European Journal of Epidemiology 10: 743-748, 1994. © 1994 Kluwer Academic Publishers. Printed in the Netherlands.

Stability of genomic DNA fragment patterns in methicillin resistant Staphylococcus aureus (MRSA) during the course of intra- and interhospital spread

W. Witte 1, C. Cuny 1, O. Zimmermann 2, R. Riichel 3, M. H6pken 4, R. Fischer ~ & S. Wagner 6 I Robert Koch-lnstitute, Wernigerode, 2 Institute of Medical Investigations and 3 Institute of Medical Microbiology at the University of Giittingen; Institutes for Medical Investigations at 4 Hannover and ~ Magdeburg; 6 Hospital Zehlendorf, Berlin, Germany

Accepted in revised form 14 September 1994

Abstract. The analysis of genomic DNA fragment patterns has revealed as a powerful tool for strain dis- crimination in Staphylococcus aureus; for use as an epidemiological marker, stability during the course of an outbreak is an essential prerequisite. Genomic DNA fragment patterns (SmaI restriction, pulsed- field electrophoresis) of four different epidemic MRSA strains were compared along with intra- and interhospital and country-wide spread over more than 12 months in Germany. Strain I was isolated from infections in 8 hospitals. In one hospital a subclone arised which differed from the original strain by 4 fragments. Strain II was spread among 4 hospitals,

Key words: Genomic DNA fragments, MRSA strains

isolates from three of these hospitals exhibited a variability of one to three fragments in the 150-200 kb range. Two hospitals in the Hannover-area were affected by strain III; in 17 isolates of this strain a variability up to three fragments was found in the 170-200 kb range. Strain IV was isolated from 19 cases of infections in 3 hospitals in Berlin. The fragment patterns were completely stable. When S. aureus strains are typed by genomic DNA fragment patterns, a variability in a definite range of molec- ular masses during the course of an epidemic should be taken into consideration.

Introduction

Genomic DNA fragment patterns derived after digestion of DNA by the rare cutting restriction endonuclease SmaI became a powerful tool for typing multiresistant Staphylococcus aureus from nosocomial infection [1-3]. Although evidence provided by [4] suggests a clonal ancestral origin of methicillin resistant Staphylococcus aureus (MRSA), subsequent evolution with regard to alloenzyme patterns is rather likely [5], and genetic rearrange- ments have probably led to different distribution patterns of SmaI restriction sites on the S. aureus chromosome [6]. Nevertheless according to similar- ities in genomic-DNA fragment patterns, MRSA obviously form a cluster of related strains when compared to methicillin susceptible Staphylococcus aureus (MSSA) [3]. Besides a broad capacity of genomic DNA fragment patterns for strain discrim- ination, their stability during the course of an outbreak is an important prerequisite for use as an epidemiological marker. This contribution reports the stability of genomic DNA fragment patterns in two overregional disseminated MRSA and in MRSA from two regional outbreaks.

Material and methods

All of the investigated MRSA were sent for typing to our laboratory during the time period from 1991 to 1993, mostly on agar slants. Prior to genomic DNA-fragment pattern analysis the strains were grown on nutrient agar plates containing 3% (v/v) of fresh sheep blood. All of the investigated isolates originate from different individual patients during the study period; 'copy' isolates from the same patient and isolates from the hospital environment were not included.

Genomic DNA fragment pattern analysis: for cell lysis, deproteinization and endonuclease digestion the protocol according to [1] was followed. The BioRad CheflI system was used for pulsed-field elec- trophoresis with the following conditions: 0.5 M Tris-Borate buffer (pH 8.5); 1% agarose (medium EEO from SIGMA); 200 V; pulse times 5 s to 60 s for 15 h and 60 s to 90 s for another 15 h for SmaI digests [7] and 2 s to 8 s for 12 h and 10 s for another 12 h for SaclI digests. Oligomers of phage ~. 0~-genome ladder) served as mol. mass standard differing in size for one k (48.5 kb).

