- Email: [email protected]
The isolation of multiple strains of Mycoplasma ovipneumoniae from individual pneumonic sheep lungs
I "eterinarr Microbiolog.l,, 29 ( ! 991 ) 349-360 Elsevier Science Publishers B.V., Amsterdam
349
The isolation of multiple strains of Mycoplasma ovipneumoniae from individual pneumonic sheep lungs G. lonas a, J.K. Clarkea and R.B. Marshall b 'Department of Microbiolog.I, and Genetics. Masse.r University. Pahnerston North. New Zealand hDepartment of I'eterinar), Pathology and Public Health. Masse), Universit),. Pahnerston North. New Zealand {Accepted 25 April 1991 )
ABSTRACT lonas, G., Clarke. J.K. and Marshall, R.B., 1991. The isolation of multiple strains of M)'coplasma oripneumoniae from individual pneumonic sheep lungs. I "et.MicrobioL, 29: 349-360. The heterogeneity ofMycoplasnla ovipneumoniae isolates from the lungs of sheep with chronic nonprogressive pneumonia (CNP) from the same flock raised the possibility that multiple isolates derived from one lung were not all identical. To test this hypothesis, thirty isolates were obtained from each of six pneumonic sheep lungs at slaughter. Four lungs had relatively severe lesions and from each of these, three or four strains ofM. ovipneumonia, distinguishable by REA and in most cases by SDSPAGE. were detected. From the lungs of each of two sheep with mild lesions, two strains of M. ovipnt'lottoniae were detected. Four isolates from one lung were further examined by restriction endonuclease analysis (REA) using many restriction endonucieases. Those which differed with EcoRi also differed when other restriction endonucleases were used. However, partial digests occurred mainly with those restriction endonucleases which recognise cytosine-rich sequences. The presence of multiple strains of one species of microorganism in individual lesions is an unusual concept which may not be limited to one disease or to one host.
INTRODUCTION
Mycoplasma ovipneumoniae can be consistently recovered from the lungs of sheep with chronic non-progressive pneumonia (CNP) and can be used to transmit the disease. However, when CNP is transmitted by isolates of M. ovipneumoniae usually only mild lesions are produced (Foggie et al., 1976; Gilmour et al., 1979; Alley and Clarke, 1980; Jones et al., 1982b). More severe lesions can be produced by inoculating pneumonic lung homogenates which contain Pasteurella haemolytica as well as M. ovipneumoniae (Jones et al., 1978; Gilmour et al., 1979; Alley and Clarke, 1979) which implies that the presence of both organisms favours the production of more severe dis0378- ! 135/9 i / $ 0 3 . 5 0 © ! 991 Elsevier Science Publishers B.V. All rights reserved.
350
G. IONAS ET AL.
ease. This concept is supported by the work of Jones et al. (1982a,b) who showed that inoculation of specific pathogen-free lambs with isolates of M. ovipneumoniae followed by P. haemolytica produced lesions of proliferative exudative pneumonia (which we regard as identical to CNP). Interestingly however, these workers found that mixed clones of M. ovipneumoniae followed by P. haemolytica produced major lesions more frequently than did individual isolates of M. ovipneumoniae followed by P. haemolytica. We have examined many isolates of M. ovipneumoniae using SDS-PAGE of total proteins and have also examined their DNA using restriction endonucleases (Ionas et al., 1991 ). Isolates from different sheep from the same farm are able to be distinguished from one another by either of these techniques. The extreme heterogeneity of isolates along with the experimental transmission work of Jones et al. (1982a) suggests the possibility that the lungs of sheep with severe CNP may be simultaneously colonised by more than one strain of M. ovipneumoniae. In order to test this hypothesis multiple isolates of M. ovipneumoniae from the lungs of six sheep were examined. MATERIALS AND METHODS
Collection of lungs Lungs exhibiting CNP lesions were collected at slaughter and placed in individual plastic bags. Four lungs with severe lesions and two with mild lesions were selected. The six lungs were from sheep on different farms.
Isolation ofM. ovipneumoniae Three procedures were used to obtain multiple M. ovipneumoniae isolates. In method one, portions of a lesion were cut into pieces about 3 mm 3, placed into 3 ml FM4 medium (Frey et al., 1968) and shaken for 1 min. A 0.3 ml aliquot was then added to 2.7ml FM4 and incubated at 37°C until a pH (colour) change was detected. This procedure was repeated ten times with fresh lung material. In method two, approximately 1 g of pneumonic lung tissue was placed in a sterile mortar with about 2 g of sterile sand. The tissue was ground and 9 ml FM4 medium was slowly added until a homogeneous suspension was produced. The supernatant of this suspension was regarded as a 10-fold dilution of the original lung. Ten 0.3 ml aliquots were independently used to prepare 10-fold serial dilutions (up to 10 -~°) in FM4 medium. The dilutions were incubated at 37°C and examined daily for a colour change. The most dilute of each, which showed a colour change was selected. In method three, ten homogenates of lung lesion were prepared as for method two and 5 /tl of 10-fold dilutions from 10 -2 to 10 - 4 w e r e spotted on FM4 agar (FM4 medium+ 1% Davis agar). These were incubated at 37°C and one well-separated colony was selected in each case, placed in FM4 liquid medium and
MYCOPL4SMAOI'IPNEUMONL4EFROMSHEEPLUNGS
351
incubated until a colour change was observed. These procedures gave 30 isolates from each of the six lungs.
Identification and cloning of isolates All 180 isolates were identified as M. ovipneumoniae using the method described by Clarke et al. (1974). All were cloned at least three times as described by Mew et al. ( 1985 ).
Examination ofM. ovipneumoniae isolates by restriction endonuclease analysis and SDS-PAGE The preparation of DNA was as described by Mew et al. ( 1985 ). The digestion of the DNA with a range of restriction endonucleases (see Table 2 ) was carried out for each enzyme as recommended by the suppliers. SDS-PAGE of the total protein content of mycoplasmas was as described by Laemmli (1970) but, with two minor modifications as described by Mew et al. (1985).
