burkitts lymphoma. Mycoplasma, a common contaminant in leukemia tissue, is rarely present in the malignant tissues of patients with Hodgkins disease.

31P-NMR studies of Mycoplasma gallisepticum cells using a continuous perfusion technique

31P‐NMR studies of Mycoplasma gallisepticum cells have been carried out using a continuous perfusion technique; these are the first such studies with this organism. Using this technique, glucose metabolism was monitored in the intact organisms, and cell extracts were prepared to identify the intermediates. Under glycolytic conditions, high levels of fructose‐1,6‐diphosphate were observed, indicating that this sugar may play a key role in the regulation of metabolism. The level of phosphoenolpyruvate was low under normal glycolytic conditions, and did not increase during starvation. From the position of the internal inorganic phosphate peak, the intracellular pH was estimated. The cells were found to maintain an intracellular pH of ~ 7.1 over an investigated external pH range of 6.6–8.6.

Lack of genetic relatedness among animal and plant acholeplasmas by nucleic acid hybridization

The present study compared the genetic characteristics of three previously unclassified acholeplasmas of plant origin with those of the eight established species ofAcholeplasma. A radiolabeled DNA probe was derived from the lemon (L1) strain acholeplasma by using the nick translation method and hybridized to the excess unlabeled DNA isolated from the eight established species ofAcholeplasma and three new isolates from plants. Nucleic acid hydridization and serological tests confirmed that the L1 acholeplasma was distinct from all eight established species ofAcholeplasma but closely related to strains isolated from the flowers of grapefruit trees (GF1) and powder puff plants (PP2). The results suggest that these new acholeplasmas isolated exclusively from plants may represent a new species ofAcholeplasma.

MIXED MYCOPLASMA-VIRUS INFECTIONS IN CELL CULTURES

The existence of mycoplasmas that could replicate in tissue cultures without causing any noticeable cytopathic effect has been known, and widespread latent contamination of tissue cultures with mycoplasmas has been well docurnentedI4 (Barile and colleagues, this Annal). It was therefore not surprising to anticipate that mycoplasma-contaminated tissue cultures used for virus research would have an effect on the subsequent virus yields from these cultures. Indeed, reports on this phenomenon began appearing in the literature in 1963 (TABLE 1 ) . Our interest in this field was heightened when bizarre results were obtained in a series of viral experiments done in our laboratory that, upon further investigation, proved to be due to the presence of mycoplasma. We therefore began a series of experiments demonstrating both the viral-enhancing and -inhibiting effects of mycoplasma on virus replication, and those factors needed to achieve this effect. We also found a relationship between the presence of mycoplasma and the production of interferon, a viral-inhibitory protein, which could help to explain how increased virus yields could be obtained from mixed virus-mycoplasma cultures. In this report we would like to review and summarize this previously reported data in order to demonstrate the varied effects that mycoplasma can have on virus interactions in tissue culture. As can be seen in TABLE 1, the presence of mycoplasma can lead to either increased or decreased viral yields. Deleted from the table are reports where the results did not appear to be statistically significant or where no observable effects were noted. This later possibility is a distinct one and has also been documented.’”lS There are several mechanisms involved in decreased virus yields that have been either suggested or described in the literature. These are: (1) destruction of the cell sheet so that there is less of a substrate in which the virus can replicate;8 (2) a fall in the pH providing an acid milieu unsuitable for the replication of a virus;s and (3) arginine utilization by the mycoplasma that competes with certain viruses requiring this chemical for the synthesis of viral coal protein and, hence, complete viruses.5J2 Examples of arginine utilization leading to decreased virus yields are shown in TABLE 1, but it is likely that any other arginine-utilizing mycoplasmas, of which there are many,20.z1 would behave similarly. In addition, it is also likely that other viruses that have been described as arginine utilizers, such as adenovirus type 1 ,22 h e r p e ~ , z ~ , ~ ~ SVq0,25 polyoma,Ze and cyt~megaloviruses,~~ would also show decreased virus yields in the presence of an arginine-utilizing rnycoplasma. (See Reference 45.) The mechanism for increased virus yields that has been postulated is more complex, and was suggested by the results of our first experiments showing increased virus yield^.^ In this series of experiments M. arginini was the mycoplasma, and vesicular stomatitis virus (VSV) was the virus used, We noted an increase in the yield of VSV that ranged from 0.5 to 1.2 l ~ g , ~ . ~ A difference in virus yield between mycoplasma-free and mycoplasma-infected cells of 0.5 loglo was significant in our