Phenotypic characterization: phage-typing and

744

crystal-violet-reaction were performed as described [8], for resistance determination the micro broth assay according to DIN 58940 [9] was used.

Resul t s

Phenotypic characteristics of the four different MRSA strains. Strain I and III are nontypable by phages, strains I, III and IV exhibit crystal violet- type C, strain II crystal violet-type A. Whereas strains I and III exhibit a broad pattern of multiresistance, strains II and IV are less multiresistant (data shown in Table 1).

Pattern stability of the 'North-German' epidemic MRSA-strain I. This strain was disseminated among hospitals in the northern part of Germany from autumn 1992 until now. Genomic DNA fragment patterns of strains isolated from of the affected hospitals are shown in Figures 1 and 2. A remarkable stability of SmaI generated fragment patterns was

observed for isolates from hospitals A, C, D, F, G and L over a time period of more than 1 year. Of special interest is the development of a subclone in patient O in hospital A where MRSA with the same resistance phenotype but two different fragment patterns were observed. Whereas all of the other fragments are of identical position, the subclone does not posses the bands in the 400 kb and in the 670 kb range but exhibits two 'new' bands, one in the 480 kb and the other in the 580 kb range. The subclone later on was disseminated in parallel to the original strain in hospital A as well as in hospital B which is located in the same city. That this subclone devel- oped from MRSA-strain I is further indicated by the genomic DNA fragment pattern obtained after digestion with restriction endonuclease SaclI which generates more and therefore smaller fragments (Figure 7). Among 11 isolates derived from a corre- sponding number of infections in hospital G, 1 isolate exhibited a loss of the 310 kb band and a new band of about 270 kb.

Table 1.

strain

I

Typing characteristics of epidemic MRSA

phage-pattern

a,b,c NT 100 F(TD

1! a,b,c NT 100 RTD t

clumping coagulase DNase crystal- factor ~

+ + C

+ + A

111 a 54,75,77,+ RTD b 92 RTD + + + C

I c 618,623,629,630,+ RTD IV a,b NT 100 RTD + + C

c 9 t 100 RTD

resistance phenotype genomic DNA fra merit attern

PEN,OXA,GEN,ERY, I i CLI TMP CIP,RIF ~ 900 kb*~ PEN,OXA, GEN,ERY, CLI.CIP ! PEN,OXA,CMP,TET, 500 kb-4 ERY,TMP,GEN,CIP i

...... i PEN,OXA,ERY,CLI, CIP m

I 200 kb--~ i ~ i 100 kb~

.... ~. P¢[I [ m ......

PEN = penicillin; OXA = oxacillin; GEN = gentamicin, ERY = erythromycin, CLI -- clindamycin, TET = oxytetracy- clin; TMP = trimethoprim-sulfamethozazole, CIP = ciprofloxa¢ine, RIF = ifampicin, Z = mol. mass standard. For phage-patterns: a = Internatinal Basic Set for phage-typing; b -- experimental phages 616-630; c = experimental phages 88-93. 1 A part of the isolates (48 from 55) resistent to RIF.

ho~ital A in town X patient O, n hospital A hospital A hospital B in town X hospital C. spring t993 Jun~ 1993 autumn 1993 July - September 1993 May 1993

Figure 1. Stability of genomic DNA fragment patterns (SmaI) during the course of intra- and interhospital dissemina- tion of MRSA-strain I among hospitals A to C.

7 4 5

hosptaf:D hospital E hospitalF ~ June 1993

hospital G

Figure 2. Stability of genomic DNA fragment patterns (SmaI) of strain I during the course of intra- and interhospital spread among hospitals D, E, F, G.

kb

1 2 3 4 5 6 7 ;L 1 2 3 4 5 ), 6 7 8 9 10 2,. I 2 3 4 5 1 2 3 4

hospital H hospital H hospital I hospital J July 1992 August 1993 November 1992 May 1993

Figure 3. Stability of genomic DNA fragment pattems (SmaI) of strain 1I during the course of intra- and interhospital spread among hospitals H to J.