Examination ofM. ovipneumoniae DNA for the presence of methylated bases The DNA from M. ovipneumoniae strain L6T3 was prepared as described by Mew et al. ( 1985 ). The dialysed DNA preparation was diluted to 29 ml, 29 g of CsCI and 2.0 ml of ethidium bromide ( 10 mg ml-~ ) was added and the refractive index adjusted to 1.388. The DNA was ultracentrifuged at 315 000 g using a TV850 (Sorval rotor) for 48 h at 15°C using a Beckman L5-75 ultracentrifuge. Approximately 5 ml of DNA/CsCI was removed from the gradient. The ethidium bromide was extracted from the DNA using isopropanol saturated with CsCI and IM Tris (pH 7.5 ). The DNA was dialysed exhaustively against 0.02 M Tris (pH 7.9), 0.001 M MgCI2 at 4 °C and spectrophotometrically assayed for purity and concentration. DNA was degraded to its nucleosides using a modification of the method of Ford et al. (1980). Briefly, 250/zg of purified DNA was diluted to 1 ml with the dialysis buffer and digested with 10 ~ulof pancreatic DNase I ( 10 mg m l - I ) for 60 min at 37°C and then with 10/ll snake venom phosphodiesterase ( 10rag ml -~ ) for 2 h at 37°C. 40/~1 of calf intestine alkaline phosphate ( 10 mg ml-~ ) was then added and incubated for 60 min at 37°C. The nucleosides were precipitated with 2.5 volumes of cold absolute ethanol, left overnight at - 2 0 ° C and collected by centrifugation at 16 500 gfor 30rain at 4°C. The pellet was washed with 1 ml cold absolute ethanol, dried under vacuum and dissolved in 250/~1 of 0.1% phosphoric acid. For use as standards 2'-deoxcyfidine,. 2'-deoxyadenosine, 2'-deoxyguanosine, thymidine-2'-deoxyriboside (United States Biochemical Corp) and 5methyl-2'-deoxcytidine, 6-methylaminopurine-9-ribofuranoside (Sigma)
G. IONAS ET AL.
352
were each dissolved separately in 0.1% phosphoric acid (specific gravity 1.685 ) to give a final concentration of 0.1%. A Shimadzu (LC-4A) high-performance liquid chromatograph with a reverse phase 8MBCI 8U radial pak liquid chromatography cartridge (Waters Associates Inc), compressed in a radial compression module-100 (Waters Associates Inc) was used to separate the nucleosides. The nucleosides were diluted to l 0 ng/tl- ~ in 0. 1% phosphoric acid and 25 ld samples were injected into the 0.1% phosphoric acid-equilibrated column and eluted with 0.1% phosphoric acid with a flow rate of I ml min-i at ambient temperature. The absorbance of the eluants was measured at 273 rim. RESULTS
Restriction endonuclease analysis of isolates Ten mycoplasma isolates were obtained by each of three methods from each of six lungs. The isolates were confirmed as M. ovipneumoniae. Isolates from each lung were examined by REA using EcoRIo The REA results from each lurg showed that the isolates from any one lung were not identical. Vhus the thirty isolates from lung one with mild lesions exhibited two different patterns as did the isolates from lung two which also had mild lesions. These patterns were assigned letters (Table 1 ) Lungs three, four and "FABLE I
The number of M. m'ipnezommiae strains and variants found in six sheep lungs Lung No,
Pattern
No,
!
A B
29 1
2
C D E F G H i J K L M N O P Q
25 5 23 4 2 I 12 10 6 "~ 27 3 ! 9 6 12 3
3
4
5
6
R S
353
M Y( "OI'LASM.! OI "IPNE~ '.IIONI.-IE FROM SH EEP LU NGS
six, which had more severe lesions each yielded isolates which showed four different patterns and lung five which also had severe lesions yielded isolates which showed three different patterns. The differences between isolates obtained from one lung are illustrated using the isolates from lung six. Figure 1 shows 10 isolates obtained by method one, nine of which were almost identical, whereas the pattern in lane eight is markedly different. The four patterns seen with isolates from this lung are illustrated in Fig. 2, lanes 1-4. The results obtained for all six lungs are summarised in Table 1. To show that differences in REA patterns were not limited to results obtained with EcoRI only, four isolates representing one of each of the four different patterns seen with isolates from lung six were examined using 12 different restriction endonucleases. In all cases the restriction patterns which differed with EcoRI also differed markedly when examined using other endonucleases (Table 2). Conversely, a selection of 10 pairs of isolates which (within the pairs) showed the same pa~terns with EcoRI were examined t,s-
I
2
3
4
5
S
7
a
9
10
Fig. 1. Restriction endonuclease analysis of ten M. oripneumoniae isolates which were recovered simultaneously from a single pneumonic sheep lung by using method one (see text). The DNA was digested with EcoRl and electrophoresed in a 0.7% agarose gel. Of the ten isolates, nine are identical or almost identical, whereas one isolate (lane 8 ) has a totally different restriction pattern.
3S4
G. IONAS ET AL.
I
R
3
4
6
6
7
a
Fig. 2. Thirty isolates of M. oripneumoniae from one pneumonic sheep lung (lung 6) showed four different clea,.age patterns when digested with EcoRI (lanes 1-4). DNA from three of the same four isolales ~vasonly partially digested by Xhol (lanes 5, 6 and 7 ), but nevertheless each show different cleavage patterns. The fourth isolate was not cleaved by Xhol (lane 8 ).
ing other restriction endonucleases and in all cases, patterns which were identical with EcoRl were also identical when compared using other enzymes. The finding that isolates which gave different REA patterns when examined with EcoRI also differed when examined by other enzymes was to some extent complicated by the observation (Table 2 ) that some restriction endonucleases gave only partial digestion or, even no digestion. For example this occurred with four M. ovipneumoniae isolates from lung six when digested with Xhol and HaeIII. This is illustrated in Fig. 2, lanes 5-8 which shows that Xhol only partially digested three of the four isolates while the remaining isolate was undigested. Even when the DNA was only partially digested, the patterns show that the isolates differ. This result and the results obtained with other restriction enzymes are summa~'ised in Table 2.
SDS-PAGE ofM. ovipneumoniae isolates The total protein of each of the 180 isolates was examined by SDS-PAGE. As would be expected from isolates of one species, the patterns of all isolates
MYCOPL,.ISM! O! "iPNEU.iION/AEFROM SHEEP LU NGS
355
TABLE 2
The examination of four simultaneous isolates, from one lung (lung six ) using a range of restriction enlymes Restriction endonuclease
Recognition sequence'
Different patterns-"
Digestion of D N A 3
EcoRI Xbal Bgll
GVAATTC TVCTAGA A"GATCT
yes yes
Hindlll Mspl
A~AGCTT C~CGG
Dral
TTT~AAA
Hpall
C~CGG
Kpnl Sall
GGTACVC GVTCGAC
BamHI Smal
GVGATCC CCC'GGG
Xhol Haelll
GG'CC
yes yes yes yes yes yes yes yes yes yes
+ + + + + +
C~TCGAG
yes
+ + + + + + +
+ + + + + + + + + + + + +_ + + + + + + + + + + + + +
+ + + + + + + + + + + -
' Recognition - ' i f " y e s " , the
sequence is read from 5' to 3' end of D N A . four restriction patterns remain different using the indicated endonuclease. 3Digestion of four M. ovipneumoniaestrains. + : complete digestion of M. oripneumoniae D N A . + : partial digestion of M. ovipneumoniae D N A . - ; no digestion of M. oripneumoniae D N , ~ .
had much in common. However, 16 different patterns were seen. This compares with 19 different patterns with REA. In general, isolates with REA patterns which were identical showed the same SDS-PAGE pattern and conversely isolates which differed in their REA patterns usually had different SDS-PAGE patterns. However, in three cases, namely REA patterns A + B, G + H and P + Q , different pairg of REA patterns gave identical SDS-PAGE results. The differences seen in SDS-PAGE are illustrated (Fig. 3 ) using four isolates from lung four. In this case three isolates (lanes 1, 2 and 3 ) differed but lane 2 is closely similar to lane 4.
Methylation ofM. ovipneumoniae DNA M. ovipneumoniae isolate L6T3 which was not digested with either Xhol or HaeIIl was selected for investigation with restriction endonucleases which could discriminate between methylated and non-methylated bases in a restriction sequence (Fig. 4a), The DNA was not cleaved by Dpni which cuts the sequence 5'-GmAvTC-3 ' only if the adenosine is methylated. Mbol which also recognises the sequence 5'-VGATC-3 ' and cleaves it only in the absence of adenosine methylation was compared with Sau3al which cleaves the same sequence regardless of the presence or absence of adenosine methylation. Both these enzymes gave identical restriction patterns. The same four isolates were
3~6
G. IONAS ET AL.
I
R
:3
4
Fig. 3. SDS-PAGE examination of four M. ovipneumoniae isolates from lung 6 which differed in their DNA cleavage patterns. Although the protein banding patterns are similar i.e. lane 1 resembles lane 3 and lane 2 resembles lane 4. Nevertheless; differences (see arrows) in the banding patterns are seen in all four lanes.
Fig. 4. a. M. ovipneumoniae strain L6T3 digested with enzymes which discriminate between methylated and non-methylated adenine. Lane I, I kb ladder (i.e. molecular weight marker); Lane 2, DNA exposed to Dpnl (cleaves 5 ' - G m A V T C - 3 ' only); Lane 3, DNA cleaved with Mbo ! (cleaves 5'-'GATC-3' only): Lane 4, DNA cleaved with Sau3al (cleaves 5'-VGATC-3 ' and 5'-'GmATC-3' ). b. AI. ovipneumomae strain L6T3 DNA digested with enzymes which discriminate between methylated and non-methylated cytosine. Lane !, I kb ladder (i.e. molecular weight marker); Lane 2, DNA cleaved with Mspl (cleaves 5'-cvmCGG-3 ' and 5'-CVCGG-3 ' ); Lane 3, DNA cleaved with Hpalll (cleaves 5'-C'CGG-3' only ). Note the uncleaved band in lane 3 (see arrow). c. The DNA of four isolates from one lung cleaved with Dral. M. ovipneumoniae isolates from lung 6 with different restriction patterns when digested with EcoRl, were digested with Dral, which has a recognition sequence lacking cytosine (5'-TTTVAAA-3 ' ). The DNA of all four isolates (lanes 2-5 ) were completely digested and gave different cleavage patterns. Lane ! is lambda DNA digested with Dral.
357
MYCOPLASMA O|TPNEUMONIAE FROM SHEEP LUNGS
I
2
3
4
I
2
:3
,4
1
B
2
:3
358
G. IONAS ET AL.
examined using Dral which recognises a sequence containing only A and T. The four isolates could be distinguished by this enzyme (Fig. 4c). To investigate the possible presence of 5-methylcytosine, L6T3 DNA was also digested with Mspl (cleaves 5'-C vmCGG-3' and 5'-CVCGG-3 ' ) or with Hpall (cleaves 5'-CvCGG-3 ' only). The results (shown in Fig. 4b) indicate the presence of a large fragment of DNA which was not digested by Mspl. These results remained unchanged when higher enzyme concentrations were used for longer periods. The DNA from M. ovipneumoniae isolate L6T3 was reduced to nucleosides and examined by HPLC. A peak corresponding to 5-methylcytosine was detected which represented approximately 20% of the total cytosine content. DISCUSSION
The classical approach to pathogenic microorganisms as producers of disease concentrated on associating one species of microorganism with a disease or diseases. More recently infections involving more than one species of microorganism have been the subject of many studies and it is generally accepted that in the case of CNP P. haemolytica can exacerbate lesions caused by M. ovipneumoniae (Gilmour et al., 1979; Jones et al., 1982a). However, the report of Jones et al. (1982b) that proliferative interstitial pneumonia was best produced by inoculating sheep with mixed clones ofM. ovipneumoniae followed by P. haemolytica takes this concept a step further in that it suggests that mixed infections with organisms of one species can help to initiate lesions. Nevertheless, results obtained experimentally with sheep inoculated ff~
[ll~..
lilLi~--ffl~ill...lll~Li
E~/ULIw
illC~
ll~JL
EUEIUI..,L
VV|I~LL
~,~UiO
Ell
II~II~E
t..,~llU|Lll./ll3.