Pattern stability in MRSA-strain H spread in the South West of Germany. The occurrence of this strain came at first to our attention in July 1992 in con- nection with an outbreak in hospital H, in which two different MRSA strains were involved [10]. The seven isolates exhibited identical SmaI-genomic DNA fragment patterns. Among 10 MRSA isolated in August 1993 from infections in the same hospital, the fragment pattern in 7 isolates was completely identical to the original strain, in 3 isolates the fragment in the 200 kb range was replaced by an about 30 kb larger fragment. In November 1992 this strain was isolated from 5 cases of infections in hospital J. In isolates 1-4 the pattern of the original strain was found, in isolate 5 two bands were present in the 200 kb range. In May 1993 4 MRSA were isolated in hospital J, their fragment patterns corre- spond to those of original strain II (Figure 3). When 9 isolates of this strain from corresponding infections in hospital K in April 1993 were analyzed for

genomic DNA fragment patterns (Figure 4), 2 exhib- ited a loss of the fragment in the 170 kb range, and in 2 isolates the fragment in the 200 kb range was about 30 kb larger. The SaclI digests of the genomes of strains 1 and 2 from hospital K (Figure 4) were identical except for the loss of a band in the 180 kb range in that strain which did not exhibit the 170 kb band of the SmaI digest (Figure 7).

Pattern stability of MRSA strain 111 during the course of an outbreak in an urological, surgical department of hospital L and spread to hospital M. The first MRSA outbreak in hospital L was recorded in June 1992. Figure 5 shows the fragment patterns of 6 isolates. For 5 isolates the Sinai-patterns were iden- tical whereas for one isolate an additional fragment was present in the 160 kb range. When MRSA were isolated again in this hospital in September-October 1993, the isolates have shown the same pattern as the original strain. This strain was also isolated from 4

146

;L 1 2 3 4 5 6 7 8 9

hospital K

Figure 4. Stability of genomic DNA fragment patterns (SmaI) of strain II during the course of intra- and inter- hospital spread among hospital K.

cases of infection in hospital M. These isolates from hospital M exhibit an additional variation in the 150-200 kb range. When isolates 1, 2, 3 and 4 from hospital M (according to Figure 5) were analyzed by SaclI digestion, also differences were as observed in few bands in the molecular mass range from about 70 kb to 180 kb (Figure 7).

kb

700-

Pattern stability of MRSA-strain IV from an outbreak of infections in hospital N and sporadic infections in 2 further hospitals of the large city B. From December 1992 until September 1993 16 MRSA- isolates were isolated in hospital N exhibiting an identical fragment pattern generated by SmaI. The same pattern was observed in MRSA from sporadic infections in hospitals 0 and P (Figure 6).

Discussion

Genomic DNA fragment patterns derived after restriction endonuclease digestion are based on the distribution of recognition sites on the bacterial genome. A single genetic event such as a point mutation affecting the recognition site or an insertion or deletion of a DNA sequence carrying the corre- sponding site can lead to the generation of a three band-difference. Acquisition or loss of transposons and prophages are examples for such kind of events. Thus, a variability of fragment patterns in the course of an epidemic should be expected. Colonizing or infecting a new macro organism, a particular strain meets a new micro environment with opportunities for acquisition of new genetic elements.

Altogether the four different strains observed retained their characteristic fragment patterns along with their clonal dissemination to different patients and to different hospitals. The pattern of clone I was remarkably stable during it's spread besides the development of subclone A with a four-band differ- ence when compared to the parent strain. This dif- ference can not simply be explained by a gain or loss of a DNA sequence since the sum of the tool. masses

400-

200-

50-

;L t x 2 3 4 5 6 1 2 3 4 L 1 2 3 4 5 6 7 ;L hospital L hospital M hospital L June 1992 September t 993 September - October 1993 (x represents strain I)

Figure 5. Stability of genomic DNA fragment pattems (SmaI) of strain III in hospital L and in hospital M.

747

-400

-200

-50

X t 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 1718 19 X

Figure 6. Stability of genomic DNA fragment patterns (SmaI) of strain IV during the course of intrahospital dissemi- nation in hospital N (1 to 16) and spread to hospitals O (17) and P (18, 19).