However, our results extend the finding of Jones to the field disease since we showed not only that multiple strains of M. ovipneumoniae are present in the lungs of the sheep with CNP but also found that the number of strains present, albeit in a small sample, correlates with the severity of the lesions in that the less severe lesions contained fewer i.e. two strains of M. ovipneumoniae, whereas the more severe lesions contain at least three or, in most cases four strains of M. ovipneumoniae. The conclusion that multiple strains of one species of microorganism can act together to increase the severity of lesions merits confirmation by further studies in sheep; especially as this phenomenon, if confirmed may be found to occur with other microorganisms and in other hosts. The concept that multiple strains of one organism can simultaneously clonise a lung is relatively novel and our conclusion depends on distinguishing isolates by SDS-PAGE of proteins and REA of their DNA. The SDS-PAGE results showed small differences in isolates but such differences are reproducible when replicate cultures are compared (lonas et al., 1991 ). The isolates were also distinguished by REA using EcoRl. However, the
M YCOPL.ISM..! O I "IPNEI 'MONL.!E FROM SH EEP LU NGS
3 59
REA differences were not merely a function of the restriction endonuclease used, because four isolates from one lung re-examined using 12 different restriction endonucleases remained different. The finding that any one of the 12 restriction endonucleases could be used to distinguish the isolates shows that the distinctions made by EcoRi are significant and that differences in the DNA of isolates are not limited to differences in just one restriction site sequence. When many restriction enzymes were used to cleave the DNA of M. ovipneumoniae isolates, it was found that some gave partial digests of the DNA of some isolates and one isolate was totally uncleaved by either Xhol or HaeIII (Table 2 ). This finding raised the possibility of the presence of methylated bases i.e. 6-methyladenosine or 5-methylcytosine. However, since the partial digests were associated with C-rich rather than A-rich recognition sequences (Table 2 ), it seemed more likely that 5-methylcytosine rather than 6-methyladenosine could be present. The absence of 6-methyladenosine was confirmed by exposing the DNA ofM. ovipneumoniaestrain L6T3 to Dpnl which cleaves the sequence 5'-GmAvTC-3 ' only if the adenine is methylated (Chart and Ross, 1984). No digestion occurred (Fig. 4a). MboI which also cleaves the sequence 5'-VGATC-3' but only in the absence of 6-methyladenosine gave complete digestion, as did Sau3AI which is unaffected by the presence or absence of adenine methylation. When the DNA of this isolate was exposed to Mspl, which cleaves the sequence 5'-CCGG-3' whether or not the internal cytosine is methylated, complete digestion occurred. In contrast to this Hpall which cleaves the same sequence but only in the absence of methylation, gave incomplete digestion l=io Ah ] thn~ i n d i e a t i n t , the. nre~ence_ a f 5 - m e t h y ! c y t o s i n e in t h e DNA_ T h i s
DNA was degraded to nucleosides and examined by HPLC which confirmed the presence of 5-methylcytosine in this isolate. The presence of 5-methylcytosine in the DNA of some other mycoplasma species has been reported (Razin and Razin, 1980; Chan and Ross, 1984), but in the present context the question arises as to whether or not the observed REA differences depend on variations in the extent of methylation of DNA, rather than differences in base sequence. However, this is excluded because differences in the isolates were observed when the DNA was cleaved by Dral which recognises the sequence 5'-TTTVAAA-3 ' and consequently is not affected by the presence of 5'-methylcytosine in the DNA. Furthermore, if differences in the DNA were assumed to be iimited to methylation ofbases this could not account for the differences in proteins as shown by SDS-PAGE. REFERENCES Alley, M.R. and Clarke, J.K., 1979. The experimental transmission of chronic non-progressive pneumonia. N.Z. Vet. J., 27: 217-220.
360
O. IONAS ET AL.
Alley, M.R. and Clarke, J.K., 1980. The effect of chemotherapeutic agents on the transmission ofovine chronic non-progressive pneumonia. N.Z. Vet. J., 28: 77-80. Chan, H.W. and Ross, R.F., 1984. Restriction endonuclease analysis of two porcine mycoplasma deoxyribonucleic acids: Sequence-specific methylation in the Mycop/asma hyopneumoniaegenome. Int. J. Syst. Bacteriol., 34: 16-20. Clarke. J.K., Brown, V.G. and Alley, M.R., 1974. Isolation and identification of mycoplasmas from the respiratory tract of sheep in New Zealand. N.Z. Vet. J., 47:591-596. Foggie, A., Jones, G.E. and Buxton, D. 1976. The experimental infection of specific pathogenfree lambs with Mycoplasmaovipneumoniae. Res. Vet. Sci., 2 I: 28-35. Ford. J.P., Coca-Prados, M. and Hsu, M.T., 1980. Enzymatic analysis of 5- methylcytosine content in eukaryotic DNA, J. Biol. Chem., 255: 7544-7547. Frey, M.L., Hanson, R.P. and Anderson, D.P., 1968. A medium for the isolation of avian mycoplasmas. Am. J. Vet. Res., 29:2163-2179. Gilmour, J.S., Jones, G.E. and Rae, A.G., 1979. Experimental studies ofchronic pneumonia of sheep. Comp. Immun. Microbiol. InfecL Dis., I: 285-293. lonas, G., Clarke, J.K. and Marshall, R.B. 199 I. A Study of the heterogeneity of isolates of M. o~'ipneumoniaefrom sheep in New Zealand. Vet. Microbiol, 29: 339-347. Jones, G.E., Gilmour, J.S. and Rae, A., 1978. Endobronchial inoculation of sheep with pneumonic lung-tissue suspensions and with the bacteria and mycoplasmas isolated from them. J. Comp. Path., 88: 85-96. Jones, G.E., Gilmour, J.S. and Rae, A., i982a. The effect of Mycoplasma ovipneumoniaeand Pasteurellahaemol.vticaon specific pathogen-free lambs. J. Comp. Path., 92:261-266. Jones, G.E., Gilmour, J.S. and Rae, A., 1982b. The effects of different strains of Mycoplasma oripneumoniae on specific pathogen-free and conventionally-reared lambs. J. Comp. Path. 92: 267-272. Laemmli, U.K., 1980. Cleavage of structural proteins during the assembly of th~ head of bacteriophage T4. Nature. 227: 680-685. Mew, A.J., lonas, G.~ Clarke, J.K., Robinson, A.J. and Marshall, R.B., 1985. Comparison of Mycoplasma oripneumoniaeisolates using bacterial restriction endonuclease DNA analysis and SDS-PAGE. Vet. Microbiol., 10: 541-548. Razin, A. and Razin, S., 1980. Methylated bases in mycoplasma DNA. Nucleic Acids Res., 8: 1383- ! 390.
349
The isolation of multiple strains of Mycoplasma ovipneumoniae from individual pneumonic sheep lungs G. lonas a, J.K. Clarkea and R.B. Marshall b 'Department of Microbiolog.I, and Genetics. Masse.r University. Pahnerston North. New Zealand hDepartment of I'eterinar), Pathology and Public Health. Masse), Universit),. Pahnerston North. New Zealand {Accepted 25 April 1991 )
ABSTRACT lonas, G., Clarke. J.K. and Marshall, R.B., 1991. The isolation of multiple strains of M)'coplasma oripneumoniae from individual pneumonic sheep lungs. I "et.MicrobioL, 29: 349-360. The heterogeneity ofMycoplasnla ovipneumoniae isolates from the lungs of sheep with chronic nonprogressive pneumonia (CNP) from the same flock raised the possibility that multiple isolates derived from one lung were not all identical. To test this hypothesis, thirty isolates were obtained from each of six pneumonic sheep lungs at slaughter. Four lungs had relatively severe lesions and from each of these, three or four strains ofM. ovipneumonia, distinguishable by REA and in most cases by SDSPAGE. were detected. From the lungs of each of two sheep with mild lesions, two strains of M. ovipnt'lottoniae were detected. Four isolates from one lung were further examined by restriction endonuclease analysis (REA) using many restriction endonucieases. Those which differed with EcoRi also differed when other restriction endonucleases were used. However, partial digests occurred mainly with those restriction endonucleases which recognise cytosine-rich sequences. The presence of multiple strains of one species of microorganism in individual lesions is an unusual concept which may not be limited to one disease or to one host.
INTRODUCTION
Mycoplasma ovipneumoniae can be consistently recovered from the lungs of sheep with chronic non-progressive pneumonia (CNP) and can be used to transmit the disease. However, when CNP is transmitted by isolates of M. ovipneumoniae usually only mild lesions are produced (Foggie et al., 1976; Gilmour et al., 1979; Alley and Clarke, 1980; Jones et al., 1982b). More severe lesions can be produced by inoculating pneumonic lung homogenates which contain Pasteurella haemolytica as well as M. ovipneumoniae (Jones et al., 1978; Gilmour et al., 1979; Alley and Clarke, 1979) which implies that the presence of both organisms favours the production of more severe dis0378- ! 135/9 i / $ 0 3 . 5 0 © ! 991 Elsevier Science Publishers B.V. All rights reserved.
350
G. IONAS ET AL.
ease. This concept is supported by the work of Jones et al. (1982a,b) who showed that inoculation of specific pathogen-free lambs with isolates of M. ovipneumoniae followed by P. haemolytica produced lesions of proliferative exudative pneumonia (which we regard as identical to CNP). Interestingly however, these workers found that mixed clones of M. ovipneumoniae followed by P. haemolytica produced major lesions more frequently than did individual isolates of M. ovipneumoniae followed by P. haemolytica. We have examined many isolates of M. ovipneumoniae using SDS-PAGE of total proteins and have also examined their DNA using restriction endonucleases (Ionas et al., 1991 ). Isolates from different sheep from the same farm are able to be distinguished from one another by either of these techniques. The extreme heterogeneity of isolates along with the experimental transmission work of Jones et al. (1982a) suggests the possibility that the lungs of sheep with severe CNP may be simultaneously colonised by more than one strain of M. ovipneumoniae. In order to test this hypothesis multiple isolates of M. ovipneumoniae from the lungs of six sheep were examined. MATERIALS AND METHODS
Collection of lungs Lungs exhibiting CNP lesions were collected at slaughter and placed in individual plastic bags. Four lungs with severe lesions and two with mild lesions were selected. The six lungs were from sheep on different farms.
Isolation ofM. ovipneumoniae Three procedures were used to obtain multiple M. ovipneumoniae isolates. In method one, portions of a lesion were cut into pieces about 3 mm 3, placed into 3 ml FM4 medium (Frey et al., 1968) and shaken for 1 min. A 0.3 ml aliquot was then added to 2.7ml FM4 and incubated at 37°C until a pH (colour) change was detected. This procedure was repeated ten times with fresh lung material. In method two, approximately 1 g of pneumonic lung tissue was placed in a sterile mortar with about 2 g of sterile sand. The tissue was ground and 9 ml FM4 medium was slowly added until a homogeneous suspension was produced. The supernatant of this suspension was regarded as a 10-fold dilution of the original lung. Ten 0.3 ml aliquots were independently used to prepare 10-fold serial dilutions (up to 10 -~°) in FM4 medium. The dilutions were incubated at 37°C and examined daily for a colour change. The most dilute of each, which showed a colour change was selected. In method three, ten homogenates of lung lesion were prepared as for method two and 5 /tl of 10-fold dilutions from 10 -2 to 10 - 4 w e r e spotted on FM4 agar (FM4 medium+ 1% Davis agar). These were incubated at 37°C and one well-separated colony was selected in each case, placed in FM4 liquid medium and
MYCOPL4SMAOI'IPNEUMONL4EFROMSHEEPLUNGS
351
incubated until a colour change was observed. These procedures gave 30 isolates from each of the six lungs.
Identification and cloning of isolates All 180 isolates were identified as M. ovipneumoniae using the method described by Clarke et al. (1974). All were cloned at least three times as described by Mew et al. ( 1985 ).
Examination ofM. ovipneumoniae isolates by restriction endonuclease analysis and SDS-PAGE The preparation of DNA was as described by Mew et al. ( 1985 ). The digestion of the DNA with a range of restriction endonucleases (see Table 2 ) was carried out for each enzyme as recommended by the suppliers. SDS-PAGE of the total protein content of mycoplasmas was as described by Laemmli (1970) but, with two minor modifications as described by Mew et al. (1985).