0

0

the pattern variability is restricted to definite frag- ments; in this case in the 150 to 200 kb range. The fragment pattern of clone IV was completely stable.

Our observations of the stability of genomic DNA fragment patterns in the investigated MRSA indicate that there are probably definite regions of the staphylococcal chromosome which are affected by rearrangements. This should be taken into consider- ation when SmaI-generated macrorestriction patterns are used for epidemiological tracing of S. aureus.

4 5: i ~ ~ 1 2 . . . . . . . . . X 1 2 3 4

A 0 K M

Figure 7. Genomic DNA fragment patterns obtained after SaclI digestion. A: hospital A; isolates 4 and 5; B: patient O in hospital A; isolates 1 and 2; K: hospital K; isolates 1 and 2; M: hospital M; isolates 1, 2, 3 and 4.

of the two fragments in question is about 1000 kb as it is for both fragments in question of the subclone. An explanation is an inversion of a rather large piece of the chromosome. As observed this should have no (or only little) influence on the banding (fragment) pattern derived after digestion with a more frequent cutting restriction enzyme (SaclI). One among the 11 isolates from hospital G revealed as a further subclone with a probable loss of a 20 kb sequence from the original 305 kb fragment.

A variability for clone II was only observed for a fragment in the 200 kb range, even when isolates were compared with had been obtained after a nearly one year difference. As in clone II also in clone III

References

I. Goering RV, Duensing TD. Rapid field inversion get electrophoresis in combination with an rRNA gene probe in the epidemiological evaluation of staphy- lococci. J Clin Microbiol 1991; 28: 426-429.

2. Prevost G, Jaulhac B, Piemont J. DNA fingerprinting by pulsed field gel electrophoresis is more effective than ribotyping in distinguishing among methicillin resistant Staphylococcus aureus isolates. J Clin Microbiol 1992; 30: 967-973.

3. Witte W, Cuny Ch, Claus H. Clonal relatedness of Staphylococcus aureus strains from infections in humans as deduced from genomic DNA fragment patterns. Med Microbiol Lett 1993; 2: 72-79.

4. Kreiswirth B, Kornblum JU, Arbeit RD, Eisner W, Maslow JN, McGeer A, Low DE, Novic RP. Evidence for a clonal origin of methicillin resistance in Staphy- lococcus aureus. Science t989; 259: 227-230.

5. Musser JM, Kapur V. Clonal analysis of methicillin- resistant Staphylococcus aureus strains from inter- national sources: phylogenetic lineages implies dissemination by horizontal transfer and recombina- tion. J Clin Microbiol 1992; 30: 2058-2063.

6. E1-Adhami W, Roberts L, Vickery A, Inglis B, Gibbs A, Stewart PR. Epidemiological analysis of a methicillin-resistant Staphylococcus aureus outbreak using restriction length polymorphisms of genomic DNA. J Gen Microbiol 1991; 137: 2713-2720.

748

7. Witte W, Grimm H, Occurrence of quinolone resis- tance in S. aureus from nosocomial infections. Epidemiol Infect 1992; 109: 413-421.

8. Witte W, Richardson JF, Marples RR. Complex typing of methicillin resistant Staphylococcus aureus (MRSA). Zbl Bakt Hyg 1988; A68" 76-82.

9. Deutsches Institut for Normung e.V. Methoden zur Empfindlichkeitspriifung yon Krankheitserregern gegen Chemotherapeutika. Berlin: Beuth Verlag, 1990: Mikrodilution DIN 58940, Teil 8.

10. Cuny Ch, Schassan H-H, Witte W. Outbreak of noso- comial infections with two different MRSA-strains involved: significance of genomic DNA fragment patterns in strains otherwise difficult to type. Epidemiol Infect 1993; 111: 55-61.

Address for correspondence: Dr Wolfgang Witte, Robert Koch-Institut, Branch Wernigerode, Burgstral3e 37, D- 38855 Wernigerode, Germany