Examination ofM. ovipneumoniae DNA for the presence of methylated bases The DNA from M. ovipneumoniae strain L6T3 was prepared as described by Mew et al. ( 1985 ). The dialysed DNA preparation was diluted to 29 ml, 29 g of CsCI and 2.0 ml of ethidium bromide ( 10 mg ml-~ ) was added and the refractive index adjusted to 1.388. The DNA was ultracentrifuged at 315 000 g using a TV850 (Sorval rotor) for 48 h at 15°C using a Beckman L5-75 ultracentrifuge. Approximately 5 ml of DNA/CsCI was removed from the gradient. The ethidium bromide was extracted from the DNA using isopropanol saturated with CsCI and IM Tris (pH 7.5 ). The DNA was dialysed exhaustively against 0.02 M Tris (pH 7.9), 0.001 M MgCI2 at 4 °C and spectrophotometrically assayed for purity and concentration. DNA was degraded to its nucleosides using a modification of the method of Ford et al. (1980). Briefly, 250/zg of purified DNA was diluted to 1 ml with the dialysis buffer and digested with 10 ~ulof pancreatic DNase I ( 10 mg m l - I ) for 60 min at 37°C and then with 10/ll snake venom phosphodiesterase ( 10rag ml -~ ) for 2 h at 37°C. 40/~1 of calf intestine alkaline phosphate ( 10 mg ml-~ ) was then added and incubated for 60 min at 37°C. The nucleosides were precipitated with 2.5 volumes of cold absolute ethanol, left overnight at - 2 0 ° C and collected by centrifugation at 16 500 gfor 30rain at 4°C. The pellet was washed with 1 ml cold absolute ethanol, dried under vacuum and dissolved in 250/~1 of 0.1% phosphoric acid. For use as standards 2'-deoxcyfidine,. 2'-deoxyadenosine, 2'-deoxyguanosine, thymidine-2'-deoxyriboside (United States Biochemical Corp) and 5methyl-2'-deoxcytidine, 6-methylaminopurine-9-ribofuranoside (Sigma)
G. IONAS ET AL.
352
were each dissolved separately in 0.1% phosphoric acid (specific gravity 1.685 ) to give a final concentration of 0.1%. A Shimadzu (LC-4A) high-performance liquid chromatograph with a reverse phase 8MBCI 8U radial pak liquid chromatography cartridge (Waters Associates Inc), compressed in a radial compression module-100 (Waters Associates Inc) was used to separate the nucleosides. The nucleosides were diluted to l 0 ng/tl- ~ in 0. 1% phosphoric acid and 25 ld samples were injected into the 0.1% phosphoric acid-equilibrated column and eluted with 0.1% phosphoric acid with a flow rate of I ml min-i at ambient temperature. The absorbance of the eluants was measured at 273 rim. RESULTS
Restriction endonuclease analysis of isolates Ten mycoplasma isolates were obtained by each of three methods from each of six lungs. The isolates were confirmed as M. ovipneumoniae. Isolates from each lung were examined by REA using EcoRIo The REA results from each lurg showed that the isolates from any one lung were not identical. Vhus the thirty isolates from lung one with mild lesions exhibited two different patterns as did the isolates from lung two which also had mild lesions. These patterns were assigned letters (Table 1 ) Lungs three, four and "FABLE I
The number of M. m'ipnezommiae strains and variants found in six sheep lungs Lung No,
Pattern
No,
!
A B
29 1
2
C D E F G H i J K L M N O P Q
25 5 23 4 2 I 12 10 6 "~ 27 3 ! 9 6 12 3
3
4
5
6
R S
353
M Y( "OI'LASM.! OI "IPNE~ '.IIONI.-IE FROM SH EEP LU NGS
six, which had more severe lesions each yielded isolates which showed four different patterns and lung five which also had severe lesions yielded isolates which showed three different patterns. The differences between isolates obtained from one lung are illustrated using the isolates from lung six. Figure 1 shows 10 isolates obtained by method one, nine of which were almost identical, whereas the pattern in lane eight is markedly different. The four patterns seen with isolates from this lung are illustrated in Fig. 2, lanes 1-4. The results obtained for all six lungs are summarised in Table 1. To show that differences in REA patterns were not limited to results obtained with EcoRI only, four isolates representing one of each of the four different patterns seen with isolates from lung six were examined using 12 different restriction endonucleases. In all cases the restriction patterns which differed with EcoRI also differed markedly when examined using other endonucleases (Table 2). Conversely, a selection of 10 pairs of isolates which (within the pairs) showed the same pa~terns with EcoRI were examined t,s-
I
2
3
4
5
S
7
a
9
10
Fig. 1. Restriction endonuclease analysis of ten M. oripneumoniae isolates which were recovered simultaneously from a single pneumonic sheep lung by using method one (see text). The DNA was digested with EcoRl and electrophoresed in a 0.7% agarose gel. Of the ten isolates, nine are identical or almost identical, whereas one isolate (lane 8 ) has a totally different restriction pattern.
3S4
G. IONAS ET AL.
I
R
3
4
6
6
7
a
Fig. 2. Thirty isolates of M. oripneumoniae from one pneumonic sheep lung (lung 6) showed four different clea,.age patterns when digested with EcoRI (lanes 1-4). DNA from three of the same four isolales ~vasonly partially digested by Xhol (lanes 5, 6 and 7 ), but nevertheless each show different cleavage patterns. The fourth isolate was not cleaved by Xhol (lane 8 ).
ing other restriction endonucleases and in all cases, patterns which were identical with EcoRl were also identical when compared using other enzymes. The finding that isolates which gave different REA patterns when examined with EcoRI also differed when examined by other enzymes was to some extent complicated by the observation (Table 2 ) that some restriction endonucleases gave only partial digestion or, even no digestion. For example this occurred with four M. ovipneumoniae isolates from lung six when digested with Xhol and HaeIII. This is illustrated in Fig. 2, lanes 5-8 which shows that Xhol only partially digested three of the four isolates while the remaining isolate was undigested. Even when the DNA was only partially digested, the patterns show that the isolates differ. This result and the results obtained with other restriction enzymes are summa~'ised in Table 2.
SDS-PAGE ofM. ovipneumoniae isolates The total protein of each of the 180 isolates was examined by SDS-PAGE. As would be expected from isolates of one species, the patterns of all isolates
MYCOPL,.ISM! O! "iPNEU.iION/AEFROM SHEEP LU NGS
355
TABLE 2
The examination of four simultaneous isolates, from one lung (lung six ) using a range of restriction enlymes Restriction endonuclease
Recognition sequence'
Different patterns-"
Digestion of D N A 3
EcoRI Xbal Bgll
GVAATTC TVCTAGA A"GATCT
yes yes
Hindlll Mspl
A~AGCTT C~CGG
Dral
TTT~AAA
Hpall
C~CGG
Kpnl Sall
GGTACVC GVTCGAC
BamHI Smal
GVGATCC CCC'GGG
Xhol Haelll
GG'CC
yes yes yes yes yes yes yes yes yes yes
+ + + + + +
C~TCGAG
yes
+ + + + + + +
+ + + + + + + + + + + + +_ + + + + + + + + + + + + +
+ + + + + + + + + + + -
' Recognition - ' i f " y e s " , the
sequence is read from 5' to 3' end of D N A . four restriction patterns remain different using the indicated endonuclease. 3Digestion of four M. ovipneumoniaestrains. + : complete digestion of M. oripneumoniae D N A . + : partial digestion of M. ovipneumoniae D N A . - ; no digestion of M. oripneumoniae D N , ~ .
had much in common. However, 16 different patterns were seen. This compares with 19 different patterns with REA. In general, isolates with REA patterns which were identical showed the same SDS-PAGE pattern and conversely isolates which differed in their REA patterns usually had different SDS-PAGE patterns. However, in three cases, namely REA patterns A + B, G + H and P + Q , different pairg of REA patterns gave identical SDS-PAGE results. The differences seen in SDS-PAGE are illustrated (Fig. 3 ) using four isolates from lung four. In this case three isolates (lanes 1, 2 and 3 ) differed but lane 2 is closely similar to lane 4.
Methylation ofM. ovipneumoniae DNA M. ovipneumoniae isolate L6T3 which was not digested with either Xhol or HaeIIl was selected for investigation with restriction endonucleases which could discriminate between methylated and non-methylated bases in a restriction sequence (Fig. 4a), The DNA was not cleaved by Dpni which cuts the sequence 5'-GmAvTC-3 ' only if the adenosine is methylated. Mbol which also recognises the sequence 5'-VGATC-3 ' and cleaves it only in the absence of adenosine methylation was compared with Sau3al which cleaves the same sequence regardless of the presence or absence of adenosine methylation. Both these enzymes gave identical restriction patterns. The same four isolates were
3~6
G. IONAS ET AL.
I
R
:3
4
Fig. 3. SDS-PAGE examination of four M. ovipneumoniae isolates from lung 6 which differed in their DNA cleavage patterns. Although the protein banding patterns are similar i.e. lane 1 resembles lane 3 and lane 2 resembles lane 4. Nevertheless; differences (see arrows) in the banding patterns are seen in all four lanes.
Fig. 4. a. M. ovipneumoniae strain L6T3 digested with enzymes which discriminate between methylated and non-methylated adenine. Lane I, I kb ladder (i.e. molecular weight marker); Lane 2, DNA exposed to Dpnl (cleaves 5 ' - G m A V T C - 3 ' only); Lane 3, DNA cleaved with Mbo ! (cleaves 5'-'GATC-3' only): Lane 4, DNA cleaved with Sau3al (cleaves 5'-VGATC-3 ' and 5'-'GmATC-3' ). b. AI. ovipneumomae strain L6T3 DNA digested with enzymes which discriminate between methylated and non-methylated cytosine. Lane !, I kb ladder (i.e. molecular weight marker); Lane 2, DNA cleaved with Mspl (cleaves 5'-cvmCGG-3 ' and 5'-CVCGG-3 ' ); Lane 3, DNA cleaved with Hpalll (cleaves 5'-C'CGG-3' only ). Note the uncleaved band in lane 3 (see arrow). c. The DNA of four isolates from one lung cleaved with Dral. M. ovipneumoniae isolates from lung 6 with different restriction patterns when digested with EcoRl, were digested with Dral, which has a recognition sequence lacking cytosine (5'-TTTVAAA-3 ' ). The DNA of all four isolates (lanes 2-5 ) were completely digested and gave different cleavage patterns. Lane ! is lambda DNA digested with Dral.
357
MYCOPLASMA O|TPNEUMONIAE FROM SHEEP LUNGS
I
2
3
4
I
2
:3
,4
1
B
2
:3
358
G. IONAS ET AL.
examined using Dral which recognises a sequence containing only A and T. The four isolates could be distinguished by this enzyme (Fig. 4c). To investigate the possible presence of 5-methylcytosine, L6T3 DNA was also digested with Mspl (cleaves 5'-C vmCGG-3' and 5'-CVCGG-3 ' ) or with Hpall (cleaves 5'-CvCGG-3 ' only). The results (shown in Fig. 4b) indicate the presence of a large fragment of DNA which was not digested by Mspl. These results remained unchanged when higher enzyme concentrations were used for longer periods. The DNA from M. ovipneumoniae isolate L6T3 was reduced to nucleosides and examined by HPLC. A peak corresponding to 5-methylcytosine was detected which represented approximately 20% of the total cytosine content. DISCUSSION
The classical approach to pathogenic microorganisms as producers of disease concentrated on associating one species of microorganism with a disease or diseases. More recently infections involving more than one species of microorganism have been the subject of many studies and it is generally accepted that in the case of CNP P. haemolytica can exacerbate lesions caused by M. ovipneumoniae (Gilmour et al., 1979; Jones et al., 1982a). However, the report of Jones et al. (1982b) that proliferative interstitial pneumonia was best produced by inoculating sheep with mixed clones ofM. ovipneumoniae followed by P. haemolytica takes this concept a step further in that it suggests that mixed infections with organisms of one species can help to initiate lesions. Nevertheless, results obtained experimentally with sheep inoculated ff~
[ll~..
lilLi~--ffl~ill...lll~Li
E~/ULIw
illC~
ll~JL
EUEIUI..,L
VV|I~LL
~,~UiO
Ell
II~II~E
t..,~llU|Lll./ll3.
However, our results extend the finding of Jones to the field disease since we showed not only that multiple strains of M. ovipneumoniae are present in the lungs of the sheep with CNP but also found that the number of strains present, albeit in a small sample, correlates with the severity of the lesions in that the less severe lesions contained fewer i.e. two strains of M. ovipneumoniae, whereas the more severe lesions contain at least three or, in most cases four strains of M. ovipneumoniae. The conclusion that multiple strains of one species of microorganism can act together to increase the severity of lesions merits confirmation by further studies in sheep; especially as this phenomenon, if confirmed may be found to occur with other microorganisms and in other hosts. The concept that multiple strains of one organism can simultaneously clonise a lung is relatively novel and our conclusion depends on distinguishing isolates by SDS-PAGE of proteins and REA of their DNA. The SDS-PAGE results showed small differences in isolates but such differences are reproducible when replicate cultures are compared (lonas et al., 1991 ). The isolates were also distinguished by REA using EcoRl. However, the
M YCOPL.ISM..! O I "IPNEI 'MONL.!E FROM SH EEP LU NGS
3 59
REA differences were not merely a function of the restriction endonuclease used, because four isolates from one lung re-examined using 12 different restriction endonucleases remained different. The finding that any one of the 12 restriction endonucleases could be used to distinguish the isolates shows that the distinctions made by EcoRi are significant and that differences in the DNA of isolates are not limited to differences in just one restriction site sequence. When many restriction enzymes were used to cleave the DNA of M. ovipneumoniae isolates, it was found that some gave partial digests of the DNA of some isolates and one isolate was totally uncleaved by either Xhol or HaeIII (Table 2 ). This finding raised the possibility of the presence of methylated bases i.e. 6-methyladenosine or 5-methylcytosine. However, since the partial digests were associated with C-rich rather than A-rich recognition sequences (Table 2 ), it seemed more likely that 5-methylcytosine rather than 6-methyladenosine could be present. The absence of 6-methyladenosine was confirmed by exposing the DNA ofM. ovipneumoniaestrain L6T3 to Dpnl which cleaves the sequence 5'-GmAvTC-3 ' only if the adenine is methylated (Chart and Ross, 1984). No digestion occurred (Fig. 4a). MboI which also cleaves the sequence 5'-VGATC-3' but only in the absence of 6-methyladenosine gave complete digestion, as did Sau3AI which is unaffected by the presence or absence of adenine methylation. When the DNA of this isolate was exposed to Mspl, which cleaves the sequence 5'-CCGG-3' whether or not the internal cytosine is methylated, complete digestion occurred. In contrast to this Hpall which cleaves the same sequence but only in the absence of methylation, gave incomplete digestion l=io Ah ] thn~ i n d i e a t i n t , the. nre~ence_ a f 5 - m e t h y ! c y t o s i n e in t h e DNA_ T h i s
DNA was degraded to nucleosides and examined by HPLC which confirmed the presence of 5-methylcytosine in this isolate. The presence of 5-methylcytosine in the DNA of some other mycoplasma species has been reported (Razin and Razin, 1980; Chan and Ross, 1984), but in the present context the question arises as to whether or not the observed REA differences depend on variations in the extent of methylation of DNA, rather than differences in base sequence. However, this is excluded because differences in the isolates were observed when the DNA was cleaved by Dral which recognises the sequence 5'-TTTVAAA-3 ' and consequently is not affected by the presence of 5'-methylcytosine in the DNA. Furthermore, if differences in the DNA were assumed to be iimited to methylation ofbases this could not account for the differences in proteins as shown by SDS-PAGE. REFERENCES Alley, M.R. and Clarke, J.K., 1979. The experimental transmission of chronic non-progressive pneumonia. N.Z. Vet. J., 27: 217-220.
360
O. IONAS ET AL.
Alley, M.R. and Clarke, J.K., 1980. The effect of chemotherapeutic agents on the transmission ofovine chronic non-progressive pneumonia. N.Z. Vet. J., 28: 77-80. Chan, H.W. and Ross, R.F., 1984. Restriction endonuclease analysis of two porcine mycoplasma deoxyribonucleic acids: Sequence-specific methylation in the Mycop/asma hyopneumoniaegenome. Int. J. Syst. Bacteriol., 34: 16-20. Clarke. J.K., Brown, V.G. and Alley, M.R., 1974. Isolation and identification of mycoplasmas from the respiratory tract of sheep in New Zealand. N.Z. Vet. J., 47:591-596. Foggie, A., Jones, G.E. and Buxton, D. 1976. The experimental infection of specific pathogenfree lambs with Mycoplasmaovipneumoniae. Res. Vet. Sci., 2 I: 28-35. Ford. J.P., Coca-Prados, M. and Hsu, M.T., 1980. Enzymatic analysis of 5- methylcytosine content in eukaryotic DNA, J. Biol. Chem., 255: 7544-7547. Frey, M.L., Hanson, R.P. and Anderson, D.P., 1968. A medium for the isolation of avian mycoplasmas. Am. J. Vet. Res., 29:2163-2179. Gilmour, J.S., Jones, G.E. and Rae, A.G., 1979. Experimental studies ofchronic pneumonia of sheep. Comp. Immun. Microbiol. InfecL Dis., I: 285-293. lonas, G., Clarke, J.K. and Marshall, R.B. 199 I. A Study of the heterogeneity of isolates of M. o~'ipneumoniaefrom sheep in New Zealand. Vet. Microbiol, 29: 339-347. Jones, G.E., Gilmour, J.S. and Rae, A., 1978. Endobronchial inoculation of sheep with pneumonic lung-tissue suspensions and with the bacteria and mycoplasmas isolated from them. J. Comp. Path., 88: 85-96. Jones, G.E., Gilmour, J.S. and Rae, A., i982a. The effect of Mycoplasma ovipneumoniaeand Pasteurellahaemol.vticaon specific pathogen-free lambs. J. Comp. Path., 92:261-266. Jones, G.E., Gilmour, J.S. and Rae, A., 1982b. The effects of different strains of Mycoplasma oripneumoniae on specific pathogen-free and conventionally-reared lambs. J. Comp. Path. 92: 267-272. Laemmli, U.K., 1980. Cleavage of structural proteins during the assembly of th~ head of bacteriophage T4. Nature. 227: 680-685. Mew, A.J., lonas, G.~ Clarke, J.K., Robinson, A.J. and Marshall, R.B., 1985. Comparison of Mycoplasma oripneumoniaeisolates using bacterial restriction endonuclease DNA analysis and SDS-PAGE. Vet. Microbiol., 10: 541-548. Razin, A. and Razin, S., 1980. Methylated bases in mycoplasma DNA. Nucleic Acids Res., 8: 1383- ! 390.